june 1999 hydrodynamics and overpressure mechanisms in the sacramento...

American Journal of ScienceJUNE 1999

HYDRODYNAMICS AND OVERPRESSURE MECHANISMSIN THE SACRAMENTO BASIN, CALIFORNIA

BRIAN J. O. L. MCPHERSON* and GRANT GARVEN**

ABSTRACT. Subsurface pore fluid pressures exceeding 70 percent of lithostaticpressure are observed in large areas of the Sacramento basin and adjacent CoastRanges, California. Basin- or crustal-scale processes are required to explain theregional distribution of overpressures. Several plausible mechanisms of overpres-suring are (1) depositional compaction by the Miocene and younger stratarecently deposited, (2) aquathermal expansion under high heat flow conditions,and (3) horizontal or lateral tectonic compression. The latter mechanism issuggested by observed deformation patterns in the Sacramento basin. Through-out the geologic history, east-west compressive stresses by eastward subduction ofthe Juan de Fuca plate, subsequent tectonic wedging, right-lateral tectonism(primarily in southern part of basin), and east-west extension in the Basin andRange together created a primarily compressive stress regime. Such a stressregime has likely contributed to the cause and maintenance of overpressuresobserved in the region.

We developed two-dimensional numerical modeling experiments to testthese hypotheses. Results of the experiments suggest that aquathermal expansionis negligible as an overpressure mechanism. Sedimentary compaction is a pos-sible mechanism but only for locally high sedimentation rates or for unusuallylow permeability and/or unusually high compressibility under lower sedimenta-tion rates.

Finally, model simulation results suggest that tectonic compression is aplausible, ongoing mechanism creating and maintaining high fluid pressures inthe Sacramento basin. We speculate that other processes such as chemicaldiagenesis may be contributing to overpressures, but probably only on a localscale and of secondary importance.

INTRODUCTION

Most, if not all, sedimentary basins throughout the world exhibit some ‘‘anoma-lous,’’ or non-hydrostatic fluid pressures, particularly overpressures (Deju, 1973; Brad-ley, 1975; Spencer, 1987; Domenico and Schwartz, 1990; Hunt, 1990; Harrison andSumma, 1991; Fertl, Chapman, and Hotz, 1994; Neuzil, 1995; Hart, Flemings, andDeshpande, 1995; McPherson, 1996). Overpressures typically are defined as pore fluidpressures that significantly exceed hydrostatic pressure. These high pressures occurbecause the pore fluid bears the weight of the overburden to a significant extent. A lowpermeability environment and conditions that reduce available pore space or increasefluid volume are necessary for overpressures to occur and be maintained.

We considered overpressures observed on a regional scale throughout the Sacra-mento Valley and the Coast Ranges of California (Fig. 1). Overpressuring is common inbasins with sedimentary sequences similar to those of this area, where marine unitsunderlie rocks of higher permeability (Garcia, 1981). However, the mechanisms respon-sible for overpressures are variable. While several processes of overpressure generationmay be active in the Sacramento basin, Berry (1973) proposed a tectonic cause. We refer

* Hydrology Program, Department of Earth and Environmental Science, New Mexico Institute ofMining and Technology, Socorro, New Mexico 87801

** Department of Earth and Planetary Sciences, The Johns Hopkins University, Baltimore, Maryland21218

[AMERICAN JOURNAL OF SCIENCE, VOL. 299, JUNE, 1999, P. 429–466]

429

to his hypothesis as the ‘‘tectonic vise hypothesis’’ of overpressure generation. Otherstudies since 1973 refer to this as ‘‘tectonic compaction.’’ It is very simple in concept:Basin and Range expansion pushing westward and slightly convergent motion of thePacific plate effectively squeeze the Great Valley and increases fluid pressures. Figure 2 isa schematic drawing of the Sacramento Basin with an interpretation of compressionalforces (block arrows) in a transpressional tectonic setting. Areas of observed overpres-sures in the Great Valley Sequence are also indicated on figure 2. Berry’s (1973) originalinterpretation extends the overpressured zone into the Franciscan basement, but theactual extent of this overpressuring is not clear.

Fig. 1. Location map of the Sacramento basin, California. Digital relief map of California (left) shows theelongate Great Valley of California in the center of the state and extending for much of its length; theSacramento basin comprises the northern half of the Great Valley. The right-hand figure is a simplifiedgeologic map of late Mesozoic arc-trench belts exposed within and juxtaposed to the basin (after Ingersoll,1982, and Moxon, 1990).

Fig. 2. Schematic cross section showing general structure at the regional scale. Also shown are thecompressional forces (block arrows) that may occur in a transpressional tectonic regime. Observed overpres-sures inferred by Berry (1973) are also indicated.

B. J. O. L. McPherson and G. Garven—Hydrodynamics and430

High fluid pressures of tectonic origin are commonly associated with major oro-genic belts, a prime example being the Andes (Deju, 1973; Villeges and others, 1994).The Sacramento basin originally formed as an Andean type arc-trench system (Dickin-son and Seely, 1979; Moxon and Graham, 1987), suggesting significant horizontaltectonic compression. More recent tectonism in the Sacramento basin is compressionalin nature, resulting in active seismicity (Wong, Ely, and Kollmann, 1988), activeshortening accommodated by thrust faults and folds, and coeval strike-slip faulting in thenorthern Coast Ranges (Unruh, Loewen, and Moores, 1995; Fox, 1983). Shorteningalong the geomorphic boundary is driven by a component of Pacific-North Americanplate motion normal to the plate boundary (Mount and Suppe, 1987; Zoback and others,1987; Wentworth and Zoback, 1989; Wakabayashi and Smith, 1994). The geomorphicexpression and structural relief associated with the boundary are similar along a greatportion of the range front (Wakabayashi and Smith, 1994), suggesting similar shorteningor deformation rates overall. Geodetic surveys of Argus and Gordon (1991) and Gordonand Argus (1993) suggest that the maximum shortening rate normal to the Pacific-NorthAmerican plate boundary in central California is approx 2 to 3 mm/yr. Given theevidence of horizontal basin shortening, tectonic stresses cannot be ignored as a potentialoverpressure mechanism. However, other mechanisms may cause or influence high fluidpressures, including depositional compaction and aquathermal pressuring (Bredehoeftand Norton, 1990).

The purpose of this study is to evaluate the origin of overpressures observed in theSacramento basin. We considered several different mechanisms, including tectoniccompression, using numerical model simulations.

GEOLOGY AND HYDROLOGY

Structure, stratigraphy, and general geologic history.The Sacramento basin in northernCalifornia originated as a forearc basin, situated in the arc-trench system of the westernU. S. convergent margin. The elongate, asymmetric, sediment-filled valley comprises thenorthernmost half of the Great Valley of California (fig. 1). The basin is bounded on theeast by the Sierra Nevada Mountains, on the west by the Coast Ranges and Franciscansubduction complex, on the north by the Klamath Mountains, and on the south by theStockton Arch, which marks the northern edge of the San Joaquin basin. The area of thebasis is approx 13,000 square km, and it extends 240 km in length (from northwest tosoutheast) and ranges from 50 to 75 km in width (Hull, 1984). As much as 16 km ofsediment, from Jurassic to the Holocene in age, record the progression from a marine toa terrestrial depositional environment in the basin fill (Page, 1986).

The Sierra Nevada ophiolitic basement slopes gently westward beneath the basin(Harwood and Helley, 1987; Godfrey and others, 1997). Harwood and Helley (1987,plate 1) published a structure-contour map of basement rocks in the Sacramento basin.Based on the interpretations of Cady (1975) and Harwood and Helley (1987), weindicate the possible location of the boundary between Sierran basement rocks to theeast and adjacent ophiolitic ocean crust to the west in map view on figure 1. Based onrecent gravity and magnetic data, Godfrey and others (1997) provide a meaningfulinterpretation of the deep structure underneath and adjacent to the Great Valley ofCalifornia, including the location of the contact between the Sierra Nevada Block andthe oceanic, ophiolitic crust. This study focuses on Jurassic and younger sedimentarystrata above the ophiolitic crust and Sierran basement rocks.

To illustrate a general interpretation of the origin and evolution of the basin, we usea series of schematic cross sections in figure 3, derived from the work of Ingersoll (1982,1988). In the mid-Jurassic (top), the formation of the Coast Range ophiolite behind aneast-facing intraoceanic arc began during early subduction along what is now northernCalifornia. This first major uplift of the Nevadan mountains (precursor to the currentSierra Nevada) provided the supply of Jurassic and Cretaceous sedimentary fill. It is mostlikely that the basin filled as subduction progressed and the volcanic arc migrated

overpressure mechanisms in the Sacramento basin, California 431

eastward throughout the Jurassic and Cretaceous. The sequence of cross sections shownin figure 3 are general (Ingersoll, 1982, 1988) and are intended to show how thesedimentary section of the Sacramento basin developed and how the tectonic setting wasconducive to developing an east-west compressional stress regime across the basin. Thegross details are consistent with, but not drawn from, more recent interpretations

Fig. 3. Schematic representation of the evolution of the Sacramento basin. Shown are cross sections ofnorthern California: formation of Coast Range ophiolite behind an intraoceanic arc (top) through terminationof Great Valley forearc by conversion to a transform margin (bottom). During evolution of the forearc basin,tectonic forces compress the basin from both east and west. Adapted from Ingersoll (1988).


(Dickinson and others, 1996; Godfrey and others, 1997) of regional and deep structure ofthe Great Valley. The interested reader is referred to several papers for discussion of thestructure, stratigraphy, and geologic history, including Redwine (1972), Ingersoll, Rich,and Dickinson (1977), Almgren (1978), Nilsen and McKee (1979), Haggart and Ward(1984), Harwood and Helley (1987), Dickinson and others (1996), and Godfrey andothers (1997).

