University of Southern California Dissertations and Theses

Heat Flow And Other Geophysical Studies In The Southern California Borderland

(USC Thesis Other)

Heat Flow And Other Geophysical Studies In The Southern California Borderland

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted. The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction. 1.The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity. 2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame. 3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material, it is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete. 4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced. 5. PLEASE NOTE: Some pages may have indistinct print. Filmed a s received. Xerox University Microfilms 300 North Zeeb Road Ann Arbor, Michigan 48106 LEE, Tien-Chang, 1943- HEAT FLOW AND OTHER GEOPHYSICAL STUDIES IN THE SOUTHERN CALIFORNIA BORDERLAND. University of Southern California, Ph.D., 1974 Geophysics University Microfilms, A XEROX Company, Ann Arbor, Michigan THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED. HEAT FLOW AND OTHER GEOPHYSICAL STUDIES IN THE SOUTHERN CALIFORNIA BORDERLAND by Tien-Chang Lee A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Geological Sciences) August, 1973 UNIVERSITY OF SOUTHERN CALIFORNIA T H E G R A D U A TE S C H O O L U N IV E R S IT Y P A RK LO S A N G E LE S , C A L I F O R N IA 9 0 0 0 7 This dissertation, w ritten by Ti.en-Chaivg..Le.e..................... under the direction of A . i . ? . . . Dissertation Com mittee, and approved by a ll its members, has been presented to and accepted by The Graduate School, in p a rtial fu lfillm en t of requirements of the degree of D O C T O R O F P H IL O S O P H Y DISSERTATION COMMITTEE •man TABLE OF CONTENTS Page ABSTRACT..... ................................... INTRODUCTION 1 General Statement 1 Previous Work on Heat Flow 2 Present Problems 5 Purposes 7 Acknowledgements 8 REGIONAL GEOLOGY AND GEOPHYSICS 9 Geology 12 Continental Borderland 12 Adjoining physiographic provinces . . 18 Geophysics 20 Seismic refraction profiles .... 21 Gravity 25 Geomagnetism 30 HEAT FLOW 34 Methods 34 Geothermal gradient 34 Thermal conductivity 39 Results 41 Geothermal gradient 42 Thermal conductivity 57 ii Page Heat flow 59 Perturbing Effects on Heat Flow .... 86 Sedimentation 86 Bottom water temperature 91 Relief and conductivity contrast .. 94 Discussion 97 San Diego Trough 97 San Pedro Basin 108 San Clemente Basin 113 Santa Catalina Basin 118 Santa Cruz Basin 122 San Nicolas Basin 126 Tanner Basin 130 Santa Barbara Basin 132 East Cortes Basin 1.33 Miscellaneous locations 134 Continental shelf off central ' California 136 Heat Flow Distribution Pattern 137 SEISMIC AND MAGNETIC PROFILES 150 Seismic Reflection Profiles 150 Magnetic Profiles 151 Presentation and Discussion of Data ... 153 San Pedro Basin 153 iii iv Page Santa Catalina Basin 154 San Nicolas Basin 156 Tanner Basin 157 San Clemente Basin 159 Animal Basin 160 Velero Basin 161 Outer Basin 161 Continental Slope 162 Santa Barbara Basin 163 Magnetic profiles across Continental Borderland .... 163 TECTONIC IMPLICATIONS FROM HEAT FLOW 183 Tectonic Model 184 Implications from Heat Flow............. 196 SUMMARY AND CONCLUSIONS 216 REFERENCES 222 APPENDICES 234 I. Calibration of Thermistor Probes . 235 II. Standard Error of Heat Flow Measurement 252 III. Heat Flow Data in the Southern California Borderland ..... 253 IV. Effects of Non-cyclic Surface Temperature Variation on Temperature Gradient ....... 272 iv V Page V. Heat-flow Refraction across Boundary of Dissimilar Media - 1. Analytical Solutions .... 278 Semi-cylinder 279 Semi-sphere 281 Semi-elliptical cylinder .... 282 Two parallel semi-cylinders ... 289 Semi-ellipsoid 297 Buried cylinder 306 VI. Heat-flow Refraction across Boundary of Dissimilar Media — 2. Finite Element Method .... 310 v ILLUSTRATIONS Figure Page 1. Physiographic provinces adjoining the Continental Borderland 10 2. Trans-borderland crustal structure based on seismic refraction profile.. 22 3. Trans-borderland crustal structure based on Bouguer gravity ......... 26 4. Magnetic anomaly pattern 32 5. Schematic electrical diagram for the thermograd recorder 35 6. Geothermal gradients in San Diego Trough 43 7. Geothermal gradients in San Pedro Basin 45 8. Geothermal gradients in San Clemente Basin 47 9. Geothermal gradients in Santa Catalina Basin 49 10. Geothermal gradients in Santa Cruz Basin 551 11. Geothermal gradients in San Nicolas Basin 53 12. Geothermal gradients in Tanner Basin. 55 13. Thermal conductivity of sediments in San Diego Trough 61 14. Thermal conductivity of sediments in San Pedro Basin 64 vi vii Figure Page 15. Thermal conductivity of sediments in San Clemente Basin 66 16. Thermal conductivity of sediments in Santa Catalina Basin 69 17. Thermal conductivity of sediments in Santa Cruz Basin 74 18. Thermal conductivity of sediments in San Nicolas Basin ............... 77 19. Thermal conductivity of sediments in Tanner Basin 80 20. Thermal conductivity of sediments in East Cortes Basin ............... 82 21. Thermal conductivity of sediments in Santa Barbara Basin ............... 84 22. Sedimentation effect on geothermal gradient 88 23. Location of basins in the northern Borderland 98 24. Heat flows through the Continental Borderland off southern California ... 100 25. Corrected heat flows through the Continental Borderland off southern California 102 26. Heat flow through San Diego Trough .. 105 27. Heat flow through San Pedro Basin .. 110 28. Heat flow through San Clemente Basin . 114 29. Heat flow through Santa Catalina Basin 119 30. Heat flow through Santa Cruz Basin .. 124 31. Heat flow through San Nicolas Basin .. 128 32. Heat flow profile 111-111' 141 vii viii Figure Page 33. Trans-borderland heat flow profile II-II' 144 34. Trans-borderland heat flow profile I-I' 146 35. Seismic reflection profiles in Santa Catalina Basin, San Pedro Basin, and San Nicolas Basin 165 36. Seismic reflection profiles in Tanner Basin, across Patton Ridge, and on shelf west of Pt. Conception ..... 16 7 37. Seismicr refleciton profiles in San Clemente Basin, Animal Basin, San Diego Trough, and at southeastren end of Southwest Bank 169 38. Seismic reflection profiles in Velero Basin, Outer Basin, Santa Cruz Basin, West Cortes Basin, and the Continental slope southeast of Outer Basin ..... 171 39. Magnetic profiles in San Pedro Basin and Santa Catalina Basin 173 40. Magnetic profiles in San Nicolas Basin, Animal Basin, and San Clemente Basin .. 175 41. Magnetic profiles in Tanner Basin, West Cortes Basin, Velero Basin, Outer Basin, Rampart and Animal Basin ............. 177 42. Trans-borderland magnetic profile CG .. 179 43. Trans-borderland magnetic profile ST and Santa Monica Basin 181 44. Relative position of the Pacific, Farallon, and North American plates ... 186 45. Relative position of ridge, trench, and transform fault 189 46. Relative plate motion as a function of time 192 viii ix Figure Page 47. Heat flows from ocean basin, continental margin, and shield ...... 198 48. Relation of heat flow and age ....... 202 49. Heat flow contour map in the western United States and eastern Pacific Ocean 206 50. Effect of lateral heat transfer ....... 211 51. Cooling rate of a slab 211 52. Paleotemperature in Tanner Basin ..... 276 53. Coordinates used in solving a problem of noncyclic surface temperature variation. 276 54. Cylindrical coordinates ............ 2 80 55. Spherical coordinates 280 56. Elliptical-cylindrical coordinates .... 283 57. Bipolar coordinates 283 58. Refraction of heat flow around two parallel semi-cylinders 295 59. Refraction of heat flow around a semi ellipsoid 304 60. Refraction of heat flow due to a buried cylinder 308 61. Perturbation of heat flow around a parallelepiped with infinite length ... 319 62. Perturbation of heat flow due to a plane-slope 323 63. Perturbation of heat flow along heat flow profile I-I' due to topography, bottom water temperature variation, and conductivity contrast 327 ix LIST OF TABLES Table Page 1. Comparison of thermal conductivity in different basins 60 2. Calibration data of a few selected thermistors 242 3. Heat flow data in the southern California Borderland 253 x ABSTRACT Heat flow measurement as well as magnetic and seismic reflection profiling were made to investigate the crustal structure and its evolution in the Continental Borderland off southern California. Fifty-three new heat flow values together with 12 published values show that the average 2 heat flow 1.9 (pea1/cm sec) through the Borderland off southern California is higher than the average 1.5 of the heat flow through the earth's surface. As a whole, heat flows increase systematically southward and slightly land ward, with probably a relatively high heat flow zone associated with troughs within the Patton Ridge. In individual basins heat flows are fairly uniform as demon strated in Santa Catalina, San Nicolas, Santa Barbara, and East Cortes Basins, where standard deviations are 0.1 or less. After removal of the perturbing effects of topog raphy and sedimentation, uniform heat flows also appear in central San Pedro Basin, northern San Clemente Basin, and San Diego Trough, but a linearly north-south increasing heat flow trend appears in Santa Cruz Basin. Thermal conductivities increase with depth at a rate of 2 to 3 % per meter. Greater variation in conductivity is found in the inner basins than in outer basins. xi xii Seismic reflection profiles support the delineation of the northwest-trending major faults of Moore (1969) and hence the structural zonation in the northern Borderland, namely, an inner fault belt, outer fault belt, central fold belt, and northern transverse structural belt. In addition, the profiles reveal transverse faults in Santa Catalina, and San Nicolas as well as San Pedro Basins. These transverse structures may be responsible for the east-trending gravity anomalies as reflected by components with intermediate wavelengths. A few places previously mapped as volcanic or basement highs are not associated with magnetic anomalies, suggesting that mapping based on few dredgings was probably inconclusive, or the rocks forming the topographic highs have low magnetic suscepti bility or were reversely magnetized. In general, the magnetic anomaly pattern trends northwestly, in contrast to the north-trending linear magnetic anomaly pattern over the eastern Pacific Ocean Basin. The southeastward increasing heat flow is attributed to a southeastward migrating ridge-trench-transform triple junction on the hasis of a plate tectonic model which suggests that the ridge-trench impingement began 29 m.y. ago at the northern end of the present Continental Border land. The relative displacement between the Pacific and North American plates was initially taken up along an xii xiii offshore transform fault assumed to be present but now defunct at the base of the continental slope. During the course of the plate interaction, the displacement was transferred more and more inland along strike-slip faults in the western U.S. until 5 m.y. ago; the San Andreas fault system has since taken up all the relative displacement. xiii INTRODUCTION General Statement This study is primarily concerned with heat flow from the sea floor off southern California and its relationship to this region's crustal structure and geologic history. The measurements of heat flow were started in the summer of 1970 and ended in 1972 on board R/V VELERO IV. Magnetic profiling and seismic reflection profiling accompanied the heat flow studies. The state of the problems on heat flow studies is in troduced in this chapter but the heat flow data are pres ented in a later chapter following the review of regional geology and geophysics. The presentation of magnetic and seismic reflection profiles is preceded by an introductory statement concerning previous work and the technique used in this study. The Cenozoic tectonic models of Atwater (1970) for the western United States are modified on the basis of heat flow data. A study on the perturbation of regional heat flow due to near surface conductivity contrast is presented in the APPENDICES IV and V. Analytical solutions for heat-flow refraction across boundary of dissimilar media are given for a number of useful configuration such as semi-sphere, -cylinder, and -ellipsoid, two parallel semi-cylinder, and buried cylinder. For arbitrary geometry, a numerical solu tion based on the finite element method is given. Previous Work On Heat Flow Numerous heat flow measurements made in the western United States during the last two decades delineate signif icant heat flow patterns (Sass and others, 1971; Roy and others, 1972). Several heat flow provinces are strongly correlated with the major physiographic provinces. It is widely known that heat flows in the Basin and Range prov ince are typically high, whereas those in the adjacent Sierra Nevada are low to normal. In the Pacific coastal region and northwestern Columbia Plateau' the heat flow is normal. High heat flow is found in the northern and south ern Rocky Mountain, with a heat flow low located in the central Rocky Mountain. Sparse data indicate that the Colorado Plateau has normal values in its north-central part but becomes higher near the eastern margin. In central' California, high heat flow is characteristic of the Franciscan block to the east of the San Andreas fault and may also continue into the Salinian block west of the fault (Sass, 1973, personal communication). Isofluxes in central California and western Nevada are parallel to the regional structural trend (Sass and others, 1971). The Los Angeles 3 Basin probably has slightly above normal heat flow and the Peninsular Ranges have low to normal values. The linear correlation between heat flow and the radio active heat production of the surface rocks has important implication on the vertical distribution of radioactive el ements (Roy and others, 1968; Lachenbruch, 1968). Studies in the Idaho batholith (Swanberg, 1972; Swahberg.and Blackwell, 1973) suggest that sources are distributed ex ponentially whereas a linear relation appears to be reason able in many heat flow provinces of the world and conse quently permits comparison of mantle heat flows on the basis of near surface measurements (Roy and others, 1972). Transition zones between adjacent heat flow provinces are usually on the order of 100 km or less (Roy and others, 1972). These sharp transitions are not well understood. Tactonically the problem, as in the case of the Sierra Nevada - Basin and Range transition, is qualitatively in terpreted as representing transient shallow-seated thermal events resulting from subduction beneath the western United States. Noteworthy is the lack of a definite heat flow anomaly associated with the San Andreas fault system (Henyey, 1968; Henyey and Wasserberg, 1971; Sass and others, 1971) believed to be an active ri.dge-ri.dge trans form fault by many workers (Wilson, 1965; Morgan, 1968). Oceanic heat flow data from the eastern Pacific Ocean floor have confirmed the suggestion that the Juan de Fuca ridge (Lister, 1970; Korgen and others, 1970) and East Pacific rise are spreading centers (Von Herzen and Uyeda, 1963; Von Herzen, 1964; McKenzie and Sclater, 1969). High heat flow in the Gulf of California indicates that a spreading center has encroached into at least part of the North American continent. Outside these heavily sampled areas, the heat flow data are comparatively rare. Prior to 1965, there were fewer than 10 oceanic heat flow measurements between San Francisco and the United States - Mexican border and to the east of meridian 124° W. This data gap included the borderland off southern California. Based on a dozen sparsely spaced data points, Von Herzen (1964) observed that heat flow in the Continental Borderland decreases systematically from south to north. In two transverse profiles across the Borderland to the Pacific Ocean, he found a periodic variation of heat flow with a characteristic wavelength of about 250 km. Other heat flow measurements on the continental shelf off the western United States have been made by Fo:ster; (1962) and Korgen and others (1971). Additional measurements at sea near the continental margin have been made by Bullard and others (1956), Foster (1962), Von Herzen (1964), Burns and Grim (1967), Vacquier and others (1967), and Korgen and others (1971). Present Problems Due to the large thermal time constant and hence the slow transfer of heat in the earth, the present surface heat flow at a given locality may be the integrated result of more than one thermal event. The property of the dif fusion equation prohibits the deduction of the original temporal source distribution from the observed surface heat flux. As difficult as it may be to identify each thermal event, interpretations can often be made with modeled source functions subjected to the physical and geological constraints. Sclater and Francheteau (1970) have demonstrated that the oceanic heat flow decreases with the age of the crust, as does the continental heat flow. It is with this time-dependent property of terres trial heat flow that one may try to test the validity of a proposed regional tectonic model by an independent method. This process thus narrows the non-uniqueness encountered in inverting geophysical and/or geological data. Based on the magnetic anomaly patttern in the eastern Pacific, it has been postulated that subduction of the East Pacific rise by a trench off California started about 29 m.y. ago (Atwater, 1970). Two triple junctions of the transform-transform-trench type and the ridge-trench- transform type evolved at the time when the North American and Pacific plates started to encroach on each other. They have since migrated northwestward and southeastward with respect to each other to break up the East Pacific Rise into two parts, one off Oregon and Washington and the other off Central America. However, the place of initial ridge-trench impingement has not been deduced owing to the unknown relative motion between the Pacific and North Amer ican plates in the past. Events such as the formation of the Continental Borderland, the opening of the Gulf of Cal ifornia, the evolution of the San Andreas transform fault, and the genesis-of the Transverse Ranges are still not well understood in terms of the plate, interactions. Because of the proposed time sequence of plate interaction, the heat flow data may serve as a constraint to the tectonic modeling. Unfortunately, the aforementioned heat flow data gap off central, southern, and Baja California is associ ated with the missing part of the East Pacific Rise. Extrapolation of continental and oceanic data to a structurally complex area such as the Continental Border land of California to fill a data gap is impossible. The Continental Borderland physiographically and tectonically is a transition zone from continental to oceanic struc tures , and must be treated in the same manner as for a major physiographic tarnsition on land. The measured heat flow is often affected by disturb ing factors near the earth surface such as relief, erosion, deposition, uplift, subsidence, water circulation, and variation in conductivity and surface temperature. For heat flow data measured on land these effects can be large ly removed but in the oceanic heat flow measurements they are often neglected because the local geology and topogra phy around the sites of measurement are not well enough known to make an appropriate correction. Such negligence may be justified in the abyssal region, but in areas such as the Continental Borderland near surface perturbations can be large enough to mask the true nature of heat flow at the deeper level of the crust. Thus comparison between the continental heat flow data and the oceanic heat flow data is biased by the quality of data. The bathymetry, marine geology, and oceanography in the Continental Borderland, especially in the northern part, are reason ably well enough known to permit removing the effects of topography, sedimentation, and temperature variation. Thus it appears possible to study the heat flow transition from continent to ocean and to evaluate its implication on tectonic models although the Continental Borderland is not a typcial continent-ocean transition. Purposes The purposes of this study are: 1) to fill the heat flow data gap in the Continental Borderland off southern California and continental shelf off central California; 2) to investigate the he.at flow in the transitional region between continent and ocean; and 3) to synthesize these heat flow data with other geophysical and geological data so as to make an integrated interpretation on the tectonic regime of the west coast of North America. Acknowledgements The author wishes to thank Dr. Thomas L. Henyey whose interest and advice made this study possible. He also wishes to thank Drs. Gregory A. Davis, Ta-liang Teng, and William G. Wagner for their criticism, and comments on the manuscript. The field work of this study has been aided by Mr. D. V. Manov of University of Southern California, Mr. E. E. Welday of Division of Mines and Geology, Dr. J. Combs of University of California at Riverside, and many students from University of Southern California, California State University at Fullerton, California Institute of Technology, and University of California at Los Angeles. Help from the officers and the crew of the R/V VERLERO IV is appreciated. In addition, the author has been benefited from discussions with Drs. 0. L. Bandy, D. S. Gorsline, and R. Kolpack, Mr. L. Lawver, Mrs. Z. M. Lee, Mr. J. P. Tung, and other members of the Geophysical Laboratory at the University of Southern California. This study was supported by the National Science Foundation under Grants GA 13008 and GA 30633. Computer time was provided by the Computing Center, University of Southern California. REGIONAL GEOLOGY AND GEOPHYSICS The sea floor between the mainland shoreline and the continental slope off southern California and northern Baja California has been named the Continental Borderland by Shppard and Emery (1941) for its distinctive basin and ridge topography. The adjoining physiographic provinces include the Transverse Range province to the north, the Peninsular Range province to the east, and the Baja California Sea Mount province to the west and southwest (Figure 1). Each province possesses a unique physiography and characteristic geologic relationships. This chapter is divided into two sections: geology and geophysics. The section on geology deals with near surface crustal sturcture and geology identifiable from the bathy metry and conventional dredging, coring, and seismic re flection profiling. However, discussion on the unconsoli dated sediments is excluded and interested readers are re ferred to the section on interpretation and presentation of reflection seismic data in this study or to the work of Moore (1969) and references therin. The geology of adja cent physiographic provinces will be briefly introduced. The section on geophysics is concerned with the general crustal structure identified by gravimetry, magnetics, 9 Figure 1 Physiographic provinces adjoining the Continental Borderland (after Moore, 1969). 11 RANGES TRANSVERSE — r 60 N. MILES 12 seismic refraction, and surface wave dispersion. Geology Continental Borderland Based on bathymetry, Krause (1965) divides the Con tinental Borderland into northern and southern parts with the boundary defined by the presumed left-lateral Santo Tomas fault (Figure 1). The northern Borderland is char acterized by broad, flat-topped, shallow banks and ridges, and flat bottomed basins. The southern Borderland averages a few hundred meters deeper than the northern Borderalnd and is characterized by either narrow and steep or low and broad ridges. In this study, the northern Borderland is a synonym to the southern California Borderland, as is the* southern Borderland to the Baja California Borderland. Moore (1969) has compiled a structural map for the Borderland based on his interpretation of reflection seis mic profiles in conjunction with the bathymetry and the work of Emery (1960) in the northern Borderland and Krause (1965) in the southern Borderland. As is the case onshore, the offshore structural pattern is composed of two sets of faults, one consisting of primary northwest-trending faults and the other consisting of subordinate east-northeast- trending fault. According to Moore, the northern Border land can be divided into four interrelated structural belts: 1) the inner fault belt, the area east of and in cluding San Clemente Island; 2) the outer belt, the area west of the Santa Rosa - Cortes Ridge and associated banks; 3) the intervening oval-shaped central belt of apparently en echelon oriented folds; and 4) the transverse structural belt, the part north of and including Channel Islands. The geology and structure in the southern Borderland has been described by Krause (1965) and Doyle (1973). The inner fault belt is an offshore extension of the northwest-striking Peninsular Range structure. The San Clemente fault, probably connected with the right-lateral Agua Blanca fault in northern Baja California (Allen and others, 1960) is the most significant fault of the border land because it has the highest level of seismicity and is the only idnetifiable right-lateral fault in the Continen tal Borderland according to Moore's (1960) structural map. Another important fault bounds the eastern side of the Santa Catalina block and appears to continue southeastly along the western edge of San Diego Trough, finally con verging with the San Clemente fault. Onshore the Newport- Inglewood fault, which presumably separates the granitic basement to the east from the Franciscan basement to the west, plays an important role in the Mesozoic borderland tectonics (Hill, 1971). Major faults in the outer fault belt include the Patton Escarpment (continental slope), one delineating the eastern edge of the Patton Ridge, and another following the western edge of the Santa Rosa - Cortes Ridge. Reverse drag folds (for example, see Figure 36) are usually asso ciated with the faults in the outer fault belt whereas they are absent in the inner fault belt (Moore, 1969, p.39). The continental slope off southern California is well defined by the steep Patton Escarpment which becomes less steep in the southern Borderland off Baja California. Along the base of the continental slope, there appears a long, narrow, shallow depression, which deepens southward and is named Cedros Trough or Cedros Tranch (Moore, 1969) off Cedros Island. Folding predominates in the central en echelon fold belt which comprises the anticlinorial Santa Rosa - Cortes Ridge and the associated synclines forming Santa Cruz, San Nicolas and West Cortes Basins. Near the Santo Tomas fault this fold belt becomes less distinct and the inner and outer fault belts converge. In the southern Borderland northwest-striking faults are the dominant features. In addition to the Santo Tomas fault, important transverse faults include the Santa Cruz - Malibu Coast fault which separates the transverse structural zone from the rest of the northern Continental Borderland, one south of the saddle between Santa Cruz and San Nicolas Basins, and another in line with Popcorn Ridge and Punta Canoas in the southern Borderland. These faults are identifiable from the bathymetric chart. The southern sides of the faults appear to be the consistently down-dropped block. On the basis of bathymetry, the Santa Tomas fault is believed to be left-lateral (Krause, 1965) but according to Doyle (1973), the sense of movement cannot be resolved from the bathymetry. The Santa Cruz - Malibu Coast fault is also left-lateral as suggested by the left-lateral offset along the fault in the short-wavelength Bouguer anomaly map (Harrison and others, 1966) and along its onshore extension, the Sierra Madre fault. The rock types in the northern Borderland have been described by Emery (1960) on the basis of dredge samples and onshore lithology. Catalina Schist and related facies, which may be equivalent to the Jurassic-Cretaceous Franciscan Group in central and northern California, are exposed in the Palos Verdes Hills, on Catalina Island (Bailey and others, 1964), on Cedros Island, and Punta Eugenia, Baja California (King, 1969). Similar metamorphic rocks have been dredged from Thirtymile, Fortymile, and Sixtymile Banks. Emery (1960, p.6 6 ) reported that schist in place on a bank about 20 miles north of Sixtymile Bank is similar to the Pelona Schist of the San Gabriel Mountains. Pre-Cretaceous dacite has been found on Santa 16 Cruz Island and Richardson Rock (Palmer, 1965). Granite similar to the granitic rocks of the Peninsular batholith is known to occur only on Santa Cruz Island but granitic basement and associated metamorphic rocks have been found in the deep wells to the east of the Newport- Inglewood fault in Los Angeles Basin (Yerkes and others, 1965). Rocks of Franciscan type have not been found to the east of the Newport-Inglewood fault in the Peninsular Range province unless the Pelona Schist is Franciscan equivalent, whereas granitic rocks have never been ex posed or dredged to the west of the fault and south of Channel Islands. Hence, it may be concluded that the Continental Borderland is underlain by a basement of Fran ciscan type and Los Angeles Basin east of the Newport- Inglewood fault is floored with granitic and associated metamorphic rocks. Location of basement highs in the Continental Borderland can be found from Moore's (1969, Plate 14) litho-orogenic map; possible refinement of the map will be suggested after the discussion of magnetic data. Possible Jurassic green chert and chloritic-altered sandstone have been dredged from several banks west of Tanner Basin (Emery, 1960) but Moore's (1969, p.43) study indicates these banks are basement highs capped by Miocene 17 sedimentary rocks. Noteworthy is the San Onefre Breccia exposed locally along the coast of southern California and northern Baja California (Woodford, 1925; Minch, 1967). This breccia consists of quartzites and glaucophane- bearing schists believed to have come from the Catalina Schist. Fossiliferous Cretaceous shale and sandstone are exposed nearshore around La Jolla and Point Loma. Eocene shale, sandstone, and limestone crop out on the San Nicolas Island and around Point Loma. Miocene andesite and basalt comprise the bulk of the volcanic rocks which are widespread around Santa Barbara, San Clemente, Anacapa, and Santa Cruz Islands; Cortes, Tanner Basin, and Northeast Banks; banks west and southwest of Tanner Basin, Emery Seaknoll, and on several hills in San Diego Trough (Emery, 1960; Moore, 1969, Plate 14). In general, andesite probably dominates the inner half of the northern Borderland whereas basalt is dominant in the outer half (Emery, 1960). Small diabase dikes are present on San Nicolas Island. Tertiary strata are predominantly Miocene sedimentary rocks which include chert, shale, sandstone, limestone and authigenic phosphorites. Most of the enclosed fauna are Luisian and Mohnian. Pliocene mudstone is found on the northeast slope of San Clemente Island, near the sill 18 of San Nicolas Basin, on west slope of the Santa Rosa - Cortes Ridge, northeast slope of the San Pedro and Santa Monica Basins. Late Pliocene to Recent mudstone occurs on the Coronado Bank. The sediments in the basins are believed to range in age from Pliocene to Recent. Adjoining Physiographic Provinces Structurally the Peninsular Ranges have been uplifted on the east along steeply dipping faults, tilted gently westward, and dissected by northwest-striking faults into sub-parallel blocks (Jahns,1954). In southern California the Peninsular Range province is dominated by right-lateral faults of the San Andreas fault system. The Elsinore and San Jacinto faults are two major branches of the system. Pre-Cretaceous marine sedimentary rocks were metamorphosed during episodic intrusion of plutonic rocks which consti tute the great batholith of southern and Baja California. Gabbroic to granitic plutonic rocks and associated meta morphic rocks form the backbone of the Peninsular Ranges. Postbatholithic sequences consist of post-Cretaceous marine and nonmarine sedimentary rocks together with Cenozoic volcanic rocks. Because of the similarity in structural trend, the Continental Borderland south of Channel Islands has been considered as part of the Peninsular Range province (Jahns, 1954). However, the appearance of 19 Franciscan rocks on the offshore islands and ridges or banks, is in contrast to the absence of such rocks in the Peninsular Ranges. It is appropriate to treat the Con tinental Borderland as an individual physiographic province. For this reason the Palos Verdes Peninsula west of the Newport-Inglewood fault, may be considered part of the Continental Borderland. The Transverse Range province., stretching from west to east across the dominant northwestly structural trend in California, is characterized by eight distinct mountain ranges and five well-defined topographic and geologic ba sins (Bailey and Jahns, 1954). Santa Barbara Basin and Chan nel Islands are actually part of the Transverse Range prov ince because of similarities in structural trend and under lying basement rocks. Noteworthy east-west striking faults include the Big Pine fault, which delineates the northern boundary of the western Transverse Ranges, the Malibu- Raymond-Sierra Madre fault system, which separates this province from the Peninsular Range province to the south, and the Santa Ynez fault and Oakridge fault within the ranges. These are cut by the northwestly right-lateral San Andreas and San Gabriel faults. The basement includes Pre- cambrian and Late Cretaceous plutonic rocks and pre-Cenozoic metamorphic rocks. Marine and nonmarine sediments of Late Cretaceous and Middle Miocene ages are widespread in the western Transverse Ranges. Post Middle Miocene sedi 20 mentary rocks are found in the western and southern parts of the Transverse Ranges, Ventura Basin, Santa Barbara Basin, Santa Monica Mountains and on the Channel Islands. The Middle Miocene was also a time of intense volcanism in the Transverse Ranges. The Baja California Sea Mount province to the west of the Borderland lies between the Murray Fracture Zone on the north and the Clarion Fracture Zone on the south. This mountainous region is characterized by abundant Cenozoic submarine volcanoes (Menard, 1955). According to Atwater's (1970) ages of the easternmost magnetic anomalies in the eastern Pacific Ocean, the sea floor immediately to the west of the continental slope ranges in age from 15 m.y. near the southern end of the Borderland to 25 m.y. near the northern end of the Borderland. Geophysics The crustal studies of previous workers are presented in this section. Notable are the seismic refraction studies of Shor and Raitt (1958) and gravity studies of Harrison and others (1966) in the northern Borderland, as well as the refraction studies of Roller and Healy (1963) across southern California. Magnetic studies of Harrison and others (1966) and Krause (1965) will be reviewed here 21 although their data have not been used to construct crustal models. Seismic Refraction Profiles A crustal model from the Pacific Ocean Basin across the northern Continental Borderland to Peninsular Ranges (Figure 2) has been deduced by Shor and Raitt (1958) using seismic refraction data. Variation of depth to the Moho is the most prominant feature of this crustal section. The depth increases eastward rapidly from approximately 10 km to 16 km across the continental slope, it then increases gently to about 32 km beneath the Peninsular Ranges. The crustal thickening toward the continent is mainly a direct response to the increase of lower crustal layer (6 . 6 to 7.0 km/sec), which is equivalent to layer 3 of oceanic crust in terms of P wave velocity. Velocities of the top layer (unconsolidated sediments) appear to range from 2.4 km/sec on the Santa Rosa - Cortes Ridge, 2.8 km/sec on Patton Ridges and in San Nicolas Basin, to 3.8 km/sec in Santa Catalina Basin. These variations are probably not real but rather the result of near surface topographic irregularities. Noteworthy is the overlapping of the 4.5 km/sec - layer and the 5.8 km/sec - layer below the San Clemente block (Shor and Raitt, 1959; also Figure 2). Both layers are within the range of P wave velocity in the layer 2 Figure 2. Trans-borderland crustal structure based on seismic refraction profile from the Pacific Ocean floor to the Peninsular Ranges (after Shor and Raitt, 1958). Numbers are P-wave velocity in km/sec; dashed curves designate interpolation; the profile is about 50 km southeast of heat flow profile II-II" (Figure 24). DEPTH ( km) DISTANCE ( km ) 00 200 3 0 0 4.5' .3.8 5.9 2L8 6.6 6.2 6.6 6.8 6.7 7.0 6.8 20 8.2 7.0 3 0 8.2 to CO 24 of the oceanic crust. The layer with velocity of 5.8 to 6 .2 km/sec is considered by Shor and Raitt to be the upper layer of the crystalline crust (granodiorite). If the material of 4.5 km/sec below the Santa Catalina Island is of Franciscan Group, the exposed Catalina Schist on the island and the Palos Verdes Peninsula may be part of a thrust sheet extending approximately 1 0 0 km over the Peninsular batholithic basement (Figure 2). This sugges tion is probably inconclusive, however, because the 4.9 km/sec layer near Point Loma (San Diego) can be extra polated to the Eocene or Cretaceous sedimentary rocks exposed near La Jolla and Point Loma according to Shor and Raitt. In a seismic refraction profile from Santa Monica Bay, California to Lake Mead, Nevada, Roller and Healy (196 3) found that the crustal thickness is approximately 29 km under Santa Monica Bay, 36 km under the Transverse Ranges, and 26 km under the Mojave Desert. Interpolation of Shor and Raitt's profile indicates that the Moho depth is also 29 km nearshore around Newport Beach, California. However, the Pn velocity differs along the two profiles; it is 7.8 km/sec below southern California, but 8.2 km/sec below the Continental Borderland. A crustal layer with a compressional velocity of 7 km/sec is present in both pro files but continuity of this layer across the Continental 25 Borderland is not yet substantiated. The upper crustal layer has an apparent P wave velocity of 6.3 km/sec near Santa Monica Bay, compared to 5.9 km/sec in the offshore area. Healy (1963) in a seismic refraction profile along the central California coast has found Pn to be 8.2 km/sec be tween Los Angeles and Camp Roberts, and the Moho depth to be 35 km at Los Angeles. A layer of 7 km/sec was not de tected. The upper crust has a velocity of 6.1 km/sec. Reinterpretation of seismic refraction data made by Prodehl (1970) showed that Pn is 7.9 to 8.1 km/sec between Santa Monica Bay and Lake Mead, and is 8.3 km/sec between Santa Monica Bay and Camp Roberts but in a reversing profile from Camp Roberts to Santa Monica Bay it is 7.9 km/sec. If the discrepancy is real, the property of the upper mantle below the Continental Borderland is probably closer to that below the Coast Ranges than to that below interior southern California. Gravity A crustal section modified from Shor and Raitt's section has been constructed by Harrison and others (1966) using Bouguer gravity computed from a density of 2.35 g/cm3 (Figure 3a). Their major modification is to place the Moho at a depth of 21.5 km instead of 17.5 km near the 26 Figure 3a. Trans-borderland crustal structure based on Bouguer gravity (from Harrison and others, 1966). Solid line: observed Bouguer anomaly; dashed line: computed Bouguer anomaly from the model; density in g/cm ; profile about 10 km northwest of seismic refraction profile. Figure 3b. Bouguer anomaly map for components with intermediate wavelength (from Harrison and others, 1966). Contour interval is 10 mgal; area with magnitude over 30 mgal is not shaded; dashed curves are bathy metry . ! 0 + l O g SSSSl -20 to-30 OiSTANCE, km OCPrn faELO# SLA U v t L . km CNAVtTY PROFILES. miiuga* o o x £ c o o S o □□ :0NT1NENTAL SLOPE § TANNER BASIN SAN NICHOLAS BSN SAN CLEMENTE i Ca ta lin a SHORE 8 to -o 28 shelf break of the continental slope. In addition they 1) interpret gravity high over San Clemente block to be the result of igneous extrusion (2.50 g/cm ) rather than 3 basement uplift (2.76 g/cm ), 2) use an isolated block of 3 2.65 g/cm for the gravity high over Santa Catalina Island, and 3) interpret the gravity high in the vicinity of 3 Newport Beach as a basic igneous intrusion (2.95 g/cm ) 3 into basement (2.76 g/cm ). 3 The layer with a density of 2.50 g/cm (4.5 to 5.1 km/sec equivalent to layer 2 of oceanic crust), which probably consists of consolidated sediments, metasediment, or extrusive igneous rocks, is present to the west of San Clemente block. Below this layer and further west, 3 there appears another layer with a density of 2.82 g/cm (6 . 2 km/sec) which is probably intermediate between oceanic layer 2 and layer 3 in terms of density and P-wave velocity.. 3 3 Both 2.50 g/cm - layer and 2.82 g/cm - layer are ter minated on the west by the continental slope, and are re** placed by a layer with 2.76 g/cm (5.8 to 5.9 km/sec) east of San Clemente block. One may attempt to interpret the 3 2.76 g/cm - layer to consist of the granitic rocks of the Penihfeular batholith. Interpretation as such would imply that the inner fault belt of the Continental Borderland is underlain by granitic basement instead of Franciscan basement. This implication is not compatible with the 29 geology previously discussed. Because the inversion of geophysical data is not unique, the crustal model in Figures 2d and 3 cannot be cited as convincing evidence that granitic basement underlies the inner Continental Border land. In addition, the quality of the data may not be sufficient to resolve the fine structure such as the bound ary between Franciscan rocks and granitic rocks in an area with complex topography and geology. From the phase ve locity of Rayleigh waves, Press (1956) found the sharp transition in the Moho depth occurred at the coast instead of occurring at the continental slope (i.e., the crustal thickness is 11 km below the continental slope, 15 km below San Clemente Island, 20 km below the coast, and 35 km below the Peninsular Ranges). These data must be regarded as only a first order approximation due to the nature of surface wave propagation in a marginal wedge. Spatial filtering of the Bouguer gravity (Harrison and others, 1966) has suggested a pronounced east-west grain in the Bouguer anomaly map for components with inter mediate wavelengths (Figure 3b). Such a structural grain must reflect deep structure because of its relatively long wavelength and its lack of correlation with surfacial features. Additional but concealed transverse structures can be recognized from the refraction profiles of Shor and Raitt (1958) and the reflection profiles to be discussed 30 later in this study. Delineation of the transverse struc ture may give some constraints on the models of tectonic evolution, but available data are not sufficient to corre late these transverse structures with the gravity anomalies. Geomagnetism A magnetic anomaly map covering most of the northern Borderland has been compiled by Harrison and others (1966) . Their measurements were made over 5 years with several fluxgate and proton precession magnetometers and thus the results inherited the problems of secular variation and calibration between different instruments. Nevertheless, the map (Figure 4) shows a large positive anomaly belt which coincides with the chain of Santa Cruz - Santa Barbara Islands but further south this belt deflects south- westward to the insular shelf of San Clemente Island and possibly continues southward over Sixtymile Bank and into the southern Borderland. To the northeast of this positive zone, there lies a negative anomaly belt along the south western side of the Santa Catalina Basin. Outside the paired linear anomaly zones, anomalies are smooth and of low amplitude. The magnetic anomaly map in the southern Borderland (Krause, 1965) also shows several linear magnetic anomaly belts some of which probably can be correlated with 31 magnetic anomalies in the northern Borderland. Correlation, however, is hindered by the method of removing the geomagnetic reference field and by the existing data gap at the intervening area of these two maps. The pat terns of magnetic anomalies in the Continental Borderland trend primarily northwest-southeast whereas the anomaly patterns over the Pacific Ocean basin west of the Border land strike north-south (Atwater and Menard, 1970). 32 Figure 4. Magnetic anomaly pattern. Traces of maxima and minima magnetic anomalies in the Continental Borderland are based on Krause (1965), Harrison and others (1966) , and this study; the anomaly patterns in the Pacific Ocean Basin are sketched after Atwater and Menard (1970). PACIFIC O C E A N BASIN HEAT FLOW Methods Heat flow at the earth surface is the product of geo thermal gradient and thermal conductivity. Its determina tion consists of measuring the gradient and the conductiv ity. The method used in this study is described below and followed by the estimation of sources of variability and error. Geothermal Gradient Temperatures are sensed with thermistors whose resist ances are determined in situ using a galvanometric film recorder designed originally by M. Langseth at Lamont- Doherty Geological Observatory and modified by J. I. Smith and others at M.I.T. and T. L. Henyey at U.S.C. A rotary switch is used to sample various thermistors successively in reference to a water probe (Figure 5) whose resistance is in turn measured absolutely against a precision re sistor (R^) on a Wheatstone Bridge configuration. Zero and full scales are marked by placing precision resistors R2 + R3 and R2 on the active arm respectively. The im balance of the bridge is amplified and recorded on film by means of a light beam galvanometer. Five thermistor probes were mounted on the outriggers 34 35 Figure 5. Schematic electrical diagram for the thermograd recorder, showing the sequence of marking zero scale and full scale, and recording of water temperature and sediment temperature. Rl/ &2 , and R3 are precision resistors. = water probe, Rp = other thermistor probes in the sediments. 36 20 K Zero Scale Full Scale Water Probe (R^) Thermistor Probes in the Sediment p ” 1» 2 f 3 r 4 f 5 37 spaced one meter apart and spiraled in increments of 72° along a 21-foot iron pipe barrel. Cables were taped to the barrel. A piston core was taken at each heat flow station. The depth of penetration was estimated from the residual mud trapped in the tape. Errors in the measurements of gradients may arise from several factors. 1) The incorrect spacing between the thermistors during measurement: The position of a thermistor in the stainless steel probe is not exactly known; vari ation of distance to the tip of the probe is believed to be less than 0.5 cm. Because of the spiral arrangement of outriggers, an error of 0.5 cm in the spacing between the tips of probes is possible during shipboard setup. These two sources of error are at most 1 to 2%. Occasion ally, slippage of outriggers occurs and the gradient is accordingly corrected assuming no slippage during pullout. Spacing change due to the bending of the core barrel is negligible in most cases. Experience indicates that if the core cutter hits the hard rock, the barrel falls over the sea bottom and there is no bending and measurement .'. Although bending may occur during the entry to or with drawal from the sediments, the latter is more likely. Bending typically occurs most severely along that portion of the barrel which did not penetrate the mud. Partial 38 penetration occurs when the core cutter encounters the strong resistance of relatively thick sandy layers. In this case the thermistor probes affected by the bending are outside the sediments and correction is not necessary because there is no measurement. Another related problem is oblique penetration. No device is used to detect corer inclination. If the gradient is uniform over the range of barrel length, the measured value in the case of oblique entry is less than the actual values. Judging from several cores studied, the layers of sediments are usually perpen dicular to the core liner and the effects of oblique penetration are negligible. 2) Improper determination of probe resistances: Error in the determination of probe resistances can arise from a) error in primary calibration, b) non-linear res ponse of Wheatstone Bridge circuit, and c) incorrect values of R , R , and R_ (ratio arm and full scale resistor, see 1 2 * * 3 Figure 5). Errors due to the primary calibration are discussed in APPENDIX I. Repeated calibration of the Wheatstone Bridge with known resistors in place of the thermistor probes verified that the galvanometer response is linear (within the limits of readability of the film) for off balance resistances up to + 200- f t . In this work off balance resistance did not exceed these values. The value of R^, R2r and R^ were set on shipboard according to 39 the particular in situ temperatures. Their values were measured with a portable Wheatstone Bridge with a resolu tion of 0.01X2.. Due to shipboard motion and electrostatic noise, errors in reading may occasionally be as large as + 0.5X1. This may cause up to a + 2.5 % error in a 100° C/km geothermal gradient. 3) Failure to reproduce film readings: The spots on the films are less than 0.5 mm in width. The reading is made by placing the cross hair of a precision travelling microscope whose resolution is one hundredth of 1 mm on the center of the spots. Reproducibility is within 0.1 mm, which is equivalent to an error of 0.3 to 0.4 XI or 0.0015°C In the worst situation, this would contribute 3 % error in gradient (1.5 % for one spot). All the errors mentioned here are not cumulative. If all odds are against an individual measurement, the error in the measured gradient could be as high as + 9 %. In reality the error is probably on the order of + 5 %. Thermal Conductivity Conductivities were measured at 10 or 15 cm intervals along the core using the needle probe technique (Von Herzen and Maxwell, 1959). All measurements were made in the laboratory no later than 1 0 days after the core was taken. The core liner was sealed at both ends against the loss of 40 water but no attempt was made to prevent desiccation through the liner. Two needle probes (Fenwal Model #K824D) were used alternatively. A constant current generator (supplying 90 mA) provided the heat source and the variation of re sistance with time was recorded on magnetic tape and processed by computer. Heating time was 90 seconds but only the central 70 data points at sampling rate of once every second were used in the linear least square fitting. The calculated value was then corrected for the difference between the laboratory and bottom water temperature and pressure effects (Ratclif fe-f , I960). The corrected value is around 4 % less than the measured value. The temperature-resistance relation for the needle probes were obtained with a quartz thermometer and a digi tal voltmeter in the similar manner as for the other therm istor probes. The same voltmeter was also used for the measurements of resistivity. Recalibration was made over a one year span. The results indicate that the calculated conductivities using new and old calibration parameters differ by less than 1 % for the same set of time-resistance data. The conductivity of a cylindrically shaped specimen of fused quartz (G.E. 102), 9 cm long and 5.4 cm in diam eter (same diameter as the core sample) was also measured with the needle probe method. The drill hole in the fused 41 quartz was stuffed with silicon grease to reduce the con tact resistance between the needle and the glass. The measured values differ from the published conductivity of fused quartz by 3 % for both needles at room temperature. Factors causing the error in measurement of conductivity include: 1 ) the calibration error mentioned above will affect all the measured values; and 2 ) disturbance of the core sample may affect conductivity locally along the core. The core liner was opaque, hence direct observation of the state of the sample was not permissible without splitting the liner. Usually the disturbance left some cracks un filled with sediments. Large cracks can be easily felt when the needle is inserted into the sediments. Further more, the resistances there decrease and level off rapidly with time after the heat is supplied. Frictional heat produced in the course of needle penetration was negli gible. Ambient temperature drift was determined before heat was supplied for the conductivity measurements. Its effect was negligible for all measurements. Results Approximately 100 heat flow stations have been occupied in the Continental Borderlnad of California from 1970 to 1972. About 70 % of them were successful. Poor 42 penetration and bending of the core barrel account for most of the failures. Geothermal Gradients Geothermal gradients for one meter intervals over a depth range up to 6 meters is shown in Table 3 and Figures 6 to 12 (see Figures 23 and 24 for locations). In case the bottom water temperature was not recroded (trace off scale), the extrapolated water temeprature has been ad justed so as to fall within the range of the recorded water temperatures in that basin. This may affect the absolute values of temperature measured in the sediments but will not affect the gradient over the observed tempera ture intervals ( APPENDIX I). Because of the uncertain ty in estimating the depth of core-barrel penetration from the residual mud on the barrel, the temperature difference between the water probe and the topmost core barrel does not represent the geothermal gradient near the water- sediment interface and has been excluded from the calcula tion of heat flow. Occasionally the bottommost thermistor does not reach thermal equilibrium as evidenced on the film. The unequilibrated value is asterisked in Table 3 and is not used in the calculation of heat flow. Values are also excluded when it was suspected that the probe had not penetrated into the sediments. In this case that probe is doubly asterisked in Table 3. 43 Figure 6. Geothermal gradients in San Diego Trough. DEPTH (METERS) 44 SAN DIEGO TROUGH i y — i— f t — i— i— i— i— i— i— \D3 \D5 D2 4.0 TEMPERATURE °C 45 Figure 7. Geothermal gradients in San Pedro Basin. DEPTH (METERS) SAN PEDRO BASIN 5.5 T E M P E R A T U R E °C 0 a \ 47 Figure 8 . Geothermal gradients in San Clemente Basin. DEPTH (METERS) 48 SAN CLEMENTE BASIN L6 TEMPERATURE °C 49 Figure 9. Geothermal gradients in Santa Catalina Basin. 50 CATALINA BASIN CO cn Lii H LsJ X h- CL UJ O A3 * 4.5 0 3 4 TEMPERATURE °C 51 Figure 10. Geothermal gradients in Santa Cruz Basin. 52 SANTA CRUZ BASIN •V\ i R4 °*-° 0 - I - 2 - 3 - 4 - 5 4.0 4.5 TEMPERATURE °C Figure 11. Geothermal gradients in San Nicolas Basin. 54 SAN NICOLAS BASIN cn cr L±J ! — L lJ X h- CL LlJ Q \ N3 4-0 TEMPERATURE °C 55 Figure 12. Geothermal gradients in Tanner Basin. DEPTH (METERS) TANNER BASIN T2 T3 T5 4.5 4.0 TEMPERATURE °C 57 The interval gradient usually decreases slightly with depth. This trend is compatible with a slight increase of conductivity with depth if the heat flow is uniform over the depth of penetration. When deviation from this general trend occurs, two neighboring gradients are usually involved, suggesting an error in the temperature recording or the reading of the common sensor. Such deviation can be easily verified by averaging the two gradients concerned and comparing the average value with other interval gradients. Thermal Conductivity The thermal conductivities of all core samples range between 1.5 and 2.6 mcal/cm sec °C with more than 85 % of the values falling within the range of 1.7 to 2.0 mcal/cm sec °C. The standard deviation in conductivity is within the range of 2 to 8 %. The average conductivities over one meter intervals are shown in Table 3. In general the conductivity increases 2 to 3 % per meter from the top to the bottom of the core, suggesting the effect of sediment compaction. A similar tendency has also been observed in the lake sediments of Lake Tahoe (Henyey and Lee, 1973) where a 20 % change was observed between the sediment- water interface and 6 meters in depth. Abrupt changes in conductivity are observed in some cores, especially those 58 from the inner basins: San Pedro Basin, San Diego Trough, and Santa Catalina Basin. Those change may reflect varia tion in quartz contents or disturbance in the core. The measured values of all core samples from each basin are shown in Figure 14 to 21. Because these composite figures are compromised by the uncertainty in depth with respect to the water-sediment interface, the conductivity V£. depth of the longest core sample in a basin is also plotted. Conductivity variations in Santa Catalina Basin is so large that no core seems representative of the basin, hence results of several cores are also presented. Three cores from San Pedro Basin indicate comparable variations. In San Diego Trough three short cores (1 to 2 m long) suggest the existence of large conductivity variation also exist there but oddly, the longest core (Figure 13, D2) does not differ from the conductivity trend observed in outer basins. A second degree polynomial is fitted to the observed data with least squares method. This poly nomial is also plotted along with the data to show the trend of conductivity variation. The slope is positive for every core sample, indicating that conductivity increases with depth. The second degree term is usually negative, suggesting that conductivity does not increase with depth indefinitively and a linear increase does not hold to a greater depth. No interpolation or extrapolation is 59 implied whatsoever from the polynomial fit. Comparison of conductivities measured in this study with those of Von Herzen (1964) and Lawver and others (1973, written communication) in the Continental Borderland is presented in Table 1. The discrepancies are within experimental error. It is noted, however, that Von Herzen's values are higher than mine except in San Diego Trough and that Lawver and others' values are consistently lower than mine. Heat Flow The product of the interval gradient and the average conductivity over the interval gives the interval heat flow. For a single station, interval heat flows are generally consistent but inconsistencies up to 10 % may occasionally occur. These are attributed to the unmatched interval gradients and conductivities, or factors cited before to be responsible for the errors in the measurements of gradi ent. Loss of the topmost or bottommost part of a core sample may cause difficulties in matching the interval gradient with appropriate conductivities, but this problem accounts for only a small part of the difference in in terval heat flows. The average of interval heat flow is taken as the heat flow value at a station. The method of estimating standard error and 95 % confidence limit is given in APPENDIX II. Table 1. Comparison of thermal conductivity (mcal/cm sec) in different basins ________ Basins_____________ ___ Santa Cruz Basin San Diego Trough San Clemente Basin East Cortes Basin West Cortes Basin San Pedro Basin Santa Barbara Basin San Nicolas Basin Santa Catalina Basin * 1 from this study, number in core samples used. *2 from Von Herzen (1964). *3 from Lawver and others(1973 *1 *2 *3 1.81 (5) 2 . 0 0 - 1.87 (3) 1.83 - 1 . 8 8 (4) 1.89 - 1.81 (1 ) 1 . 8 8 - 1.84 (1 ) 1.99 - 1.93 (5) - 1.82 1 . 6 8 (1 ) - 1.62 1.77 (7) - 1 . 6 8 1 . 8 8 (9) - 1.85 parenthesis indicating the number of , written communication). Figure 13. Thermal conductivity of sediments in San Diego Trough. A: Three core samples. B: One core sample at station D2. 62 CONDUCTIVITY (MCflL/CM SEC C) 1 .0 0 2. 00 0. 00 1.00 L. ~ 2.00 c r , C£ L lI ! — L j J 3. 00 _ LU Q 4.00 . 5.00 C . 0 0 i 3. 00 S R N U 1E . G 0 0. 00 CONDUCTIVITY <MCRL/CM SEC C) 1. 00 2. 00 3. 00 145 14 H. 00 5. 00 6. 00 64 Figure 14. Thermal conductivity of sediments in San Pedro Basin, five core samples. 65 CONDUCTIVITY (MCAL/CM SEC C) 1. 00 0 . 00 1. 00 co a: UJ 2.00 2. 00 i ___ 1 ' 3. 00 IE CL LiJ Q 00 - . 0 0 j 0 L S R N PLURQ PR Ni Figure 15. Thermal conductivity of sediments in San Clemente Basin. A: Five core samples. B: One core sample at station L3. CONDUCTIVITY (MCflL/CM SEC C) 1.00 2. 00 3. 0 . 00 1.0 0 ~ 2.00 co c r . lu b~ UJ z ~ 3 , 00 X 1— CL. LlI { — ~ l 4.00 5. 00 S.RM CLEMENTE 3.RSIN <sa3i3W) Hid 30 CONDUCTIVITY (MCflL/CM SEC C) 1. 00 2. 00 3. 00 0. 00 1. 00 2. 00 3. 00 H. 00 5. 00 6. 00 L3 14486 Figure 16. Thermal conductivity of sediments in Santa Catalina Basin. A: Eight core samples. B: One core sample at sattion A4 C: One core sample at station A5 D: One core sample at station A7 DEPTH (METERS) CONDUCTIVITY CMCRL/CM SEC C) 1.00 2. 00 0. 00 1. 00 2. 00 3. 00 4. 00 _ 5. 00 . 00 3. 0 CFITRLINR B R S IN 71 CONDUCTIVITY CMCflL/CM SEC C) 1. 00 2. 00 0. 00 1. 00 /■n 2 . 00 C f ) a: lu I x l 21 m 3. 00 CL LU Q 4.00 5. 00 A4 •1 14 3 12 3.00 CONDUCTIVITY (MCflL/CM SEC C) 1. 00 2. 00 3. 0 ✓"N CO oc UJ I — UJ 21 v X CL LU Q 0. 00 A5 H 3 H 1. 00 2. 00 3. 00 4. 00 5. 00 6. 00 DEPTH (METERS) CONDUCTIVITY <MCAL/CM SEC C> 1. 00 2. 00 3. 0. 00 A7 m315 1. 00 2. 00 3. 00 4. 00 5. 00 6. 00 Figure 17. Thermal conductivity of sediments in Santa Cruz Basin. A: Six core samples. B: One core sample at station Rl. 75 CONDUCTIVITY (MCAL/CM SEC C) 1. 00 2. 00 3. 00 0.00 1.0 0 ~ 2.00 C D C£ UJ • , Li X 1— CL LJ f — - j 4.00 5. 00 G. 0 0 SANTA CRUZ BRSIN CONDUCTIVITY CMCRL/CM SEC C) 1. 00 2. 00 3.0 C£ LU f- LU s: I — CL LU Q 0. 00 1. 00 2. 00 3. 00 5. 00 6. 00 Figure 18. Thermal conductivity of sediments in San Nicolas Basin. A: Seven core samples B: One core sample at station N7. CONDUCTIVITY (MCflL/CM SEC C> CONDUCTIVITY (MCAL/CM SEC C) 1. 00 2. 00 0.00 1. 00 /N O) cn LU 1— Lxi SZ v 2. 00 3. 00 . CL UJ Q H. 00 5. 00 14 3 5 1 N 7 3. _I 6. 00 Figure 19. Thermal conductivity of sediments in Tanner Basin (two core samples). 81 CONDUCTIVITY (MCAL/CM SEC C) 1.00 2. 00 3.00 0. 00 1. 00 2. 00 01 LU I — LU H 3. 00 t . X CL UJ o 4. 0 0 _ 5. 00 l- G. 00 TANNER BASIN Figure 20. Thermal conductivity of sediments in East Cortes Basin (one sample at station El). CONDUCTIVITY (MCflL/CM SEC C) 1. GO 2. 00 3. 00 0. 00 1. 00 ^ 2.00 co QL UJ 1— LU ^ 3. 00 X I — Q_ LU Q 4. oo 5. 00 e. oo ERST CORTES BRSIN Figure 21. Thermal conductivity of sediments in Santa Barbara Basin (two core samples). 85 CONDUCTIVITY (MCflL/CM SEC C) 1. 00 0. 00 1. 0 0 L . co C£ L lI I — LU 2.00 3. 00 Cl Ll! O 4. 0 0 _ 5. 00 _ . 00 2. 00 3. 00 SRNTR BRR3RRR BRSIN 86 The standard error for a heat flow measurement is between 10 and 14 %, based on the standard deviation of thermal resistivities and the maximum likely error 8.5 % in the measurement of geothermal gradient. All the measured values are listed in Table 3 and shown in Figures 23 and 25 to 30. The values corrected for topography and sedimentation effects are also shown in Table 3 and Figures 24 and 25 to 30. Perturbing Effects on Heat Flow Heat flow measured near the earth surface has been, to varying degrees, affected by the local topographic, geologic, sedimentological and water conditions around the site of measurement. To use successfully the heat flow data as a constraint in geophysical modeling, these per turbing effects should be removed. Sedimentation The thermal disturbances resulting from cases of con stant sedimentation and instantaneous slumping have been discussed by Von Herzen and Uyeda (1963). The ratio (normalized geothermal gradient) of the disturbed gradient to the undisturbed geothermal gradient in a homogeneous half space is presented in Figure 22 based on the formula (Jaeger, 1965). 87 ( f r V s = 1 + t t!) where g = the undisturbed gradient, p = Ut/(Kt)^/^, U = sedimentation rate, t = duration time of sedimentation, and K = thermal diffusivity. Because the function is nonpositive, the measured gradient is always less than the undisturbed value. In the following discussion, the sedi mentation correction is referred to as ; that is, the observed gradient has been reduced by of the true gradient. Variable sedimentation rates in the Borderland have been suggested in view of the unpredictable occurrence and thickness of turbidite deposits in the sedimentary column. The suggestion is further substantiated by dating different parts of core samples. For example, non- linearity between ages and depths of the sediments has been found by Emery and Bray (1962). Nonzero age of the surface sediments also raises a problem in estimating the sedimentation rates (Emery and Bray, 1962; Moore, 1969, p. 66-73) and no agreement on zero corrections of dates has been reached. Some workers (e.g., Gorsline and others, 1968) prefer to delete sandy layers from the estimation of sedimentation rates, which poses a problem in converting 88 Figure 22. Sedimentation effect on geothermal gradient; thermal diffusivity: upper 2 figure, 0 . 0 0 2 cm /sec; lower figure, 2 0.003 cm /sec. Duration time: 0.5, 1.0, 2.0, 4.0, 6.0, 8.0, and 10.0 million years from upper right to lower left of the figure. 89 SEDIMENTRTI ON RATES (CM/ 1000 YEARS) 1. L±J 0 . a C E ce CD a UJ M 0.002 0 200 100 0 'i. C E EE C E O 0.003 5 0 ._i____ i— 90 the deposition rate from a weight unit into a thickness unit. Extrapolation of the rates determined from the near surface sediments cannot go far into the past if the emplacement of sediments is pulsating as suggested by Moore., In view of these uncertainties in the estimation of sedimentation rates, certain mandates have to be made. Based on the total volume of the most recent sediments and annual sediment supply and some other lines of evidence, Moore (1969) gives one million years for the duration of depositing what he calls the post-orogenic sediments. This duration will be used to estimate the sedimentation effect. The thermal diffusivity is assumed to be 0.003 cm /sec and a constant sedimentation rate around a heat flow station will be also assumed but the rate may vary from station to station within a basin. Clearly one-million years is too short for the accumulation of the total sedimentary column represented in the borderland. Sedimentation presumably commenced prior to the time of block faulting which started during middle to late Miocene and recurred in Pliocene and early Pleistocene in the inner Borderland. However, our records of pre-orogenic (pre 1 m.y.B.P.) sediments are not complete enough to permit accurate calculations of this effect. Other factors also contribute to the uncertainty in the sedimentation effects. Effective sedimentation rates are not the same as the actual sedimentation rates 91 because of sediment compaction. The thermal diffusivity will increase as a result of compaction and loss of water. Variable sedimentation rates in the past are not well known. Data on basin uplift and subsidence are also lacking. Stations located in areas underlain by pre- orogenic sediments add another limitation to the applica tion of equation 1 . Bottom Water Temperature The deposition of biogenic carbonates in Tanner Basin has been generally uniform during the past 1 2 , 0 0 0 years. This reflects the fact that the California Current over the outer Borderland has been nearly constant in position and direction, as has the associated countercurrent over the inner Borderland (Gorsline and others, 1968). Hence we can assume that drastic changes in bottom water temperature due to the variation of current patterns have not occured. However, slow circulation in the deep basin areas has been suggested as an explanation for the lateral movement of the suspended-load into depressions where hemipelagic sed iments are much thicker than in the shallow parts of the basin (Gorsline and others, 1968). Measurements of bottom currents have been made by Revelle and Shepard (1939) in the central part of Santa Cruz Basin, Barnes (1970) on the western basin slope of Santa Cruz Basin, Moore (1969) and 92 others in the San Diego Trough. The measured current speeds range from nil to 10 cm/sec. The origins of these currents are not known. If they are deep tidal currents, their effects on bottom water temperature are negligible. Temperature difference in the water of different basins may induce lateral flow of water from basins to basins. The fact that the water temperature gradient below the sill depth is nearly uniform suggests that the mixing processes are effective enough to retain the characteristic water temperature of an individual basin. Short term effects of such flows on the bottom water temperature are probably negligible. The amplitude of annual variation in surface water temperature is less than 3° C (Emery, 1960, Figure 84). Observations made at depth 200 m indicate that the water temperature is nearly uniform throughout the year (Emery, 1960). According to Emery, the temperature gradients in the water column below sill depths are very small. The negligible gradients are further substantiated from data taken by the California Cooperative Oceanic Fisheries Investigation in Santa Cruz Basin. The large volume of isothermal water below sill depth also requires very large surface temperature change to produce significant change in bottom water temperature. The buffer of the large volume of water may keep the bottom water from the effect of seasonal temperature variation. However, the apparent 93 bottom water temperature variation among the stations shown in Figures 6 to 12 need some comments. They reflect the difference in the depths of stations rather than the existence of horizontal temperature gradient near the basin floor. As an example, a water temperature of 4.36 5° C at station Rl (Santa Cruz Basin) is measured at a depth of 1500 m. This gives a gradient of 0.09° C/lOOm with respect to 4.177° C at a depth 1710 m (Station R2). This value is about one third smaller than the gradient for the first hundred meters below the sill depth (Emery, 1960). It is thus demonstrated that the bottom water temperature is immune from seasonal temperature variation at surface and that the basin floor is nearly an isothermal surface. Although water temperatures below sill depths are at present annually steady, long-term climatic changes may have affected bottom water temperatures. Gorsline and Barnes (1972) have discussed the climatic change during the past 37,00 0 years for Tanner Basin, using data of 1 8 I /T 0 /O ratio in the benthonic and/or planktonic foramin- ifera. The calculated result from benthonic paleotempera- ture shows that the measured geothermal gradient has been decreased by 2.6°C/km for thermal diffusivity of 0.004 cm2/sec, 3.0°C/km for 0.003 cm2/sec, and 3.4°C/km for 0.002 cm2/sec near the surface (APPENDIX IV). These con tribute to a 2 to 3 % reduction of the undisturbed heat 94 flow. Disturbances of similar magnitude probably prevails in other basins of the Continental Borderland subjected to the validity of the assumption that oxygen isotope tempera ture represents paleotemperature of bottom water. Paleor. temperature determined from planktonic foraminifera yield similar correction. Relief and Conductivity Contrast Relief perturbs the heat flow. More heat will flow toward the topographic lows because of the relatively shorter thermal conduction path. In addition, the spatial temperature distribution in the marine environment is such that heat converges toward the basins where the bottom water is cold and to diverge away from the ridges where the bottom water is relatively warmer. These two effects are compensated to some extent by the variation of thermal resistance which is greater for basin sediments than for ridge material. Birch's (1950) method of topographic correction is applied to all heat flow measurements in the Borderland for the case of steady-state topography. A lapse rate of 1.5°C/km is used. This value is not crucial because of the shallow corer penetration. The area within 30 km of a heat flow station is divided into 19 concentric rings and 224 compartments. The relief in each compartment is visually averaged from bathymetric maps of the U.S. Coast and Geodedic Survey (1306N-20, 1306N-19, 1206N-15, and 1206N-16, contour interval 50 m, scale 1:250,000). However, the relief within one kilometer of the station is neglected for the reasons that the scale of available maps is not large enough and the navigation is not so precisely known. On the basis of theoretical calculations, Lachenbruch (1968) suggested that microrelief on the order of the probe length in close proximity to the point of penetration may cause variations in the oceanic heat flow data over a short dis tance. In the basins of the Continental Borderland, micro relief has been detected but no correction will be made for lack of details. Warm rim effects are also neglected because the topography above sea-level is rarely within 30 km of a heat flow station and where it is, it occupies a small portion of a ring only. The topographic corrections for all station ranges from less than 1 % to 19 % (observed values are greater than undisturbed values). Most of them are less than 5 %. Repeated readings around station R1 (highest topographic effect) indicate that the error in reading is around 5.3 % of the topographic correction. The heat flow was first corrected for topography, then for sedimentation because the topographic correction can be achieved with confidence subject to the uncertainty men tioned above. Sedimentation corrections require extra polation of sedimentation rates determined at other local- 96 ities as well as the assumption of duration time. Heat-flow refraction across the boundary of dissimilar media may cause significant variation in the surface heat flow, depending on the thermal conductivity ratio, the con figuration of the boundary and the distance to the anoma lous region (APPENDICES V and VI). If the dissimilar media are exposed, a discontinuity in surface heat flow across the boundary may occur and the relative heat flow may change as much as the corresponding conductivity ratio. In case of a sediment-bedrock contact, this ratio is 3. Since most heat flow stations were occupied at large dis tance from bedrock outcrop, the refraction effects are probably negligible, and the effect due to basement relief is damped by the thick sedimentary cover. However, a few stations were located on the basin slope or in saddle re gion where the refraction effect may be significant. The topographic effects along trans-borderland heat flow profile I-I' have been estimated with the finite element method (APPENDIX VI) assuming two dimensional to pography and uniform conductivity. In general, the observed surface heat flow in the basins may be enhanced up to 10 % and over the ridges the reduction is around 5 to 6 %. Horizontal heat flow through basin slope may reach 4 %. If a conductivity ratio of 1 to 3 is assumed for the sediments versus the ridge material, the relative heat flows between 97 the ridges and basins are reduced but greater variation appears near the edge of the basin. For lack of detailed knowledge about sediment and bed rock configuration, the refraction effect will not be quantitatively evaluated for each station. Discussion The heat flow data will be discussed basin by basin. This section is concerned with the individual measure ments and their topographic and sedimentation effects. Since the sedimentation rates vary from place to place within a basin, the correction is based on the best available data around a station. San Diego Trough The San Diego Trough is an elongated basin which has been filled to the sill depth with sediments. The trough is bounded to the northeast by the Coronado Bank and the southwest by the Thirtymile Bank. Turbidity-current channels have been recognized within the trough and on the mainland shelf and basin slope north of the Coronado Bank (Figures 23 and 26) . Five heat flow measurements have been made here in this study. The measured values are from south to north 98 Figure 23. Location of basins in the northern Borderland: 1. San Pedro Basin 2 . San Diego Trough 3. Santa Catalina Basin 4. Santa Barbara Basin 5. Santa Cruz Basin 6 . San Nicolas Basin 7. Tanner Basin 8 . San Clemente Basin 9. West Cortes Basin 1 0 . East Cortes Basin 1 1 . Velero Basin M. Outer Basin 0. Animal Basin Bathymetric contours at 400, 1000, and 2000 fathoms (modified after Moore, 1969). Fine lines showing the magnetic profiles presented in Figures 39 to 43; coarse portions of the lines showing reflection seismic profiles presented in Figures 35 to 38. a \ cn 118° 34 1 1 7 35° 119 33 120 121 1 1 6 32 LOS ANGEL IIO°W 20° W 36' Q CP, 122 117 1000 100 km 35 118' 31° 119 120' 123 121 34 33 32 122 100 Figure 24. Heat flows through the Continental Borderland off southern California 9 in unit of jucal/cm sec. • : from this: study ° : from Von Herzen (1964) o : from Foster (1962) a : from Lawver and others (written communication) a v : from Henyey (196 8 ) and Sass and others (1971) , normal triangles designate values above 1.5, reversed triangles designate values below 1.5. Stations without numbers are shown in Figures 26 to 31. 