Figure 4 depicts a more detailed cross section of profile A-A�, an east-west transectfrom the Coast Range to the Sierra Nevada. Figure 5 is a generalized stratigraphic sectionfor Sacramento basin strata. Cretaceous and Tertiary formations consist of alternatingsandstone and shale units (figs. 4 and 5). The thick upper Cretaceous sequence is marinein origin, whereas the overlying Tertiary strata originate from both marine and terrestrialdepositional environments.

The generalized stratigraphic column (fig. 5) also shows the distribution of indi-vidual subgroups of the Cretaceous and Tertiary units. The chronostratigraphic units aredemarcated and described by Moxon (1990) and further discussed in Williams (1993).Units labeled NG, EP, and UK play a fundamental role in our analysis of compactionprocesses and are discussed in more detail in a subsequent section. The stratigraphy ofunderlying units JK and LK play a less significant role in our hydrogeologic analyses.Unit JK ranges from Tithonian to Valanginian in age and consists of gravity-flowdeposits of mudstone and siltstone with local layers of conglomerate and sandstone(Moxon, 1990; Williams, 1993). Unit JK is interpreted as a submarine fan complex thatprograded southward over basin plain mudstones. Unit LK ranges in age from Hauteriv-ian to Albian and consists primarily of mudstone and thin-bedded sandstone with somediscrete conglomerate layers (Moxon, 1990). Unit LK is interpreted as localized distalsubmarine fans (Moxon, 1990).

General hydrogeology.Two broad hydrologic generalizations can be made about thesediments in the Sacramento basin (figs. 4 and 5). First, the deeper sediments, mostlymarine in origin, are of relatively low permeability whereas the shallow strata arenon-marine units with comparatively higher permeability (Page, 1986; Bertoldi, Johnson,

Fig. 4. Generalized structural cross section of profile A-A� (see location of profile on fig. 1). Shaded regionsindicate areas of overpressures observed along the cross section. Vertical exaggeration of cross section approx20. Abbreviations used on figure: UDSV � undifferentiated strata and volcanics, UDS � undifferentiatedmarine and nonmarine strata, MF � Markley Formation, MCF � markley Canyon Fill, WI � WintersFormation, SS � Sacramento Shale, FO � Forbes. Unit thicknesses and faults shown interpreted from well logdata by California Division of Oil and Gas (1997 written communication).


Fig. 5. Generalized stratigraphic section of Sacramento basin strata.


and Evenson, 1991). Second, sedimentary units lap onto the west-sloping basement rockof the Sierra Nevada batholith while folds and thrusts deform the western sediments andthe Northern Coast Ranges. This general structure defines two regional aquifer systems,one shallow and one deep.

Shallow aquifers in the Sacramento basin hold the primary supplies of fresh waterfor drinking and agricultural needs. As such, they have been extensively studied (Page,1986; Williamson, Prudic, and Swain, 1989; Belitz and Heimes, 1990; Bertoldi, Johnson,and Evenson, 1991). However, these shallow aquifers make up less than 10 percent of thesedimentary section throughout the region (Bertoldi, Johnson, and Evenson, 1991). Inthe deepest portions of the basin, the less permeable marine facies extend to 16 km depthand are largely unstudied (Page, 1986). While water resource studies provide most of theshallow data, most information from depth has come by way of petroleum exploration.Deep basin fluids tend to be saline and have signatures suggesting they have remained inplace for a substantial period of time, possibly since deposition (Unruh, Loewen, andMoores, 1995; Davisson, Presser, and Criss, 1994). The fresh water/saline water inter-face generally coincides with the boundary between marine and continental deposits(Bertoldi, Johnson, and Evenson, 1991). Isotopic signatures of perennial springs in thesouthwestern part of the basin (Unruh, Loewen, and Moores, 1995), in addition toisotopic evidence for mixing of meteoric and basinal groundwater (Davisson and Criss,1993; Davisson, Presser, and Criss, 1994), suggest hydrologic communication betweenthe shallow and deep systems. Additionally, Unruh, Loewen, and Moores (1995),Davisson, Presser, and Criss (1994), and Melchioore, Criss, and Davisson (1999) discussa possible origin of these perennial springs: overpressures, probably created by activetectonic compression, driving discharge. Overpressures are observed throughout thebasin and the adjacent Coast Ranges (Berry, 1973).

Observed overpressures.Local overpressures (amount of pressure above hydrostatic) inthe Sacramento basin range from 0.06 to 23.8 MPa, with an average of 8.1 MPa. Forreference, 10 MPa of overpressure is roughly equivalent to an increase in hydrostaticpressure produced by adding 1 km thickness of freshwater on top.

Overpressures are observed over a substantial range of depths. The shallowestoverpressures are observed near the top of the marine units in the Sacramento basin(Garcia, 1981). Additionally, inasmuch as fluids from depth are expelled from ridgetopsin the Rumsey Hills suggests that overpressures locally extend to the surface (Unruh,written communication, 1999). The shallow depth of overpressuring also is evident inour compilation of overpressure data, consisting of 171 pressure-depth values. The data,derived from oil and gas well drill stem tests from 171 different petroleum wells withinthe basin, were provided by John D. Bredehoeft of the United States Geological Survey(1995, written communication). The greatest density of observed overpressures occursbetween 1500 to 2800 m depth (fig. 6).

The dark patches drawn on the structural-stratigraphic cross section (fig. 4), repre-sent 30 localized areas of observed overpressure for profile A-A�, based on data takenfrom 12 wells. It is not apparent from the data of figures 4 and 6 that overpressures arecorrelated to any specific depth or lithology. We also examined pressures in specificstratigraphic horizons (formations) and did not find any systematic correlation ofoverpressures.

The areal (map view) distribution of fluid pressures provide more information aboutthe possible causes of overpressures. Berry (1973) observed that overpressures in thebasin are regional in nature but concluded no obvious correlation to particular structureswithin the basin. Our overpressure data are perhaps more in numbers and show bettersome systematic correlations with geographic features. In map view (fig. 7), our dataillustrated that overpressures are widespread through much of the basin but are not


ubiquitously or uniformly distributed. Drilling activity is minimal along the perimeter ofthe basin, and thus basin margin fluid pressure information is scarce. Otherwise, fordepths less than 5 km, our pressure data suggest that overpressures are regional in nature,and some correlation with geographical area may be inferred—we discuss these correla-tions below.

Given the regional nature of the observed overpressures, basin-wide processes seemto be the most relevant mechanisms to explore.

OVERPRESSURE MECHANISMS IN THE SACRAMENTO BASIN

Casting the tectonic history in the context of overpressures.Understanding the tectonichistory of the basin is critical to determining the role of compressional tectonic forces inthe hydrodynamic regime. The objectives of this study include the answers to severalquestions: Does a tectonic ‘‘vise’’ act on the basin as Berry (1973) hypothesized? Arecompressional tectonic forces the sole cause of overpressures, do they play a secondaryrole to other overpressure mechanisms, or do they influence the hydrodynamic regimeat all? We now cast the tectonic history of the basin in the context of overpressures todevelop the premise of the numerical model simulations.

Evidence of tectonic deformation trends.The Sacramento basin sedimentary fill appearsto be an independent block that has accommodated relatively small scale contractionalstrain in response to both large scale right lateral transform tectonism in the San AndreasFault zone to the west and major east-west crustal extension in the Basin and Range to theeast (Harwood and Helley, 1987; Wong, Ely, and Kollmann, 1988).

Fig. 6. Observed pressure data plotted versus depth. Also indicated on the plot are reference lines ofhydrostatic pressure and lithostatic pressure.


Harwood and Helley (1987) evaluated patterns of late Cenozoic deformationthroughout large areas of the basin. They identified deformation structures (folds andfaults) and interpreted the most likely stress regime that formed them. Harwood andHelley (1987) and workers they cited also estimated the time of deformation, using datedvolcanic units to constrain absolute ages of deposition and tectonic events. Historiclow-magnitude earthquakes were associated with specific structures and interpreted asevidence that strain currently is being accommodated by those structures (Marks andLindh, 1978; Bolt, 1979; Harwood and Helley, 1987).

Harwood and Helley (1987) inferred that patterns of deformation were timetransgressive, and grouped structures with similar kinematic patterns into 6 structuraldomains (fig. 8). Most late Tertiary and Quaternary faults and folds appear to haveformed in an east-west or northeast-southwest compressive stress regime. Additionally,

Fig. 7. Contour map of observed overpressures (pressure above hydrostatic in MPa) within the Sacra-mento basin. Contour levels include 0 to 25 MPa, with an interval of 5 MPa. Symbols indicate location of wellsfrom which pressure data were taken. Inset map of California indicates the coverage of the contour map withinthe basin. Background base map adapted from Harwood and Helley (1987).


Harwood and Helley’s (1987) structural domains are progressively younger northwardin the basin, with the exception of the Dunnigan Hills domain, showing that thedeformation was time transgressive. As Harwood and Helley (1987) concluded, north-ward progression of deformation indicates a northward-migrating compressive stressregime associated with the northward migration of the Mendocino triple junction (Silver,1971; Atwater and Molnar, 1973; Dickinson and Snyder, 1979; Harwood and Helley,1987).

Horizontal crustal shortening indicates a compressive stress regime. Active crustalshortening in this region is inferred by patterns of seismicity (Wong, Ely, and Kollmann,1988), tectonic-geomorphic development, and high fluid pressures at shallow depths,attributed to tectonic compression (Unruh and others, 1992; Unruh, Loewen, andMoores, 1995; Davisson, Presser, and Criss, 1994).