36° 122° 1 1 6 ° 120 | I’ * 1.6 \ 116' ✓ 32 3.2 1 .8* 2 3 T6 * l . f ,-o.X/.3.2 v-V> '\o ; » i 1.7^ v A • *o •1.4 2.2 • 1 .7 2.5 cp ^ ^ S 2. 2^ V J j Q -9 - 7 ^ 2.0 . '2. 1^ IQ 2.2 100 118 0.8 30' 120 30 34 122 32 102 Figure 25. Corrected heat flows through the Continental Borderland off southern California (topographic and sedimenta tion effects removed). See Figure 24 for figure caption. 103 0£ 021 >0£ 8 0, 8M \ 001 y 61-3 R-- 92 >22 <P 52 2 2 , Z 2 V V -0 o 021 ,911 o22l o9£ 104 1.63, 2.30, 2.12, 3.24, and 1.81 at stations D1, D2, D3, D4, and D5, respectively. The topographic corrections range from 2.8 % to 3.5 %, and the corrected values are 1.56, 2.24, 2.05, 3.15, and 1.75, respectively. South of station D2, Von Herzen (1964) reported 4 measurements 2.03 (MEN-39), 2.58 (GU-1), 2.05 (T-l), and 1.98 (T-2) (Figure 26). Their corrected values are 1.97, 2.40, 1.99, and 1.92. Noteworthy is the fact that the two stations with largest values (D4 and GU-1) were located in a turbidity- current channel which has a depth of about 20 to 30 m relative to the floor outside the channels (Figures 26 and 37). Station D4 (3.11) was located in an area of relative ly high relief while station GU-1 (2.54) was in a flat region. The two lowest values, 1.56,and 1.76, were observed at the edge of the channels. Such a correlation of heat flow and position with respect to the channels may reflect the effects of microrelief and/or scour and fill but it may be fortuitous in view of the fact that two intermediate values 1.97 (MEN-39) and 1.92 (T-2) were also observed in the channels. We will assume that the re presentative heat flow values in San Diego Trough are those measured outside the channels: 2.25, 2.05, and 1.99, with an average of 2 .1 0 . Emery and Bray (1962) estimated the sedimentation rate Figure 26. Heat flows through San Diego Trough. Bathymetry from U.S. C.&G.S. 1206N-16 contour interval 50 m, shaded area showing turbidity current channels (after Moore, 1969). Corrected values (sedimentation + topography) are asterisked. 107 from a core in the central San Diego Trough to be 18 cm/1000 years. This value is comparable to those in the outer basins and may be too low to represent the sediment ation rate in the central portion of the San Diego Trough as a whole. Moore's (1969) chart of the volumetric dis tribution of post-orogenic sediments indicates that the deposition rate is about 7 to 42 cm/1000 years around the two northernmost and one southernmost stations, and is about 42 to 110 cm/1000 years or more around the rest of the stations. From an experimental Mohole drilling site off La Jolla, California, in San Diego Trough, Inman and Goldberg (1963) gave an average rate of 200 cm/1000 years at depth 70 m (age 34,000 + 1,000 years B.P.) below the water-sediment interface. This rate probably is too high to persist for one million years or more because the sediments in the San Diego Trough are less than 2 km thick. The corrected heat flows based on Moore's average figures and a duration time of 1 m.y. are from south to north, 1.71 (9 % for 25 cm/1000 years), 3.40, 3.11 (34 % for 110 cm/1000 years), 3.44, and 1.93 (9 %). The re presentative heat flow in the central San Diego Trough is thus 3.18 (34 % correction). 108 San Pedro Basin San Pedro Basin is bounded on the northeast by San Pedro Escarpment and on the southwest by Santa Catalina Escarpment, Both escarpments have a slope of 8°. To the north the basin is terminated by a volcanic seaknoll. Its southern margin is not distinguished by a clear bathy metric break (Figures 23 and 27), but rather gradually shoals southward into a region of irregular topography. Five heat flow measurements were made in San Pedro Basin. From south to north the station are PI (3.08), P2 (1.72), P3 (1.66), P4 (1.40), and P5 (2.36). The topog raphic corrections range from 1.7 % to 4.9 %. The cor rected values are 2.94, 1.65, 1.60, 1.37, and 2.32. Lawver and others (written communication, 1973) have obtained a value of 1.6 2 from two repeated measurements in the central part of the basin, and its topographically corrected value is 1.58. Station P4 falls within the closure of 40 cm/1000 yrs sedimentation rate contour (Emery and Bray, 1962). Nearby piston core stations yield rates of 42, 55, 6 8 , and 70 cm/1000 years. This station is also located near the center of 420 m to 1100 m zone of Moore's (1969, Plate 15) post- orogenic sediments. The average value of Moore1s data yields a rate 76 cm/1000 yrs for a duration time of 1 m.y. The corrected value (topography + sedimentation) is thus 109 1.83 (25%). Two stations (Pi and P3) are located within the terrain of Moore's pre-orogenic sediments (see also Figure 35) and a rate of less than 7 cm/1000 years is implied. But Emery and Bray's data indicate a rate 30 cm/100 0 years (between the contour of 20 to 40 cm/1000 yrs) which gives the corrected values,3.32 at PI and 1.79 at P3 (10%). Other stations fall between the rates of 20 cm and 40 cm contour. In this region Moore's data suggest a rate of 24 cm/1000 years. A 10 % correction is made for these stations, i.e., 2.60 at P5, 1.83 at P2, and 1.76 at Lawver and others' stations. Thus, the corrected heat flows are from south to north 3.32, 1.83, 1.76, 1.79, 1.82, and 2.60. Relatively high heat flow at both ends of the basin (PI and P5) may be caused by 1) the difference in sedi mentation rate used is less than the actual difference; 2) the duration time used is too short; or 3) unmapped microrelief within one kilometer of the northernmost and southernmost stations is unduly neglected in the topog raphic correction. If the minimum thickness (70 m) of the post-orogenic sediments (or a rate of 7 cm/1000 years) is used for the northernmost station P5, the sedimentation effect is negligible there. To make the heat flow values in the central part of the basin to be the same as the value 2.33 at station P5, a 40 % correction is needed for F.igure 27. Heat flow through San Pedro Basin. Bathymetric chart from U.S. C.&G.S. 1206N-15. Contour interval 50 m. Corrected values are asterisked. 10 KM O " C P 4- ■o © ■<£ III UIMma i Mi 112 station P4, and 30 % correction for stations P3 and P2. The former requires a depositional rate of 130 cm/1000 yrs and the latter 100 cm/lDOO years for a duration time of 1 m.y. If the duration time is 2 m.y., the former corre sponds to 85 cm/1000 years and the latter to 53 cm/1000 yrs. In both cases the implied sedimentary column exceeds the maximum allowable thickness of Moore's post-orogenic sediments. The discrepancy becomes greater for longer duration times. On the other hand, if a shorter duration time is used, say 0.5 m.y., the implied thickness of sediments fits Moore's data but the rates, 190 cm/1000 yrs and 135 cm/1000 years are unusually high, and not compati ble with Emery and Bray's data. The southernmost station PI was located within the terrain of deformed sediments but close to the boundary with the undeformed sediments. If the conductivity of the deformed sediments is twice that of the undeformed sedi ments at a depth greater than the depth of penetration, re fraction of heat flow may be significant enough to account for the anomalous value here. Station P3 was also located within the terrain of deformed sediments but at greater distance away from the undeformed sediments and hence the refraction is negligible there. Excluding station PI, the average of the corrected (sedimentation + topography) heat flows is 1.98. 113 San Clemente Basin San Clemente Basin lies to the southeast of San Clemente Island. Basin closure occurs at the 1830 m isobath. The basin is bounded on the east by Fortymile Bank and on the west by Sixtymile Bank. The basin can be separated into northern and southern parts with a narrow trough tying them together. The northern part has a northwest trend whereas a north-south trend dominates the southern part. An elliptically shaped volcanic high stands 540 m above the floor in the northwest quadrant of the southern basin (Figures 23 and 28). Six heat flow measurements were made in San Clemente Basin. The measured values in the northern basin floor are in order from south to north 2.00, 2.06, 1.87, and 2.82 for stations L3, L4, L5, and L6 , respectively. The topo graphic corrections range from 5 % to 10 % and the correct ed values are 1.82, 1.91, 1.78, and 2.61 (Table 3). Excluding the value 2.61, the average is 1.84. Von Herzen (1964) reported two values 1.96 and 2.78 in the same area (Figure 28), for which the corrected values are 1.70 and 2.50. Station Ll located in the southern basin has a value of 1.71 and its corrected value is 1.60. A value of 3.23 was observed at station L2 on the eastern slope of the trough. Emery and Bray (1962) obtained a sedimentation rate 114 Figure 28. Heat flow through San Clemente Basin. Bathymetric chart from U.S. C.&G.S. 1206N-15, and 1206N-16. Correct values are asterisked. 115 W>i 01 116 of 30 cm/1000 years in the northeastern corner of the north ern basin and 29 cm/1000 years in the northern half of the southern basin. The equality of the rates is contrary to Moore's (1969) data which suggest that the post-orogenic sediments are less than 70 m thick in the northern basin, whereas in the southern basin the thickness may reach at least 420 m. In terms of sedimentation rates the two sections are respectively equivalent to 7 and 42 cm/1000 yrs for a duration time of 1 m.y. Two reflection seismic profiles (Figure 37) indicate that the sediments in the northern basin may be locally thicker than 1 00 m and in the central basin my data are consistent with Moore's, except in its southernmost part. Using Emery and Bray's sedimentation rates, the correction is 8 % in the northern basin and 7 % in the southern basin and corrected values are 2 . 0 0 (excluding the two largest values) and 1.72, respectively. If Moore's data are used, the correction in the northern basin is less than 3 % and more than 15 % in the southern basin and the corrected values will be 1.90 and 1.88, respectively. Thus corrections applied according to Moore's data bring the two values into closer accord. The true correction is probably somewhere in between. After removing the topographic and sedimentation effects, the variation of heat flow needs further comment. Station L6 was located in a valley near the toe of a basin 117 slope and the above normal value (2.82) is probably caused by microrelief. Von Herzen's value 1.96 was measured at the base of a steep slope, hence, high value is expected there but the measured value is normal within the San Clemente Basin. His high value 2.78 may be the result of microrelief too. The existence of microrelief in the northern basin is supported by a seismic profile (Figure 37) about 1 km southeast of station L5 (Figure 28). The profile also indicates that the variation in the thickness of sediments below the basin floor may be as much as 25 % from the mean. Such variation is accompanied with basement relief but heat flow variation due to the buried basement relief is less than 10 %. The representative values for the northern basin are probably 2.00, 2.06, and 1.87, with an average of 1.97 for the uncorrected values, or 1.82, 1.91, and 1.78 with an average of 1.84 for the topog raphically corrected values. Station L2 is located near the mid-point on the northeastern basin slope. The topographic correction is about 3 %. Hence the high value of 3.2 3 cannot be a con sequence of large-scale relief alone. Microrelief in the vicinity of the station is shown on seismic profile QR (Figure 37). Erosional effects or effects due to bottom water temperature variation can be ruled out since interval gradients are uniform. A local heat source together with 118 the effect of microrelief could account for the anomaly. Three faults occur close to the station (Moore, 1969, Plate 13). The anomalous heat flow may be related to the seismically active San Clemente fault, possibly to hydro- thermal circulation along it. Santa Catalina Basin Santa Catalina Basin is bounded on the northeast by Santa Catalina Island, the northwest by Santa Barbara Island, and the southwest by San Clemente Island. Its southeastern end gradually shoals toward Thirtymile and Fortymile Banks. Two sills occur at the northwest and the southeast margins at depths of 980 m. Within the 1150 m isobath the basin floor is essentially flat but tilted slightly toward the northwest (Figures 23, 29 and 35). Emery Seaknoll, an oval shaped andesitic volcano (Gaal, 1966) , stands 500 m above the basin floor in the southeast ern quadrant. A slope-apron complex appears along the west-facing escarpment of Santa Catalina Island but is absent along most of the San Clemente escarpment. Land slide debris appears to be present at the base of the San Clemente escarpment and the west slope of Emery Seaknoll (Gaal, 1966). Eight heat flow measurements were made in the basin floor. One high heat flow of 2.81 was recorded at station 119 Figure 29. Heat flow through Santa Catalina Basin. Bathymetric chart from U.S. C.&G.S. 1206N-15. Corrected values are asterisked. 120 o 121 A1 near the southeastern end of the basin. Excluding that station, the uncorrected values average 1.75 with a range from 1.62 to 1.93 and with a standard deviation of 0.10. Foster (1962) reported a value of 1.8 in the central part of the basin. Lawver and others (written communication) also obtained an average value of 1.75 from four repeated measurements. The topographic corrections range from less than 1 % to 4.2 % and the corrected values are given in Table 3 with an average of 1.73. Because the heat flow is fairly uniformly distributed over the basin floor, the value at station A1 appears anomalous. Emery and Bray (1962) give sedimentation rates in the northeastern portion of the basin ranging from 20 to 40 cm/1000 years. A value of 25 cm/1000 years is also re ported from the central part of the basin floor (Gaal, 1966). Moore's (1969) data suggest the rate is between 7 arid 42 cm/1000 years. Using 30 cm/1000 years for the average sedimentation rate, the corrected average heat flow is 1.92 (10 % correction). A reflection seismic profile (Figure 35) indicates that the sediments are not uniformly distributed along the elongated trend of the basins. Differential sedimentation effects and refraction of heat flow are expected but the variation in the observed values are within the experimental errors. Due to the areal variation in sedimentation rate and the nearly uniform 122 values of the corrected heat flow, the undisturbed heat flow probably increases slightly from southeast to northwest. Santa Cruz Basin Santa Cruz Basin is a syncline in the northwest oriented en echelon fold belt of the California Continental Borderland (Moore, 1969). To the north it is separated from the transverse structure of the Channel Islands and Santa Barbara Basin by an east-trending fault believed to be an offshore extension of the Santa Monica Mountains - Malibu fault system. Santa Rosa - Cortes Ridge forms its western margin while the Anacapa - Santa Barbara ridge bounds it on the east. On the south the basin is ter minated by the San Nicolas - San Clemente ridge. Basin slopes from shelf break at 200 m to basin floor at 1750 m have an average slope of 8°. Steeper slopes (15° - 20°) are present to the west of Santa Barbara Island, south of Santa Cruz Island, and at the southern end of the basin. Within the 1750 m isobath, the basin floor is essentially flat and featureless except in the north western quadrant where hillocks 10 to 40 m high may extend out 5.5 km from the western wall (Barnes, 1970). Submarine fans and their feeders (canyons) appear also in the northern and southern ends of the basin floor. Seven heat flow measurements were made in Santa Cruz Basin (Figures 23 and 30). The topographic corrections range from 4 % to 18.7 % (Table 3). Sedimentation rates are 31 to 51 cm/1000 years from determinations made at four localities (Emery and Bray, 1962). Moore's (1969) data indicate that the deposition rate of post-orogenic sediments may be as high as 42 cm/1000 years. Applying an average of 10 % correction (30 cm/1000 years) to stations R4 and R3 in the southern part of the basin floor will raise them from 1.61 and 1.47 to 1.79 and 1.64, respective ly. The former value is compatible with the value 1.82 at station R2 where the sedimentation effect is probably negligible according to Moore's and Emery and Bray's data. The relatively low value 1.64 at station R3 is probably due to the under estimation of sedimentation correction. By the bathymetry and Moore's (1969) route of sediment transport, the sedimentation rate is likely to be greater at station R3 than at R4. At the southernmost station Rl, the measured value 1.77 is topographically corrected (18.7 %) to 1.47. Effect of microrelief is probably negligible based on the bathymetry in the vicinity around the station (Figure 30). The sedimentation effect is also negligible. Since this station was located near the base of a 17° slope, the low heat flow measured there may be caused by refraction. Hence the representative Figure 30. Heat flow through Santa Cruz Basin Bathymetric chart from U.S. C.&G.S 1206N-15. Corrected values are asterisked. 125 126 value in the southern Santa Cruz Basin is 1.8. In the northern part of the basin, a 14 % correction (40 cm/1000 years) is applied to stations R6 and R5; the corrected values are 1.37 and 1.33, respectively. Both values are compatible, with the topographically corrected value 1.40 at station R7, where the sedimentation correction is negligible because it was located on the basin slope. Here, the representative value appears to be 1.4. Von Herzen (1964, station MEN-2A) reported an un corrected value of 1.43 to the southeast of station R6 . After making a 5 % topographic correction and the same sedimentation correction (14 %), the value becomes 1.58 which is intermediate between the representative values in the northern (1.4) and southern (1.8) parts of the basin. San Nicolas Basin San Nicolas Basin, southeast of San Nicolas Island, is a syncline in the norhtwest oriented en echelon fold belt (Moore, 1969). It is bounded on the northwest by an east-northeast striking escarpment, and on the southwest by the Santa Rosa - Cortes Ridge and Tanner Bank. Its southwest margin shoals gradually across the divide south eastward to San Clemente Basin and southwestward to East Cortes Basin. The lowest sill is located on the southeast 127 margin at a depth of 1110 m. The basin has the largest area at sill depth and largest volume below the sill of the basins in the Continental Borderland of California. The depth difference of San Nicolas Basin between the sill and basin floor is second only to that of Santa Cruz Basin. The basin floor, below the 1625 m isobath lies in the northern half of the basin (Figures 23 and 31). Six heat flow measurements were made in San Nicolas Basin (Figures 23 and 30). Among them, four were located below the 1625 m isobath (N3, N4, N5, and N6 from south to north), one (N7) about 50 m below the effective sill depth on the steep basin slope of the divide between San Nicolas Basin and Santa Cruz Basin, and one (N2) on the gentle basin slope south of the basin floor. The topographic corrections range from less than 1 % to 3.5 %. Excluding station N7 where the uncorrected heat flow is 1.44 and the topographically corrected values is 1.41, the average is 1.67 with a range from 1.58 to 1.84, and the average of corrected values is 1.62 with a range from 1.55 to 1.75. Lawver and others (written communication) report a value of 1.45 from two repeated measurements; its corrected value is 1.43. In addition to these 6 stations, a measurement (1.68 at station Nl) which was made in the broad saddle region be tween San Nicolas Basin and San Clemente Basin is included Figure 31. Heat flow through San Nicolas Basin. Bathymetric chart from U.S. C.&G.S. 1206N-16. Corrected values are asterisked. 129 \ 10 KM 130 here. The value at this station represents a heat flow transition between these two basins. It should be noted that station N1 was located near the effective sill depth, although we find no evidence of water temperature varia tions. The topographic correction is negligible here. Emery and Bray (1962) estimated the sedimentation rate to be 20 cm/1000 years. This amounts to 7 % correction, and the corrected average heat flow for the basin is accordingly 1.74 with a range from 1.67 to 1.88. Tanner Basin Within the 1000 m isobath, the Tanner Basin floor can be separated into three parts: a shallow western portion at a mean depth of about 1150 m, and two narrow, deeper zones along the eastern edge (herein called eastern Tanner Basin) with maximum depths of 1480 m and 1420 m in the northern and southern depressions, respectively. The floor is predominantly a basin apron, with a small area of basin plain in its northern depression (Gorsline and others, 1968). Six heat flow measurements were made: two (Tl and T2) on the western basin floor, one (T4) in the northern depression of the eastern basin, one (T3) half way between the northern and southern depressions of the eastern basin, and two (T5 and T6 ) in an elongated depression extending 131 northwest from Tanner Basin to San Miguel Gap. This de pression is not usually considered part of Tanner Basin but is included for convenience. The topographic corrections range from 2 to 7.5 %. The mean sedimentation rate of Holocene hemipelagic components has been estimated by Gorsline and others (196 8 ) using radiocarbon and paleontologic dates on 26 2 piston cores to be 10 mg/cm year over the past 7500 years and 9.5 mg/cm2year between 7500 and 12000 years B.P. Based on their isopach map of Holocene hemipelagic sedi ments, the lateral variation of sedimentation rates for the past 12,000 years range from 8 to 25 cm/1000 years (exclud ing sandy layers). In general the sedimentation rates in the deepest eastern depressions are two to three times faster than in the western floor. According to Gorsline and others' Figure 8 , near station T4 sand constitutes about 10 % of the post-12,000 years B.P. sediments. Thus the sedimentation rate here is approximately 28 cm/1000 years. Near station T3, the pelagic sedimentation rate is approximately 15 cm/1000 yrs and the sand contribution is negligible. Near stations Tl and T2 the rate is 10 cm/1000 years. No data are avail able for stations T5 and T6 , but on the basis of bathymetry the rate is probably 10 cm/1000 years. For a duration time of 1 million years, the corrected values are (from Tl 132 to T6 ): 1.96, 2.23, 1.65, 1.34, 1.62, and 1.71 (Table 3). These values are essentially the same as the measured values. The difference in the values at stations T3 and T4 is greater than the experimental error. For a duration time of 4 m.y., the two values are 1.76 and 1.49 and the difference is within experimental error. For a duration time of 10 m.y., the values are 1.86 and 1.70. The difference in values at station Tl and T2 is within ex perimental error. The heat flow difference between the eastern and western basin is significant. Santa Barbara Basin Santa Barbara Basin is within the Transverse Range province of southern California. It is bounded on the north by the east-west trending coast line and on the south by the Channel Islands. Because of the small depth dif ference (120 m) between the bottom (630 m) and effective sill depth (510 m), the water temperature gradient near the bottom is relatively large (^1.6°C/km) compared to those ('*'0.5°C/km or less) of other offshore basins (Emery, 1960). Its bottom water temperature (6.43°C) is the highest among the basins in the Continental Borderland. Two heat flow stations were centrally located near the deepest part of the basin. Station B1 gives an uncorrected value of 1.12 and station B2 yields 1.20. To the east of 133 station B1, Lawver and others (written communication, 1973) report an average value of 1.19 from 5 repeated measure ments at a single locality. The topographic correction is less than 1 %. Both stations fall between the 80 cm and 180 cm contour of sedimentation rates (Emery and Bray, 196 2, unit cm/1000 years) and between the 420 m and 1100 m isopach of post-orogenic sediments (Moore, 1969). Emery and Bray's sample localities suggest that the rate near the two stations is approximately 123 cm/1000 years, which is about 10 % higher than the maximum allowable rate based on Moore's data assuming a duration time of 1 m.y. This value gives a correction of 47 %,whereas the correction according to Moore's data does not exceed 35 %. Using a 35 % correction, the corrected values are 1.72 at station Bl and 1.84 at B2. Emery and Bray's data also indicate that the rate near station Bl is slightly higher than near B2. This can account for the 0.12 difference although it is within the experimental error. East Cortes Basin East Cortes Basin is one of the synclines in a north west striking en echelon fold belt (Moore, 1969). It lies to the southeast of Cortes Bank and northwest of Sixtymile Bank. It has a sill shared with West Cortes Basin. 134 Topographic corrections are 4.3 % at station El (2.01) and 3.0 % at station E2 (1.77), and the corrected values are 1.97 and 1.72, respectively. The sedimentation rate in the East Cortes Basin is 25 cm/1000 years determined from one core by Emery and Bray (1962). The results of the two corrections are 2.16 at El and 1.89 at E2. Von Herzen (1964) obtained a value of 1.78 about 5.5 km west of station E2. Its corrected value is 1.89. Miscellaneous Locations For areas discussed in this section, the bathymetry is less well known than for the areas previously discussed. Topographic corrections are estimated from Moore's chart C1969) of which the scale is about 1:1,000,000, or ten times smaller than the USCGS charts off southern California. Station Ml was located to the west of the southeastern extension of the Patton Ridge (Figure 23 and Table 3) . This is the deepest station (2470 m) made in this study. The measured heat flow is 1.90. The sedimentation effect, though unknown, is probably less than 3 % and the topo graphic correction is roughly 3 to 4 %. No correction has been ..made. Station M2 was located in the so-called Outer Basin which lies between Southwest Bank and the Rampart. No core was recovered and a conductivity of 1.84 mcal/cm sec°C is 135 assumed. Based on data from Prensky (1973) the weighted 2 average rate of sedimentation is 4.2 mg/cm year or 5 cm/1000 years, comparable to the deep sea deposition rate. The rate provides a 2 % correction. The topographic correction is also small but the effect of microrelief (Figure 38) may be significant here. A value of 2.21 was measured at station M3 in Animal Basin. About 20 km southeast of this station, Von Herzen (1964) reported two values: 2.25 and 3.35. The sedimenta- tion rate here is about 7 mg/cm year (Prensky, 1973) or 8 cm/1000 years. This rate requires a 4 % correction but the topographic correction is about 3 %. Hence no correction is made. The highest heat flow 3.83 so far recorded in the Borderland comes from station M4 occupied in the northwest ern slope of Velero Basin. This value is based on two gradients: 208 and 207 °C/km. A weighted average sedi- 2 mentation rate of 6 mg/cm year (Prensky, 1973) or 7 cm/1000 years requires a 3 % correction, which is not made because it is probably balanced: by the topographic effect. No explanation can be given for the high heat flow at this locality. In the northwestern depression of West Cortes Basin, only one temperature gradient was obtained (M5) giving a heat flow of 2.59. This value is considerably higher than 136 two values measured in East Cortes Basin (El and E2). Correction due to a sedimentation rate of 12 cm/1000 years (Emery and Bray, 1962) increases heat flow by 4 % which is counterbalanced by the topographic correction 5 %. Von Herzen (1964) obtained a value of 1.89 in the southeast margin of the basin about 53 km southeast of station M5. Continental Shelf off Central California This area is not part of the Continental Borderland of California but is included here for the convenience of presentation. Four heat flow measurements were made on the shelf between Pt. Conception and Monterey Bay. The observed values are in order from south to north 2.46, 1.42, 1.58, and 2.18 at stations Cl, C2, C3, and C4, res pectively. Topographic corrections range from 1 % to 3 %, and the corrected values are 2.46, 1.39, 1.57, and 2.16. Station Cl was located in a channeled area (Figure 36). Although the large-scale topographic correction is neg ligible, microrelief may cause anomalous heat flow here. Except for station Cl, heat flows appear to be correlative with water depth and increase northward. The stability of bottom water temperature under the influence of the southeastward flowing California Current is not well understood. From March to July, one of the upwelling centers is located at 35° N (Sverdrup 137 and others, 1942). The upwelling water probably rises from depths less than 2 0 0 m, below which a countercurrent flows close the coast. Toward the end of summer, the upwelling gradually ceases and evolves into a number of eddies through which the coastal and oceanic waters are interchanged. In the fall upwelling ceases, a surface countercurrent (Davidson Current) develops along the coast but the subsurface countercurrent still exists. Because of the presence and absence of upwelling, the average surface water temperature is lower (/v 1 . 5 2.0°C) in March to July than in December to January. Stations Cl, C2, C3, and C4 were occupied in early July around the center of upwelling (35° +40' N), but the disturbing effect on the gradient is not known and the quality of the data cannot be evaluated at present. The quality of data is further compounded by the unknown sedimentation rates except at station C4 where the sedimentation rate is 28 mg/cm2 year (Prensky, 1973) or 30 cm/1000 years. Heat Flow Distribution Pattern The heat flow from the Continental Borderland off southern California can be summarized by the following facts: 1 ) the average heat flow is slightly higher than normal; 2) the heat flow decreases systematically northward; 138 3) the heat flow increases slightly landward; and 4) the heat flow is generally uniform in a given basin but variation greater than experimental error is found in the inner most basins. The average of 69 heat flow measurements in the north ern Borderland is 1.86 including 53 from this study, 12 from Von Herzen (1964), 4 from Lawver and others ( written communication, 1973), and 1 from Foster (1962). After topographic correction, the average is 1.78. If the sedi mentation effects (best estimates) are removed, the average becomes 2.05, assuming a duration time of 1 million years 2 and a value of 0.003 cm /sec for the thermal diffusivity of the sediments. These averages are biased by the sample distribution, since the majority of the stations were occupied on basin floors and the number of samples in each basin is not the same. If the representative value of each basin is used (Table 3), the average is 1.74 for un corrected data, 1.67 for topographically corrected data, and 1.99 for data with topography and sedimentation corrections. In any case the average is higher than the world average value of 1.5. Uniformity in heat flow is best demonstrated in Santa Catalina Basin, San Nicolas Basin, and Santa Barbara Basin. The average of uncorrected heat flows in these basins are 1.75, 1.67, and 1.17 with standard deviations of 0.10, 139 0.09, and 0.05 for 9, 6 , and 3 measurements, respectively. Heat flows in San Diego Trough may be locally perturbed by the turbidity-current channels but away from the channels the heat flows are uniform. Three values in East Cortes Basin indicate that heat flows tend to be uniform there too. After removing the effects of topography and sediment ation, the heat flows in northern San Clemente Basin are uniform (excluding two values), as are the heat flows in the central part of San Pedro Basin. In Santa Cruz Basin the measured values are not uniform and the corrected values show that the heat flows decrease linearly north ward from 1.8 at the southern end to 1.4 at the northern end, over a distance of 60 km. Profile III-III' shows heat flow variation along the trend of regional structure from Santa Barbara Basin across Santa Cruz Basin, San Nicolas Basin, and East Cortes Basin to Animal Basin (Figures 24, 25, and 32). In general the observed heat flow (solid curve) confirms Von Herzen's (1964) conclusion that heat flows decrease systematically northward in the Continental Borderland. Superimposed upon this general trend is a maximum over the divide between Santa Cruz Basin and San Nicolas Basin. The general heat flow trend is consistent with increasing sedimentation rates to the north whereas the maximum appears in an area where the thermal blanketing due to 140 the cover of sediments is likely to be least effective. These two apparent correlations necessitate the removal of sedimentation effects. One such corrected heat flow trend is shown as a dotted curve in Figure 32. As expected, these two curves converge to the south east but diverge to the northwest. Differential sediment ation rates have significantly altered the characteristics of heat flow trend from Santa Cruz Basin across Channel Islands to Santa Barbara Basin. The heat flow low is shifted from Santa Cruz Island, a heat flow low appears in the "corrected" profile. This heat flow low coincides with the change in structural trend from the northwestly oriented Peninsular Range structure to the east-striking Transverse Range structure. Closely spaced data are not presently available for extending the heat flow profile into the southern end of the Borderland. However, from a few measurements made by Von Herzen (1964), the linear heat flow trend between East Cortes Basin and Animal Basin cannot continue to increase very far into the southern Borderland. Superimposed upon the regional southward-increasing heat flow trend are com ponents with an "apparent wavelength" on the order of 75 km. Extension of the profile to the north is also limited for lack of sufficient data. However, 3 measure ments made on the continental shelf off central California 141 Figure 32. Heat flow profile III-III' along the trend of regional structure, solid circle: this study, open circle : from Von Herzen (1964). open triangle: corrected value, solid curve : trend of measured values, dotted curve: trend of corrected values. 100 142 HEAT FLO W DEPTH PO o o 3 SANTA BARBARA BASIN CRUZ ISLAND SANTA SANTA CRUZ > • SAN NICOLAS B. EAST CORTES B. NO NAME B. ANIMAL B. 143 suggest that heat flows may increase northward in the area north of the Transverse Ranges. Von Herzen (1964) found a periodic variation in heat flow with an apparent wavelength of 260 km along two east-west profiles across the Continental Borderland to the Pacific Ocean floor. In this study only one measurement (Ml, heat flow 1.9) was made in the Pacific Ocean floor and hence no confirmation of his reported periodic variation can be reached, but a distinct heat flow trend across the Continental Borderland is recognized with the addition of present data. Profile II-II1 extends from the Patton Escarpment across Tanner Basin, San Nicolas Basin, Santa Catalina Basin, and San Pedro Basin to the continent (Figure 33). All data lie within a band of 40 km along the profile. On land, heat flow values are obtained from Henyey (1968) and Sass and others (1971). High heat flow appears in a trough associated with the Patton Ridge system (called western Tanner Basin in previous sections). This high is adjoined by a heat flow low at a distance of 24 km in the eastern Tanner Basin. From here, the heat flows increase slightly landward to San Pedro Basin. On land the heat flows appear to decrease away from shore. A significant transition in heat flow apparently occurs somewhere in San Pedro Basin Figure 33. Trans-borderland heat flow profile II-II' (see Figure 32 for caption). 4 5 * O O O D E P T H r o o o o ______i ___ HEAT FLOW 2 u cal/cm sec o _ ro J 3 w '9 R tOGE PATTON TANNER BASIN SAN NICOLAS B. SANTA CATALINA B. SAN PEDRO B. sn 2.8H Figure 34. Trans-borderland heat flow profile I-I1 (see Figure 32 for cation). r o o DEPTH o HEAT FLOW j x cal/cm^ sec LPT o PATTON LONG WEST CORTES B. EAST CORTES B. SAN CLEM ENTE B. SAN DIEGO TROUGH 148 but the data are too scattered to locate the heat flow maximum. The average value of the corrected heat flow data in each basin is also plotted. The corrected and un corrected data give essentially the same heat flow trend. In San Pedro Basin the highest value at station PI is ex cluded from the average due to the probability of local perturbation. A similar heat flow trend occurs in Profile I-I1 (Figure 34) which is essentially in line with Von Herzen's (1964) profile 1-1' about 180 km south of profile II-II'. Observed heat flows decrease from Long Basin across West Cortes Basin to East Cortes Basin, and then increase land ward across San Clemente Basin to San Diego Trough. Data from other places in the northern Borderland also suggest a relatively high heat flow zone associated with troughs within the Patton Ridge system. If Von Herzen's profile 2-2' in the southern Continental Borderland is extended further landward (using his data), it can be seen that the heat flow appears to increase landward in the southern Borderland too. The transition zone from high to low heat flow in the outer Borderland becomes wider southward. Part of the sharp transition in profile II-II' may be attributed to the rapid change in elevation across the continental slope and to the refraction of heat as a consequence of variable 149 sediment cover. However these two effects are not respon sible for the broad transition zone in profile I-I1. The continental slope is not well defined topographically in the southern Borderland and thus refraction does not affect profile 2-2' of Von Herzen (1964). The linear increase of heat flow from the seaward edge of the central Boderland to the inner Borderland is positively correlated with a shoaling basin floor toward the continent along profile 1 1-1 1 ', whereas such a correlation is not apparent along the other two profiles. SEISMIC AND MAGNETIC PROFILES Seismic Reflection Profiles Seismic reflection data were obtained in the northern Borderland along the traverses shown in Figure 23. Part of each traverse (heavy line) is shown in Figures 35 to 38. The acoustic energy was provided with a Bolt air-gun (model 600B) operated at pressures ranging from 1700 to 2000 psi. The air chamber volume used was either 10 to 20 cubic inches. The echoes were received with an 8-meter hydrophone array (five hydrophones connected in parallel) towed about 80 m behind the ship. The signals were ampli fied and filtered (pass band: 80 to 320 hz) and recorded on a Gifft recorder. The recording was maintained at one second per sweep and the trigger programming was operated at integral multiples of 0.5 sec. All profiling was accomplished at a ship speed of 7 knots. The seismic reflection profiling was used to gain a better understanding of the local basin sediment structures and topography adjacent to heat flow stations and thus to investigate effects which might perturb the local thermal regime. Some profiles were run on NW-SE lines to supple ment the EW profiles of Moore, which cross the regional structural trend of the Continental Borderland. The 150 151 transverse profiles have the advantage of providing an in tegrated knowledge of the regional structure but suffer the drawback of missing the transverse structure which has been unduly neglected in the literature. Moore (1969) divided the sedimentary deposits in the Continental Borderland into pre-orogenic sedimentary rocks and post-orogenic sediments. The pre-orogenic sedimentary rocks were deposited before or during the time of intense folding, faulting, and volcanism which formed the principal topographic features of the Borderland, and the post- orogenic sediments include all deposits which have since accumulated. Major tectonic activity occurred during Miocene, hence post-orogenic is in general equivalent to post-Miocene. Because tectonic activity continued locally into Pleistocene, pre- and post-orogenic sediments refer actually to deposits accumulated before and after the formation of a given basin, respectively. Moore's criteria for distinguishing these two types of deposits will be used in the interpretation of the following data although ambiguity in such a division occasionally arises. Magnetic Profiles Total magnetic field intensity was measured routinely with a proton magnetometer (Geometries G801) along the 152 traverses shown in Figure 23. Calculation of the geo magnetic reference field was based on the spherical harmonic coefficients extrapolated from the 1965 Interna tional Geomagnetic Reference Field (IGRE, 1969) . Anomalies,, obtained by subtracting the reference field from the meas ured values, are plotted in Figures 39 to 43. No correc tion for diurnal variation has been made. Along each traverse, the topography was also plotted on the basis of seismic reflection data when available. For convenience, the magnetic anomaly along a given profile is discussed in terms of a long wavelength com ponent reflecting deep structure (referred to here as "background anomaly") and a short wavelength component reflecting near surface structure (referred to here as "local anomaly"). Usually the local anomaly is correlative with the juxtaposition of near surface rocks with contrast ing magnetic susceptibility. Magnetic profiling was used to aid in mapping the basement relief and the distribution of volcanic rocks. A detailed correlation between the local magnetic anomalies and features revealed by the seismic reflection profiles has been made in conjunction with the litho-orogenic map of Moore (1969, Plate 14). The resulting analyses provided additional insight into the general structure of the Borderland. 153 Presentation and Discussion of Data The data are presented basin by basin, according to the following format. First, the major structures revealed by reflection profiles are introduced and if available,the refraction data of Shor and Raitt (1958) are added. Second, the magnetic data is introduced to evaluate the significance of faults and for correlation with features revealed by the the reflection profiles. Finally, features in sedimentary section are presented. Two trans-borderland magnetic anomaly patterns are also presented at the end of this section. San Pedro Basin Two profiles cross San Pedro Basin (Figure 35, upper right, CA and AB). The undeformed sediments along the northern profile CA and the deformed sediments along the southern profile AB suggest that a major structural dis continuity exists in the region between the two profiles. The deformed sediments are probably pre-orogenic and the undeformed sediments are post-orogenic. The background magnetic anomaly increases toward the continent, suggesting a deeper burial of the basement rocks toward the Santa Catalina block (Figure 39). San Pedro Escarpment possesses peak-to-peak local anomalies of 100 154 and 40 gammas (f) in the northern and southern profiles respectively, whereas the fault northeast of the Santa Catalina block is marked by an inconspicuous anomaly e of only 15 gammas. Anomaly h is within the San Pedro Sea Valley. The sharp features of anomalies b, c_, and d in dicate steeply dipping shallow-seated tabular bodies such as dikes. Some minor features in the sedimentary section are worth noting. The intensity of synclinal folding in the post-orogenic sediments becomes more pronounced away from the continent. Landsliding and slumping on the northeastern edge of the basin becomes less conspicuous to the south. A relatively strong reflector exists in the pre-orogenic sediments at a depth of 0.65 sec, suggesting a discon- formity or a significant change in the type of sediments. Santa Catalina Basin The basement underlying Santa Catalina Basin has high relief and numerous faults (Figure 35,upper left, DC), but the near surface sedimentary structures do not appear to reflect these features and are separated from them by an intervening transparent layer. The basement appears to crop out at the base of the ridge between Santa Barbara and Santa Catalina Islands. A refraction profile of Shor and Raitt (1958, Figure 7) indicates that a depression with a 155 relief of >vi2 km in the basement occurs in the northwestern part of the basin. Correlation between that depression and the depressions revealed in profile DC cannot be es tablished with the available data. It seems, however, that a significant transverse structure may exist in the basement of Santa Catalina Basin. As a whole, Santa Catalina Basin is an area of magnet ic low in the Continental Borderland (Figure 39, lower DC). Anomaly a (50 gammas) coincides with the andesitic Emery Seaknoll (Gaal, 1966). Anomaly b can be correlated with a buried basement high and anomaly c appears 'over an ex posed basement high. The elevation difference between the buried and exposed basement high, is only ~ 1 0 0 m and the corresponding anomaly difference is 2 0 0 gammas, whereas anomalies a and b have similar magnitudes despite an elevation difference of 510 m. Lack of correlation between magnetic anomaly and relief suggests that the difference in magentic susceptibility of the volcanic rocks and base ment rocks is significant if both have similar direction in magnetization. The post-orogenic sediments appear to grade downward from well-stratified and essentially undeformed sediments into gently folded sediments. Pre-orogenic sedimentary rocks as observed by Moore (1969, Plate 5) about 10 km to the northeast of this profile could be absent here. 156 San Nicolas Basin Three types of sediment configurations can be dis tinguished in San Nicolas Basin (Figure 35, lower left FE): 1 ) gently dipping but unfolded sediments in the northwest portion, 2 ) flat-lying sediments in the central portion, and 3) folded and faulted sediments in the southeast portion. Type 1 and 3 are presumably pre-orogenic and type 2 is post-orogenic. Type 1 sediments abut a steep, east-trending escarpment (south of San Nicolas Island) at the base of which there is a trough, suggesting the presence of an active fault associated with the escarpment. The contact between type 1 and type 2 sediments is a fault, as is the contact between type 2 and type 3. On the assumption of plane-layered structure, Shor and Raitt (1958, Figure 6 ) have shown that the basement below San Nicolas Basin tilts northwestly. This is com patible with the fact that the background magnetic anomaly decreases northwestward (Figure 40, lower half, FE). Anomaly a is associated with the fault between type 2 and type 3 sediments but no local anomaly is associated with the presumed fault between type 1 and type 2 sediments. Probably a faulted plane-layer structure is a better approximation to the structure of San Nicolas Basin than a tilted plane layer structure as shown by Shor and Raitt. 157 Tanner Basin (Transverse Profile) A profile across western Tanner Basin is shown in Figure 36 (upper left, CG). This basin is a graben whose edges are marked by a pair of strong positive magnetic anomalies c and e (Figure 42, CG). The ridge between the eastern and western depressions of the basin is associated with magnetic anomaly d. The fault at the western edge is one of the major faults bounding the Patton Ridge. Some features in the sedimentary section are note worthy. The "W-shaped" folds are supratenuous. At the base of each escarpment there is a trough, below which the drag folds grade downward from reverse to normal. In the western half of the section there are five distributary channels, sitting directly above the synclines of the W-shaped folds. The sediments appear to drape over the eastern flank of the ridge between eastern and western Tanner Basin. Since no magnetic anomaly is associated with the W-shaped folds, diapiric igneous intrusion can be ruled out as a possible mechanism for the formation of the W-shaped folds in the outer fault zone. Some 6>f the troughs within the Patton Ridge are barren of sediments while others are filled, indicating different stages of evolution (Figure 36, upper middle, GH). Here, the sedimetnary structures are supratenuous, as are the sedimentary structures in the depression northwest of 158 Tanner Basin proper. Tanner Basin (Northwest-Southeast Profile) Near the middle part of profile HI (Figure 36, upper right and lower left), a 1 0 0 -m fault escarpment is also associated with magnetic anomaly £ (Figure 41, upper, HI). This is the western boundary fault of southern Tanner Basin. It has been traced northward across the basin to the eastern boundary fault of Tanner Basin by Emery (1960) and Moore (1969) based upon bathymetry. If these two faults are actually correlative, they have opposite senses of apparent vertical displacements. If connected, a rota tional movement along the fault plane would be implied. If connected as a strike slip fault, it would be left lateral judged from topography, but its sense of re lative movement would be contrary to those of onshore northwest-trending faults. A likely course of the eastern boundary fault in the northern basin would be to continue along the eastern side of the southern basin. A transverse fault scarp bounding the northern end of Tanner Basin is associated with magnetic anomaly b. Anomaly a is associated with a hill northwest of the Tanner Basin proper. This hill was mapped as sedimentary rocks by Moore (1969, Plate 14) but the strong positive anomaly (+180 gammas) suggests that it may be a basement 159 or volcanic high covered with a thin veneer of sediments. Three minor structures are worth noting. 1) An anticline occurs about 2 km from the northern basin wall. Between the wall and the anticline, the flat-lying sedi ments are underlain by the down-buckled sediments which at depth greater than 1.05 sec are replaced by a sequence of sediments parallel or subparallel to the wall. The anti cline also loses its identity at that depth. 2) A possible fault occurs between the dipping sedimentary rocks associated with the anticline and the flat-lying sedimentary rocks in the basin proper although no magnetic anomaly was detected. And 3) a monocline exists in the middle part of the basin. San Clemente Basin In the northern San Clemente Basin (Figure 37, upper left, ED), the sediments conform to basement topography and the basin floor reflects the basement relief. In central San Clemente Basin (Figure 37, upper middle QR), post-orogenic sediments overlie (unconformably?) a pre- orogenic syncline. Noteworthy is the monocline near the the base of the northeastern slope. The area between the edge of the basin floor and a 300-m escarpment about 12 km east of the basin is presumably within the San Clemente - Agua Blanca fault zone. Here, the sedimentary structures 160 are highly disturbed by faulting. The basin floor in the southern San Clemente Basin (Figure 37, lower right, PQ) is essentially devoid of flat-lying post-orogenic sediments. This observation is inconsistent with Moore's (1969) in ference that in this area post-orogenic sediments may range from 0.1 to 0.5 sec (two way travel time) in thickness. The background magnetic anomaly increases from the southern part of San Clemente Basin to the central part (Figure 40, upper, PQR) . Anomaly f_ is peaked over the monocline at the base of the northeast slope. This probably indicates the presence of the San Clemente fault here. Thirtymile Bank is an area of magnetic low but it was mapped as volcanic rocks by Moore (1969, Plate 14). The topographic high on the western side of southern San Clemente Basin is flanked by small anomalies d and e suggesting that it is composed of rocks similar in magnetic property to those in Thirtymile Bank. This topographic high was mapped by Moore as either volcanic or basement rocks. Animal Basin Animal Basin is characterized by abundant fault scarps (Figure 37, lower middle, OP). The sediments are gently folded. In this section a topographic low or tilted graben, which is formed by a series of normal faults, is marked by a fairly symmetric anomaly a (Figure 41, upper, 161 OP). No Name Basin (Figure 41, upper, at P) is an area of magnetic low with respect to San Clemente Basin and Animal Basin. Velero Basin In Velero Basin (Figure 38, upper left, KL) the sedi ments are gently folded but fault scarps are rare. Only a pair of minor horsts and grabens appear in the middle part of the section. The magnetic anomaly decreases toward the center of the basin (Figure 41, lower K). Outer Basin Notable in Outer Basin (Figure 38, upper middle, LMN) are the humps (anticlines) and the intervening flat-lying or slightly down-buckled sediments. According to U.S. ONR chart BC 1206 (1969) , the Outer Basin shown in the bathy metric map of Moore (1969) does not exist. It opens to the Pacific Ocean floor and is named Santo Tomas Gap by Doyle (1973). Based on bathymetry and reflection seismic profiles, Doyle has delineated the Santo Tomas fault across the Borderland to the Gap. However, there is no clear evidence from profile LMN to show that the Santo Tomas fault could go through the central part of Outer Basin or Santo Tomas Gap. A profile across the eas tern end of the Southwest Bank (Figure 37, lower left, 162 NO) appears to confirm Moore's (1969, Plate 13) and Doyle's extrapolated trace of the Santo Tomas fault, where low mag netic anomaly is also recorded (Figure 41, lower, ; j _ ) . No local magnetic anomaly occurs over the anticlines in Outer Basin except a magnetic low d appears at its northeast edge (Figure 41, lower, M). The Southwest Bank here is flanked by a pair of magnetic highs b and c, while its southern tip is associated with anomaly k. This bank was mapped as volcanic in its southern half but volcanic or basement in its northern half (Moore, 1969). The mapping is compatible with the magnetic anomaly. Continental Slope To the southeast of Outer Basin, the continental slope is marked by a step-like fault scarp (Figure 38, upper right, MNO). Here, the sediments at the base of the con tinental slope probably do not exceed 2 0 0 m in thickness. The Rampart, equivalent to Patton Ridge in the northern Continental Borderland, was mapped as volcanic rocks (Moore, 1969). The magnetic anomaly in this area is very complex, reflecting the complex topography and thus suggesting that the Rampart is composed of volcanic rocks (Figure 41, lower, MN). 163 Santa Barbara Basin A profile across the western part of the transverse structural zone is shown in Figure 36 (lower right, XY). The shelf to the west of Pt. Conception and Pt. Arguello is underlain by folded sedimentary rocks which have been truncated and covered with a 1 0-m veneer of sediments. There is no evidence from bathymetry or sedimentary structures to suggest that the Santa Ynez fault extends offshore. The fault at the foot of the basin slope, as marked by a sediment-free trough is a westward extension of the north-bounding fault of Santa Barbara Basin. Near the southern end of the profile, a strong reflector appears at depth of 0 . 6 sec (one way) and rises south- eastly and probably forms the basement of San Miguel Island. Noteworthy in this profile is the abundance of canyons and/or distributary channels. Magnetic Profiles across Continental Borderland A magnetic profile (CG) across northern Continental Borderland, about 5 km northwest of heat flow profile II-II', is presented in Figure 42; the portion over San Pedro Basin is given in Figure 39. The background anomaly appears to decrease seaward until it reaches the southest edge of Santa Rosa - Cortes Ridge. In the outer fault belt the background anomaly seems to be fairly flat. It 164 is noted that the peaks of local anomalies (a, b, and c) lie to the seaward side of the corresponding ridges. Noteworthy is the flat-topped anomaly f^ above the base of the continental slope. This feature also appears along a profile to the north (the profile is not presented here). Another trans-borderland profile about 20 km south of Santa Cruz Island is shown in Figure 43 (ST) in which a magnetic profile following the trend of Santa Monica Basin is also presented (AS). The ridge between Santa Monica Basin and Santa Cruz Basin is marked by a peak-to-peak anomaly of 200 gammas (£ and d). As is true in the trans borderland profile CG, the background anomaly decreases seaward. Santa Cruz Basin is an area of magnetic high. The Santa Rosa - Cortes Ridge is marked by magnetic low t_, as is the topographic high to the west of Santa Rosa Ridge (between anomalies h and i). 165 Figure 35. Seismic reflection profiles in Santa Catalina Basin (DC), San Pedro Basin (CA,AB), and San Nicolas Basin (FE). Upper left, DC: a SE-NW profile follow ing the trend of Santa Catalina Basin from Emery Seaknoll to the base of the ridge between Santa Barbara and Santa Catalina Islands. Upper right, CA: a SW-NE profile from the Santa Catalina ridge across northern San Pedro Basin to the San Pedro shelf. Upper right, AB: a NE-SW profile from the SanLPedro shelf across southern San Pedro Basin to Santa Catalina Island. Lower left, FE: a NW-SE profile follow ing the elongated trend of San Nicolas Basin. Lower right, DC: part of the profile DC but from Fortymile Bank to Emery Seaknoll. Note the contrast of dipping sediments here and north of Emery Seaknoll. 166 167 Figure 36. Seismic reflection profiles in Tanner Basin (CG, HI), across Patton Ridge (GH), and on shelf west of Pt. Conception (XY). Upper left, CG : a NE-SW profile across the western Tanner Basin from western edge of Tanner Basin proper to Patton Ridge. Upper middle, GH: Lower right, XY a SW-NE profile obliquely across Patton Ridge to the depression northwest of Tanner Basin proper. a N-S profile to the west of Pt. Conception and Santa Barbara Basin. Upper right H-— : a NW-SE profile follow- lower left I ing the elongated trend of Tanner Basin. in r - ~ o o « o in in Mill i i i n II I I 4 - I 1 .1. I I M ' ' ‘ ' z r l —I - s y fel* ... - I I ' ' I I I F-M M4-4" i i r jm i i i I M I —M I I II II I UL-Ui-l I - U J . -H I 1 * 1 I in i —I £ •oas i I - - I i ^ u : -Ste-y j Ml i —i—i i iW'ti f- i i J t . ' 1 j' i i i -M 1 1 1 1 1 1 1 1 1 1M 1 1 4 i i \ . i I i i t i 1 1 ’w v m i ® : I 1 1 rw i i i n ' i I I 4 1 1 I N I M M •ip ^ ' , i S fi i ’1 i l l l l l l i M M LLi M i l 1.1 1 1 i •I ..' 1— 1 —i i I I iJJ i l l /£ I 1 fv/ I t k 1 j F- ' r I f '-•j - M i l I I I I L_J_4- i_ I i” i i m i i i l a 1 i i i'll j < i ■ s i r a . ^4-1 1 1 I I I I I m i w ; m i ) f i 1 4 rr-1 '‘• I •' Ji I - ..I - ' n i l : _.i, i— i_i ' T t r t j i i n • I f ‘ ‘ ‘M i ll • W JI M - T t 1 1 1 1 1 1 U 1 —I — — /h) III ! i - f ft 1 •• 1 "K ii i: n . 1 -I M f f H 'V I M i ! u i - I — M M III! ; i t fo-'lVU! ! ! " 4 1 U l 1 I 1 ! f j i i t , M - M M I I M . 1 - i I 1 1 llit!5 ? i i n i . ! ! i i o o •oas in CN H 169 Figure 37. Seismic reflection profiles in San Clemente Basin (ED, QQ1, PQ), Animal Basin (OP), San Diego Trough (Q'R), and at southeastern end of Southwest Bank (NO) . Upper left, ED Upper middle and right, QR a SW-NE profile across northern San Clemente Basin. a SW-NE profile from central San Clemente Basin (QQ1) across Thirtymile Bank to San Diego Trough (Q'R). Note the shifting of vertical scale. Lower middle, OP: a SE-NW profile follow ing the elongated of northern Animal Basin. Lower right, PQ : a SW-NE profile across southern San Clemente Basin. 171 Figure 38. Seismic reflection profiles in Velero Basin (KL), Outer Basin (LWN), Santa Cruz Basin (UW, ST), West Cortes Basin (IJ), and the continental slope southeast of Outer Basin (MNO). Upper left, KL Upper right, MNO: Upper middle, LMN Lower left, UW Lower middle, ST: Lower right, IJ a NW-SE profile in the northern Velero Basin. a composite profile (NW-SE and SW-NE) across the continental slope southeast of Outer Basin. a profile (NE-SW and NW-SE) across Outer Basin. line tracing of a E-W profile across the southern end of Santa Cruz Basin. a NE-SW profile across northern Santa Cruz Basin. a NW-SE profile follow ing the elongated trend of West Cortes Basin from northern to southern de pressions . CN rH SZ'T 0 0 *1 - U I > [ "1 i i 0 ~tos ' I m W r n m m i i a l l i i l g l w i i i s H B u 9Z'T 09 ’ T 9Z'T n SL'T 09 ’ I 9Z.‘T - 09*1 N W mm - I 0 0 *Z SL’ I cn W n 173 Figure 39. Magnetic profiles in San Pedro Basin (upper CB) and Santa Catalina Basin (lower DC). Coarse line: bathymetry from seismic reflection profile. Magnetic scale 100 gammas, bathymetric scale 1 0 0 m. 50 KM T 0 ±10 174 175 Figure 40. Magnetic profiles in San Nicolas Basin (lower FE), Animal Basin (upper OP) and San Clemente Basin (upper PQR). Magnetic scale 100 gammas, bathymetric scale 1 0 0 m. cr Ui e o a. Ln 9LT 177 Figure 41. Magnetic profile in Tanner Basin (HI), West Cortes Basin (IJ), Velero Basin (JKL), Outer Basin (M), Rampart (MN), and Animal Basin (0), Magnetic scale 100 gammas and bathy metric scale 1 0 0 m. 178 m •A 179 Figure 42. Trans-borderland magnetic profile (CG) . Magnetic scale 100 gammas, bathy metric scale 1 0 0 m. SAN NfC/DLAS BA 5/A/ \ TANJNER. BASIN to 081 Figure 43. Trans-borderland magnetic profile (ST) and Santa Monica Basin (AS). Magnetic scale 100 gammas, bathy metric scale 1 0 0 m. SANTA MONICA BASI N 50 TECTONIC IMPLICATIONS FROM HEAT FLOW A linear correlation between heat flow and radioac tive heat production has been well documented in the Sierra Nevada province, Basin and Range province, and eastern United States (Roy and others, 1968; Lachenbruch, 1968). Because the radioactive heat production per unit volume is higher in the continental crust: than in the oceanic crust, landward increasing heat flow in the Continental Borderland may suggest that the abundance of radioactive crustal rocks increases toward the continent; however, no data are pre sently available to subtantiate or negate this suggestion. If this suggestion is confirmed, it still cannot explain the systematic increase of heat flow toward the southeast. On the basis of rock types described by Emery (1960) , Krause (1965), Doyle (1973), and other workers, there is no positive reason to believe that the content of radioac tive elements should also increase southeastward. An alternative explanation of the heat flow pattern is to consider a transient thermal regime (the major radioactive heat production elements, U, Th, and K have long half-lives and their contribution to heat flow can be considered as steady state). Regional heat flow has been shown to decrease with age of the most recent regional 183 184 tectonic activity or age of the crust (Sclater and Franche- teau, 1970). Since the heat flow increases southeastward, perhaps the formation of the Continental Borderland has pro ceeded from north to south. This hypothesis can be tested. Tectonic Model Hypotheses on the formation of the Continental Border land include regional and local warping (Emery, 1960) , thrusting of the Peninsular Ranges over the Pacific Ocean floor (Rusnak and Fisher, 1960), oblique rifting and com pression (Hamilton and Meyer, 1966 and 1968), shearing be tween the American and the Pacific plates (Atwater, 1970) , rhombochasm-rifting (Suppe, 1971), east-west dilation coupled with northwest-southeast shearing (Krause, 1965), east-west rifting (Yeats, 1968 and 1973), and differential east-west extension (Davis and Burchfiel, 1973). Only the models of Suppe and Atwater provide a time frame work of borderland development. Atwater (1970) hypothesized that before 29 m.y.B.P. a plate (Farallon) lay between the North American and the Pa cific plates (Figure 44a). The Farallon plate was generated by a spreading center shared with the Pacific plate but was consumed by a trench then off present western North America. Because the consumption rate was greater than the generation rate (half rate 5 cm/year) the ridge eventually impinged upon the trench and the Pacific plate came into contact with the North American plate at 29 m.y.B.P. (Figure 44b). 185 This contact was resolved into two migrating triple junc tions with an intervening transform fault serving as the Pacific and North American plate boundary (Figure 44c). Two models of plate motion were examined by Atwater. A model with changing motion predicts that the San Andreas fault system has taken up all the relative motion between North American and the Pacific plates since its activation 5 m.y. ago and that the place of initial impingement was near San Francisco. A second model of constant relative motion over the past 29 m.y. predicts that motion between the Pacific and North American plates has been taken up not only by the San Andreas fault system but structures across the western United States and that the place of initial impingement was near Guaymas, Mexico. In this study, I propose as an alternative to Atwaterfe models, that the place of initial ridge-trench impingement was located 29 m.y. ago at the base of the continental slope off the present western Transverse Ranges. The Mendocino Fracture Zone was then in line with the western Transverse Ranges and the Murray Fracture Zone was in line with the southern end of the Continental Borderland (Figure 45a). If the relative motion between the Pacific and North American plates has been constant (6 cm/year) in the past 29 m.y., then a total strike slip displacement of 1740 km has occurred between the two plates. It is assumed 186 Figure 44. Relative position of the Pacific, Farallon and North American plates before, at, and after ridge-trench impingement. a. Before ridge-trench impingement, relative position derived from anomaly 13. b. At 29 m.y.B.P., hypothesized time of initial impingement; dot des ignating place of initial impinge ment. c. After initial impingement. North Mendocino F.Z. L American Pacific Plate Plate Farallon Plate Figure 44a. — Place of Initial Impingement Mendocino F. Z. Murray F. Z Figure 44b North American Plate Pacific Plate Figure 44c. 188 here that 850 km of this displacement has been taken up along an offshore transform fault at the base of the con tinental slope and that the Murray and Mendocino Fracture Zones had moved to their present positions relative to coastal California by 5 m.y.B.P. (Figure 45c). The rest of the total displacement has been dispersed along the San Andreas fault and other strike slip faults in the west ern United States. In other words, a continental sliver bounded by an offshore transform fault and the San Andreas fault once existed. Because Baja California and coastal California west of the San Andreas fault may have been part of the Pacific plate since the Gulf of California was opened, the relative plate motion along the offshore transform fault can be inferred to have ceased 5 m.y. ago. On the basis of this hypothesized model, the tectonics of the westernmost North America between 29 and 5 m.y.B.P. was influenced by five plates, the North American, Pacific, southern and northern Farallon plates, and a continental sliver. According to the hypothesized model, the relative motion V between the Pacific plate and the continental ps sliver was constrained by the initial velocity V = 6 x 10-^ km/year, (1) ps terminal velocity V =0, (2) ps 189 Figure 45. Relative position of ridge, trench, and transform fault. a. At time of initial impingement, 29 m.y.B.P. b. At time 0, 4, 8 , 12, 16, 20, and 24 m.y. after the initial impinge ment. c. At present. Mendocino Fracture Zone Murray F. Z place of initial impingement "Continental Borderland" Figure 45a. 24 Figure 45b. Mendocino Murray Borderland Figure 45c. 190 I I 191 and relative displacement along the offshore transform fault Sb = j[vpsdt where S j - , is the length (850 km) of the Continental Borderland along the continental slope and T = 24 x 10 years (i.e., from 29 to 5 m.y.B.P.). Without further constraints, the relative motion VpS as a function of time t cannot be uniquely determined. If the function can be approximated by V„c = A + Bt + Ct2 (4) Pt o g where 0 £ t ■ £ 24 x 10 years. Substituting (4) into (1), (2), and (3) yields — ^ A = 6 x 10 km/year, B = -1.146 x 10 km/year2, and C = -5.642 x 10- d 2 km/year2. This solution has the property that the velocity decreases with time whereas the deceleration increases with time (Figure 46a). This also represents the northwestward movement of the trench-transform-transform triple junction. For the convenience of discussion, the continental sliver is fixed in position and other plates move with respect to the sliver. The Pacific plate had moved northwestward since the initial impingement by the amount S = f V„0dr = At + Bt2/2 + Ct3/3 (5) ps j0 ps 0 < t < 24 x 10 The North American plate with respect to the con tinental sliver moved southeastward at a rate of 192 Figure 46. Relative plate motion as a function of time. a. Velocity-time curve. b. Plate motions with respect to the continental sliver in which the dot indicates the place of initial impingment. Figure 46b. Relative Velocity cm/year a\ oo 194 V _ = 6 x 10“ 5 - V 0 for 0 < t < 24xl06 . (6 ) aS p S — — Since 5 m.y. ago, the sliver has been part of the Pacific plate thus V = 6 x 10"^ for t > 24 x 10^ (7) clS (Figure 46a). An interior point in the North American plate has moved southeastward by S fv dT 0 < t < 24 x 106 (8 ) as Jo as ^ — — Sag = Ji“ ' M0 Vasdr + (t - 24xl06) x 6 x 10~ 5 o 6 t > 24 x 10 . (9) In the last 29 m.y., = 890 km. Presumably this offset cLS has been taken up by the San Andreas fault zone and other strike slip faults in the western United States. Another triple junction (ridge-trench- transform) was moving southeastward with respect to the continental sliver at a rate of V' = 1/2 V ^ esc oC - V e (10) pf ps where the relative speed of the Pacific and Farallon plates is Vp^ = 10 cm/year according to Atwater (1970), of(37°) is the angle between the spreading center and trench, assumed to be the angle between the trends of magnetic anomaly (e.g., anomalies 5 and 6 ) and the con tinental slope. This rate was increasing until 5 m.y. ago (Figure 46b). With respect to the place of initial impingement on the sliver, this southeastward migrating triple junction has moved by the amount S' = 1/2 Vpf t cscot - SpS ,0 < t < 24 x 106 195 The time sequence of the positions of the ridge, fracture zones, and triple junctions are shown in Figure 45b. At this rate V1, the ridge segment between the Mendocino and Murray Francture.Zones was completely subducted within 10 m.y. (by 19 m.y.B.P.), and the triple junction had traversed the length of the Continental Borderland by the end of 20 m.y. (9 m.y.B.P.). If the Miocene block faulting and volcanism in the Continental Borderland were related to the migration of triple junction and the subduction of a spreading center, the Borderland should be formed successively from north to south. The actual timing of the formation is not known but an estimation based on the proposed tectonic model may be suggested. Subduction of a spreading ridge could result in anomalous temperature in the crust and upper mantle. After the overridden ridge had moved beneath the borderland to a certain distance from the trench, the temperature could be so high as to partially melt the rocks in part of the crust or upper mantle. Thus there was a time lag between the onset of volcanism and the passage of the southeastward migrating triple junction. This time lag, At, can be estimated by At ~ (D/V'tam*. cosp) + (D tanp/Vm) (11) where p is the dip of Benioff zone, D is the mean distance of volcanic belt measured in a direction perpendicular 196 to the trench, and Vm is the average velocity of rising magma. Because V 1 became faster with time, the time lag was smaller at the northern end than at the southern end of the Borderland. As an example, if / } »18° (Schlotz and others, 1971) and D ~ 200 km (the distance from the base of continental slope to the central part of inner Border land) , the time lag is at least 12 m.y. (neglecting the second term in equation 11 for unknown V ) in the northern end and 4 m.y. in the southern end. Thus, on the basis of these assumptions, the Borderland was formed not earlier than 17 m.y.B.P. at the northern end and 5 m.y. at the southern end and the time difference was 12 m.y. instead of 20 m.y. which was the travel time of the triple junction along the length of the Borrderland. Implications from Heat Flow The hypothesized model suggested that the ridge- trench-transform triple junction migrated southeastward away from the place of initial impingement (Figure 45b). In other words, subduction of the ridge proceeded from northwest to southeast. The overridden ridge was essen tially a heat source which migrated southeastward beneath the Borderland. Thus, the model is compatible with the fact that heat flow in the Borderland increases southeast ward. The next step is to determine from heat flow data 197 when the heat source or overridden ridge had passed be neath the Borderland and to explain why the place of ini-, tial impingement was located off the present western Transverse Ranges. To determine when the ridge had passed the Continental Borderland, one can estimate the transient effect due to the passage of the heat source by removing the equilibrium value from the observed values. If at some time in the past the ridge had been far from the trench, the heat flow from the oceanic plate near the trench had reached an equilibrium value of 1.1 or 1.2 (Sclater and Francheteau, 1970). This value is a combination of ~0.2 from the radioactive heat production within the oceanic lithosphere and 0.9 to 1.0 from the oceanic asthenopphere. On the other hand, in shield areas the equilibrium heat flow of 1.05 consists of 0.55 from radioactive heat production in the lithosphere and 0.5 from the asthenosphere (Sclater, 1972). At the continental margin, the equilibrium heat flow can be considered to come from the down-going slab, the crust and the part of the mantle above the slab, and the part of the mantle below the slab (Figure 47). The radioactive heat production per unit volume in the down- going slab is the same as in the oceanic plate. The contribution from the crust and the part of mantle above the slab suffers from the uncertainty in estimating the 198 Figure 47. Heat flows from ocean basin, con tinental margin, and shield. a. When the ridge was far from the trench. b. When the ridge was near the trench. Shaded area indicates oceanic plate. 