Uplift, folding, and faulting of the Pliocene Tehama Formation in the DunniganHills region (fig. 8) represents propagation of uplift and crustal shortening eastward intothe southwestern Sacramento basin in middle to late Pleistocene time (Unruh, Loewen,and Moores, 1995). Unruh, Loewen, and Moores (1995) interpret modern topographicand structural relief in the Dunnigan Hills domain as the result of renewed movement on

Fig. 8. Structural domains in Sacramento basin. Numbers under domain name provide age or duration ofdeformation. Adapted from Harwood and Helley (1987).


the Rumsey Hills thrust fault and growth of an overlying fault-propagation fold.Structural relief on the base of the Pliocene Tehama Formation may be used to estimatethe total Quaternary slip on the Rumsey Hills thrust and a range of average quaternaryslip rates. Assuming the Unruh, Loewen, and Moores (1995) structural model (fig. 2 ofUnruh, Loewen, and Moores, 1995), approx 1.6 km of slip is necessary to produce 800 mof relief on the base of the Tehama Formation. The east-west width of the east-dippingpanel also gives a minimum estimate of total post-Tehama slip. Unruh, Loewen, andMoores, 1995 determined that post-Tehama slip ranges between 1 and 1.6 km, and thusthe average Quaternary slip rate on the Rumsey Hills thrust ranges between 1 and 3.5mm/yr. This corresponds to a horizontal shortening rate of about 0.9 to 2.6 mm/yr(Unruh, Loewen, and Moores, 1995).

Other estimates of slip rates and shortening have been published for the basin (table1). The Harwood and Helley (1987) data tabulated are near-vertical to vertical move-ment on thrust faults. They indicate a compressional stress regime but do not constrainrates of horizontal shortening. The corresponding rates (mm/yr) we calculated for theHarwood and Helley (1987) faults are the minimum values, calculated by assuming thedeformation developed uniformly over the longest possible duration of time. The actualslip/shortening rates in most instances may be higher than our calculated minima,inasmuch as we probably overestimated the maximum duration of deformation. Har-wood and Helley (1987) publish slip and slip rates for several other faults in the basin (notlisted in table 1). The last three horizontal shortening rates listed in table 1 correspond tointegrated motion across larger areas than the Harwood and Helley (1987) data. TheArgus and Gordon (1991) and Gordon and Argus (1993) rates are based on currentgeodetic observations, constraining the direct convergence rate between the Pacific Plate

TABLE 1

Selected slip and tectonic convergence rate information for the Sacramento basin

ReferenceName of faultor structure

Dip of fault ordirection ofshortening

Amountand/or timingof offset (m)

Estimatedminimum

slip rate (mm/yr)

Harwood andHelley (1987)

Willows Fault 74° east (reversefault)

475 m (pre-Pliocene)

0.014 (near verticalslip)


Corning Fault north-striking,steeply east-dip-ping (reversefault)

�1500 m (lateCretacene); post-Pleistoceneoffset is small



Chico MonoclineFault

north-striking,steeply east-dip-ping (reversefault)

365 m (between2.6 and 1.1 my)



Battle Creek Fault north-striking,steeply east-dip-ping (reversefault)

45 to 440 m (�0.5my to recent)


Argus and Gordon(1991); Gordonand Argus(1993)

Pacific–NorthAmerican PlateBoundary(including Sacra-mento basin)

Overall conver-gence or short-ening normal toplate boundary

Current conver-gence rate

�5 (Argus andGordon, 1991);1 to 3 (Gordonand Argus,1993) (horizontalconvergence)

Unruh, Loewen,and Moores(1995)

Rumsey HillsThrust

Thrust fault mod-erately dippingto east

1600 m (Quater-nary)

0.9 to 2.6 (horizon-tal shortening)

Wakabayashi andSmith (1994)

Fold and thrustbelt along entireCoast Range–Great ValleyBoundary

Convergencenormal to NorthAmerican–Pa-cific Plateboundary

Current conver-gence rate

1.8 to 8 max; 1.8 to3 likely/pre-ferred range(horizontalshortening)


and the North American Plate (or Sierra Nevada—Great Valley microplate). Similarly,the Wakabayashi and Smith (1994) data pertain to the larger-scale plate boundaryconvergence rate. The Unruh, Loewen, and Moores (1995) data correspond to severalfaults in the southern half of the Sacramento basin. All these data indicate that acompressive stress regime has acted on the basin during different periods of its geologichistory. All references listed in table 1, in addition to other work (Wong, Ely, andKollmann, 1988; Davisson, Presser, and Criss, 1994; Unruh, Loewen, and Moores, 1995)suggest that parts of the basin are under tectonic compression today.

Correlating overpressures with structures and geographic areas.The contour map ofoverpressures in the Sacramento basin (fig. 7) is superimposed on an index map of localstructures, towns and surface water features (redrawn from fig. 1 of Harwood and Helley,1987). In general, high pressures are observed in the very northwestern portion of thebasin (approx 40° latitude), in the central part of the basin just northwest of Sutter Buttes(approx 39.2° latitude), and in the south-southwestern end of the basin (approx 38°latitude).

The high pressures observed in the northwestern and south-southwestern portionsof the basin may correlate to the recent tectonic deformation described by Harwood andHelley’s (1987). As discussed previously, structural deformation patterns are interpretedto young northward in the basin. The exception to this sequential pattern is theDunnigan Hills domain (south-southwestern; fig. 8), where deformation is dated between0 and 1 Ma. This deformation and the Battle Creek domain deformation (northwestern;fig. 8), dated between 0 and 0.5 Ma, are the only domains where deformation is clearlyrecent in age, and recent earthquake activity has occurred in the Chico and Sacramentodomains (Unruh, written communication, 1999). Thus, the high pressures observed inthese areas may be associated with active tectonic compression.

One can also correlate local high pressures to specific structures. For example, thehigh pressures observed at approx 40° latitude are proximal to the Orland Buttes(southern boundary of Battle Creek domain, fig. 8). Additionally, high pressures areobserved between the Sutter Buttes and the adjacent dome near Colusa to the west (figs.1 and 7). Although the Sutter Buttes area is somewhat close to the Dunnigan Hillsdeformation domain, active deformation of the Buttes ceased about 1.4 Ma (Harwoodand Helley, 1987). Thus, this is an area where other mechanisms, such as aquathermalexpansion and depositional compaction, may play an important role in the developmentof high fluid pressures. While high temperature gradients are not observed in the SutterButtes area today (Sass and others, 1981), it is possible that higher heat flow in the recentpast may have influenced fluid pressures by fluid density decreases and buoyancy-drivenflow.

In the southern portion of the basin, overpressures are observed in the Dunnigan,Rumsey, and Montezuma Hills. These features lie in the Harwood and Helley’s (1987)Dunnigan Hills deformation domain (fig. 8). This area in particular is underlain by activethrust faults (Unruh, Loewen, and Moores, 1995), similar to much of the San Joaquinvalley to the south. Many workers (Bloch and others, 1993; Unruh, Loewen, andMoores, 1995; Wong, Ely, and Kollmann, 1988) discuss blind fold and thrust systemsrelated to transpressional plate boundary kinematics. Active horizontal shortening (table1) and observed seismicity in this area clearly indicate a horizontal compressive stressregime, possibly prone to overpressures.

In summary, deformation patterns in the Sacramento basin indicate compressivetectonism. Such a stress regime may contribute to the cause and maintenance of highfluid pressures observed in the region.

THEORY OF GROUNDWATER OVERPRESSURES

The forces driving subsurface fluid migration in sedimentary basins are attributableto gradients in topographic elevation, gradients in temperature and chemical potential,changes in porosity associated with deformation, and the generation of pore fluids by


chemical reactions and phase transformations. Garven (1995) reviews the physiochemi-cal coupling between these forces, many of which can at best be only approximated bymathematical models based on simplified hydrogeologic theory and estimated materialparameters for sedimentary formations. Bredehoeft and Norton (1990) enumerate themechanisms affecting pore pressures in hydrothermal systems, and Neuzil (1995)provides a detailed review of the history of research related to overpressures anddescribes how they are best characterized in a hydrodynamic framework.

Departure from hydrostatic conditions is common in all subsurface environments,and so-called ‘‘abnormal pressures’’ are actually quite normal. Overpressures andunderpressures occur naturally as equilibrium phenomena for topographically-drivengroundwater flow, osmotic flow, and density-driven flow systems, because they adjustover time to their geologic and hydrologic framework. On the other hand, disequilibrium-type overpressures are caused by ongoing geologic processes, not adjusted to theirframework. These include processes such as the compaction and diagenesis taking placeduring sediment burial, hydrocarbon generation during thermal maturation, fluid produc-tion associated with the transformation of clay and metamorphic mineral phases,thermal expansion and fluid degassing with intrusion of magma, and changes in porevolume caused by crustal deformation. All of these processes have the effect of creatingdistributed fluid sources or sinks, the magnitude of which can be quantified as ‘‘geologicforcings’’ that can be incorporated into the partial differential equations for fluid flow anddeformation in porous media.

In the context of the governing equation for groundwater flow in a porous medium,the geologic forcing appears as a combined source/sink term � (after Neuzil, 1993):

�� · (�q) � Ss

�h

�t � (1)

where � is fluid density, q the Darcy velocity vector, h � p/�g z, the hydraulic head, ppore pressure, g the gravity constant, Ss specific storage, and � the geologic forcing thatincludes the effects of stress changes and diagenesis/metamorphism on porosity, theeffects of thermal transients on pore and fluid volumes, and actual fluid volume sourcesor sinks related to chemical reactions. The specific storage is represented by compress-ibilities of the fluid and pore skeleton. It can be defined rigorously using bulk modulii asintroduced in Biot’s Theory for linearly elastic porous media (Bear, 1972). Eq (1) can besolved analytically for many applications in geotechnical engineering, but numericalsolutions are increasingly used in geologic applications because of the long time scales,permeability heterogeneity, inelastic rheology of sediments and rocks, and nonlineardependency of porosity, specific storage, and hydraulic conductivity (or permeability) oneffective stress (Person and others, 1996). Additional partial differential equations can bewritten to define the force balance and stress equilibrium within the deforming porousmedia, from which stress and strain rates can be predicted for elastic or plastic rheology.For example, the force balance equation takes the form:

� · � �F (2)

where is the internal stress tensor for the poroelastic medium, and F is the force vectorwhich can include the body force of gravity and the external loads imposed by tectonicsand loading by sedimentation. Constitutive rheological equations are also needed torelate the stress tensor to strain rate that includes the effects of pore pressure, such asthose formulated in Biot-type proelasticity or Mohr-Coulomb behavior with plasticity(Ord and Oliver, 1997). The interested reader is referred to the works of Lee (1968), Riceand Cleary (1976), van der Kamp and Gale (1983), Green and Wang (1990), or Ge andGarven (1992) for additional discussion of these equations. In many studies of sedimen-tary basins, a rigorous representation of multidimensional deformation is bypassed by


assuming the strain history is known a priori and unidirectionally (Bethke, 1985) ordefined in a kinematic fashion (Screaton, Wuthrich, and Dreiss, 1990; Waltham andHardy, 1995; Wieck, Person, and Strayer, 1995).