199 « s c « u c r \ q.CeoJ-i.1 <u 1.2. = + 0.1 • .2 1.05 -r ^ « f o.£5 Oceanic mantle 0.5 coniinenta.1 "mantle Figure 47a. Q J Q c I p Figure 47b. radioactive heat production and the frictional heating. Clastic sedimentary rocks comprise approximately 90 % of the eugeosynclinal assemblages in the Franciscan Group which presumably is the basement of the Continental Borderland. The radioactive heat production in the Mesozo ic clastic roof-pendants and country rocks of the Sierra 3 Nevada batholith is ~0.35 p.;ical/cm sec based on Wollenberg and Smith's (1970) data. This value is nearly equal to 3 the average value of 0.38 ppcal/cm sec in the granitic rocks of the batholith (based on data from Lachenbruch, 1968). The radioactive heat production per unit volume in the Franciscan Group probably does not exceed these two values. If the heat production in the rocks underlying the Borderland is 0.3 pjacal/cm sec and is uniformly dis tributed in a "hypothesized slab" of 1 0 km thick, the contribution of radioactive heat production is 0.3. The "hypothesized slab" deduced from the linear relation of heat flow and radioactive heat production is 10 km thick in the Sierra Nevada province, 9.4 km thick in the Basin and Range province, and 7.5 km in the eastern United States (Roy and others, 1972). The contribution of frictional heating probably does not exceed 0.2. If, to a first approximation, the contribution from the part of mantle below the down-going slab is the same as that from the asthenosphere below shield areas, the mantle heat 201 flow at continental margin is 0.2 + 0.2 + 0.5 = 0.9 and the equilibrated surface heat flow is 0.9 + 0.3 = 1.2. As a comparison, the mantle heat flow is 0.4 from the Sierra Nevada province, 1.4 from the Basin and Range province, and 0.8 from eastern United States (Roy and others, 1972). The heat flow from the continental margin will in crease as the ridge approaches the trench due to the resid ual heat in the oceanic lithosphere (Sclater and Franche- teau, 1970). To estimate when (apparent age) the ridge passed the Continental Borderland, one reconstructs the relation of heat flow and age by removing the equilibrated oceanic heat flow from the oceanic heat flow, i.e., Aq (t) = q (t) - q(oo) (1 2 ) where q(t) is the heat flow - age relation for the northern Pacific Ocean floor (Sclater and Francheteau, 1970, re produced in Figure 48a), and q(«>) is the the equilibrium value/ -~1.15, near the trendh.when the ridge was -~100 m.y. from the trench. The anomalous heat flow Aq1 from the Continental Borderland is Aq' - q 1 - q'(«) (13) where q' is the observed heat flow (or corrected heat flow) and q ' (°° ) is the equilibrium value, 1 .2 , when the ridge was far (~100 m.y.) from the trench. The apparent age is obtained by equalizing Aq to Aq1 and inverting Aq' for 202 Figure 48. Relation of heat flow and age. a. Relation of heat flow and age in the northern Pacific (after Sclater and Francheteau, 1970). q and aq are defined in equation (1 2 ). b. Theoretical heat flow profile across a trench*, solid curve from Oxburgh and Turcotte (1971) and dotted curve from McKenzie and Sclater (1968). c. Speculated heat flow distribution pattern based on Figures 48a and 48b assuming no lateral heat transfer across the transform faults. Bigger symbols represent relatively high heat flow but no relation between different symbol is implied. Solid circle designate point of initial impingement. 50 100 Figure 48a. ! 3 o O * ■ I —i h -p ( d Q - B 1 150 Age xlO years < i Oce an 500 Figure 48b, 500 km Figure 48c. t with the aid of Figure 48. For example, the average heat flow q' in Santa Cruz Basin is 1.8 and Aq1 = 1.8-1.2 = 0.6. Since Aq = A q' = 0.6, the apparent age from equation (12) or Figure 48a is 27 m.£. Similarly the apparent ages are 22 m.y. in San Nicolas Basin, 20 m.y. in East,Cortes Basin, and 15 m.y. in Animal Basin. These ages appear to be too old in comparison with the ages based on the hypothesized tectonic model and the assump tions made in evaluating time lag in equation (11). The discrepancy may result from over-estimation of the radio active heat production or frictional heating. It may be due to the possibility that the heat flow from spreading center became smaller when the ridge approached the trench. Another unknown factor in the estimation is the effective ness of convection heat transfer in the early stage of the tectonic evolution. So far no evidence has been given to indicate that the place of initial impingement was located west of the western Transverse Ranges other than that this position is consistent with the heat flow trend in the Continental Borderland. To investigate whether heat flow along the entire continental margin of western North America is consistent with this suggestion, the heat flow data in the western United States and the eastern Pacific Ocean floor are contoured on the basis of the interpolation formula where q^ is the heat flow data at site : L compiled from literature and this study, r^^ is the distance from heat flow site i^ to grid point j_, a is a weighting radius, and q is the average heat flow at grid point j. Figure 49 " j is a contour map for a = 1° of arc and the grid interval of 0.5°. 2 2 Because the weighting factor exp(-r_^_./a ) favors the values measured near a grid point, the average value at a grid point may be biased to a high or low value if the number of data points is not large enough around the grid point. The complex contours south of 28° N are probably the result of a few stations with very high values. It is noted that the contouring along the coast between 38° N and 34° N is largely based on four data points south of 35° 40' N. For sampling distribution in the western United States the readers are referred to Sass and others (1970) and Roy and others (1972). Distinct heat flow distribution patterns are recog nized in the area north of 28° N. The isofluxes in the Continental Borderland are nearly perpendicular to the continental slope whereas the predominant trend of the isofluxes elsewhere is parallel to the continental slope. The isofluxes centered around the Sierra Nevada are also 206 Figure 49. Heat flow contour map in the western United States and eastern Pacific Ocean. — ------: Showing the base of continental slope. 207 -F f \ j 1 . 4 2.0 2.0 32*N soo KM HEAT FLOW CONTOUR INTERVAL 0 .2 pcal/cm 2 sec 20*N a > c v i 208 parallel to the continental slope. The trend of the iso fluxes (1.4 and 1.2) surrounding the heat-flow low to the west of the continental slope is at an oblique angle to the continental slope. Thus, there is a component of north increasing heat flow along the continental margin north of the Transverse Ranges. Assuming no lateral conduction heat transfer across the transform fault, the heat flow distribution pattern can be depicted and are schematically shown in Figure 48c. Although still highly speculative, McKenzie and Sclater (1968) and Oxburgh and Turcotte (1970) consider that rela tively higher heat flow (Figure 4 8b) is to be expected on the continental side of the trench due to the frictional heating from the down-going slab. Because the triple junctions migrated from the place of initial impingement, the subduction processes became inactive gradually away from that place. This suggests that the heat flow on the continental side of the trench increases away from the place of initial impingement and consequently, a heat flow minimum occurs around that place. On the other hand/ i-n the oceanic plates relatively higher heat flows are located at increasingly closer distances to the spreading ridge (Figure 48a), and the heat flows on the oceanic side of the trench decrease to the northwest because the oceanic crust is older to the northwest. Thus, the expectation is 209 in conflict with the northward increasing heat flow along the continental slope north of the Transverse Ranges. This conflict can be resolved by appealing to the effects of lateral heat transfer and frictional heating along the offshore transform fault. Since the Pacific plate moved northwestward with respect to the continental sliver and the ridge-trench- transform triple junction also was migrating southeast ward, the heat flow contrast across the transform fault became greater with time (Figure 48c). Significant lateral heat transfer should eventually occur. This is estimated in a simple model (Figure 50). Assume that the surface is maintained at temperature zero and initial temperature is f(YfZ) = q2z/k for y > 0 , continental side of the transform fault, (15) and f(y,z) = q-^z/k for y < 0 , oceanic side of the transform fault where q1 is the heat flow on the oceanic side of the trans form fault, q2 is the heat flow on continental side, and k is the thermal conductivity. The temperature distribution under these conditions is given by Carslaw and Jaeger (1959, p.277), 210 where lc is the thermal diffusivity, and t is the time. The effect of lateral heat transfer can be expressed by the ratio m.y., the heat flow on the oceanic side will increase 25 % at 1 km, 16 % at 10 km, and 5 % at 50 km from the trans- c^/cil = 1.5 (Figure 50). This suggests that owing to the lateral heat transfer the heat flow within a band of less than 1 0 0 km on the oceanic side of the transform fault may increase northwestward instead of decreasing as anticipated from Figure 48c. In addition, shear heating along the transform fault may contribute part (aq) of the anomalous heat flow. At the time when the offshore transform fault ceased being active, aq was equivalent to an anomalous temperature distribution of aqz/k (initial condition) where 0 £ z < 1^ and 1 is the depth to which frictional heating had affected (Figure 51). For time t > 0, T = 0 at z = 0, and dT/az = 0 at z = 1. The anomalous temperature as a function of t (assuming no lateral heat Conduction transfer away from the fault) is (17) where Uq = y/2 (let) and erf(-x) = -erf (x) . Within 10 form fault for J C . = 0.01 cm^/sec and an initial ratio of 2.<aq “ <-») JL k n=o 0{n (18) 211 Figure 50. Effect of lateral heat transfer,initial ratio q2/q^. Figure 51. Cooling rate of a slab whick has an initial temperature of aqz/k. 92 /9, 212 1.5 1.0 O T=3.* o u t t - 0 2 Figure 50. CH d a I S I i . o o.s o 30 20 10 ■AT - Q 50 Km ~ S t / Z Figure 51. 213 and the ratio of anomalous heat flow to the original Aq is - K.«*t "«ff ^ T T (19) where °(n = (*h+i,)7c//. 2 For 1 = 50 km, 1C = 0.01 cm /sec, 90 % of the original Aq is still retained if the movement along the offshore trans form fault ceased 5 m.y. ago (Figure 51). Thus the effects of lateral heat transfer and frictional heating may yield a high heat flow zone associated with the continental slope at the base of which the transform fault once existed. Since the lowest heat flow on the continental side of the transform fault is located off the western Transverse Ranges, it may be concluded that that was the area of initial impingement. If the position of initial impingement based on the heat flow data is acceptable, it appears that the formation of the Continental Borderland was related to the southward passage of the ridge-trench-transform triple junction and the subduction of a spreading center. Absence of basin- and-range structure off central California can be attribut ed to the presumption that no active spreading ridge has been subducted along this portion of the Pacfic margin. The southeastward migrating triple junction (ridge- trench-transform) was accelerated while the relative motion between the Pacific and North American plates was 214 shifting more and more from the offshore transform fault to the inland transform faults, especially the San Andreas fault. The ridge south of the Molokai Fracture Zone in tercepted the trench at smaller angles (2 0° compared to 37°north of the Molokai Fracture Zone). The difference in time lag between the volcanism and the passage of the tri ple junction was smaller along the southern Baja California than along southern California and northern Baja California Hence the tectonic development along southern Baja Califor nia due to the subduetion of a spreading ridge could be nearly contemporaneous, and instead of forming basin and range structure, the Gulf of California was opened. Based on three K/Ar dates on basalt from different localities in the southern Borderland, Doyle (1973) con cluded that the Continental Borderland was formed after 4 m.y.B.P., and concurrently with the opening of the Gulf of California. On the basis of the hypothesized model, the formation of the Continental Borderland began from not earlier than 17 m.y.B.P. at the northern end to 5 m.y.B.P. at the southern end. Thus, Doyle's conclusion is in con flict with the time sequence suggested here. Since the volcanoes in the Pacific Ocean may remain active for more than 10 m.y. (Menard, 1969), it is believed that the vol canism in the southern Borderland may have lasted for 5 m.y. or longer. If this belief is real, the apparent 215 conflict may be removed. The Mohnian clastic sediments atop some banks in the southern Borderland (Doyle and Bandy, 1972; Doyle, 1973) indicate that the block faulting there may be post-Mohnian but it is not necessary to be post 4 m.y.B.P. The rifting in the northern Borderland was pre-Mohnian (Yeats, 1973). According to Berggren (1971), the Mohnian ranges from 12 to 10 m.y.B.P. (early Late Miocene). Thus, the timing of tectonic development inferr- red from the model is compatible with the conclusion of Yeats and is not in conflict with the data of Doyle. SUMMARY AND CONCLUSIONS The deep crustal heat flow of the Continental Border land is being strongly masked by rapid sedimentation in part of the Borderland. Surface heat fluxes in some places are only 2/3 of the steady state values. Due to the topog raphic effect, the measured values in the basins are in general higher than the deep crustal heat flow values by 5 %. Seasonal bottom water temperature variations are insignificant but long term climatic change in the last 37,000 years may have reduced the geothermal gradient by 2 to 3 °C/km. Corrections from different sources have been treated independently of one another for lack of available methods to estimate the combined effects of sedimentation, relief, refraction, and surface temperature variation. The uncertainty for-the measured heat flow is ~ 1 0 % but for the corrected value it is -v20 %. The thermal conductivities of all core samples range between 1.5 and 2.6 mcal/cm sec °C with more than 85 % of the values falling within the range of 1.7 to 2.0 mcal/cm sec °C. Standard deviation in conductivities of a core sample is within the range of 2 to 8 % with the smaller values being found in the outer basins and larger values in the inner basins. In general the conductivity 216 217 increases around 2 to 3 % per meter from the top to the bottom of a core. Heat flow from the southern California Borderland is characterized by (1 ) the average heat flow being higher than normal, (2 ) the heat flow increasing systematically southeastward, (3) the heat flow increasing slightly land ward with a possible high heat flow zone associated with the troughs within the Patton Ridge system, and (4) the heat flow being generally uniform in a given basin except for a couple of values scattered beyond the experimental error. The average of 69 heat flow measurements is 1.86 peal/cm sec for uncorrected data, 1.78 for topographically corrected data, or 2.05 for data with topography and sedi mentation corrections. If the representative values in each basin are taken, the average is 1.74 for uncorrected data, 1.69 for topographically corrected data, or 1.99 for data with topography and sedimentation corrections. Uniformity of heat flow in an individual basin is best demonstrated by data in Santa Catalina Basin, San Nicolas Basin, and Santa Barbara Basin. Corrected heat flows are fairly uniform in northern San Clemente Basin, San Diego Trough, and Central San Pedro Basin. A few measured heat flows in San Diego Trough are probably perturbed because of their proximity to turbidity current channels. A linear heat flow trend is present from 1.8 at southern Santa Cruz 218 Basin to 1.4 at northern part of the basin after the ef fects of sedimentation and topography are removed. Uni formity in heat flow suggests that the sedimentary cover within a basin is thick enough to damp the perturbation arising from the irregularity on the surface of basement. A few values which differ significantly from the represent ative value of a given basin are attributed to slumping effect or refraction of heat flow due to conductivity con trast; several other anomalous values remain unexplained. Seismic reflection profiles confirm Moore's (1969) mapping on the major northwest-trending faults and thus indirectly confirm the structural zonation in the northern Borderland proposed by Moore: inner fault belt, central fold belt, outer fault belt, and transverse structural belt. A possible exception is the delineation of a major fault along the eastern edge of northern Tanner Basin. This fault may continue southward on the eastern edge in stead of cutting across the basin and lying to the west of southern Tanner Basin. A significant transverse structures may be present in San Nicolas Basin and another in Santa Catalina Basin. Both are compatible with the results of refraction profiling made by Shor and Raitt (1958). A third transverse structure in San Pedro Basin is also inferred from two reflection profiles. These three buried transverse structures together with others 219 yet to be discovered probably can account for the east- wast trending gravity anomalies reflected by components with intermediate wavelengths (Harrison and others, 1966). All major faults are coincident with sharp magnetic anomalies. Tanner Basin as a graben in the outer fault belt is well defined by the magnetic anomaly whereas it is poorly defined on the west by bathymetry. Moore's (1969) secondary faults may or may not be associated with magnetic anomalies. Background magnetic anomalies (components with long wavelength) along two trans-borderland profiles increase landward from the Santa Rosa - Cortes Ridge but they level off from Tanner Basin seaward until they reach the Patton Escarpment. The former trend coincides with the landward increasing heat flow. No satisfactory ex planation can be given for the coincidence. A few topog raphic highs mapped as volcanics, or basement or volcanics by Moore (1969) are not associated with magnetic anomalies. The relationship suggests that mapping based on dredging is probably inconclusive. It may also suggest that the rocks forming the topographic high have low magnetic susceptibility compared to their surroundings or that they are reversely magnetized. A trans-borderland seismic refraction profile (Shor and Raitt, 1958) indicates that the crust of the Continen tal Borderland is differentiable into 3 layers and each of 220 them might be equivalent to the oceanic layers 1, 2, and 3 based on density and P wave velocity. The equivalence/how ever, does not imply that the Continental Borderland is of oceanic crust. The most significant transition in crustal structure occurs across the continental slope from a Moho depth of 10 km to 16 km landward. Crustal thickness of the borderland increases landward gradually and reaches 32 km below the Peninsular Ranges. This crustal section has been modified by Harrison and others (1966) on the basis of Bouguer gravity. The crustal sections of Shor and Raitt and Harrison and others do not reveal crustal transition from Franciscan to granitic basements across the Newport- Inglewood fault zone, instead, a basement transition below San Clemente Island is implied. The implication is not con sistent with the geological inference which suggests that the Continental Borderland is probably floored by the Fran ciscan basement and Los Angeles Basin east of the Newport- Inglewood fault is floored by the granitic basement;however, it is consistent with Yeats' (1968) suggestion that the gra nitic rocks of the Borderland may lie in a thrust sheet above the Franciscan rocks. Pn velocity (8.2 km/sec) in the Borderland is closer to Pn (8.2 km/sec) along the coast from southern to central California than to Pn (7.8 to 8.0 km/sec) in the interior part of southern California. This may suggest a transitional upper mantle structure along the coastal area. 221 A tectonic model is proposed on the assumptions that 1) the relative motion of the Pacific and North American plates has been at a constant rate of 6 cm/year since a ridge impinged upon a trench 29 m.y. ago at a place off the present western Transverse Ranges? and 2) the relative dis placement was taken up along a hypothesized offshore trans form fault at the base of the continental slope and along inland strike slip faults, especially the San Andreas, but that the motion along the offshore transform fault was de celerated to zero 5 m.y. ago. The relative speed along the offshore transform fault is approximated by a second degree polynomial in time by bringing the Murray Fracture Zone which was initially at southern end of the Borderland to its present position in 24 m.y. (by 5 m.y.B.P.). The model suggests that the subduction of the East Pacific Rise pro ceeded from northwest to southeast. In other words, a heat source had migrated southeastward beneath the Continental Borderland, hence the heat flow in the Continental Border land increases southeastward and tectonic development also proceeded southeastward. North of the Transverse Ranges the heat flow along the continental margin probably increases northwestward. This trend is also compatible with the tec tonic model because frictional heating along the once ex isted offshore transform fault and lateral heat transfer from the continental side of the fault to the oceanic side were likely significant. REFERENCES Allen, C. R., Amand, P. St., Richter, C. F., and Norquist, J. M., 1960, Relationship between seismicity and geologic structure in the southern California: Bull. Seism.. Soc. America, v.55, p.753-798. Atwater, T., 1970, Implications of plate tectonics for the Cenozoic tectonic evolution of western North America: Geol. Soc. America Bull., v.81, p.3513-3536. Atwater, T., and Menard, H. W., 1970, Magnetic lineations in the northeast Pacific: Earth Planet. Sci. Lett., v.7, p.445-450. Bailey, T. L., and Jahns, R. H., 1954, Geology of the Transverse Range province, Southern California, in Jahns, R. H., ed., Geology of Southern California: California Division of Mines and Geology Bull. 170, p.83-106. Barnes, R. W., 1970, Marine geology and oceanography of Santa Cruz Basin off southern California: Ph.D. Dis sertation, Univ. of Southern California, Los Angeles, California, 175p. Bennett, A. S., 1972, The calibration of thermistors over the temperature range 0°-30°C: Deep-Sea Res., v.19, p.157-163. 222 Berggren, W. A., 1971, Tertiary boundaries and correla tion, in Funnel and Riedell, eds., Micropaleontology of the oceans: Cambridge Univ. Press, p.693-809. Birch, F., 1950, Flow of heat in the Front range, Colorado Geol. So. America Bull., v.61, p.567-630. Bosson, G., Gutmann, F., and Simmons, L. M., 1950, A rela tionship between resistance and temperature of therm istors: Jour. Appl. Phys., v.21, p.1267-1268. Bullard, E. C., Maxwell, A. E., and Revelle, R., 1956, Heat flow through the deep sea floor; Advan. Geophys. v.3, p.153-181. Burns, R. E., and Grim, P. J., 1967, Heat flow in the Pacific Ocean off central California: Jour. Geophys. Res., v.72, p.6239-6247. Carslaw, H. S., and Jaeger, J. C., 1959, Conduction of heat in solids: Oxford Univ. Press, 510p. Davis, G. S., and Burchfiel, B. C., 1973, Garlock fault: an intracontinental transform structure, southern California: Geol. Soc. America Bull., v.84, p.1407- 1422. Doyle, L. J., 1973, Marine geology of the Baja California Continental Borderland, Mexico: Ph.D. Dissertation, Univ. of Southern California, 193p. Doyle, L. J., and Bandy, 0. L., 1972, Southern Continental Borderland, Baja California: its tectonic and en- 224 vironmental development: Geol. Soc. America Bull., v.83, p.3785-3794. Emery, K. 0., and Bray, E. E., 1962, Radiocarbon dating of California basin sediments: Am. Asso. Petroleum Geologists Bull., v.46, p.1839-1856. Emery, K. 0., 1960, The sea off southern California, a modern habitat of petroleum: John Wiley & Sons Inc., New York, 366p. Foster, T. D., 1962, Heat flow measurements in the north east Pacific and in the Bering Sea: Jour. Geophys. Res., v.67, p.2991-2993. Gaal, R. A. P., 1966, Marine geology of the Santa Catalina Basin area, California: Ph.D. Dissertation, Univ. of Southern California, Los Angeles, Calif., 267p. Gorsline, D. S., and Barnes, P. W., 1972, Carbonate varia tions as climatic indicators in comtemporary Califor nia flysch basins: 24th International Geological Congress, Section,6 , p.270-277. Gorsline, D. S., Drake,D. E., and Barnes, P. W., 1968, Holocene sedimentation in Tanner Basin, California Continental Borderland: Geol. Soc. America Bull., v.79, p.659-674. Hamilton, W., and Myers, W. B., 1968, Cenozoic tectonic relationships between the western United States and the Pacific Basin, in Dickinson, W. R., and Grantz,A., 225 eds., Proceedings of Conference on Geologic Problems of San Andreas Fault System: Stanford Univ. Publica tions, Geol. Sci., v.ll, p.342-357. Hamilton, W., and Myers, W. B., 1966, Cenozoic tectonics of the western United States: Review of Geophysics, v.4, p.509-550. Healy, J. H., 1963, Crustal structure along the coast of California from seismic-refraction measurements: Jour. Geophys. Res., v.6 8 , p.5777-5787. Henyey, T. L., and Lee, T. C., 1973, Heat flow from Lake Tahoe: in preparation. Henyey, T. L., and Wasserburg, G. J., 1971, Heat flow near major strike-slip faults in California: Jour. Geophys. Res., v.76, p.7924-7946. Henyey, T. L., 1968, Heat flow near major strike-slip faults in central and southern California: Ph.D. Dis sertation, California Inst, of Technol., Pasadena, California, 415p. Hill, M. , 1971, Nevyport-Inglewood fault zone and Mesozoic subduction, California: Geol. Soc. America Bull., v.82, p.2957-2962. IGRE (1965), 1969, The international geomagnetic reference field 1965.0: Jour. Geomagnetism and Geoelectricity, v.21, p.569-571. Inman, D. L., and Goldberg, E. D., 1963, Petrogenesis and 226 depositional rates of sediments from experimental Mohole drilling off La Jolla, California: Am. Geophys. Union Trans., v.44, p.68-69. Jaeger, J. C., 1965, Application of the theroy of heat con duction to geothermal measurements, in Lee, W. H. K., ed., Terrestrial heat flow: Am. Geophys.,Union, Monography 8 , p.7-23. Jahns, R. H., 1954, Geology of the Peninsular Range prov ince, southern California and Baja California, in Jahns, R. H., ed., Geology of Southern California? Calif. Div. Mines and Geol. Bull. 170, p.29-52. King, P. B., 1969, Tectonic map of North America: United States Geol. Survey. Korgen, B. J., Bodvarsson, G., and Mesecar, R. S., 1971, Heat flow through the floor of Cascadia Basin: Jour. Geophys. Res., v.76, p.4758-4774. Krause, D. 0., 1965, Tectonics, bathymetry, and geomag netism of the southern Continental Borderland west of Baja California, Mexico: Geol. Soc. America Bull., v.76, p/617-650. Lachenbruch, A. H., 1968, Preliminary geothermal model of the Sierra Nevada: Jour. Geophys. Res., v.73, p.6977-6989. Lachenbruch, A. H., 1968, Rapid estimation of the topog raphic disturbance to superficial thermal gradients: 227 Rev. Geophys., v.6 , p.365-400. Lachenbruch, A. H., and Marshall, B. V., 1966, Heat flow through the Arctic Ocean floor and the Canada Basin - Alpha Rise Boundary: Jour. Geophys. Res., v.71, p.1223-1248. Lachenbruch, A. H., 1957, Thermal effects of the ocean on permafrost: Geol. Soc. America, v.6 8 , p.1515-1530. Lister, C. R. B., 1970, Heat flow west of the Juan de Fuca ridge: Jour. Geophys. Res., v.75, p.2648-2654. McKenzie, D. P., and Sclater, J. G., 1969, Heat flow in the eastern Pacific and sea floor spreading: Bull. Volcanology, v.33-1, p.101-118. McKenzie, D. Pi, and Sclater, J. G., 1968, Heat flow inside island arcs of the northwestern Pacific: Jour. Geophys. Res., v.73, p.3173. Menard, H. W., 1969, The deep ocean floor: Scientific America, no.9. Menard, H. W., 1955, Deformation of the northeastern Pacific Basin and the west coast of North America: Geol. Soc. America Bull., v.6 6 , p.1149-1198. Minch, J. A., 1967, Stratigraphy and structure of the Tijuana - Rosarito area, northwestern Baja California, Mexico: Geol. Soc. America Bull., v.78, p.1155-1178. Moore, D. G., 1969, Reflection profiling studies of the California Continental Borderland: Structure and 228 Quaternary turbidites basins: Geol. Soc. America Special Paper 107, 142p. Morgan, W. J., 1968, Rises, trenches, great faults, and crustal blocks: Jour. Geophys. Res., v.73, p.1959-1982. Morse, P. M., and Feshbach, H., 1953, Method of theoretical physics (2 vols.): McGraw Hill, New York, 1978p. Oberhettinger, F., and Magnus, W., 1949, Anwendung der elliptischen Funktonenen in Physik und Technik: Springer-Verlag, Berlin, Germany. Oxburgh, E. R., and Turcotte, D. L., 1970, Thermal struc ture of island arcs: Geol. Soc. America Bull., v.81, p.1665-1684. Palmer, H. D., 1965, Geology of Richardson Rock, northern Channel Islands, Santa Barbara County, California: Geol. Soc. America Bull., v.76, p.1197-1202. Press, F., 1956, Determination of crustal structure from phase velocity of Rayleigh waves, Part I, Southern California: Geol. Soc. America Bull., v. 67, p.1647- 1658. Prensky, S. E., 1973, Carbonate stratigraphy and related events, California Continental Borderland: M.S. Thesis University of Southern California, Los Angeles, Calif., 208p. Prodehl, C., 1970, Seismic refraction study of crustal structure in the western United States: Geol. Soc. 229 America Bull., v.81, p.2629-2646. Ratcliffe, E. H., 1960, The thermal conductivities of Ocean sediments: Jour. Geophys. Res., v.65, p.1535-1541. Revelle, R., and Shepard, F. P., 1939, Current measurements off the California coast in 1938 and 1939: Records of Observations, Scripps Inst. Of Oceanography, San Diego, Calif., v.l, p.83-88. Roller, J. C., and Healy, J. H., 1963, Seismic refraction measurements of crustal structure between Santa Monica Bay and Lake Mead: Jour. Geophys. Res., v.6 8 , p.5837-5849. Roy, R. F., Blackwell, D. D., and Decker, E. R., 1972, Continental heat flow, in Robertson, E. C., Hays, J. F., and Knopoff, L., eds., Nature of the earth: McGraw Hill Inc., New York, 677p. Roy, R. F., Blackwell, D. D., and Birch, F., 1968, Heat generation of plutonic rocks and continental heat- flow provinces: Earth Planet. Sci. Lett., v.5, p.l- 12. Rusnak, G. A., and Fisher, R. L., 1964, Structural history and evolution of Gulf of California, in Van Andel, Tj. H., and Shor, G. G., Jr., eds., Marine geology of the Gulf of California: Am. Assoc. Petroleum Geologists Mem., no.3, p.144-156. 230 Sass, J. H., Lachenbruch, A. H., Munroe, R. J., Green, G. W. , and Moses, T. H., Jr., 1971, Heat flow in the western United States: Jour. Geophys. Res., v.76, p.6376-6413. Scholtz, C. H., Barazangi, M., and Sbar, M. L., 1971, Late Cenozoic evolution of the Great Basin, Western United States, as an ensalic interarc basin: Geol. Soc. America Bull., v.82, p.2979-2990. Sclater, J. G., 1972, Heat flow and elevation of the marginal basins of the western Pacific: Jour. Geophys. Res., v.77, p.5705-5719. Sclater, J. G., 1972, New perspectives in terrestrial heat flow: Tectonophysics, v.13, p.257-291. Sclater, J. G., and Francheteau, J., 1970, The implications of terrestrial heat flow observations on current tectonic and geochemical models of the crust and the upper mantle of the earth: Geophys. Jour., v.20, p.509-542. Shepard, F. E., and Emery, K. 0., 1941, Submarine topog raphy off the California coast: canyons and tectonic interpretation: Geol. Soc. America Special Paper 31, 171p. Shor, G. G., and Raitt, R. W., 1958, Seismic studies in the southern California Borderland: Congreso Geo- logico Internac., XXa Session 1956, Mexico IX, Segundo Tomo, P.243-259. 231 Smythe, W. R. , 1950, Static and dynamic electricity, 2nd ed., McGraw-Hill, New York. Steinhart, J. S., and Hart, S. R., 1968, Calibration curves for thermistors: Deep-Sea Res., v.15, p.497-503. Suppe, J., 1971, Offset of late Mesozoic basement terrains by the San Andreas fault system: Geol. Soc. America Bull., v.81, p.3253-3258. Sverdrup, H. U., Johnson, M. W., and Fleming, R. H., 1942, The ocean: Prentice-Hall, Englewood Cliffs, New Jersey, 1087p. Swanberg, C. A., and Blackwell, D. D., 1973, Areal dis tribution and geophysical significance of heat generation in the Idaho batholith and adjacent intrusions in eastern Oregon and western Montana: Geol. Soc. America Bull., v.84, p.1261-1282. Swanberg, C. A., 1972, Vertical distribution of heat generation in the Idaho Batholith: Jour. Geophys. Res., v.77, p.2508-2513. Vacquier, V., Sclater, J. G., and Correy, C. E., 1967, Studies of the thermal state of the earth: heat- flow, eastern Pacific: Bull. Earthquake Res. Inst., v.45, p.375-393. Von Herzen, R. P., 1964, Ocean floor heat-flow measure ments west of the United States and Baja California: 232 Marine Geol., v.l, p.225-239. Von Herzen, R. P., and Uyeda, S., 1963, Heat flow through the Eastern Pacific Ocean floor: Jour. Geophys. Res., v.6 8 , p.4219-4250. Von Herzen, R. P., and Maxwell, A. E., 1959, The measure ment of thermal conductivity of deep sea sediments by a needle probe method: Jour. Geophys. Res., v.64, p.1557-1565. Wilson, J. T., 1965, A new class of faults and their bearing on continental drift: Nature v.207, p.343. Wollenberg, H. A., and Smith, A. R., 1968, Radiogenic studies in the central part of the Sierra Nevada batholith, California: Jour. Geophys. Res., v.l3, p.1481-1498. Woodford, A. O., 1925, The San Onofre Breccia, its nature and origin: Calif. Univ. Pubs. Geol. Sci., Berkeley, Calif., v.15, p.159-280. Yeats, R. S., 1968, Rifting and rafting in the southern California Borderland: Stanford Univ. Pubs. Gedl. Sci., v.ll, p.144-165. Yeats, R. S., 1973, Poway fan and submarine cone and rifting of the inner southern California Borderland: Abstracts with Programs, Geol. Soc. America, Cordilleran Section, p.125. Yerkes, R. F., McColloh, T. H., Schoellhamer, J. E., 233 and Vedder, J. G., 1965, Geology of the Los Angeles basin - an introduction: U.S. Geol. Survey Prof. Paper 420A, 57p. Zienkiewicz, 0. C., 1971, The finite element method in engineering science: McGraw Hill, New York, 521p. APPENDICES 234 APPENDIX I Calibration of Thermistor Probes Thermistor probes were calibrated over the temperature o o o o range from 0 C to 30 C at an interval of ~ 3 or 5 C. Temperature calibrations were performed with U.S. Bureau of Standards certified guartz thermometer (Hewlett Packard model 2801A) with a resolution of 0.0001°C, an absolute accuracy of 0.02°C and a precision on the order of a thousandth of a degree C. 0°C calibrations were made with a distilled water ice bath in a large stainless dewar with the entire probes submerged. Other points were measured in a Tamso thermostatic bath (model TEV 70, Neslab Instrument Co.) with temperature consistancy of + 0.002°C and uniformity of 0.002°C. Readings were re corded to the nearest 0.001°C. All thermistors used in this study were isocurve elements with + 2 % deviation from 2 K-fh in the temperature range from -5°C to 35°C (Fenwal part K2049). Calibration was made after the thermistor was installed at the tip of a stainless steel tube which was then impregnated with epoxy (beads were pressure relieved prior to impregnation). During initial calibration, a copper block was used to host a batch of thermistor probes before the probes were 235 236 used in the fi^ld. Recalibration in the laboratory was done without copper block after the field use. Relative calibrations were made in the field by sustaining the corer above the sea bottom until the thermistors reached equilibrium (equilibration took two minutes). The resis tances of the thermistor probes were measured with a portable precision Wheatstone Bridge. Detailed instrument ation has been described by Henyey (1968) but in this study the lead compensation was neglected because the lead wires were not more than 7 m. The dependence of resisitivity p on temperature in a semiconductor is usually represented by p(T) e*p (T)/2 Kt ] where A E(T) is the energy gap between the valence and conduction band energies, K is Boltzman's constant, and F(T) is a function of Kelvin temperature T. Fitting of calibration data to such a relation remains a difficult task because the thermistors used in precision temperature measurements do not have simple, sharply defined energy bands and lack of simple explicit expressions for F(T). Over a small temperature range F (T) and A E(T) can be treated as constants and a curve of the form U R = A 1 b/t -t c / r z (2) has been used by Henyey (1968) to fit calibration data, 237 where R is the resistance, A and B are parameters to be fitted and C term represents non-linearity correction. The resulting residuals are relatively large as compared to the function _L = e x -t & J U l R + < $ ’£-£*70 T 1 (3) proposed by Steinhart and Hart (196 8 ) based on criteria pertaining to goodness of fit. Formula (3) is suitable for many kinds of thermistors over a temperature range of several degrees to a few hundred degrees and was used in this study. In application, the temperature vs. resistance conversion table is pre pared. Linear least squares fitting procedures were used to find the coefficients oi, ji, and Table 2 shows the observed and fitted temperatures for a few thermistor probes along with the values of the coefficients. In addition to the criteris suggested by Steinhart and Hart (1968) for the choice of an interpolation func tion, Bennett (1972) stated that the form of an interpola tion function should not be affected by a change of units. Equation (3) violates this criterion by its omission of the second order term in ln(R). Bennett (1972) therefore suggests the function to preserve independence of the system of units used, 238 where RQ is another parameter to be fitted (resistance at temperature T0) and P is the order of fit. Equation (4) is equivalent to J - = o( -t Jtw R •+ T ( 4 v l 0 % T (5) with the order P = 3. Calibration data of our thermistor probes were fitted to equations (3) and (5) by the linear least squares method. The results are presented in Table 2. For comparision with previous work Table 2 also lists the results of published data (BGS from Bosson and others, 1950; T3 from Steinhart and Hart, 196 8 ). The relative standard error is defined by t = (Z (C Tobs - T.,ty / r « ] ) l/z ( 6 ) (Steinhart and Hart, 1968). For 12 out 15 thermistors, the relative standard error for equation (5) is smaller than that for equation (3). Thus the statistics supports 2 Bennett's observation that the inclusion of the (InR) term improves the fitting, contrary to Steinhart and Hart's statement that the second term degrades the fit. To a large extent this is a function of the quality of calibration and operation of the primary standard as well as computing machine rounding errors. However, an examination of the difference in the estimated and ob served temperature reveals that both equations provide 239 the precision required in heat flow measurements with our present technique. The fact that we are only interested in the temperature differences further supports the use of either function. To investigate how well the interpolation functions (3) and (5) predict the temperature at specified resist ances , some data points from the calibration data of each thermistor were deleted in the least squares fit. A good interpolation function whether it is derived from a few or many data points of high quality should be able to predict the temperature at a given value of resistance over the calibration range. The number and position of the deleted data points were arbitrarily selected, but at least 5 data points for each thermistor were retained. The predicted values of T vs_ R determined from the partial set of data yield residuals comparable to the maximum residual obtained with the complete data set. This justifies use of a few data points to establish the interpolation functions (3) and (5) Extrapolation based on equations (3) and (5) is in general not permissible. If the extrapolation is not made beyond one calibration interval above or below the calibra tion range, the difference between the extrapolated and observed values was found to be less than 0.01°C. 240 The similarity of the R-T slopes of all thermistor probes was also checked since this was crucial in our method of underwater calibration. From equation (3) , we have (7) At a given temperature, the variability in resistance among thermistors did not exceed 0.4 % in a temperature range o o 0 to 5 C. The variation in the slope of T-R curve for a thermistor probe with respect to the slope for a reference thermistor is approximately cF r ^ ^ ^ / 3t3*<5YAtO.) (8) The error in reading the temperature recorded by a thermis tor probe from a T-R table prepared for the reference thermistor probe is therefore / 5 + 3£ U U R ? (9) The largest error is 0.4 % for the temperature range from 0° to 5°C. This justifies the use of a calibrated T-R table for all the thermistors used in this study. Re- claibrations were periodically made to check for thermistor malfunction. New tables prepared after each recalibration. Justification for preparing a T-R table in the 0° to 5°C range used in this study are explicitly summarized. Since calibration formula (3) is good for a thermistor over a wide temperature range from -6 8° to 134°C (Steinhart and Hart, 1968, see also Table 2 BGS data), there is no reason to believe that abnormal response may occur to a thermistor in the 0° to 5°C range. The table was prepared by interpolation and extrapolation based not only on 0°, 5°, and/or 3°C but also on higher than 5°C calibrations. As previously noted, interpolation based on partial set of data is as good as that based on full set of data, and extrapolation within one calibration interval may be made for pratical purpose. If the quality of calibration points are well controlled, the prepared T-R conversion in the 0° to 5°C range is as good as those at the calibra tion points. The 0°C calibration is most crucial in the quality control of the table since it serves to check "extrapolation" and interpolation. 242 Table 2. Calibration data of a few selected thermistors PROBE 1 to 13: PROBE A : BGS : R OBS : T OBS : T EST : (3rd column) T EST : (5th column) T DIFF = T EST selected thermistors used in this study A thermistor calibrated in USC Geophysical Laboratory but not used in this study data from Bosson and others (1950) measured resistance calibration temperature estimated temperature from calibration formula (5) estimated temperature from calibration formula (3) • T OBS PROBE 2 PROBE R OBS T CBS T EST T DIFF ---56 98 < > - 8 - 0--- —0.003 ----- — 0-.G0 3 - -0.0002 4 9 07.00 3 . 34 0 3.341 0.0007 4300.60 6.347 6.347 0.0 4013.10 7.918 7.917 -0.0007 3 6 04.90 1 C-. 45 6 1 0 .456 0.0005 — 3110.80---.1 . 3.968 - ...13.96,9 0.0012 2761.00 16.866 16.866 0.0002 2566.80 16.665 IS.66 2 -0.0024 2327.40 21.104 2 1.10 4 0.0005 --2043.20 .- 2 4.407 .— .2-4.40 7 --- 0.0005 1914.30 26.08 3 26 . 08.4 0.0017 1707.70 2 9.063 29 . 064 0.0007 1548.00 3 1.67 0 .31 .667 -0.0034 --14 90.20----— 32 .-6-84----- - 32. 686 0.0022- ALPHA 1 .2598930-03 BETA 2.7323070-04 GAMMA -1.1669200-06 - - - - - DEL T-A ----1.9385500-07 RELATIVE ERROR 4.84E-06 R OBS T GPS T EST T O F F 5696.60 0.008 4 3 .3 B .A 0 .....— - 6 . 1 6 2- 36 69.0 0 10.0c3 28A4.00 16.183 ‘23 78.4 0 2 0.610 2-0-66-^40-___ 2 4.175 1590.30 31.021 0.008 0.0002 --6. -161 -0.0010.... 10.065 0.0022 16.18 1 -0.0022 2 0.610 0.0 -2-4 .1 76 - - 0 .00-15.... 31.021 -0.0005 ALPHA 1 .3222650—03 BETA 2.4870760—04 GA M M-A---- 1 .9402850-06 DELTA 6.5964520-08 RELATIVE ERROR 4.755-06 T EST T DIFF -0.003 0.0002 3.341 0.0007 6.347 -0.0002 7.91 7 ---0.0010 1 0.456 0.0002 13.969 0.0010 16.866 0.0002 18.663 ---0.0022 21.104 0.0007 24.4*0 8 '0.0007 26.085 0.0020 29.064 0.0007 31 .666 -0.0037 32.686 0.0017 ALPHA 1.284448D—03 BET A 2.639514D—04 D E L T A 1.4502050-07 RELATIVE ERROR 4.83E-06 T EST T DIFF 0.008 -0.0002 ----- 6a l61------0.0005 -------- 10.066 0.0027 16.181 -0.0022 20.610 -0.0002 - 24 . 176------0.0010 31.021 0.0 ALPHA 1•281002D—03 BETA 2.642182D—04 DELTA 1.4673800-07 RELATIVE ERROR 4.82E-06 to .... - -- • - ... - ----- ' CJ PROBE 4 PROBE m P . 0 3 s ....-... T CBS__ ____T- - E-ST T DIFF 5 6 92.50 -0 .003 -0 . 00 A -0.0007 4 9 03.40 3 . 3 3 9 3 .34 3 0.0017 --4 299.80 - —.6.3.4 7----..-.,6.-34 6 ...-0.0012 4 016.60 7.92 1 7.°2 1 0.0005 3604.20 3 0.A 62 3 0 . 4f. 3 0.0015 3112.20 13.972 ! 3 . G7 0 -0.0020 ---- 2 763.50. .16.664.. 1 6.86 4- -0.0002 2 5 69.0 0 1 8 . 6 6 6 18.6 66 -0.0002 2329.90 21.10 7 21.10 8 0.0007 2046.00 24.410 24.411 0.0007 . _ 1917.40 - 2 6. O H 4------.—26. 085 -- ---0.0 0 07 1710.50 29.069 2 9.06 0 0 .0 , 1550.50 31 .677 31 .67 7 -0.0002 ALPHA 1 .3295450- : ________________-------- . — - - r-.FTA.... 2.4 66 8.670— GAMMA 2. 101860 D—1 DELTA 6.3623480- RFLATTVF EFROR 3.52 R OBS o 694.00 43 48.80 3 636 * 90 2937.90 23 34.80 1955.40 16 12.80 T OPS - 0 . 0 0 1 6 .080 !0.238 15.344 8 0.488, 25.530 3 0.5 c5 T EST - 0.001 6 . 079 1 0.239 15.345 -2 0 . 48 8- 2 5.53 0 T DIFF 0.0005 - 0.0010 0.0010 0.0007 - 0.0002 - 0.0002 30.5 6 5 0.0002 ALPHA 1 .3279400-03 Be: T A - 2 . 4896-030 — 04- . GAMMA 1.6937900-06 DELTA 3.211935D—08 RELATIVE e r r o r 2.26E-06 T EST T D I F F ---- -0.004 -0.0012 3. 341 0.0020 6.346--- -0.0007--- 7.922 0.0010 1 0. 464 0.0017 13.970 -0.0020 16.864 -0.0002 18.665 -0.0005 21.107 0.0002 24.410 0.0002 26.034--- 0.0002--- 29.069 0.0 31.677 0.0005 ALPHA 1.284975D-03 BETA 2•634637D—04 DELTA 1.512665D—07 RELATIVE ERROR 3.72E-06 T EST T DIFF - 0.001 0.0002 6.079 -0.0007 10.239 0.0012 15.344 0.0005 20.487 -0.0007 25.529 -0.0005 30.566 0.0007 ALPHA 1.2916720—03 BETA 2.625673D—04 DELTA 1.527052D—07 RELATIVE ERROR 2.50E-06 to PROBE R OBS T CBS T E?T T DIFF LO 5690.90 -0.003 -0.002 0.0007 __ __ 4.9-02-^5-Q— — 3.34-0- __3-.33SL____-0.0012 4297.90 6 . 34 6 6 .34 5 —0.0 005 4013.60 7. 931 7.931 0.0002 3602.10 10 .467 10.469 0 . 0020 ---34-44-. 7-0 1 3-»-9-64- -----1-3.965 - 0.0010-- 2762.30 16.865 16.865 0 . 0 2568.20 18.666 18.664 -0.0020 2329.60 21.102 21.102 -0.0002 1916.80 26 .08 3 26.084 0.0015 pjn.no 2 0, nt- o .... . ? 9 . 06 9 ... 0.0002__ . . . . 1 493. 10 32.683 3 2.683 -0.0002 ALPHA 1.311381D-03 - .....- -... RET A - 2•52 72 67D—94 gamma 1 .4501 SAD-06 DELTA 8.640295D—08 RELATIVE ERROR 3.65E-I T EST T DIFF --- -0.003 0.0005 3.339 -0.0012 6# 3 4 6-----’ O.-0 0 0 2---- 7.931 0.0005 10.469 0.0022 13.965 0.0010 16. 865 — 0.0002------ 18.664 -0.0022 21.101 -0.0005 26.084 0.0012 29. 069------0.0 32.683 0.0002 ALPHA 1.280752D—03 BETA 2.6427960-04 DELTA 1.469797D—07 RELATIVE ERROR 3.82E-06 245 PROBE --s_.Q & S . T CDS . . . 7 r . ST. - - T.DIFF-----._ T EST T DIFF 5696.50 -0.003 -0 .002 0.0007 --0.003 ----- 0.0002 4905 .4-0 3.34 0 3.339 -0.0010 3.339 -0.0007 --AgOQ.1-0 - -6. 34-6 -------6 »3A6 -0.0002----i 6.346 0.0005 4016.10 7.520 7.920 -0 . 0002 7.920 0.0005 3-603.2 0 ..10.458 1-0.45 9 0.0010 1 0.459 0 . 0 0 1 7---- 3108.50 13.97 8 13.979 0.0012 13.979 0.0015 - --- 2-7-60.4-&- 3 6.86-6 ---- .1.^865. ---0.0010------- 16.865 0.0007 2565.80 18.668 18.666 -0 .0022 18.666 -0.0024 2326.90 21.103 2 1.104 0.00 12 21. 103 0.0005-------- 2042.80 24.408 24.407 -0 .0007 24.406 -0.0015 ---- 1913.50 --26.085---- 2 6.065 — --0.000 2 26. 084 -0.0010 1707.00 2 9.06 9 29.069 0.0 29.069 -0.0002 1546.50 3 1.678 31.679 0.0010 31.679 0.0015 1490.00 3 2.68 3 32.682 -0.0005 32.684 0.0007 ALPHA---- _1 .337QO40—03.. ALPHA • 1 — V' 1 . • V V X • C O *r 7 •? w \J U O BETA 2 .3627750-04 BETA 2.6384990-04 gam ma 3.467559D—0 6 DELTA T . 458046D — 07- DELTA 6.83 2 54 OD—10 REL ATIVE ERROR 3.96E-06 ........ .... ..... . . ........... . . . - ABLATIVE , E T .R90« 3.34E-06 246 PR<[)BE 8 j I PROBE R OBS T Op S T P ST t DIFF 5695.00 -0.003 -0.00 3 0 . 0 4 9 06.40 3.336 3 . 33 6 0.0 0 02 4301.30 6.34a 6.34 3 -0.0007 --4013.40 -----3.-9 1-9-. .. -.7.92 0 0.0007 .. 3 6 04.^0 1C.47] 10.471 0 . 0 3114.20 13.962 13.96 3 0.0007 ' 2764.00 16.868 1 6 . 86 7 -0 .0005 - - .2570.30 -. 16.66 3___ ..... 16.661 .-. -0.0015 23 30.80 21.10 6 21.10 6 0.0005 2047.00 24.406 24.407 0.0012 1918.10 26.084 26.084 0.0005 ---1711 .90 ---29.053--- -..-2 9.05 7 -0.0010 ... 1551.4 0 3 1 .670 31.671 0.0010 14 94.00 32.683 32.68 2 -0.0005 ALPHA. 1.3424050-03 ___________ . ____ __ .BETA___ --2 . 41 38 63D-04 gamma 2.82 0 936D— 06 DFLTA 3. 124 6 2 3 D— 0 8 RELATIVE ERROR 2.60 E— P OBS T OBS T EST T DIFR 6304. 30------0.004ft... ...-0.0 046 0 .0___ 5716.80 2.1855 2.1858 0.00024 5394.00 3.5002 3.5000 -0.0002* 5325.50 3.79 03 3.7903 O.C 520*7.40------3-.-P-968--------3.996ft — . - 0 . 0... ... 5215.10 4.2673 4.2673 0.0 5168.60 4.4714 4.4717 0.00024 5046.70 5.0173 5.0176 0.0002* — 466 3.8 0----- 6. -83 -57 ......-6.635 7 0.0_______ 4616.00 7.0745 7.0747 0.00024 ALPHA —4.094546 D— 0 4 BETA 8.453185D—04 ---------------- GAMMA - . —-6 • 776 8370-05 .. DELTA 2.7803720-06 RELATIVE ERROR 6.219E-07 T EST T DIFF -0.004 -0.0010 3.336 0.0005 6.344 0.0 7.920 0.0012 10.472 0.0007 13.963 0.0010 1 6. 867 -0.0007 18.661 -0.0020 21.106 0.0 24.406 0.0005 26.084 -0.0002 29.057 -0.0012 31.671 0.0015 32.683 0.0005 ALPHA 1•28 30 3OD—03 BETA- 2.6382100-04 DELTA 1.492859D-07 RELATIVE ERROR 3.27E-06 fo 9J - T fc.ST------T D I F F ---------------- — ............... -0.0046 0.0 2.1 855 0.0 3 .5 0 0-0-----0 .00024---------- 3.7 90 3 0.0 3.9968 0.0 4.2 673 0.0 4.4719----- -0.0 0 04 9------------- 5.0178 0.00049 6.8357 0.0 7.0745 0.0 -ALPHA 1 . 2 562 28D— 03-------------- BFTA 2.633 615D— 04 DELTA 1.5010110-07 RELATIVE ERROR 8.333F-07 BGS R OBS T DBS T EST T DIFF T EST T DIFF 10558700.00 -6 8.65 00 -68.4750 -0.02498 -68 .4 722 -0 . 02222 7368300.00 —63.9900 -63.9168 0 . 07323 -63 .9154 0 . 07458 4130100.00 -56.2100 -56.2428 -0.03276 -56.2433 -0.03326 24654 00.00 -49.0000.-__-49.022 8 ...-0.022 84 ..- -49.0245 ...-0. 08452 .. 1398800.00 -40.6200 -40.6413 -0.02126 -40.6437 -0 . 02 3 73 864970.00 -33.1100 -33.1297 -0.01965 -33.132 4 -0.02235 540510.00 -25.42 00 -25 .3924 0.02759 -25.3950 0 .02499 - . ^ 150.40-. 00... -15.98 02______-15 .2912 . . . . -0.01 099 - --r 1-5.993 4..----0.01318--- 190430.00 -6 .6802 -6.6772 0 . 00293 -6 .6 785 0.00171 115210.00 3.1799 3.1 99 7 0 . 01978 3.1997 0.01978 45383.00 23.1899 23.2480 0.05811 23.2505 0.06055 71585^.00 ___ 13.1099___ ----1 3.. 1 396 0.02979 13.1 409- -----0.0310 1--- 27163.00 35.3799 3 5.3943 0 .01489 35.3984 0.01855 17104.00 47.0798 47.1069 C.02710 47.1116 0 . 03174 11289.00 58.4099 5 8.3137 -0 .09619 58.3138 -0.09106 .....72 97.4 0 .___ 70.9099...____70 • £:£6 0 --0.05396 __-...— 70 .8603 - .. -0.04907-- 4761 .10 84.0000 83.9812 -0 . 01 880 83.9849 -0.01514 3282.50 96.2000 96.1687 -0 .03125 96.1704 -0 .02954 2263.80 1 09. 1299 109.1245 -0.00537 109.1226 -0.00732 - 1615.70 --121.5798 121.625 7 0.04590 121.6 194 . . . 0.03955 1177.60 134.0098 134.0525 0 .04272 134.0403 0. 03052 ALPHA 7. 887298D—0 4 ; ALPHA 7.92 0 0 81D—04 BETA 2. 319852D-04 BETA 2. 31076 ID-04 GAMMA__ -9. 081774D-08 DELTA____ a. 4.96 1 02 n— 08 DELTA 8. 657304D—08 RELATIVE ERROR 1.377E-04 RELATIVE ERROR 1.393E-04 248 PROBE 13 | | PROBE ' R OBS T DBS T EST T DIFF —2-56-8 246 0.1 0 2222. 80 2046.30 1773.30 -14 48.90 1 1 37. 1 0 1 031 . 90 -2.40 f t . 3 .468 5.781 7.682 10.976 1 f t . SB 7 2 0 . 8 1 0 24.356 - - - - 2 .-496 3.463 5. 77 8 7.666 10.973 1 5.38 8 2 0.81 0 24.35 6 ALPHA P.7TA - - GAMMA DELTA RFLATIVE 0.0002 0 *0005 -0.0024 0.0042 -0.0024 0 . 0 0 1 2.. - 0.0002 0.0 1.4206640-03 2.8476300-04 2.813250D-06 2.991226D—07 ERROR 6.995-06 .R—O-B-S________ T- COS_________ -T-E-ST T D I F c .. 6 3 07.2 o 6.6 0 99 5937.50 8.0086 53-2 7*40-----10. .5513 - - 4 9 2 5.40 12.4153 4471.10 14.7439 4025.30 17.3098 3665.90 19.6289 33-75.10-----21.70-29___ 2974.40 24.cp58 - 2 763.8 0 - 26.8259 2526.30 29.1819 6.6 099 0.0 8.0093 0.0004° -10.5515 -0.00024 . 12.4158 0.0 14.7449 0.00098 17.3103 0.0004° 19.6279 -0.00098 2 1 .7 029 . 0.0....... 24.9258 0.0 2 6.8 269 0.00098 2Q • 1816 -0.00024 ALPHA ..1.2240980-03 - BETA 2.4585980-04 GAMMA 2.222141C-06 DELTA 4.3324530-08 RELATIVE ERROR 1 . 91 8F-06 T EST T DIFF 2.495 0.0002 3.469------0.0007 5.778 -0.0027 7.686 0.0039 10.973 -0.0024 15.888 0.0015 20.810 0.0 24.356 -0.0002 ALPHA 1.471947D—03 BETA 2.6394840-04--- DELTA 1.725033D—07 RELATIVE ERROR 7.02E-06 T EST T DIFF 6.6099 0.0 8.0093 0.00049 10.5515 -0.00024 12 .4-1-6 0 ....0 .00024 14.7449 0 .00098 17.3103 0.00049 19.6279 -0 .00098 21 .7029 0.0 24.9250 ____0.0------ 26.8267 0.00073 29.1819 0.0 ALPHA 1 .173 23 5 D— 0 3 -FETA-----2 . 642 80 60-04---------------- DELTA 1.326125D—07 RELATIVE ERROR 1.807E-06 £ V O PRQBE 10 PROBE R OBS T DBS T EST T DT-F 5700.20 -0.003 -0 . 0C3 0 . 0 -----4910.40 ---- 3.33 7 - —... --3.33-8 - 0.0007 4304.50 6.347 6.3-4 6 -0.0012 4021.20 7.922 7.92 3 0 .0007 3 6 08.40 10.463 10.463 0.0002 -----3 116.1Q --- 13.54 5 ....13. 96 6 . - . - -0 ..-0-0-1 o~ 2766.60 16.R64 16.66 2 -0.0015 2 0 48. 10 24.410 2 4.410 0 . 0 1919.30 2 6.084 26.084 0.0005 ----1-7 12.50 ---25.06-3 ------?9.06 3 - 0.0002 1552.20 31.67 3 3 1 .67 2 -0.0005 1494.'8 0 32.683 32 .68 3 " "" 0 .0 0 02 ALPHA 1 . 34 2 0 0 8 D— BETA 2.4165 080— ------------- — ----------------------------------------- -...-GAMMA.. 8.7 734 90D— DELTA 3•33922OD REL.ATI VF ER R 0 R 2.76! R ORS T OBS T EST ■IF -56-85.3-0- 4344. 1 o 3634.80 2937.00 -2 3S5.-1 0- 1956.20 1614.10 -0 .- 0- 0 2~ 6.080 10.237 15.34 6 -2 0 . 487 25.53 1 3 0.565 — — 0 . 0-0-2- 6.080 10.237 1 5.346 -2-0.-487 2 5.531 30.565 . 0.0 .......... 0. 0005 - 0.0002 - 0.0002 0.0 005- 0.0002 0.0 ALPHA- 1.313242D-03 BETA 2 .5399790-04 .GAMMA----1 . 064 7 000-06 ■ DELTA 1.1154900-07 RELATIVE -PROP 1.07E-06 t e s t t oi f f -0.004 -0.0007 3.333 0.0007 6.346 -0.0007 7.923 0.0012 10.463 0.0005 13.966 0.0015 16.862--- -0.0017 24.409 -0.0007 26.084 -0.0002 29.063 0.0 31.67 3 0.0 32.684 0.0010 ALPHA 1•283352D—03 BETA 2 « 63761 5D —04- DELTA 1« 491617D-07 RELATIVE ERROR 3.17E-06 T EST T DIFF - 0.002 0.0 , 6.081 0.0007 10.237 0.0 15.346------0.0002 20.487 0.0002 25.531 0.0 30.565 0.0002 ALPHA 1 . 29051 3D-03-------- BETA 2.625252D—04 DELTA 1•557897D—07 RELATIVE ERROR 1.13E-06 to Ui o PRQBE 12 PROBE 11 R OBS T CBS T EST T DIFF 5705.60 0.000 0.009 0.0 4 326.2 0 6.251 6.251 0.0 3672.30 10.062 10.063 0.0015 2-846.^-0---- 1 6. 1 f>3 16.181 -0.0 022 2378.80 20.610 20.610 0.0 2066.30 24.175 24.177 0.0022 1580.80 31.021 31.020 -0.0007 A-LP-H-A - 1 .-31 73 2 80-0 3 BETA 2 • 504844D—0 4 GAMMA 1 • 7766-4 2D—06 DELTA 6•8 0233 8D— 08 EFLAllVE-FRRHS 4.555-06 . —R 6-RS - . . -T-OBE ---______T f - ST.. --T OI f f • 5694.60 -0.003 -0.003 —0.0002 49 05.20 3.339 3 .340 0.0010 _ 4.3 00.9 0 - ___e .-3-4 3- ______ 6.34 3.-.- 0.0002 4016.60 7.^27 7. 927 0.0002 3 6 04.60 10.466 10.4-6 5 -0.0005 ■ ••3112.90 13.967 13.968 0.0010 ----- 2 7 63.50 --- 1 ( - ■ » 866 ----- 1 6 • 866* - ...0.0.. - 2569.60 1 8.663 I 8.661 -0.0015 23 30.50 2-1.102 21.102 0.0005 2046.20 24.409 2 4.40 9 0.0 -1-9 1 7-.-5-0---- 2-6-.-0-S-3 — — .. 2 6.064 0.001 2 1711.20 29.059 29.059 0 . 0 1551.00 3 1 . 670 31.66 9 —0.0005 1493.50 32.683 32.683 0.0002 _____________________— .- _---ALPHA 1 .289693D-1 BETA 2.61251 2D—i GAMMA 3 . 403994D-I DELTA 1 . 3 40 02 9D-I ..»f L-ATJ VE -T.RRnR-2.-35l T EST T DIFF 0.009 -0.0002 6.251 0.0005 10.064 0.0020 16.180 -0.0024 20.609 -0.0005 24.177 0.0017 31.021 -0.0002 ALPHA 1.279536D—03 ----- BETA 2•646886D— 04 DELTA 1.419747D—07 RELATIVE ERROR 4.77E-06 T EST T DIFF -0.004 3.340 6.343 7.927 10.465 13.968 16.866 18.661 2 1. 102 24.409 26.084 29.059 31 .669 32.683 ALPHA BETA DELTA RELATIVE -0.0005--- 0.0010 0.0002 0.0005 -0.0005-------- 0.0010 0.0 -0.0015 0.0005 0.0 0.0012 0 . 0 -0.0005 0.0002 1 .282528D-03 2.639583D—04 1.482470D—07 ERROR 2.42E-06 fo ( ji 252 APPENDIX II Standard Error of Heat Flow Measurement The standard error S^ of a measured heat flow value q is defined by ,/ sF = 9 [ ( - IK + (-f-)2] where G is the weighted gradient obtained by dividing q by the average conductivity, SG is the standard error of gradient, R is the average thermal resistivity, and SR is the standard error in resistivity given c, = f 2 ( R i - R ? 1 R L 7? (7) - / 3 J V z- ~T~ n being the number of measurements. The 95 % confidence limit of a measured heat flow is calculated by f Cr where t-^ and t2 are Student's multipliers for n-1 and n-2 degrees of freedom, respectively. In the calculation, is assumed to be 1 and Sg/G assumed to be 0.085, the maximum possible error discussed in the text. APPENDIX III Table 3. Heat flow data in the southern California Borderland A . Summary: Basin No 1 2 3 * ** * ** * ** San Diego 5 2 . 2 2 2 . 1 0 2.15 2.04 3.28 3.25 San Pedro 5 2.04 1.75 1.98 1.70 2.26 1. 99 San Clemente 6 '2.28 2 . 0 0 2.13 1. 83 2.25 1.93 Santa Catalina 8 1.89 1.76 1 . 8 6 1.73 2.06 1.92 Santa Cruz 7 1.57 1.60 1.44 1.60 1.55 1.56 San Nicolas 7 1.64 1.67 1.61 1.62 1.69 1.74 Tanner 6 1.75 1.65 1.75 Santa Barbara 2 1.16 1.16 1.78 East Cortes 2 1.91 1.84 2 . 0 2 No: Number of measurements. 1 : Measured values. 2 : Values corrected for topography. 3 : Values corrected for topography and sedimentation, * : All values. **: Best values. B. Data: 253 San Diego Trough Station No. Dl 15796 32°241 117°30' Location Depth Temp. Grad. k qi q Aq <3t qts m cm °C C/km — — 1190 0 3.378 1.63 ±0.19 1.56 2.36 5.25 120 3.528 ±0.25 2 2 0 3.621 93 320 3.711 91 420 3.782 73 D2 14514 32°36' 117°331 D3 14515 32044' 117°361 1145 0 3.609 5.30 130 3.808 1.79 4.80 230 3.940 132 1 . 8 8 2.45 33Q 4.061 1 2 1 1.93 2.33 430 4.175 114 1.96 2.23 530 4.288 113 1.96 2 . 2 1 1040 0 3.772 4.10 10 3.786 2 . 0 0 11 0 3.891 105 2 . 0 0 2 1 0 4.007 116 2 . 2 0 310 4.112 105 2 . 0 0 410 4.232 1 2 0 2.28 2.30 2.12 ±0.21 ±0.25 ±0.25 +0.30 2.24 3.40 2.05 3.10 D4 14517 32°561 117°421 970 0 4.009 3.24 ±0.38 3.50 050 4.082 ±0.65 2.30 150 4.270 188 1.80 3.38 250 4.443 170 1.87 3.18 350 4.559 156 2.03 3.17 3.15 4.91 254 San Diego Trough (continued) Station Depth Temp. Grad. k No. Location in cm °C °C/km — D5 1050 0 3.977 14518 32°571 4.00 4.000 117°501 2 . 2 0 100 4.106 106 2 0 0 4.141* 36 1.73 300 4.279 137 1.92 400 4.378 99 1.82 San Pedro Basin Station Depth Temp. Grad. k No. Location m cm °C/km °C/km _ PI 840 0 14353 33°291 5.1 110 4.789 1.76 118°20' 4.7 210 4.954 165 1.85 310 5.116 162 1.89 410 5.276 160 1.97 510 5.485* 209 2.16 P2 870 0 5.034 16999 33°301 5.00 100 5.090 118°241 3.15 200 5.163 70 1.80 300 5.248 85 1.97 400 5.343 95 1.87 500 5.450 107 2.05 P3 824 0 Aq St qts 1.83 0.69 2.74 1.98 1.81 ±0.23 +0.45 1.75 2.65 St Sts 3.05 3.06 3.15 3.08 ±0.30 ±0.40 2.94 3.32 1.26 1.67 1.78 2.19 1.72 ±0.18 +0.26 1.65 1.76 1.66 +0.21 1.60 1.79 m - ui San Pedro Basin (continued) Station_______ Depth Temp. Grad. k No. Location m cm °C °C/km 14488 33°321 5.00 10 0 2.875 1.83 118°181 2 . 2 0 2 0 0 2.956 81 2 . 0 1 300 3.040 84 (1.98) 400 3.123 83 (1.98) 500 3.208 86 (1.98) P4 860 ? 4.545 14354 33°341 4.00 90 4.594 52** 1.76 118026' 2.40 190 4.671 72 1.78 290 4.771 77 1.91 390 4.825 76 (1.91) P5 836 0 5.465 14524 33°38' 5.00 1 0 0 5.625 118^30' 3.30 2 0 0 5.743 118 1.95 300 5.861 118 2.03 400 5.976 115 2.08 500 6,144 168* San Clemente Basin Station Depth Temp. Grad. k No, location' m cm °C/km °C/km _ LI 1850 0 2.566 1.62 1.66 1.64 1.70 1.28 1.47 1.45 2.30 2.40 2.39 q A q +0.37 1.40 ±0.14 +0.20 2.36 +0.24 +0.35 ts 1.37 1.83 2,32 2.60 *t <Jts 1.71 ±0.18 1.60 1.88 t o m ( T l San Clemente Basin (continued) Station Depth Temp. Grad. No. Location m cm °C °C/km 15794 32°121 6 . 0 0 2 0 0 2.851 117°58' 5.60 300 2.947 96 400 3.029 82 500 3.122 93 600 3.216 94 L2 1490 0 3.042 15797 32°26' 5.00 1 0 0 3.195 117°461 4.50 2 0 0 3.363 168 300 3.541 178 400 3.712 171 500 3.929 2 1 2* L3 1960 0 2.608 14486 32026' 6 . 0 0 2 0 0 2.858 118°07' 4.20 300 2.970 1 1 2 400 3.081 1 1 1 500 3.188 107 600 3.291 103 L4 1830 0 2.604 14487 32°281 4.70 70 2.726 118°00' 4.00 170 2.829 103 270 2.951 1 2 2 370 3.052 1 0 1 470 3.163 1 1 1 k (1.88) (1.88) 1.79 1.82 1.84 1.87 Cl.87) 1.79 1.89 1.85 1.87 1.92 q A q qt qts +0.30 3.23 +0.35 3.08 3.08 ±0,52 2.00 ±0.17 1.82 1.92 ±0.19 .04 .04 .00 .93 L. 95 2.26 L. 89 2.13 2.06 +0.18 ±0.20 1.91 2.01 257 San Clemente Basin (continued) Station Depth Temp. Grad. No. Location m cm °C °C/km L5 1840 0 2.595 14485 32°301 3.50 50 2.709 118°061 3.00 150 98 250 2.929 98 350 3.002 98 L6 1820 0 2.604 14484 32°26' 5.20 1 2 0 2.829 118°11' 4.00 2 2 0 2.975 146 320 3.135 160 420 3.274 139 520 3.424 150 Santa Catalina Basin Station Depth Temp. Grad. No. Location m cm °C/km °C/km A1 1090 0 4.105 14309 33°101 4.00 50 4.155 118°181 2.60 150 4.316 161 250 4.462 146 350 4.615 153 450 4,849* k 1.86 1.96 1.77 1.84 1.87 1.94 (1.84) k 1.82 1.91 1.77 Aq St <Jts 1.87 ±0.20 1.78 1,87 ±0,30 2.69 2.99 2.70 2.91 2.82 +0.25 ± 0 * 28 2.61 2.74 qi St Sts 2.93 2.79 2.71 2.81 ±0.27 + 0.35 2.76 3.07 258 Santa Catalina Basin (continued) Station Depth Temp. Grad. No. Location m cm c . C/km A2 14313 33°08' 118°32' 1 2 0 0 3.00 2.30 0 50 150 250 4.076 4.133 4.336 102 1 0 1 A3 1 1 0 0 0 15805 33°11' 5.00 1 0 0 4.062 118°29 4.50 2 0 0 4.139 74 300 4.233 94 400 4.328 95 500 4.412 84 A4 1180 0 4.084 14312 33°121 5.00 1 0 0 4.150 118°27 4.90 2 0 0 300 400 500 4.341 4.479* 96 95 138* A5 1270 0 4.021 14314 33°151 4.50 50 4.033 118°37 * 4.00 150 4.121 88 250 4.209 89 350 4.304 95 450 4.389 85 k. 1.64 1.72 2.18 1.99 1.87 2,18 1.86 1.86 1.94 1.77 1.82 1.87 1.82 qi A q St ts 1.75 1.82 1.70 ±0.16 1.68 1.87 ±0.20 1.61 1.87 1.78 1.83 1.78 1.77 1.77 ±0.19 1.77 1.97 ±0.28 1.78 ±0.17 1.75, 1.94 ±0.21 1.62 ±0.16 1.60 1.78 1.56 1.62 1.78 1.55 259 Santa Catalina Basin (continued) Station Depth Temp. Grad. No. Location m cm °c °C/km A6 150 4.133 14522 33°18' 1250 250 4.231 98 118°39' 5.50 350 4.323 92 3.30 450 4.369 46 550 4.500 131 A7 1250 0 4.013 14315 33°22' 4.20 20 4.040 118°461 2.70 1 2 0 4.145 105 5.10 2 2 0 4.251 106 two 320 4.338 87 cores 420 4.457 119 A8 1250 0 4.016 14317 33°241 3.80 0 4.020** 118°51' 3.25 80 4.145 180 4.230 85 280 4.318 88 380 4.407 88 k 1.82 2.10 2.20 1.93 1.68 1.82 1.87 2.01 1.87 2.01 2.07 1.59 1.77 1.82 a q 9t <JtS 1.78 1.93 1.01 2.53 1.81 ±0.24 ±0.42 1.77 1.97 1.76 1.93 1.63 2.39 1.93 ±0.20 ±0.28 1.85 2.05 1.73 ±0.17 ±0.24 1.68 1.87 260 Santa Cruz Basin Station Depth Temp. grad. No. Location m cm °C C/km Rl 1500 0 4.365 14348 33°30' 5.50 150 4.430 119°18' 5.30 250 4.526 96 350 4.624 98 450 4.717 94 550 R2 1710 0 4.177 14349 33°311 4,30 30 4.194 119°24' 2,40 130 4.298 104 230 4.397 99 330 4.503 106 430 4.610 107 R3 0 4.195 14346 330331 1840 1 0 0 4.291 95 119°281 3.00 2 0 0 4.379 88 2.30 300 4.604 225* R4 33°371 1850 0 4.199 14345 119°28' 4.50 50 4.265 4.00 150 4.380 115 250 4.449 69 350 4.548 99 450 4.638 90 k 1.70 1.79 1.85 1.87 1.87 1.81 1.82 1.99 1.87 1.62 1.73 1.78 1.76 1.87 1.81 1.80 a q Sts 1.72 1.81 1.76 1.76 ±0.15 ±0.17 1.47 1.47 1.88 1.80 2.11 2.00 1.95 ±0.18 ±0.24 1.82 1.82 1.54 1.52 1.53 +0.14 ±0.18 1.46 1.64 2.02 1.29 1.79 1.62 1.68 ±0.17 ±0.23 1.61 1.79 261 Santa Cruz Basin (continued) Station Depth Temp. Grad. No. Location m cm °C °C/km R5 1870 0 4.188 17000 33°48' 4.00 100 4.276 8 8** 119°361 20 0 4.319 43 300 4.387 68 350 4.423 36 x 2 400 4.456 33 x 2 R6 1710 0 4.188 14343 33048' 4.90 90 119°391 4.45 190 4.431 290 4.510 79 390 63 490 4.636 63 R7 1570 0 4.166 14342 33°521 4.70 70 4.228 119°441 3.75 170 4.319 91 270 4.409 90 370 4.493 84 470 4.574 81 *i a q St 1.22 ±0.12 1.14 ±0.13 1.74 1.82 1.84 1.87 1.44 1.16 1.18 1.26 ±0.11 ±0.13 1.17 1.73 1.82 1.87 1.89 1.57 1.64 1.57 1.53 1.58 ±0.14 ±0.17 1.40 <3ts 1.33 1.37 1.40 262 San Nicolas Basin Station Depth Temp. Grad. No. Location m cm °C C/km N1 1 1 2 0 0 3.370 14483 32°341 5.40 120 3.559 118°29' 2.95 220 3.657 98 320 3.742 85 420 3.807 62 520 3.903 96 N2 1450 0 3.753 14481 32°481 6.40 240 3.940 118°56' 4.00 340 4.017 77 440 4.108 91 540 4.203 95 640 4.303 1 0 0 N3 1670 0 3.756 14480 32 57* 5.40 140 3.905 118°581 4.40 240 3.994 89 340 4.089 95 440 4.179 90 540 4.281 10 2 N4 1700 0 3.757 14479 33°02' 4.80 80 3.862 119°041 4.20 180 3.966 104 280 4.045 79 380 4.160 115 480 4.247 86 k 1.78 2.04 1.06 (2.06) 1.74 1.77 1.79 1.79 (1,79) 1.77 1.74 1.72 1.75 1.77 1.71 1.75 1.66 1.76 Aq St Sts 1.74 1.73 1.27 1.98 1.74 ±0.16 ±0.17 1.74 1.74 1.36 1.63 1.70 1.79 1.62 + 0.14 ±0.14 1.59 1.71 1.55 1.63 1.58 1.80 1.64 +0.14 ±0.14 1.60 1.72 1.78 1.38 1.91 1.51 1.64 ±0.14 ±0.15 1.61 1.73 263 San Nicolas Basin (continued) Station Depth Temp. Grad. No. Location m cm °C °C/km N5 0 3.754 14478 33°061 1630 40 3.815 119°121 4.40 140 3.902 87 2.60 240 3.995 93 340 4.071 76 440 4.207* N6 14352 33°071 119°091 1700 0 3.755 4.50 50 3.814 4.50 150 3.911 97 250 4.009 98 350 4.110 1 0 1 450 4.224 114 N7 1180 0 3.830 14351 33014' 5.10 11 0 3.921 119°031 5.05 2 1 0 3.988 67 310 4.059 71 410 4.150 91 510 4.236 86 k 1.77 1.82 1.97 1.74 1.80 1.85 1.80 1.72 1.82 1.86 1.85 1.86 qi Aq Sts 1.54 1.69 1.50 1.58 ±0.14 ±0.17 1.55 1.67 1.84 ±0.17 1.75 1.88 ±0.19 1.69 1.76 1.87 2.05 1.44 +0.13 1.41 1.41 +0.14 1.22 1.25 1.68 1.60 264 Tanner Basin Station____ No. Location 16207 32 46' 119°53' T2 16.206 32°491 119°541 T3 16203 32°541 ligo44• T4 16202 33°02' 119°47' Depth Temp. Grad. m cm °C °C/km 1 1 1 0 0 **** 3.20 20 3.906 — 1 2 0 3.980 74 2 2 0 4.122 142 320 4.234 1 1 2 0 3.905 1 1 1 0 30 3.931 4.50 130 4.060 129 — 230 4.142 82 330 4.306 166 430 4.430 1 2 2 137Q 0 3.915 4.00 100 4.013 99 — 2 0 0 4.100 87 300 4.185 85 400 4.281 96 1460 0 3.944 3.20 20 3.947 2.25 1 2 0 4.051 97 2 2 0 4.127 76 320 4.194 67 k q± q 1.99 (1.82) 2.27 (1.82) 1.63 (1.82) 1.30 1.81 1.86 1.41 1.80 1 . 2 0 A q qt qts +0.18 1.87 1.96 + 0.21 +0.21 2.14 2.23 ±0.24 ±0.15 1.57 1.65 ±0.18 ±0.12 1.21 1.34 +0.14 265 Tanner Basin (continued) Station Grad. No. Location m cm °C °C/] T5 1190 0 3.824 16201 33°06" 5.00 10 0 3.944 1 2 0°0 1 1 3.60 2 0 0 4.033 89 300 4.120 87 400 4.185 65 500 4.276 91 T6 1 1 0 0 0 18016 33°16 1 2 . 1 0 1 1 0 3.728 120°041 — 2 1 0 3.826 98 k 1.85 1.79 1.79 1.84 (1.84) (1.82) 4q *ts L. 59 L. 56 L. 20 L. 67 1.51 ±0.13 1.46 ±0.13 1.78 ±0.16 1.64 ±0.19 1.62 1.71 NJ cn < j \ Santa Barbara Basin B2 16316 34°141 120°08' Station Depth Temp. °C Grad. °C/km No. Location m cm B1 0 6.468 16314 34°13' 170 6.651 1 2 0°0 1 1 270 6.718 67 370 6.786 68 470 6.851 65 570 6.903 52 560 0 6.400 5.70 170 6.608 4.50 270 6.705 97 370 6.766 61 470 6.846 80 570 6.885 39 East Cortes Basin Station Depth Temp. Grad. No. Location m cm °C/km °C/km El 1630 0 15791 32°12' 5.00 100 3.66 118 38' 4.90 200 3.83 117 300 3.86 103 400 — 114 500 3.613 113 Aq ts 1.59 1.63 1.68 (1.72) (1.75) 1.09 1.14 (1.11) (0.90) 1.12 ±0.11 +0.16 1.12 1.72 1.66 1.71 1.77 1.72 1.83 1.65 1.08 1.38 0.71 1.20 ±0.11 ±0.12 1.20 1.84 ts 1.73 1.80 1.89 1.87 1.79 2.11 1.95 2.13 2.02 2.05 ±0.18 ±0.19 1.97 2.16 267 East Cortes Basin (continued) Station Depth Temp. C Grad. C/km No. Location m cm E2 1500 0 3.2211 15792 32°15' 6 . 0 0 2 0 0 3.514 118°241 — 300 3.619 105 400 3.716 97 500 3.827 1 1 1 600 3.906 79 Pacific Ocean Station Depth Temp. grad. No. Location m cm C/km C/km Ml 2470 0 18022 30°451 4.10 1 1 0 3.436 118°11' 2.90 21 0 3.541 105 310 3.640 99 410 3.740 1 0 0 Outer Basin M2 2150 0 18Q21 31°071 4.20 20 3.653 118°331 12 0 3.766 113 2 2 0 3.880 114 320 3.994 114 420 4.105 1 1 1 k (1.81) k 1.96 1.83 (1.89) (1.84) qi q Aq qt qts 1.77 ±0.15 1.72 1.89 ±0.17 qi q q qt qts 1.90 ±0.18 ±0.24 2.01 1.81 1.89 2,08 ±0.19 ±0.24 268 Animal Basin Station No. Location M3 18023 31°211 117°50' Velero Basin M4 18019 31°371 118°401 West Cortes M5 18017 32°17' 119 18' Depth Temp. m cm °C 1970 0 3.00 100 2.822 — 200 300 3.056 2330 0 4.20 220 3.395 — 320 3.603 420 3.810 1740 0 2.50 150 3.620 3.10 250 3.761 Grad. °C/km 117 117 208 207 141 k (1.89) (1.84) 1.92 1.86 1.75 Si q Aq qt qtg 2.21 +0.20 +0.26 3.83 ±0.35 ±0.44 2.59 ±0.24 +0.30 269 Continental Shelf off Central California Station Depth Temp. Grad. n k No. Location m cm °C C/km C 1 o 790 0 4.501 17671 34^20' 4.50 50 4.559 120 51' 5.00 150 4.676 117 1 . 8 8 250 — 130 1.85 350 4.936 130 1 . 8 8 450 5.081 145 1.94 C2 590 0 5.753 17674 35900' 5.40 40 5.779 1 2 1 02 5.65 140 5.856 77 1 . 8 8 240 5.871 15 1.98 340 5.992 1 2 1 2 . 0 1 440 — — 2.06 C3 970 0 17675 35°221 5.50 150 4.781 121°331 2.40 250 4.843 62 2.36 350 4.924 81 2.25 450 4.990 66 2.17 550 5.061 71 (2.17 qi 2.20 2.40 2.44 2.81 Aq 2.46 +0.22 ±0.25 t 2.45 Sts 1.45 0.30 2.49 1.42 +0.13 ±0.16 1.39 1.46 1.82 1.43 1.60 1.58 ±0.10 +0.22 1.57 N> O Continental Shelf off Central California (continued) Station Depth Temp. §rad k q ^ q qt qtg No. Location m cm C . C/km J L . C4 1 0 1 0 0 17678 35°40' 4.50 50 4.784 121°491 4.10 150 4.835 5q** 250 4.938 103 1.99 2.05 350 5.053 115 1.98 2.28 450 5.168 115 1.91 2 . 2 0 2.18 +0.19 2.16 ±0.20 Station No. : Station Location: Depth m : k <*i q Aq *t * ts ** cm Numbers prefixed with alphabet are used in the text and five digit figures are Allan Hancock Foundation station number, north latitude and west longitude water depth below sea level (first row), corer penetration (second row), and core length (third row) position of thermistor probes below the water-sediment interface. conductivity, m cal/cm*sec °C interval heat flow, >tcal/cm2 • sec average interval heat flow, ^ical/cm^. sec standard error (upper row), 95% confidence limit (lower row) corrected heat flow (topography only) corrected heat flow (topography and sedimentation) measured interval gradient was not an equilibrium value and was not used. one probe probably did not penetrate into the sediments, hence this interval was not used. 271 APPENDIX IV Effects of Non-Cyclic Surface Temperature Variations on Temperature Gradient Solutions to the disturbance of geothermal gradient have been obtained for surface temperature variations of step-function type (Jaeger, 1965) and for the sea- or lake- level change in the permafrost region (Lachenbruch, 1957). A step-function may be adequate to represent paleo- temperature variation in the land area owing to the scarcity of available data. In the marine environment detailed paleo-temperature date can be deduced from the sediments and associated fossils, hence better approximation than step-function is needed. Bottom water temperature vari ation, cyclic or non-cyclic, can be approximated by a Fourier series and the contribution from each Fourier component can be summed up to estimate the distrubance. This method has been described previously by Carslaw and Jaeger (1959) for a homogeneous medium. In this section we shall derive a solution by replacing temporal variaion of paleo-temperature with a set of linear segments. Such approximation is conventionally found in the presentation of paleo-temperature data. Consider the case that the semi-infinite region zM) is initially at temperature G + gz and the plane z = 0 is 272 273 maintained at (t) for T < T. The question is to find o — the temperature distribution v(z,t)- and thus the perturba tion of the temperature gradient from the initially un disturbed gradient g at timfe T as the result of temporal surface temperature variation < f > (t) . The temperature v satisfies = K v = G + gz when t ^ TQ and V = < j > (t) at z = 0. The solution can be obtained by a combination of u and w v(z,t) = u(z,t) + w(z,t) where u satisfies 9%t = K 9 % l x /bzz u = 0 when t T o u = <p (t) when z = 0 and w satisfies = k w = gz + G when t TQ w = 0 at z = 0 respectively, and K is the thermal diffusivity of the medium. The solutions of u and w ahve been obtained by Carslaw and Jaeger (1971, p.63 and p.61): u and W 274 CZ,t)~~=Z f (/>&} **P (- Z */?X (jiTK tfi^K J% ft - X )3/z (z,t> - Q erf ( 'z/zjx (t-To {] -+ £z The spatial derivatives of u and w are 9a 92 __!_ fTrfwo exp(-i‘Ak(-r-»'>l ci* ZfzK Jr ’ C-T - * 1 V* , y _ x f r t o ) e > r l - z ‘/« * C T - > ) 7 J A ■§£ = $ 1 (<i//^/((T-T,-))exe[-‘ zV^C-r-%~)J . Suppose < ^ > Ct) can be approximated by m linear segments, i .e, <£(t) = ax + ^(t - tx) tx 5 t < t2 C^>(t) = a2 + b2 (t - t2) t2 < t « tj • " * • • <pm = am + bm (t - tm) tm S t < tm+1 where aj+i = aj + bj ^j+l “ bj) (1 <- j S m) ti = Tq and tm + 1 = T . Near surface, the perturbation from true gradient is ,al/ i „ G 1 fT cM therefore = JxKCr-r.) ^ J, C t - M ’ -* = ^ + _ L y f a , - * bj (> -xj 3 ^ /^7£K'C t -To) ^Tr ) t J*. ^-r- ^J)3 / ^ 2 - tj+i fTnjFcT CT-T.-i J B fi, I CT-A)* J J *J This solution is applied to the geothermal measure ments in Tanner Basin where the bottom water temperatures 275 in the last 37,000 years B.P. have been derived on the * 1 O I C . basis of 0 / 0 in the tests of benthonic foraminifera (Gorsline ■ : and others, 1972) . Function <£(t) is obtained by subtracting the present water temperature from the paleotemperature shown in Figure 52. The temperature prior to 37,000 years B.P. is assumed to be the same as present bottom water temperature 4.9°C, thus G = 2.5°-4.9° = -2.4° C, where 2.5°C is the temperature at 37,000 years B.P. The results indicate that the near surface geothermal gradient has decreased by 2.6°C/km for a thermal diffu- sivity of 0/004 cm /sec, 3.0 C/km for 0.003 cm /sec, and 3.4°C/km for 0.002 cm^/sec. 276 Figure 52. Paleotemperature in Tanner Basin deter mined from oxygen isotope ratios in benthonic foraminifera tests (Gorsline and Barnes, 1972). Ages of sediments: at depth of 2.5 m about 17,6000 years B.P. (carbonate-carbon) or 16,300 years B.P. (organic carbon); at depth of 6.5 m, about 32.000 years B.P. (carbonate-carbon) or 33.000 years B.P. (organic carbon). Figure 53. Coordinates used in solving a problem of surface temperature variation. 277 0 £ X I- 4 Q. ui a 8 Figure 52. v(z,o) =. q + 32 H Figure 53. APPENDIX V Heat-flow Refraction across Boundary of Dissimilar Media 1. Analytical Solutions The perturbation of regional heat flow around the contact of dissimilar media is sensitive to the configura tion of the media and their conductivity constrast. Usually the perturbation must be determined numerically or by ex perimental analogs for each configuration. Analytical solutions can be obtained in a simple manner only for a few simple geometric configurations by the method of images under the stringent condition that the ground surface is a plane of symmetry with respect to the anomalous body. This section presents several cases of heat refraction across the boundaries of dissimilar media involving bodies of simple geometry. The boundary conditions include (1) zero surface tem perature, (2 ) constant vertical heat flow (kg) at large distance from the anomalous body, and (3) cohfcinuity of temperature and normal heat flux across the boundary of different media, where g is the undisturbed temperature gradient and k is thermal conductivity. The procedures used involves selecting an appropriate coordinate system for the solutions to Laplace's equation and then matching the boundary conditions to determine the unknown coefficients. 278 279 Semi-cylinder (Figure 54) By the method of images, a semi-circular cylinder im mersed in a semi-infinite medium is equivalent to the prob lem in static electricity of placing a dielectric cylinder in a different dielectric medium in which the electric field is initially uniform (cylindrical axis being perpen dicular to the electric field). The temperature inside the cylinder is T - 2 3 K > V k, + Kz and outside the cylinder is At the surface, the normalized vertical heat flow inside the cylinder is 2 k, 0 5 a Outside the cylinder the normalized heat flow is 9L, = — ~ (--- J = I + ------- — - a£x<a> a > * 0 Ka + k, of which the second term represents the perturbation from the regional heat flow, q is constant everywhere inside the cylinder but q2 is a function of position outside the cylinder. A discontinuity in the vertical heat flow occurs at the boundary of cylinder where qreaches its extremum (9J = -2 K2 / ( K, + K*) Figure 54. Cylindrical coordinates. ( . v . r . : ) ( * . v . 