The full derivation of eq (1) is provided by Neuzil (1993). Comparable derivationscan be found in many papers, including those by Domenico and Palciauskas (1979),Walder and Nur (1984), Shi and Wang (1986), and Bethke and Corbet (1988) forcompacting sedimentary basins in both Eulerian and Lagrangian coordinate systems.Neuzil (1995) conducted a nondimensional analysis of (1) to show that overpressures arecontrolled by the strength of the forcing � (magnitude of the source/sink terms), the sizeof the flow domain L, and the effective hydraulic conductivity of the regime K. Definingthe dimensionless geologic forcing as �D � (L/K)�, strong disequilibrium overpressuresform whenever �D � 1, implying � � K/L. Neuzil’s analysis confirmed the longrecognized observation that overpressures are favored in low-permeability environ-ments, yet the domain size and strength of geologic forcing are equally importantcontrols.

It is important also to recognize that eq (2) is fundamentally based on poroelasticitytheory, and therefore limited in theory to linearly elastic, fully saturated media experienc-ing small strains. Processes in sedimentary basins, involving burial compaction, lateralcontraction, and diagenesis, clearly involve large three-dimensional strains and nonlin-ear elastic-plastic deformation over long geologic time scales (Luo and Vasseur, 1995).Nevertheless the adoption of linear poroelasticity theory as a first approximation doesnot greatly impair our analysis of overpressuring in sedimentary basins or significantlyaffect the hydrogeologic results, and simplifications can be relaxed when necessary. Forexample, Ge and Garven (1994) allowed for plastic deformation along thrust faultswithin a 2-D poroelastic basin to quantify overpressures and fluid expulsion in the RockyMountains thrust belt, while Gordon and Flemings (1998) allowed for thermal expansionand clay dehydration effects within a 1-D poroelastic basin to simulate compaction in theU. S. Gulf Coast basin. In fact, almost all hydrogeologic models constructed to-date formodeling processes such as compaction are based on porosity-depth relations thatintegrate the long-term history of deformation. These relations therefore avoid the needfor more explicit equations (usually unknown) relating deformation, effective stress, andtime with coupled geochemical and thermal processes. Empirical porosity-depth rela-tions appear to have worked successfully in hydrogeologic research, and ‘‘effective’’compressibility parameters can be used to include complex rheology and diageneticeffects by allowing the parameters to vary through space and time (Neuzil, 1993).Furthermore, numerical models minimize the problem of large strain by adopting smalltime steps. Strain during a single time step will generally be small, and thereforedeformation can be reasonably approximated by infinitesimal-strain poroelastic theory.We adopted this approach in our analysis of overpressures for the Sacramento basin.

HYDROGEOLOGIC MODELING OF THE SACRAMENTO BASIN

Overpressures in the Sacramento basin occur on a regional scale, suggesting a largescale process is needed to explain the magnitude and distribution of pore pressures.Equilibrium phenomenon such as topography-driven flow are unfavorable, because theelevation heads associated with the Coast Ranges and Sierra Nevada are too small andhydraulically isolated (Berry, 1973). Given the geologic history of the basin, activeprocesses such as sedimentary compaction and geologic forcing by lateral crustaldeformation are probably most important with respect to overpressure development.Our objective is to test the relative roles of these two overpressuring mechanisms in theSacramento basin. Our approach is based on developing and comparing hydrodynamicmodels with numerical solutions of eq (1) subject to approximate boundary conditions,effective compressibility and permeability parameters, and estimated rates of sedimenta-tion and lateral crustal shortening. The hydrodynamic models are heuristic and onlysemi-quantitative in that they are restricted to two-dimensional flow and strain rate fields.


Instead of trying to reproduce the exact hydrogeology of the basin, we use simplifiedboundary conditions to help isolate and understand different processes. Many of thephysical parameters are not regionally constrained, but we feel they provide an impor-tant first step in understanding a complex hydrogeologic problem. Diagenetic processessuch as clay dehydration, sandstone cementation, pressure solution, and petroleumgeneration may contribute to the overpressure history in the basin, although our sense isthat they are of secondary importance. For example, Wilson, Garven, and Boles (1999)studied the San Joaquin Basin, just south of our study area, and found that claydehydration had a major effect on pore-fluid salinity, but only a minor effect, incomparison to sediment loading, on the basin’s overpressure history. Overpressuremechanisms such as aquathermal pressuring are thought to be less important, based onobservations from other sedimentary basins (Domenico and Palciauskas, 1979; Shi andWang, 1986; Luo and Vasseur, 1992; Gordon and Flemings, 1998). Thermal effects,however, may have been important locally near magmatic intrusions such as near SutterButtes, and we address this aspect later in the paper.

Of all the diagenetic processes that could have operated in the basin, petroleumgeneration and accumulation may have had the greatest effect on overpressuring. Forexample, in the Uinta Basin of Utah, McPherson (1996) determined that oil generation isthe primary source of observed overpressures. If one superimposes maps of observedoverpressures and oil production in the Uinta Basin, a clear correlation between oilsource beds and high fluid pressures is evident (McPherson, 1996). Our Sacramentobasin overpressure data reflect some coincidence of overpressures with gas production inseveral places. Oil generation has a more profound effect on fluid pressures than does gasgeneration, but the coincidence of gas and high pressures is still interesting. Unfortu-nately, quantitative analysis of these geochemical processes for the Sacramento basin isbeyond the scope of our hydromechanical study, and we must defer definitive judgmenton their role for a later communication.

Numerical formulation of the hydrodynamics.The hydrodynamics of sedimentationcompaction during burial and lateral crustal compression are analyzed with two existingand published numerical codes. The former is based on an integrated finite differencecode while the latter is based on a finite element code. Both models solve the fluid flow eq(1) in two-dimensional profiles to predict the temporal evolution of pore pressures. Thesenumerical models are used to estimate the generation and longevity of overpressuresgenerated by vertical compaction during sedimentation and by subsequent horizontalloading of the basin by the external transpressional plate boundary and Basin and Rangeextension.

Basin subsidence.Sediment compaction is mostly a mechanical process resulting inthe reduction of porosity by increases in vertical stress due to loading on the sedimentarypile. Lateral deformation is commonly ignored, and many workers have followed thenumerical approach of Bethke (1985) and Shi and Wang (1986), where the fluid flow andheat flow domains are restricted to a vertical cross section which grows with sedimenta-tion. We followed a similar approach using the numerical algorithm of McPherson(1996), a model of basin evolution utilizing the TOUGH2 code developed by Pruess(1991). TOUGH2 solves the coupled equations of multiphase fluid flow and heattransport with thermoelastic deformation in porous and fractured media. McPherson(1996) developed a separate package of routines that construct a dynamic and growingintegrated finite difference mesh subject to transient material parameters and boundaryconditions associated with the geologic history of a basin. Finite difference cells aregradually added or removed from the numerical mesh, moved, and deformed verticallywith respect to each other to simulate differential uplift and subsidence. Although theMcPherson (1996) algorithm is based on poroelasticity theory, empirical porosity-depthrelations are used in our study to characterize the compaction history of the basin andrepresent the dependency of porosity and permeability on effective stress, and so


inelastic and nonlinear processes are incorporated in the numerical formulation. Like-wise, bulk compressibilities of the basin formations are scaled up in the fashion observedby Neuzil (1986) to account for the long-term behavior of deformation and pressurediffusion. Our numerical simulations of the Sacramento basin solve (1) with geologicforcing terms representing sediment loading and aquathermal pressuring only. Themagnitude of separate forcing terms explicitly accounting for diagenetic processes suchas clay dehydration, petroleum maturation, cementation, and pressure solution areunknown here, and therefore we have excluded them for the present time to focus onhydromechanical issues first.

Tectonic compression.Modeling the deformation and overpressure fields associatedwith horizontal shortening of the Sacramento basin required application of the numeri-cal code THRUST2D, the mathematical details of which are given by Ge and Garven(1992, 1994). This finite element code assumes plane-strain in x-z space, and theporoelasticity problem is formulated with an integral form of the stress equilibrium eq.(2) coupled with the fluid flow equations (eq 1). Both sets of equations are solved in aninteractive fashion. Small time steps are used to ensure stability and adherence to smallstrains within the poroelastic formulation. When large displacements occur, such asalong faults, special slip-type finite elements are used along with Mohr-Coulomb theoryto simulate elastic-plastic deformation coupled to pore pressure effects (Garven andothers, 1993). In this study, however, tectonic loads are gradually imposed on the lateraledges of the Sacramento basin mesh as specified stresses to understand the poroelasticeffects of crustal shortening on overpressuring at the crustal scale. The basin containsnumerous faults that accommodate local strain, but we made no attempt to representthese fault zones individually within the mesh as slip elements, as this would beimpractical. As in sediment compaction experiments, the effective compressibilityparameters are scaled up from laboratory-based measurements by one or two-orders ofmagnitude to account for the effects of inelastic deformation over geologic time.Heterogeneity and/or anisotropy in permeability, porosity, and effective compressibilityare allowed in order to make the model geologically interesting and account for theeffects of spatial variability on overpressuring. THRUST2D only considers the hydrome-chanical aspects of deformation-induced fluid flow and does not incorporate thermal-elastic or advective-conductive heat transport effects into the numerical scheme for thehydrodynamics.

NUMERICAL EXPERIMENTS AND RESULTS

Depositional compaction: model domain, boundary conditions, and parameterization.Thesubsidence and depositional histories are critical for evaluating sedimentary compactionas a mechanism of overpressures. We considered ind etail two aspects of the basin’sstructural evolution: the tectonic evolution and the resulting depositional history.According to Moxon and Graham (1987), the western and eastern sides of the basinreflect different tectonic histories. The western side of the basin records subsidence, thenuplift, in apparent response to the angle and rate of descent of the subducting Pacificplate (Moxon and Graham, 1987). In contrast, the eastern parts of the basin reflectsubsidence histories that suggest thermal contraction compatible with continentwardmigration of Sierran arc magmatism in the Late Cretaceous.

For the purposes of this study, we are interested mostly in the qualitative aspects ofthe basin’s depositional history, not necessarily what caused subsidence. Specifically,critical to this study are the timing, spatial variability, and, most importantly, themagnitude of sedimentary depositional rates. Sedimentation rate magnitude is mostimportant because compaction rate, and therefore associated overpressuring, is greaterduring high depositional rates (Schneider, Burrus, and Wolf, 1993). The timing (in otherwords, time of onset and duration) of sedimentation is critical also because overpressurescaused by compaction dissipate almost immediately (in a geologic time frame) except in


extremely low-permeability strata. Therefore, active or very recent deposition is re-quired for compaction to account for observed (present-day) overpressures.

A first order test of whether compaction is a significant cause of high fluid pressureswould be to superimpose the overpressure distribution (fig. 7) on a map of active orgeologically recent sedimentary deposition. We did not attempt this comparison. Unruh(written communication, 1999) pointed out that areas of highest pressures, such as theMontezuma Hills and Dunnigan Hills, are associated with youthful, elevated topogra-phy. These areas are high overpressures are more likely areas of erosion, not high rates ofdeposition (Unruh, written communication, 1999). In the absence of detailed study ofgeomorphic surfaces and associated depositional patterns throughout the basin, weperformed a simple backstripping analysis to estimate sedimentation rates. This analysis,discussed below, did not reveal the high rates of sedimentation necessary for overpres-sure development.

The spatial distribution of depositional rates is important too, inasmuch as if weknow where the most rapid deposition occurred, then we can predict the region mostsusceptible to compaction-induced overpressures. The processes that cause spatialvariability of depositional rates are ancillary to understanding induced overpressures.Thus, instead of taking the approach of Moxon and Graham (1987), who examinedisolated 1-D sections to evaluate the history and controls of subsidence (as they vary inspace), we elected to backstrip (Miall, 1990) an entire cross section for the sedimentationrate history across the basin.

Model domain.For this model and subsequent modeling efforts, we sought a muchsimpler cross section. A simple structural section makes it easier to isolate individualprocesses affecting the hydrodynamics. The cross section we selected is associated withprofile B-B�, farther north than profile A-A� and close to the center of the basin withrespect to north-south (see location on fig. 1). It is based on well-log data (Moxon, 1990).The numerical mesh associated with cross section B-B� is shown in figure 9. Also shownare the chronostratigraphic units, demarcated by shading.

Backstripping, porosity, and permeability.Backstripping is a process for determiningsedimentation rates using measured stratigraphy, including thicknesses and ages of each

Fig. 9. Cross section B-B� (location shown on fig. 1), illustrating the mesh used in the numerical modelingexperiments. Also shown are the individual units JK, LK, UK, NG, and EP. See figure 5 for the stratigraphicdistribution of these chronostratigraphic units.


sedimentary unit, and the present-day porosity distribution. Miall (1990) discusses thegeneral method, and McPherson (ms) discusses the details of our specific approach. Theporosity distribution used for all strata is shown in figure 10A and was used for bothbackstripping the sedimentary section (by depth) and for forward evolution of porosity(by effective stress). The solid curve on figure 10A is termed the ‘‘inelastic’’ loading curveand represents non-elastic or permanent changes in porosity. The dashed lines are‘‘unloading/elastic’’ curves, and they represent elastic changes in porosity (Domenicoand Mifflin, 1965).

Additionally, permeability was assumed to vary as a function of porosity: therelationship used and the data used to calibrate it are shown in figure 10B. We acquired asmall set of porosity and permeability data, but of questionable quality, for selected oil

Fig. 10. Porosity-permeability data. (A) porosity versus effective stress (right axis) and versus depth (leftaxis); Uinta basin porosity data plotted against depth and effective stress. Solid line represents compactiontrend used to backstrip sedimentary section as well as the functional relationship used to govern forwardporosity development in the basin evolution model. Open squares represent porosity data from Pitman andothers (1982). The remaining solid circles represent porosity-depth data for unfractured samples from the Uintabasin, Utah.


wells in the basin (California Oil and Gas Commission, 1997 written communication). Inthe absence of a substantial and reliable porosity-permeability database for the Sacra-mento basin, we elected to use comparable data from the Uinta Basin (fig. 10) rather thanassign homogeneous values or use generic relationships (for example, the Kozeny-Carman relationship for shale porosity-permeability).

We backstripped and modeled compaction processes for Cretaceous and youngerstrata only. The Jurassic strata of the section are assumed to exist at the beginning of eachmodel simulation. Table 2 summarizes the average and maximum sedimentation rate,among all columns, determined for each layer (fig. 9) of the model. The average rate foreach chronostratigraphic unit varies by layer in that unit because of variable thickness.Because of the coarse scale of the chronostratigraphic layers, estimated depositional ratesare broad averages. While this model is intentionally simple, the approach is justified, ifone considers the uncertainty associated with subsurface structure, stratigraphic architec-ture, and rock physical properties.

Fig. 10. (B) permeability versus porosity data. All data are measured samples from the Uinta basin, Utah(McPherson, 1996), except as indicated on the plot. See text for further discussion.


Compressibility.Compressibility in the forward depositional model is assumed todepend on effective stress, as illustrated in figure 11. We used curve A on figure 11 for our‘‘base case’’ parameterization and used the other curves in a sensitivity analysis toevaluate the effects on fluid pressures of higher or lower compressibility at depth. Using‘‘inelastic’’ and ‘‘reloading/elastic’’ porosity curves (fig. 10A) necessarily imply differentvalues of compressibility. In all simulations conducted in this study, we use thecompressibility distributions shown on figure 11 as ‘‘inelastic’’ compressibilities. How-ever, if effective stress decreased during a simulation, we used an ‘‘unloading’’ compress-ibility 2 orders of magnitude lower than the current ‘‘inelastic’’ compressibility. Thissemi-elastic behavior is routinely observed for soils and low permeability sediments(Domenico and Mifflin, 1965). During the course of this study we adopted this approachin our simulations to account for semi-elastic behavior suggested by other studies (forexample, Shi and Wang, 1988).

Forward model of basin deposition.Our approach was to simulate the depositionalhistory of the cross section, building it layer-by-layer through the depositional historyand evaluating the overpressures induced by compaction. We did not attempt tosimulate evolution of the basin’s structure from an initial horizontal state, with portions ofthe basin subsequently subsided to form the ‘‘bowl’’ shape (as was done in McPherson,ms). Rather, the initial condition included the basin cross section in its ‘‘bowl’’ structure.This artificially increases the overpressure potential of the basin inasmuch as flat-lyinglayers will permit more rapid dissipation of fluid pressures. The structural geometry of a‘‘bowl’’ shape in cross section will serve to increase fluid pressures at depth becausedissipation paths are updip. Our intention is to test compaction as an overpressuremechanism under maximizing conditions.

We ran dozens of simulations, using the model parameterization, boundary, andinitial conditions summarized in table 3 (applied to the cross section model) as a ‘‘base’’case. Two important properties not listed in table 3 are the bulk density of the sediments,�bulk, and the hydrologist’s specific storage, Ss. For all simulations, including compactionsimulations and tectonic compression model simulations, the grain density of allsediments was assigned a value of 2650 kg m�3. Heterogeneity in bulk density isdetermined by a weighted arithmetic mean: �bulk � �(�water) (1 � �)(�grain), where � isporosity, and �water is water density. Also for all simulations, specific storage is a functionof elastic modulii and porosity, which varies with effective stress or depth (fig. 10): Ss �

TABLE 2

Average and maximum sedimentation rates for each layer in the compaction model.The average sedimentation rate for all layers is 63 m my�1


�water g(� � ), where g is gravitational acceleration, � is rock compressibility (fig. 11)and is water compressibility, assumed constant at 4.4 � 10�10 Pa�1.

Depositional compaction: analysis and results.The base compaction model simulationproduced no overpressures anywhere in the cross section. All parameters assigned in thebase model are reasonable in a geologic context. For overpressures to develop bycompaction, extreme values of parameters and boundary conditions were necessary, asdetermined by a sensitivity analysis.

Our sensitivity analysis consisted of many more simulations, in each case leaving allparameterization the same as the base case except one variable, such as permeability,compressibility, or sedimentation rate. For example, in one simulation we increased thepermeability distribution an order of magnitude relative to the base distribution by usingthe dashed line on figure 10B rather than the base case (solid line, fig. 10B). After allsensitivity analysis simulations, we concluded that fluid pressures (specifically, genera-tion of overpressures) are most sensitive to permeability, depositional rate, and compress-ibility. The permeability controls the rate of fluid escape during compaction, whereas thelatter two variables control the amount of compaction and thus the fluid expelled bycompaction. In the context of the mathematical model, sedimentation rate and compress-ibility determine the magnitude of the fluid source term in eq (1). The interested reader isreferred to McPherson (ms) for mathematical details of this source term.

It is not surprising that overpressure occurrence is dictated by individual thresholdsof these parameters. To induce significant overpressures similar to the observed overpres-

Fig. 11. Experimental and in situ compressibility data as a function of effective stress or depth. Adaptedfrom Neuzil (1986) and Ge and Garven (1992). The curves are derived from observed porosity-depth trends insedimentary basins. Curve A corresponds to Cenozoic mud and sand in the Gulf Coast; curve B corresponds toCenozoic mud and sand in the North Sea; curve C corresponds to Paleozoic shale and sandstone in northcentral Oklahoma. See text for discussion of curve D.


sures of 20 to 30 MPa greater than hydrostatic (figs. 6 and 7), values of local compressibil-ity at depth must be greater than 10�8 Pa�1, or local confining permeability must bemuch lower than 10�18 m2, or the local sedimentation rate must be 800 m my�1 orgreater. This sedimentation rate is over an order of magnitude greater than estimatedactual rates. Whereas the permeability range is reasonable, the necessary compressibility(10�8 Pa�1) is unreasonable for any depth except the upper 1 to 2 km (fig. 11). However,if the permeability is reduced, the compressibility can be correspondingly reduced withsimilar results. We attempted to relate the ratio of permeability change (increase/decrease in permeability values) to compressibility change that would provide similaroverpressure regimes, but this was impossible because permeability is indirectly anonlinear function of compressibility: if the compressibility is changed, this changes therate of compaction and therefore the rate of permeability evolution (permeability tracksporosity).

Figure 12 shows the time-evolution of a model simulation using conditions favor-able to sedimentary compaction overpressures: we assigned a single compressibility-depth distribution depending on effective stress, with the restriction that the minimumcompressibility � � 10�8 Pa�1 (curve labeled D on fig. 11). This compressibilitydistribution is much greater than known experimental or in situ values (compare tocurves A, B, C on fig. 11). The permeability distribution assigned to all strata is an orderof magnitude lower than the base-case distribution (indicated by the dashed curve on fig.10B). The panel of images illustrates the temporal evolution of the cross section,structurally and hydrodynamically.

Between 90 and 72 Ma, overpressures begin to develop in the center of the basin.This corresponds to both a high sedimentation rate at the surface and permeability inthat area of the model domain (value) being low enough to significantly restrict flowduring compaction. The results shown in the 30 Ma time-frame show largely hydrostaticconditions. This results because the periods between 67 to 64 and 40 to 20 Macorrespond to hiatuses in deposition. Figure 13 charts the depth history for each layer ina column within the high pressure zone and shows the periods of relatively high

TABLE 3

Base case model parameterization, boundary and initial conditions, sedimentary compactionnumerical experiments and aquathermal pressuring numerical experiments

Parameter Values

Porosity PermeabilityPoisson’s

Ratio

Young’sModulus

(Pa)Compressibility

(Pa�1)

Figure 10A: backstripping bydepth; forward evolution byeffective stress

Figure 10B; solid line. 0.25 1.5 � 108 Figure 11.

Rock GrainDensity(kg m�3)

Rock MatrixThermal Conductivity

(W m�1 K�1)

Rock MatrixSpecific Heat(J kg�1 K�1)

Sandstones 2650 4.1 1080Shales 2750 1.9 850Mixed Sandstone/Shale 2700 3.0 950Igneous 2800 3.2 1380Boundary and Initial ConditionsEast boundary No flow boundary; justified by juxtaposition to Sierra blockWest boundary No flow boundary; justified by juxtaposition to Coast RangesBottom boundary No flow boundary; justified by basement rocks beneath package JKTop boundary Constant head � topographic elevationInitial conditions Layers JK present with hydrostatic conditions at onset; other layers

deposited in time


Fig.

12.

Pane

lofp

lots

illus

trat

ing

the

geol

ogic

evol

utio

nan

das

soci

ated

deve

lopm

ento

fove

rpre

ssur

esw

ithin

the

basi

n.C

ondi

tions

fort

hiss

imul

atio

nth

atde

part

from

the

base

sim

ulat

ion:

com

pres

sibi

lity

vers

usef

fect

ive

stre

sspa

ram

eter

ized

for

all

stra

taus

ing

curv

e‘D

’on

figur

e11

,pe

rmea

bilit

ydi

stri

butio

nas

sign

edto

alls

trat

ais

anor

der

ofm

agni

tude

low

erth

anth

eba

se-c

ase

dist

ribu

tion

(indi

cate

dby

the

dash

edcu

rve

onfig

ure

10B

).T

hesh

adin

gin

tens

itysh

ows

cont

ours

ofov

erpr

essu

res,

orex

cess

pres

sure

sab

ove

hydr

osta

tic.


sedimentation rate (high slope of depth curve) and how they correlate to the time-framesthat show higher overpressuring (for example, the 67 Ma time-frame in top panel of fig.12). Unless permeability is orders of magnitude lower or fluid viscosity is very high, suchas that associated with petroleum fluids (McPherson, 1996), overpressures dissipaterapidly with respect to geologic time (less than 100,000 yr) during these periods ofdepositional hiatus. Thus, the model results suggest that present-day overpressureconditions really depend on what is the current, ongoing depositional state. If sedimenta-tion ceased much longer than 105 yr ago, it is unlikely that compaction is responsible foroverpressures. This result implies that modern overpressures must be associated withareas of active late Quaternary sedimentation, if they are to be attributed to sedimentarycompaction. However, our observed overpressure data correlate more to areas of activeuplift and erosion, not deposition (Unruh, written communication, 1999).

Overpressures and sedimentation rate.Although we dispel sedimentary compaction as asignificant overpressure mechanism, the model results are interesting and warrant somefurther discussion of associated effects. The primary locus of high overpressuring (fig. 12)lies in the center of the basin, about 20 km wide by 4 km thick. This result was initiallysurprising inasmuch as we expected the highest overpressures to occur in the deepestpart of the basin where permeability is lowest. This focusing of overpressures isattributed to sedimentation rate variation. It is also possible that the locus is artificiallyaffected by our initial condition of non-horizontal layers, or the ‘‘initial bowl geometry,’’but the main influence is sedimentation rate. This conclusion is inferred from figure 14, aplot of average sedimentation rate of the uppermost Neogene layer (the layer depositedlast) for each column. The highest sedimentation rate occurs over the center of the basin,coincident with the highest overpressures at depth. Throughout the basin’s history, thehighest sedimentation rates happen to occur close to the center of the basin. This islargely consistent in a geologic context, except it is believed that the locus of greatest

Fig. 13. The depth history for each layer in a column within the high pressure zone.


Fig.

14.

Plot

ofm

odel

edpr

essu

redi

stri

butio

ndu

ring

pres

ent-d

ay(b

otto

m;r

esul

tsof

the

basi

nev

olut

ion/

sedi

men

tary

com

pact

ion

mod

el).

The

top

plot

illus

trat

esth

ese

dim

enta

tion

rate

dist

ribu

tion

ofth

eN

eoge

ne(N

G)l

ayer

,clo

sest

toth

esu

rfac

ean

dde

posi

ted

last

.


sedimentation rate may have migrated eastward with time (Unruh, written communica-tion, 1998).

Despite these results and the fact that most of our observed overpressure data lieclose to the center of the basin, we would still be reluctant to attribute more than only aportion of observed overpressures to compaction; the high in situ compressibilityrequired for overpressuring is unrealistic. Our observed data lie close to the center of thebasin because that is where most oil/gas wells (from which our data are derived) weredrilled. We do not have pressure data from the periphery of the basin. Additionally,Berry (1973) documents the occurrence of high overpressures in the Franciscan rocks ofthe Coast Ranges, illustrating that observed overpressures are not limited to the center ofthe basin. This observation alone suggests that sedimentary compaction is not a strongcandidate for explaining the dominant overpressure mechanism in the basin.

All compaction analyses described above were isothermal simulations. In thefollowing section, we describe non-isothermal compaction history experiments designedto evaluate how heat flow may influence hydrodynamics through ‘‘aquathermal expan-sion.’’

Thermal aspects of overpressuring.Elevated temperatures reduce water density. Porefluid pressure will increase if permeability is low enough to preclude efficient expulsion.First termed ‘‘aquathermal pressuring’’ by Barker (1972), many studies since haveexamined aquathermal pressuring as a possible overpressure mechanism, includingBradley (1975), Magara (1975), Chapman (1980, 1982), Barker and Horsfield (1982),Daines (1982), Shi and Wang (1986), and Gordon and Flemings (1998). Using a 1-Dnumerical model of compaction and aquathermal processes, Luo and Vasseur (1992)performed a general sensitivity analysis of a aquathermal pressuring and its competingrole with mechanical compaction as an overpressure mechanism. Luo and Vasseur(1992) concluded that, except in conditions of extremely low permeability, aquathermalpressuring is a negligible source of overpressures. However, we tested this mechanismfor the sake of completeness and because of the coincidence of observed overpressureswith Sutter Buttes, an area of relic high heat flow.

Sass and others (1981) and Sass and others (1971) report heat flow for the westernUnited States, including the Great Valley and the adjacent Coast Ranges. Only twovalues are reported for the Sacramento basin proper, both approximately the samelatitude as Sutter Buttes. No known high heat flow anomaly is associated with SutterButtes today. The value of surface heat flow directly adjacent to Sutter Buttes is �35 mWm�2, and heat flow directly adjacent to the Coast Ranges (same latitude) ranges from 36to 48 mW m�2 (Sass and others, 1981). For reference, conductive heat flow at the Earth’ssurface ranges from about 30 to 120 mW m�2 (Chapman and Pollack, 1975), notincluding extreme heat flow associated with magmatic sources and other geothermalareas. Thus, the 35 to 48 mW m�2 observed in the Sacramento basin is a relatively lowrange of surface heat flow. Also, judging from the conclusions of Luo and Vasseur (1992),the observed heat flow in the basin is not great enough to induce significant overpres-sures by aquathermal pressuring. However, these are modern heat flow values. Althoughdetermining the history of Sacramento basin heat flow through geologic time is beyondthe scope of this study (see Dumitru, 1988), it is nevertheless useful to conductcompaction simulations with different thermal regimes to evaluate the effect of aquather-mal pressuring on fluid pressures.

For non-isothermal analyses, we used the ‘‘base’’ compaction model, including themesh, boundary and initial conditions (fig. 9 and table 3), but thermal aspects of theproblem, including conductive and advective heat transport, were coupled to groundwa-ter flow. Water viscosity and density are assigned as a function of temperature andpressure, as determined by the IFC (1967). The only change of boundary conditions forthe non-isothermal simulations was a constant heat flow boundary along the bottom of


the model domain. For our ‘‘base case’’ non-isothermal model, we assigned a regionalbackground basal heat flow of 60 mW m�2, slightly higher than the high end of thepublished range.

To test the effect of aquathermal expansion as an overpressure mechanism, weconducted several simulations with varying conditions of basal heat flux, comparing fluidpressure results of these simulations to the results of a uniform 60 mW m�2 boundarycondition. One way to evaluate aquathermal expansion effects on a regional scale is torun a simulation with elevated basal heat flow. We tripled the basal heat flux to approx180 mW m�2, and the results were consistent with Luo and Vasseur’s (1992) conclusions:aquathermal expansion did not increase fluid pressures. On the contrary, fluid pressuresfor the higher heat flux case were actually slightly lower due to reductions in both fluidviscosity and density by higher temperatures.

We also evaluated the local effect of Sutter Buttes (fig. 1), a possible source of relichigh heat flow between 2.4 and 1.4 Ma, when magma injection at depth induced domingof Upper Cretaceous and Tertiary rocks (Williams and Curtis, 1977). To test this localsource of aquathermal pressuring, we conducted another simulation of our profile B-B�model domain with a uniform, constant basal heat flow boundary condition of 60 mWm�2. The profile lies immediately south of Sutter Buttes (fig. 1). Assuming that themagma injection episode that created Sutter Buttes was relatively short in geologic time,we artificially raised the basal heat flux to simulate the thermal effect of magma injectionfor 100,000 yr. For the approx 10 km interval of the profile adjacent to Sutter Buttes, wearbitrarily raised the basal heat flow to 420 mW m�2 between 2.0 and 1.9 Ma. We do notknow the actual value of the heat flow anomaly caused by Sutter Buttes emplacement.However, we chose this value because it is consistent with heat flow observed instructural domes associated with young calderas and active volcanism within the lastapprox 105 yr (Sass and Morgan, 1988; Sass and others, 1981). Figure 15 illustrates thecontrast in temperature field caused by the simulated magma injection at depth. With theincreased heat flux condition, temperature at the base of the model domain in the area ofSutter Buttes is approx 400°C; for the base case, the temperature in this area is only

Fig. 15. Temperature field predicted by a model that simulated elevated heat flow (420 mW m�2)coincident with the formation of Sutter Buttes.


around 200°C. The effects of the high basal heat flow on overpressures are small, asdemonstrated in figure 16, a plot of pressure history for the base of the UK unit (figs. 5and 9) during 2.0 and 1.8 Ma. While pressures initially rise during the first few thousandyears of the simulation, due to aquathermal expansion of water, excess pressures begin todecay by re-equilibration. Immediately after the high heat flow sources are reduced from420 to 60 mW m�2, fluid pressures re-equilibrate toward the ambient value associatedwith the original heat flux condition.

We conclude that aquathermal expansion is not an important source of present-dayoverpressures in the Sacramento basin. Some overpressures may have developed duringthe heat flow pulse associated with formation of Sutter Buttes. However, without highbasal heat flow maintained through present day, aquathermal expansion has a negligibleeffect on present-day pore pressures. Overpressures by this mechanism could only bemaintained over geologic time under conditions of unrealistically low permeability, andtherefore extremely low hydraulic diffusivity.

Tectonic compression: model domain, boundary conditions, and parameterization.Tectoniccompression is an appealing candidate for overpressure generation in the Sacramentobasin because of the regional distribution of overpressures. Tectonic forcing can act overthe regional scale, affecting the entire basin system, including the adjacent mountainbelts. Thus, overpressures observed throughout great portions of the basin and the CoastRanges support the notion of a regional mechanism.

We tested the compression hypothesis using another numerical model of the basin.Following the lead of earlier studies (Ge and Garven, 1989, 1992), we used a finiteelement method that uses poroelastic coupling of tectonic stresses and fluid flow. Weassembled a finite element mesh of the same cross section used for the compactionanalysis (fig. 9). Table 4 summarizes the parameterization, boundary, and initial condi-tions imposed for the ‘‘base,’’ or most representative, tectonic compression modelsimulation. Parameterization is similar to the previous compaction models. Porosity iscontrolled by effective stress (fig. 10A), permeability tracks porosity (fig. 10B), andcompressibility is a function of effective stress (curve A, fig. 11). If effective stressdecreased during a simulation, we used an ‘‘unloading’’ compressibility 2 orders of

Fig. 16. Pressure history model results for the base of Upper Cretaceous units (see fig. 5 for relativestratigraphic location). Bottom dashed line is fluid pressure history without compaction and without heat flow(isothermal simulation). The next line above indicates the pressure history with compaction and with a basalheat flux of 60 mW m�2. The top line indicates the pressure history with compaction and with a basal heat fluxof 420 mW m�2.


magnitude lower than the current ‘‘virgin’’ compressibility. Therefore, basin-shorteninginvoked by the model is semi-elastic and underpredicts overpressure potential. Forexample, if purely inelastic deformation such as pressure solution were invoked, theoverpressure potential would be greater because elastic accommodation of high pres-sures would not be possible.

In the sedimentation-compaction analyses, deformation and compressibility wereassumed to be vertical. For the tectonic compression simulations, we use the samefunction of effective stress (fig. 11), but compressibility was isotropic. Tectonic stress wasinvoked by specifying a horizontal shortening. Stress was applied along the western sideof the basin cross section, with the east side assigned a fixed, non-moving boundary. Thebasin cross section was shortened due east-west.

A simple comparison between the parameterization for this model and that for thebase compaction model is the net loading rate, or stress rate. In the process ofsedimentary compaction, sediment loading effectively imposes an accumulation ofvertical stress over time. This may be thought of as a ‘‘stress rate.’’ Tectonic compressionalso imposes a stress rate over time; in this tectonic setting it is a horizontal stress rate. Forour base case model of tectonic compression, we applied a horizontal stress rate(associated with the applied strain rate) equivalent to the average vertical stress rate bysedimentation. The average sedimentation rate for Jurassic and younger strata in thebasin is 63 m my�1. This sedimentation rate is equivalent to an imposed stress rate of 1.2MPa my�1, using the formula � � (� g z)/�t, where � is the stress rate, z is thickness ofdeposited material, � is an assumed average density equal to approx 2000 kg m�3

(assumed average surface porosity of 40 percent), g is the acceleration of gravity, and �tis the duration of sedimentation. Changing our perspective to the horizontal, this stressrate translates to a corresponding equivalent horizontal basin shortening rate of 0.04 mmyr�1. This is the value of horizontal shortening rate applied in our base case tectoniccompression model.

Tectonic compression: analysis and results.Overpressures did not develop anywhere inthe cross section of the base tectonic compression model. In one of our previoussedimentary compaction analyses, we assigned the entire sedimentary section conditionsof porosity, permeability, and compressibility consistent with a shale, and that modelrequired a minimum sedimentation rate of approx 800 m my�1 for significant overpres-sures to develop. Therefore, it is no surprise that the base tectonic compression model,with a horizontal compression rate equivalent to 63 m my�1 sedimentation, did not

TABLE 4

Base case model parameterization, boundary and initial conditions, tectonic compressionnumerical experiments

Parameter Values

Porosity PermeabilityPoisson’s

Ratio

Young’sModulus

(Pa)Compressibility

(Pa�1)

Figure 10A: porosity dictatedby depth.

Figure 10B;solid line. 0.25 1.5 � 109 Figure 11.

Boundary and Initial ConditionsEast boundary No flow boundary; justified by juxtaposition to Sierra block with range front

faults; constant horizontal tectonic compression rate of 0.04 mm yr�1

applied uniformly along boundary.West boundary No flow boundary; justified by juxtaposition to Coast Ranges (Franciscan)

with range front faultsBottom boundary No flow boundary; justified by basement rocks beneath package JKTop boundary Constant head � topographic elevationInitial conditions All layers hydrostatic.


induce significant overpressures. Unfortunately, we do not have independent informa-tion regarding what range of values of loading rate may be considered high versus whatmay be considered low. As a consequence, we use plate convergence velocities andshortening rates to constrain the loading rates in all tectonic compression simulations.

We compared many sets of simulation results to the base simulation in anothersensitivity analysis. The results suggest that tectonic forces provide a plausible overpres-sure mechanism: compression induces overpressuring only under specific conditions,albeit less restrictive conditions than those for sedimentary compaction. Figure 17illustrates the hydrodynamic history corresponding to a model using boundary condi-tions and rock properties favorable to overpressuring. The only difference between thismodel and the base model is that the horizontal shortening rate applied is 3.5 mm yr�1,approx 88 times greater than the shortening rate associated with the base model. Wefound that a minimum range of basin shortening of 2.9 to 3.5 mm yr�1 is necessary toinduce overpressures consistent with observed data. This range is consistent withtectonic convergence rates determined by baseline interferometry (Wakabayashi andSmith, 1994; Argus and Gordon, 1991) and east-west shortening rates estimated bystructural analysis of thrust faults in the southern portion of the basin (Unruh, Loewen,and Moores, 1995). For comparison, table 1 summarizes selected fault slip, tectonicconvergence, and horizontal shortening rates published by other workers.

At 3000 m depth, the deeper range of our observed overpressure data, the 3.5 mmyr�1 model reflects development of overpressures around 10 to 15 MPa above hydro-static (fig. 17). This level of overpressuring is comparable to the majority of observedoverpressures in the area of Sutter Buttes (fig. 7). The highest observed overpressures,approach 25 MPa above hydrostatic (fig. 7). This level of overpressuring occurs at around 6km depth in the model, but not shallower. However, these models are parameterizedwith smooth distributions of permeability and compressibility (figs. 10 and 11). Weconducted some simulations with localized regions of higher compressibility or lowerpermeability, such as caused by local geological heterogeneities, and these simulationsproduced locl overpressures greater than 25 MPa at shallower depths. We did notattempt to simulate faults, which also may localize overpressures and underpressures.

At the lower depth range of observed overpressures (1500 m), smaller values ofoverpressure developed, only 4 to 5 MPa above hydrostatic. These overpressures areconsistent with observations for that depth (figs. 6 and 7) and are lower primarily due topermeability being greater at 1500 m compared to 3000 m. Other model simulationsshowed that if permeability is assigned uniformly low values (approx 10�18 m2 or lower),then overpressures develop throughout the section.

Shortening rates.If we assume that fault creep over time is small, rapid and finite faultslip may release stress, but up until that slip event, stress may accumulate. Because thesemodels utilize an elastic/semi-elastic rheology, we are essentially simulating thoseperiods of time between finite fault slip events. We did not attempt to simulate faults ortheir propensity to release stress and alter hydrodynamics.

Over the duration of the basin’s geologic history, progressive tectonic shorteningoccurs as permanent strain accumulates in response to cumulative stress. The actualamount of basin shortening is unknown, but fault slip and other data provide minimumestimates. Our simulation results made it immediately apparent that crustal shorteningrates are extremely important to potential overpressure development.

In a series of simulations, we simulated the base model, shortening the basin thesame amount (5 km) in each run but using a different shortening rate in each simulation(up to 13 mm yr�1). In each simulation, overpressures formed in the same places in themodel, but the maximum overpressure reached in a given area of the model domainvaried with respect to shortening rate. Plotted in figure 18 are the results of this sensitivity


Fig.

17.

Dev

elop

men

tofo

verp

ress

ures

inth

eba

sin

duri

nga

sim

ulat

edhi

stor

yof

appl

ied

late

rals

hort

enin

gby

tect

onic

com

pres

sion

.T

heon

lydi

ffere

nce

betw

een

this

mod

elan

dth

eba

sem

odel

desc

ribe

din

tabl

e2

isth

eap

plie

dho

rizo

ntal

shor

teni

ngra

teis

3.5

mm

yr�

1 ,ap

prox

88tim

esgr

eate

rth

anth

eba

sem

odel

rate

.


analysis among maximum overpressure and shortening rate. The line on figure 18represents the maximum overpressure reached at 3 km depth in the center of the crosssection, with 5 km maximum shortening in each simulation, over a range of shorteningrates from 0.9 to 13 mm yr�1. The maximum overpressure plotted is normalized,because if we were to change other attributes of the base model the variation inmaximum overpressure would be similar, but the magnitude would scale up or downdepending on the specific variable and how it was changed. For example, if allpermeabilities in the model domain are increased, lower values of maximum overpres-sures resulted. However, the variation of maximum overpressures with respect toshortening rate, expressed by the shape of the curve on figure 18, was the same.

In general, greater shortening rates translate to potential of greater overpressures inthe same places in the model (fig. 18). The shaded area on figure 18 represents the limit ofminimum possible shortening rates observed in the Sacramento Basin (table 1). Wedeliberately evaluated shortening rates much higher than the observed range to test theeffect on overpressuring. Maximum overpressures with respect to shortening rate areasymptotic in nature, reflecting limits of hydrologic storage and permeability. Thedistribution of storage and permeability remains the same among these simulations, asdescribed by the base model (table 4). In this elastic model, the only ways to increasemaximum overpressure beyond this threshold would be to (1) reduce hydraulic diffusiv-ity, in other words reduce permeability or increase hydrologic storage, (2) increase theamount of shortening beyond the prescribed limit of 5 km, or (3) include inelasticprocesses such as chemical diagenesis. We conducted simulations with uniformly lowerdiffusivities and verified this effect, again justifying the use of normalized maximumoverpressure on figure 18.

Figure 18 also illustrates that significant overpressures may result from observedranges of crustal shortening rates in the Sacramento Basin. Of course, these results are

Fig. 18. Normalized maximum overpressure versus tectonic shortening rate, based on several numericalexperiments using a range of rates.


specific to this model of the Sacramento Basin. For the general case, the relationshipbetween stress rate and overpressure also depends on the in situ permeability, fluidviscosity, and geologic structures such as faults. Additionally, other overpressure-inducing mechanisms such as sedimentary compaction or chemical diagenesis contrib-ute to the potential for overpressure development.

Discussion and conclusions.Overpressures are common to many, if not most, sedimen-tary basins. Judging from the results of many published studies, it is clear that overpres-sures originate from many different processes. Also, this and other studies suggest that itis typical for more than one mechanism to contribute to high fluid pressures in a singlegiven area. Overpressures in the Sacramento basin are observed over the regional scale,with limited correlation to local geography and structures. We suggest that since theFranciscan rocks of the Coast Ranges adjacent to the basin are overpressured (Berry,1973) to some extent, sedimentary compaction, geochemical diagenesis, and aquather-mal pressuring are not clearly favored, or at least not dominant, overpressure mecha-nisms. Other evidence sheds uncertainty on these mechanisms, too. Clay diagenesis is aminor overpressure mechanism compared to sediment loading in the San Joaquin Basin(Wilson, Garven, and Boles, 1999), just south of our study area. Albeit petroleumgeneration can cause regional overpressures (Timko and Fertl, 1971; Spencer, 1987;Burrus and others, 1992; Bredehoeft, Wesley, and Fouch, 1994; McPherson, ms), weoverlaid our observed overpressure data and found only limited correlation with knownpetroleum source beds. These aspects provide motivation for testing tectonic compres-sion, an overpressure mechanism particularly regional in nature.

Our modeling experiments have led to a series of strong, but not absolute,conclusions. Comparing the model results (fig. 17) to observed overpressures (figs. 6 and7), we conclude that tectonic compression cannot be ruled out as an important overpres-sure mechanism for the Sacramento basin. The shortening rates we modeled are realisticor at least are consistent with observed rates. Additionally, the ranges of permeability,compressibility, and other model parameters are geologically realistic. We made noconcerted attempt to replicate the observed distribution exactly, because such attemptswould be impractical. We only used the simulations to test if the tectonic mechanism isplausible, and if the results are compelling in this context. Clearly, tectonic compressionis a viable overpressure mechanism.

One major difference between the sedimentary compaction models and the tectoniccompression models is the bulk compressibility distributions used in the models.Sedimentary compaction required either unrealistically high sedimentation rates (anorder of magnitude greater than observed), extremely low permeability of confiningunits, or unrealistically high compressibility distributions (curve D on fig. 11). For thetectonic compression model, published compressibility distributions were used in con-junction with basin shortening rates consistent with observed tectonic convergence rates.Thus, the modeling exercises suggest that regional overpressuring at relatively shallowdepths (�3000 m) can be caused by active tectonic compression without the need ofextremely low permeability distributions or unrealistically high rock compressibility.

The modeling also suggests that sedimentary compaction is a plausible overpressuremechanism, but only for locally high sedimentation rates or unusually low permeabilityand/or unrealistically high compressibility over a regional scale. We conclude thattectonic compression is the most important, ongoing mechanism driving overpressuresin the Sacramento basin. It is possible that locally high sedimentation rates argument(local) pressures and contribute to a noisy distribution of overpressures, but we did nottest this assertion. Low permeability zones may make compaction more heterogeneousas well. We speculate that other overpressure mechanisms such as hydrocarbon gasgeneration may be operating as well, but probably only in local areas.


The numerical cross section built for this study was deliberately simple in structure,with no faults or fractures. Also, rock properties were assigned as a simple function ofdepth or effective stress and not dictated by local structure or depositional environment.This parameterization biased the tectonic compression model into producing overpres-sures regional in scale, with no really extreme variation in local pressure. This is incontrast to the modeled sedimentary compaction overpressures, which are highest inlocal areas of high sedimentation rates. Other mechanisms such as chemical diagenesismay ultimately contribute to local variations in overpressure, too. Additionally, varia-tions in structure may contribute to local variations as well as to pockets of observedunderpressuring in the basin. We made no attempt to discern the cause of localvariability in pressures but rather tried to identify the major causes of overpressures atthe basin scale.

Also, given that in all modeling exercises, fluid pressures returned to hydrostaticconditions quite rapidly (�105 yr) during depositional or compressional hiatuses, weconclude that overpressures are attributable to ongoing and dynamic processes. How-ever, under conditions of unusually low permeability or unusually high fluid viscosity,our results suggest that overpressures would dissipate more slowly.

In summary,1. Observed regional scale overpressures in the Sacramento basin and adjacent

Coast Ranges suggest regional scale mechanisms.2. Tectonic compression is a likely cause of overpressures in the Sacramento basin.3. Sedimentary compaction is a plausible cause of overpressures in the Sacramento

basin, but only for local, extremely high sedimentation rates. This may be animportant factor in causing local variations in overpressures.

4. Aquathermal pressuring is negligible as a cause of overpressures in the Sacra-mento basin.

5. We did not test the effects of chemical diagenesis or other mechanisms ofoverpressures but suggest that other