0 ) Figure 55. Spherical coordinates. 281 and the ratio of q, to q is (*1i /Qi j = / K 2 x=a. The net perturbation over the surface is a o as expected. Semi-sphere (Figure 55) The temperature inside the sphere is T = -1^2:— g* r ^ a 1 2K2 + K, and outside the sphere is T * = ('+ 75 ) gz r ^ a The normalized surface vertical heat flow inside the sphere is 3 K ( r * «■ ~ 2KZ + K, «- and normalized surface heat outside the sphere is q ~ i K*- K| Ji r ^0 L 2 2Kz+K, r3 - 2- = 0 q-^ is a constant inside the sphere while q2 is a function of position outside the sphere. The ratio of the two at the boundary is equal to the corresponding ratio of con ductivity. Again, the net perturbation over the free surface vanishes identically. Semi-elliptical Cylinder (Figure 56) This case can be divided into two subcases, one with the minor axis perpendicular to the free surface, the other with the major axis perpendicular to the surface. The steady state temperature distribution in a homo geneous isotropic medium is governed by Laplace's equation where (u,v,z) are the elliptic cylindrical coordinates re lated to the rectangular coordinates by x = a cosh u cos v The boundary surface of elliptical cylinder is described y = a sinh u sin v (2) z z. by x2/(a2 cosh2 uQ) + y2/(a2 sinh2 u q ) = 1 (3) where a cosh uq = b = length of major axis a sinh uQ = c = length of minor axis uQ = coth ^ (b/c) 233 o <»<■?■ ,u=0 V = —a Figure 57. Bipolar coordinates. 284 The solution to Laplace's equation without z- dependence is of the general form r +mu) f +imv ) T = [ e - j ( s - j Case £ (minor axis perpendicular to the ground surface) The temperature inside the elliptical cylinder is oo = 5~ s' m^ mU sin mv 0 := • u $ U. m = l and the temperature outside the cylinder is (5) T 2 = q y + £ 1 B m e Sin m\f m={ U, S U < « T^ and satisfy automatically the following boundary con ditions T = T~ = 0 at v = 0 or TT 1 ^ T^ _> gy as u — » 00 (6) Applying the conditions that T = T_ at u = u 1 2 o and ( ( . . i l l = k, ~ at u = u, (7) 1 a u 2 su ° we have u., o K , s in h U + K. c o s h U , B = t K j3 0 - g U° sinh 2 U e 2 . ( k z sinh U0 -+ (C, cosfi (J0 ) (8) 285 and A = B =0 for m £ 2. m m At ground surface (y=0) the normalized vertical heat flow inside the semi-elliptical cylinder is a, = K. / ST, \ * V 'v, = 0 K2g (9) K, e t K, cosh ite + k . z sinh U# ) The evaluation of the temperature gradient outside the elliptical region is not straight forward because the co variant and contravariant transformations between the rec tangular and elliptical-cylindrical coordinates are differ ent. The components of a vector in the two coordinate systems are related by: \ / \ C L sinh u c o s v - a coshu sinV a coshu sinV a s»nh a cosv 0. (s in h u + sin v) (10) ^ft( sinh*U - < ■ S in 'v ) /z where y> is the gradient of a scalar < j > , i.e. y? = V(f> In the present case, f = e sin v -u. r = ------ «------- j- a( siWu + sin* v ) ' 2 L A sinv e„ + cos v e (11) With this we can evaluate the normalized vertical heat flow 286 outside the semi-elliptical cylinder K4 / aT2 S = -li- ( ) 2 Ka g I. a * Jy=c = |+ (■.!=»- M a e - si"hau>-f.!.(e-Msl,y)1 2 ( Ka s in flli0 + K, cosh U„) L3 ^ J u- • ■ - (12) v =0 - i + J J C i - K ,) e U° s m h 2 U „ e ' U 2 ( (C 2 sin h U0 + K, cosh UB) s i n h M where sinhu = (x^/a2-l) and e-u = x^/z2 - (x^/a^-1) . At the contact (x = a) q _ | + ( K. 2 ~ K>) cosh u 0 2 Ka sin h U0 ■ + k., c o s h a o __________ Ka €.a° K , sin h a + <. cosh U„ The ratio of surface heat flux across the contact is ql//q2 = kl//k2* (14) and q2 have been derived by Lachenbruch and Marshall (1966) in different form, which can be obtained through a simple transformation from our results. Let D = c/b = (sinh uQ)/(cosh uQ), (15) and £ = k-,A2 then A^ = ga(D + 1)/(D + k) and = ga(l-k)(D+l) sinh uq cosl; UQ/(D+k) (16) The heat flow inside the elliptical area is therefore q^ = k (D + 1)/(D + k) (17) and outside the elliptical area 287 = = I + = I + ( I - Kj( p + i ) sinh U0 cosh Uc D + l c - U0 . - I - K e S inh U0 D + fc e u s in h U - -LI I - < o e D + i c I - D sirlh u -u. sinh u (18) Because x = + a coshu when y = 0 and sin h u /_xj _ , \ '/i \ a* / - I I - a7xi)~'/2 - I = [l - lb*-c‘Vx*]"l / 4 - I = [l + b* ( D* + l>/x‘j'/:L -I (19) We have arrived at Lachenbruch and Marshall1s result (1966, p.1240) g = , + J j l E . — £ _ f [ i + b 2( D ‘- + i y x ‘ ] ’ i/2- , i <2 °> D + K I “ D L J Case 2 (major axis perpendicular to the ground surface) The temperature inside the semi-elliptical cylinder is 288 0 0 T = ZL A ,_ cosh (iim +i)u. cos ( 2 m + ijv / ?1 \ M = 0 2m + l V ' and outside the cylinder oo -(am + i^u (22) T2 = gx •+ 2H ®xm+( e cosUm+OV' and T2 satisfy the following boundary conditions = T2 = 0 at x = 0 or v = K/2, 3JT/2 (23) T —» gx as u 00 2 T 1 = T 2 I a t u = uo . K JLL =. k J t L. j N 5U 1 d U Applying the last two boundary conditions to and T2 yields l q a e A, =:--------------Li---------------- cosh U0 + K, s fn h u 0 {24) „ . (K.a - K ,) g a e U o sm hUo D| * ■ ■ . at^coshu,, + K, sinh U0 J and A„ = B» ., = 0 for m 4 0. At the ground surface 2m+l 2m+l T ^ ,(x = 0) the normalized vertical heat flow inside the semi elliptical cylinder is J± (Hi) 9 \9XJx=0 I f ei° *1 (25) K 2 cosliUa + K ., sinh u0 In the present case, < f > = e-u cos (v) 289 — v<£ & ( s i n h U + s i n 1 ' / ) [■ A € u COS V A , ■ » - smv I - u = - T - S in h u to s 1^ + cosh u s in 1! / ] atsin^U - < - sin1^; (26) We have the vertical heat flow outside the elliptical area Ka I *Ta = ^ Ktg \ 3x/x=0 - | + ( M 6 s m h 2 U 0 e 2 ( Ka coshH + K . sinln U „ ) t o s h a (27) where cosh u = (1 + y'Va2)"*'^2 and e-U = (1 + y2/a2)^-/2 - y/a Two Parallel Semi-circular Cylinders (Figure 57) This case can be applied to basin and ridge configura tions with minor relief. The steady state temperature distribution in a homo geneous, isotropic region is governed by Laplace's equation > = (c o s h fr - C K > s j./ r a*T ] a‘ T „ o ? L 9 ^ j (i) where (£,2,z) are the bipolar coordinates related to the rectangular coordinates (x,y,z) by 290 a sink 7 X. =• L cosh 7 - c o s £ a s!n. £ :o sh T [ ~ c o s 4 and ^ = 2 0 ^ < 2/t, ~°° ^\< 00 . (2) The cylindrical boundaries are described by ( x - a coth 7 / 4 S z = a csch^n lo' j t0 (3) where a coth^ = b = one half of bipolar distance a csch*7 = c = radius of circular cylinder * • O = cosh ^ b/c , 2 2.1/2 and a = (b - c ) Also, on xz-plane, i.e., y = 0 (or % = 0), we have cosh 7 = (x2 + z2)/|x2 - a2). (4) The solution to Laplace's equation with no z-dependence is <bf the general form . ( . = ■ - > } p « i. Specifically the temperature inside the cylinders is oo - m (?) T, = E e »;»■»! \ m=i (6) while outside the cylinders it is 00 T 4 = sy + ZZ Bw cosh ir>7 sinml- i l l 4 (7) 291 T and T satisfy the first two of the following boundary 1 2 conditions = Tj = 0 at y = 0 or 5=0 T2 —+gy when ? —* 0 or i j —*1 T-l - T2 | at I = + ?o £1 = K £2* I (8) while the coefficients A and B„ are determined by the last m m J two conditions. Thus MI?J 4 = 23ae SiVih vn? + e cosh my and B = 2 3°- ( e) m sinh m\ + e c o s h my (9) where £ = k^/k^ . In the derivation of A and B use has been made of m m the formula y £ a cosh y - c o s % 00 .-will _ 2 H e 1 ( 1 s i n n £ n=| (10) (the derivation of this formula is given later). For large m A ^ m ~ i + e and B ~ e;e * (11) m i + e 292 The temperature gradient T. = cost>,2- a ( 5 m=i * • -"HI -will c e s m $ - e Z A n»te sin ( W =f in ^ Ti — 9 ey + f £ Z>lB cosh m? cos »}? CL l I w=i rn * > o o + e. L B m. sinhrriv st'nmt / t m=l ^ \ (12) can be transformed via the realtion Y, I- / cosh 7 - cos% > ? sin| i - c-osli^ cos % -I + cosh^c©s£ - sinfipsint t % into i ST, t cosh V -1 ^ (— 7-1 = - r w e ^ 5=, n* « - W I l ) (13) (2i) = /?=o cosh? I £ g m cosh vn^ Y l f l ^ 0. M = l The normalized surface heat flow is therefore -"•(III - ?.) i ~ i e ( c o s h n - o £ ^ e III >7 (14) m=i sinh m V + ecoshwj? co © “ m ?o qx = I+2(i-e)(cosh?-i;£ COShrH ly, m =( sinh nj^ -f- g cosh III (15> ® • Differentiation of equation (10) with respect to yields 293 cosh ? c.os£ -| 0 £? “n 111 -1--- 5--- - = 2 H ne tosn£ (cosh ri - cos %) n=i 7 (16) which can be used to show that becomes unity if 6 = 1 , oo - mini q = 2 (coshi - i j H m e = i hi — J * (17) remembering that is evaluated at £ = 0. At 1. = 2 6 (cosh > ? -l) ZI ! j ------------ to m=, + e coshw? = I + (I- e)(coshy _,)£ 0 Wiaf S<«h »»? + £ cosh 00 and 9^9, = I +(c°sh^-|)XI r , -2 )nJ . . [ ( 1- e ‘*)(,-ej-2 e] (18) m =i sinh ny - + ■ e coshw? O ' ’ o Hence, a discontinuity in vertical heat flow occurs across the boundary. The discontinuity disappears if G= 1. Expressed in exponential form, equations (15) and (16) become, respectively, y y , r -(.yh- 0 1 - (w -h )? t -m y co * [ e + e J - m e . L q = ML —------------ ITSTn------> 1 m =i — (I — e (19) and 294 = I + 2(1-6) £ | 2*[- e-m(i’ 20-J Z) + ? + c-»(*?0t?J + S -aw*. ] (1 + 6 ) - (I - e) e *. >H) (20) Because of the exponential factor e-mY in and in q^, the two series converge rapidly and only a few terms are needed in the calculation of q^ and q2 » The results of a few numerical cases are shown in Figure 51. Equation (10) has been misprinted in the text by Morse and Feshbach (1953, p.1321). Rederivation is thus in order. Let Sin £ . = £ C„ sin n % Cosh \ - cos | n=| then ^ = _L f sin £ sin nfe j ^ n ti J cosh » 7 - cos f c 2 71 iTi cosh ^ - cos % f2TL cos(n-iJl - cos Cn -f Q \ X cosh >2 - cos | nt i( n + i; | rill e _ e = j l p ^ :_____dt 2rc I cosh* 7 - cos £ cosh^ — cos J - in which the contribution from the imaginary part of the integrand vanishes. To evaluate the integral, we make a 295 Figure 58. Refraction of heat flow around two parallel cylinders. Radius of cylinder = 1.0. Left : Bipolar distance = 1.0. Middle: Bipolar distance = 2.0. Right : Bipolar distance = 3.0. At the edge of the cylinder, the upper curve corresponds discontinuously to the lower curve. Conductivity ratio for solid curves 1/2, 1/3, and 1/4. Conductivity ratio for dotted curves 2/1, 3/1, and 4/1. CO CO CM t —I CO CD CD CM CD CO 296 CO m o o o rv c r v 771 297 contour integration around a unit circle. Let z = e1^ then d % = dz/iz and cos £ = (e^ + e~^)/2 = (z^ + l)/2z. Hence «-t n+i C„ = — { ---- = ------- dt 2* * ii( cosh7 - «liL) -_L_r -gn~' + z n+l . LK J 2* - 2 ( e’ + e~*) + I _ I I *" + 1 - z n"' , ** J (€ -e^ ) ( * - c-*1) _ 9 For positive ^ / the pole is located at z = e . We have r -^n + o -n Cn-i; I c. = - [ e - 2 e_n^ e c - e For negative ^ , we have similarly Cn= 2 e11^ For any real , we have therefore C„ = 2 e Semi-ellipsoid The steady state solution.for an ellipsoid immersed in an otherwise homogeneous infinite medium has been deter mined for a prolate or oblate spheroid (Carslaw and Jaeger, 1959, p.427; Jaeger, 1965). Von Herzen and Uyeda (1963) have used the case of an oblate spheroid to estimate the effect on the regional heat flow of a depression filled with low conductivity sediments. An exact solution for a tri-axial semi-ellipsoid with its minor axis perpendic- 298 ular to the ground surface has been applied to estimate the effect of conductivity contrast on the heat flow measure ment in a lake (Henyey and Lee, 1973). Here, the solution of Carslaw and Jaeger will be inte grated exactly for a semi-ellipsoid with one axis perpen dicular to the ground surface in order to see the behavior of surface heat flow inside and outside the semi-ellipsoid. This solution can be used to give a more versatile approx imation to a sedimentary basin than that for a semi-oblate spheroid. The solution can also be used to estimate the thermal disturbance near vertical fault zone by making one of the axis vanishingly small. Consider an ellipsoid of conductivity K' immersed in an otherwise homogeneous infinite medium of conductivity K. Let the ellipsoid be defined by x2/a2 + y2/b2 + z2/c2 = 1 . (1) Suppose that at large distance from the ellipsoid the tem perature tends to To» = vix + v2 Y + V3z (2) then the temperature inside the ellipsoid is ^ Vt y t V3 % I + A0(i- O I + B0(i-e> | + c0(|_ e) (3) and outside the ellipsoid is __ V, x + V* ) ! + V3 2 - (*- O V. AxX 1 + M e ~ 0 Ce~0 V4 Bx» _ (e-OVa Cxg I + b0 ( i - €) i + C. ( e - i; (4) 299 (Carslaw and Jaeger, 1959, p.427) whfere V^, V^, and V 3 are constants, € = K'/K A — a . b e r 00 1 dll 2 >^ (a^ + ujAtu) B* = abc f 00 1 du 2 J, ( fc^H- U) A(U) I I u* abc r du 2 (C* + U) A lu; A(U) » [( <kz + U)( b * + ujtc2 - . - 1 ^2 -+ ujJ and \ is the positive root of (6) (7) (8) (9) x2/ (a2+X) + y2/(b2+X) + z 2 / ( c 2 + A ) = 1 (10) A , Bq, and C0 correspond respectively to Ax, Bx , and Cx with X = 0. For a prolate (or oblate) spheroid the integrals Ax, Bx , and Cx can be expressed in terms of elementary func tions (Carslaw and Jaeger, 1959, p.427). For a tri-axial ellipsoid, the integrals can be evaluated with the trans formation a2 + u = (a2 - c2 )/sin2 y/ (1 1 ) the integrals become _ a b e ________ f * s i n * '4' W J (l - k^sin^fc (1 2 ) b>c ■ !> _ i. sir, 2*7 (12-D “ a - b > c 300 (V-b'7**- [ 5ec * - si" +] a ^ =c a b c V, (a 1 - c * / * [ F (k. , . < £ ) - E ( k., <pj k1 a. > b >c Bx = a be sin1 y d y r sin Y Ca1 -cl)>A J (,_ sin1 /c 2(fc1 be i - I - c')*/2 [ ^ ~ 2~ sin 2 ' = ( V) ^ J a =b > c 2 ( a2 - b 2.)3/i f c " * S ec< £ " U (s e c 0 + iao£)J a. > b = ■ c a b c E C k,0) - K 1 FOcft) _ sin# cos ft j ' Z v / l — k 2s i n J f t J K ' V - c^ L a > b > c Cx = abc sin2^ dty ta- abc ^ 1 - I c o s 1 V' ( l - K1 sin 1 b 1 C ( b ‘ ~ c > ) * A afc' a = ■ b >c ) 2 ( a 1 - b' 1 ) ^ ^ 3^ sec^ - Lrt(sec<P +**'’*)] a >- b =c ab c H'V where k2 = (a2 - b2)/(a2 -c2) k'2 = 1 - k2 (12-2) (12-3) (13) (13-1) (13-2) (13-3) (14) (14-1) (14-2) (14-3) (15) (16) 301 sin2< j > = (a2 - c2) / (a2 +\) (17) J * ^z. df/(i ~ kxsin (18) <7 = elliptical integral of the first kind and E (k,<^>) = ( t - k* Si* f ) d (19) Jo = elliptical integral of the second kind. The solutions (12-3), (13-3) , and (14-3) have been derived by Oberhattinger and Magnus (1949, p.8). The integrals AA, , and C*, are related by AA + BA + CA = abc/A(A) . (20) It can be shown that equations (12-1), (13-1), and (14-1) are equivalent to the solutions for a prolate spheroid, and that equations (12-2), (13-2), and (14-2) are equivalent to those for an oblate spheroid obtained by Carslaw and Jaeger (1959). Also, with the subtitution u=a^-w^, the integral r o o CK = <xbc d-Vj/[( \AJX — , ' (21) is identical to equation (8) of Von Herzen and Uyeda (196 3, except for a misprint in their equation) for A= 0. Since elliptical integrals are tabulated, the steady state tem perature distribution can easily be evaluated at any point. We shall discuss the perturbation of regional heat flow around a semi-ellipsoid with its plain surface being coincident with the ground surface. Furthermore, we require that the temperature at the ground surface is zero and that the magnitude of heat flow at large distance from the semi- 302 ellipsoid is constant and its direction is perpendicular to the ground surface. It is obvious that equations (3) and (4) satisfy the boundary conditions if any two quanti ties of V^r ^2' or V 3 vanish V2t and V3 can be consid ered as the product of the undisturbed temperature gradient and the direction consine of the regional heat flux with respect to the principal axes). Case 1: The major axis is perpendicular to the ground sur face (V2 = V 3 = 0 ), the normalized surface heat flow inside the ellipsoid Case 2_: The intermediate axis is perpendicular to the ground surface (V-^ = V 3 = 0), the normalized surface heat flow inside the ellipsoid Q: X -o x =0 I - * • AoU- e; (22) and outside the ellipsoid (€ -0 Ax 1 + te - 1 ) A (23) e I + (6 - 1) Bc (24) 303 and outside the ellipsoid a To a * “ (t H , - . __ | _ (g ~ 0 Bx_________ l+(6-«)B# (25) Case T^e minor axis is perpendicular to the ground sur face (V1 = V2 = 0), the normalized surface heat flow inside the ellipsoid 9;, K. (s? jJ=0 K (Ha) v a * /* =o €________ 1 + ( e - i) ca (26) and outside the ellipsoid q — K ( T # .p oc IT /• 3T( K ( f r ) . ,* l.e -0 cx 1 I + Ce - 0co (27) Equation (26) for a = b > c has been used by Von Herzen and Uyeda (1963) and for a > b > c by Henyey and Lee (1973) to evaluate the perturbation of surface heat flow caused by the conductivity contrast. The results of a few numerical examples for ciase 3 are shown in Figure 52. 304 Figure 59. Refraction of heat flow around a semi-ellipsoid with minor axis being perpendicular to the ground surface. Normalized heat flow vs. distance from the center of ellipsoid; half b-axial length = 1.0. Upper left : K'/K = 1/4, a:b;c = 4:2:1. Upper middle: K'/K = 1/4, a:b:c = 8:4:1. Upper right : K'/K - 2/1, a:b:c = 8:4:1 Lower left : K'/K = 1/2, a:b:c = 4:2:1. Lower middle: K'/K = 1/2, a:b:c = 8:4:1. Lower right : K1/K = 2/1, a:b:c = 4:2:1. Curve from right to left at angular increment of 15° from the major axis a, i.e., 0°, 15°, 30°, 45°, 60°, 75°, and 90°; dotted curve showing surface heat flow inside the ellipsoid. 306 Buried Cylinder (Figure 56) Let the cylindrical axis be parallel to the ground surface. We shall use the bipolar coordinate system to solve the buried cylinder problem (see the case of "two parallel semi-cylinders" for the coordinate transformation), The temperature inside the cylinder is T, = AQ + Z. Ame-m?cos wi£ (1) m=i and outside the cylinder the temperature is aO T- = gx + Bm sinhm^ cosm^ (2) Wa| It is apen that satisfies the first two of the following bounda/ry conditions T2 = 0 at x = 0 or £ = 0 T2 —> gx as x —» 00 Tx = T2 at 1= ?# (3) and K, §y = If at \ = \o The coefficients are determined from the last two boundary conditions, aq = ga Am _ (2gak2em^, > ) / (k2c°shn,,Z0 + kjsinhwi^) (4) — mo and Bm = C2ga(k2~ki) e O / (k2COsh my + k^sinhwp^ ) In the derivation of Ajn and Bm , use has been made of the relation (Morse and Feshbach, 1953, p.1321) sinh? v2 - -n»7 — ------ — - c e u«cos n£ (c) coshp - cos£ n=0 " * Differentiation of equation (5) with respect to yields 307 1 - cosh? cos£ v - _ --nv Tcosh? - ,cos|)^ = - S l nepe ^ o s ^ (6) where €a = 1 and £„= 2 for n ^ 0. The temperature gradient outside the cylinder is T2 = g ex + — — ^-g--COSg | ~e% II ^sinhm? sinm£ 00 + % £, Bm cosh m?cos m*j (7) (* V * X ) x=0 = (dT2/*X\ =0 CO = g + [(l-cosg)/aJ ZI^B^os m| (8) The normalized heat flow at surface is therefore q2 = (k2/k2g) (3T2/aX)^=Q = 1.+ 2(k2 - k^)(1 - cosg)' oo _m n E. (e * cos m|) / (k2coshm?( + k^sinhm^) (9) It is noted that the net perturbation over all the ground surface vanishes identically. That is KA ' ~ £ B^cos»i£ dy )0 (X n \=i = 4 ^ (*0“ cos$) C0SYnH a *»=i - j0 = -n, re.!' TC ' - 0 The results of a few numerical examples are shown in Figure 54. (10) 308 Figure 60. Refraction of heat flow due to the presence of a buried cylinder. Normalized heat flow vs. distance; radius of buried cylinder = 1.0. Left : Depth of burial to the center of cylinder = 1.0. Middle: Depth of burial 2.0. Right : Depth of burial 3.0. Conductivity ratio for solid curves 1/2f 1/3, and 1/4. Conductivity ratio for dotted curves 2/1, 3/1, and 4/1. 309 CD CO CD CXI CD CD CM CD CM 1 CD APPENDIX VI Heat-Flow Refraction across Boundary of Dissimilar Media 2. Finite Element Method Conduction heat transport can be described by the equation V(kvT) + Q = ji (3T/at) (1) where temperature T is a function of spatial coordinates (x,y,z) and time t, k is conductivity, Q is heat production per unit volume, and is 1/heat capacity = 1/pc, p= density, c = specific heat. On the boundary of the region either the temperature is specified with T = Tfa (2) or the first derivative of temperature is constrained with ktfT * n + q = 0 C3) where n is the outward unit normal to the boundary surface. At the boundary between different materials the tangential derivative of temperature and the normal component of heat flow are continuous. Equation (1) together with the prescribed boundary conditions (2) and/or (3) provide a unique solution which can be obtained by either analytical methods or the finite difference method. An alternative to the solution of the differential equation is the finite element method based on the calculus of variations. The formulation of finite 310 311 element method has been extensively discussed by Zienkiewicz (1971). Application of Calculus of Variation According to Euler theorem, the necessary and suf ficient conditions for the integral to be minimized is that the unknown temperature should satisfy the following differential equation within the same region V, provided T obeys the same bound ary conditions in equations (4) and (5),.where Tx=3T/ax*••, etc. It is easy to verify that equation (1) is at any instant equivalent to the minimization of the functional subjected to T satisfying the same boundary condition. The prescribed temperature on the boundary can be easily implemented in the functional. However, the imposition of the prescribed heat flow on the assumed function form is impractical. Instead of constraining the boundary heat flux, we add a term A to the functional J£(u). The integration is made over the surface A where boundary condition (3) is to be satisfied. j)C(u) = Jff f (x,y,z,t,T,Tx,Ty,Tz,Tt)dxdydz (4) (5) (u) = JJf jk/2 VT-VT - (Q - pdT/dt) Tjdxdydz (6) V (7) 312 On minimization the term automatically yeilds the required boundary condition. Its proof is given below. Suppose we want to minimize the functional X(u) = ff dV + fqT dS (8) v i where A is a portion of the boundary on which prescribed heat flux is imposed and f = (1/2)ki(VT) 2 - (Q - }1BT/at)T . (9) An arbitrary small variation in tempprature and its first spatial derivatives yields S/C= J (af/aT) T+(3f/aTx)fTx+(af/aT )STy+(af/aTz)S'Tz dV V + fq£T dS . (10) A Since £ T = ^aT/ax) = — (St) , ... etc. (11) X O. A equation (10) may be written in the form (12) Jv i d r L $ Ty a * r aTy r + j < 1 S'T dS A Integration by parts, equation (12) becomes ^ “ I s 'T { g - & U ) - f f i r ) - £(&.)} J V = - j S t { + v(K7T))dv + I St (.9 + ; + * 4 # f )ds (u) A ** y * where 1 . 1 . and 1 are the direction cosines of outward X y Z normal. The surface integral applies only to the boundary where the temperature is not specified because on the boundary where the temperature is fixed the temperature 313 variation &T = 0. At stationary point the variation of the functional vanishes, i.e., Sj(= 0. (14) This is true for arbitrary £t if and only if equations (1) and (3) are satisfied. Discretization and Extremization The region under consideration is separated by imagi nary surfaces into a number of finite elements. The ele ments are interconnected with the nodal points situated at the element boundary surface. The temperature T within and on the boundary of an element is expressed in terms of the temperatures T^, T^, ... at nodes associated with the element concerned, T Tj} = M ^ t]® (15) T = [N1,Nj,...J where [n J is the element shape function to be discussed later and superscript e designates the quantities specified in element e. By differentiating equation (8), we have for any node I f = 1 1 K [ S k W * +S k & l + >“ Cit f£ + t s r ^ J + f q ds Se dTi (16) where the surface integral applies only if the element has an external boundary on which condition of equation (3) is 314 specified. Using the relation that at ax = [ ir r > J {tie etc., and that -2X = m- a hii arc 1 > tt- = 0 at jL ( U ) = v ax / ax (17) we have from equation (16) M-' = f k ( [ ~ > +r£^,«r lii!; J v« \ ** a* * Jax +La> ’ ~] + r 3I^ , 3H) 1 ) f-rie , L =>* ' T^>..] ^ y i Ti civ - J o-N.dv + j M N^Nj^trj'dv V + f 1 M. ds s" L (18) Generalization of equation (18) to the whole element gives with h e = K* f ( ™ } ± , £ * fJ^L ) dx dv dl lv j J \ a x a x + 3^ if + a z a ? 7 3 v and F‘ = - a j Ni.lv + jJ UfdS V® Jse ^Cj = >1 | Ni Mj dV (19) Me With an appropriate choice of the shape function it may be stated that the total functional is equal to the sum of the functional of each element, i.e., X » z: jxe € Minimization of the total functional X over the whole region results in with summation over all elements which possess i and j as nodes. If = 0/ ^ e nodal temperatures are obtained from the set of simultaneous equation (20) {Tj = -[h]-1 F (21) which together with equation (15) provide the temperature distribution. Consequently we can express the heat flow in the element e in terms of its nodal temperatures f q„ .= ~ K e < lZ r 3 T a ax ax aT ay > = —Ke _2_ ay aT a i az ^az (22) The transient temperature distribution can be obtained by solving the set of first order linear differential equations in (20). The programming is relatively more involved and will not be pursued further in this study. Element Shape Function Two convergence criteria should be observed in the selection of element shape function £N]. First, with a suitable choice of nodal temperature {Tje, any constant value of {Tj or its derivative in the functional should 316 be able to be reached in the limit as element size der- creases to zero. Second, the temperature should be con tinuous at the element boundaries. This implies the finiteness of first derivatives in the functional JC . If the shape function is linear in spatial coordinates, these two criteria are met. In this study we shall limit the finite element method to two dimensional problems and the region of interest is approximated by a series of arbitrary triangles. For three dimensional problems or other configuration of elements, the readers are referred to the text by Zienkiewcz (1972). For a triangle with nodes at (x^,y^), (Xj,yj), and (xj,,yk) / ordered in the counterclockwise sense, the shape function is simply the area coordinates Ni = ^ai + bix + Nj - (aj + bjX + Cjy)/2 Nk = <ak + bkx + cky}/2 (23) in which 1 xi a = 1/2 1 xj yj = area of the triangle and (24) C. = X, - X . x k 3 with a . and a obtained by cyclic permutation of the inr> 3 k dices i, j, and k, etc. 317 An alternative way to derive the shape function for triangular elements is to assume a linear variation of tem perature within an element such that T = Ot + 0x + Y y (25) Solving ct, £ , and T in terms of the nodal temperature and coordinates, one immediately finds the shape function by comparing equations (15) with (2 3). Substituting equation (23) into (19) yields in two dimension e A t j = ( b; bj + C; Cj ) / A (26) To simplify the volume integral in equation (19), it is convenient to choose the origin of local coordinates at the centroid of a triangle such that (x± + Xj + xk)/3 = (y± + yj + Yk)/3 = 0 The first volume integral in equation (19) becomes in two dimension * - n ; J J (ai + b=* + d > = - a V 3 = F„j = F v® for 5 5 X d * C iy = (5 } dx d } = 0 and 0-1 - 0-j = = 2. A / 3 (28) The second volume integral in equation (19) is in two dimension 318 % j = aJ + ~ T T ^ + XJ* + + - t t * u * + » ; + v ; + k>c CS + c c b j I 2 (X;)1 ; + Xj y, + ^ *K )J (29) To evaluate the surface integral in equation (19), let the ith and j_th nodes lie on the boundary where type-2 boundary condition is specified, and set the origin of the local coordinates at ith node. The surface integral (or line integral in one dimension) thus becomes Fe . = qeb.x?/4* + qea,x,/2A si x j x j and Fe . = qeb.x^/4a (30) S3 3 3 Fe = 0 sk where a. = x . y, x 3 k b . = -y x k b = y j k and A = (1/2) x.y, 3 k a few examples on the application of the finite element method are given in Figure 61, 62, and 63. 319 Figure 61. Perturbation of heat flow around a parallelepiped with infinite length. A. Triangular elements, conductivity of shaded area to unshaded 1/4. vertical exaggeration 3. B. Isotherms. C. . Normalized heat flow. * Vertical heat flow, solid line being the local linear least square fit of the vertical heat flow. + Horizontal heat flow. Figure 63)A. 10 . 0 ISOTHERM C Figure 6'IB. 321 NORMALIZED HEAT FLOW -I 2. 0 0 _i I - \ -i - I - i Figure 61C. 322 323 Figure 62. Perturbation of heat flow due to a plane slope (14°), assuming uniform conductivity, and isothermal surface. A. Triangular elements. B. Isotherms. C. Heat flow Upper curve with Topography. Middle curve : Local least squares fitting to vertical heat flow *. Bottom : Horizontal heat flow, flowing to the left. Solid dot : Exact solution from Lachenbruch (196 8). 324 Figure 62a . ISOTHERM Figure 62B. u > to Ln •j? o ■J u_ a: aJ “ ■ Cl LU !N] g: s r . or o M tt 00 Figure 62C. u> ro < S \ 327 Figure 63. Perturbation of heat flow along profile I-I1 due to topography, bottom water temperature variation, and conductivity contrast. A. Triangular elements, conductivity of shaded area to unshaded area 1/3. vertical exaggeration 20 times, horizontal distance 1 cm is equivalent to 25 km. B. Isotherm, assuming uniform conductivity C. Normalized heat flow, assuming uni form conductivity (see caption in E). D. Isotherm, assuming conductivity ratio 1/3. E. Normalized heat flow, assuming conductivity ratio 1/3. * Vertical heat flow. + Horizontal heat flow, positive values flowing to the right, negative values flowing to the left. Upper curves: Local least squares fitting to vertical heat flow. Coarse curve with cross : Local least squares fitting to vertical heat flow. Figure 63A. 328 Cn i 3 o > \ 1 0 0 0 10 1 i j ISOTHERMS Figure 63B. 329 N O R M A L IZ E D HEAT FLOW CD CP CD CD CP CD oee Figure 63D. ISUTHERI 331 0 0 0 0 _* + ♦ + n Figure 6 3E. 332 TRACKING POLLUTION IN THE SOUTH COAST BASIN Leland M. Vaughan Alexander R. Stankunas SUMMER OF 1972 METRONICS ASSOCIATES, INC. A S ER NCO CO M PANY 3 2 0 1 Porter Drive • Palo Alto, California 9 4 3 0 4 FOREWORD This Citizen's Report is based on a study* conducted for the State of California Air Resources Board, The Technical Report contains the facts and figures. What we would like to do here is briefly describe the project in more general terms for the benefit of those of you in fire stations, schools, the U.S. Forest Service and private homes who volunteered to operate sampling devices during our tests. Through your efforts we were able to provide the large network of stations required to give an adequate descrip tion of the transport and dilution of polluted air in the South Coast Basin. You have our sincere thanks for help ing make this study possible. *"Field Study of Air Pollution Transport in the South Coast Basin, " Technical Report No. 186. TRACKING POLLUTION IN THE SOUTH COAST BASIN An air pollution study in the Los Angeles Basin was sponsored by the State of California Air Resources Board during the summer of 1972. As part of that study, Metronics Associates, Inc. of Palo Alto conducted a series of tests to track the movement of polluted air in the Basin and outlying areas. To do this, a small amount of super-fine, colored dust ("FP" air tracer) was blown in the air along the freeways and highways of three areas in the western Los Angeles Basin. Different colors were released simultaneously in Downtown Los Angeles, Torrance and Santa Ana. The tracer dust was tracked by a network of nearly eighty samp ling stations throughout the area.- These stations included sites in Greater Los Angeles, the San Fernando Valley, San Bernardino, Riverside, the San Gabriel Mountains, Palm Springs, Indio, Lancaster, Oceanside and San Diego. The tracer dust is so fine that even the slightest breeze keeps it airborne. A million particles wouldn't cover the period at the end of this sentence. Because it blows with the wind, we can label a mass of air and follow it wherever it goes. The concentration of tracer is diluted the same way that pollutants would be. So, by following the easily identified tracer, we can determine where the air goes, and by measuring the concentration of tracer we can calculate how much the original mass of air has been diluted. Thus, we can tag a particular parcel of air and watch exactly what happens to it. Page 2 The tracer dust was collected on a rotating rod coated with a special adhesive. The rod was then microscopically examined under ultraviolet light. The particles of tracer dust glow with their own bright colors, like a "black light" poster, so they are easily distinguishable from other dust on the rod. Strangely enough, a major difficulty during last year's tests was that there were not enough smoggy days. Although there were unpleasant periods, the season as a whole was not representative of a severe smog season. The temperature inversions, which usually keep the pollutants from being diluted, were not as strong or as frequent as usual, so it was difficult to arrange to measure conditions at their worst. A total of four tests were conducted . The first test was on 26-27 July 1972. In the morning, conditions were relatively stagnant with a shallow layer of marine air. By afternoon, a sea breeze started moving the tagged air inland. But, as the air moved inland, the temperature inversion was broken up by the sun heating the intermediate valleys. Tracer concentrations were low through out the eastern part of the test area. The second test was conducted on 29-30 August 1972 under relatively neutral stability conditions with considerable cloudiness. A tropical storm about 300-500 miles south of Los Angeles influenced Page 3 the whole area. The sea breeze carried some tracer into the eastern end of the Basin and the mountains, but it was diluted considerably in the deep layer of air under the inversion. The third test was conducted on 20-21 September 1972. Conditions were relatively stagnant within the surface layer. Northeasterly winds high above the ground deflected the wind and the tracer in the lower atmosphere to the east and southeast with relatively little movement to the northern portions of the Basin. The fourth test was conducted on 24-25 October 1972 under relatively stagnant conditions with considerable fog and low clouds at the beginning of the test period. Santa Ana winds developed in the San Bernardino area about 3:30 that afternoon. By evening the area was swept clean of both tracer and pollutants. The maps show an overall pattern of movement and dilution of the tagged air. The solid, dashed and dotted lines represent levels of tracer concentration averaged over the first 24 hours of a test. The arrows indicate the main pathways the air masses followed as they spread and mixed in the Basin. Note that tracer from each source spreads out in many directions. Part of this spreading on the map is due to the fact that the defined source areas are large enough to have the wind blowing in different directions within their borders. Part is due to the fact that we are showing the overall results of many hours of travel all on one map. Page 4 The complexity of these flow patterns indicate some of the problems faced in trying to predict exact pollution levels at any given place. Because our tracer material is unique, that is, there is nothing else like it in the air over Los Angeles, we can follow it much farther than the pollution itself and still be sure where it came from. The outer edges of the tracer pattern represent pollution contributions by the specific source areas, too small to be measured any other way. The tracer tests were used to compare the arrival time of various pollutants with the arrival of the tagged air. Good correlations of the morning rush hour carbon monoxide peak with high tracer concentration were typical for stations within ten miles of a release site. Correlations of tracer with ozone were best for stations much further away and much later in the day (for example, Fig. 1). It takes time for sunlight to act on automobile exhuast and other pollutants to produce ozone. While these chemical reactions are taking place, the polluted air moves away from where it was first tagged. Another use of the tracer tests was to determine how much pollution remains in the Basin from one day to the next. We released the tracer only once for each test. Any tracer dust seen the next day must have been left over from that release. If we just tried to measure pollutants we couldn't tell what was leftover and what was made new that day. Even with better than usual ventilation during Page 5 our tests last year, an average of 14% of the tracer was retained from one day to the next. For 1973 we are planning a similar series of tests and have made several improvements in the system. We will have better measurements of short-term variation in tracer levels; more stations on the Basin floor; a better weather prediction system so we can arrange to test during heavy smog periods and we will have a faster data handling system to get the results sooner. In summary, the tracer tests of 1972 showed some of the complexity of air movement patterns in the Los Angeles Basin. They showed that pollutants generated in the western Basin travel very long distances in relatively short times and contribute to the degradation of air quality outside the immediate area. For example, tracer dust released in the Torrance area in the early morning was seen in Riverside later that afternoon. However, for many types of pollutants the additional pollution contributed by the tagged areas was small compared to local sources under the conditions of these tests. We will continue our study in the South Coast Air Basin and hope to arrange our tests during more typical weather conditions. O z o n e (PPHM) FP-Particle-Minutes/Liter x I0~1 6 24 22 PASADENA 20 Downtown L.A. Tracer Ozone /_ Torrance Tracer 1600 1600 2 4 October 1972 2200 1000 25 October 1972 Figure 1 Source; Area , DOWNTOWN L.A M m M ' Source Area TORRANCI: Source Area S A N T A A N . ' ...... f i - 1 1 1 7 *0 0 U S 0 0 n ' i - s o « R & 1 Source Area S A N T A A N A TRANSPORT AND DIFFUSION MAPS / ^ ) ARROW S INDICATE MAIN PATHWAYS OF AIRFLOW. SHO W APPROXIMATE LEVELS OF TRACERCONCENTRATION AVERAGED OVER THE FIRST 24 HOURS OF A TEST.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Applications of artificial neural networks in real-time seismology and surface-wave profiling of the lithosphere beneath the Mojave Desert

PDF

Marine Geology Of The Baja California Continental Borderland, Mexico

PDF

The study of temporal variation of coda Q⁻¹ and scaling law of seismic spectrum associated with the 1992 Landers Earthquake sequence

PDF

A seismic source inversion study and non-stationary nonlinear seismic signal processing

PDF

Early Diagenesis In Southern California Continental Borderland Sediments

PDF

Petrology And Depositional History Of Late-Precambrian - Cambrian Quartzites In The Eastern Mojave Desert, Southeastern California

PDF

The Aleutian-Kamchatka Trench Convergence: An Investigation Of Lithospheric Plate Interaction In The Light Of Modern Geotectonic Theory

$Boundary integral equation method for the diffraction of elastic waves using simplified Green's functions$

PDF

Boundary integral equation method for the diffraction of elastic waves using simplified Green's functions

PDF

Composite constitutive theory for a solid earth of strong lateral viscosity contrast

PDF

Neogene to recent Naticidae (Mollusca : Gastropoda) of the eastern Pacific

PDF

The Geochemical Cycle Of Mercury And The Pollutional Increment

PDF

Sedimentology And Pleistocene History Of Lake Tahoe, California - Nevada

PDF

Ultrastructure Of Tests Of Some Recent Benthic Hyaline Foraminifera

PDF

Oceanography And Late-Quaternary Planktonic Foraminifera, Southwestern Indian Ocean

PDF

Full-scale experimental and analytical studies on high -strength concrete columns

PDF

Distribution And Transport Of Suspended Matter, Santa Barbara Channel, California

PDF

Carbon isotopic heterogeneity and limited oxygen isotopic homogeneity in the cretaceous Cucamonga terrane, southeastern San Gabriel Mountains, California

PDF

The relationship of stress to strain in the damage regime for a brittle solid under compression

PDF

Synoptic Surf Zone Sedimentation Patterns

PDF

Preservation Of Fossil Fish In The Miocene Monterey Formation Of Southern California

Asset Metadata

Creator Lee, Tien-Chang (author)

Core Title Heat Flow And Other Geophysical Studies In The Southern California Borderland

Degree Doctor of Philosophy

Degree Program Geological Sciences

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag geophysics,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Henyey, Thomas L. (committee chair), Davis, Gregory A. (committee member), Teng, Ta-Liang (committee member), Wagner, William G. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-687481

Unique identifier UC11356463

Identifier 7417357.pdf (filename),usctheses-c18-687481 (legacy record id)

Legacy Identifier 7417357.pdf

Dmrecord 687481

Document Type Dissertation

Rights Lee, Tien-Chang

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA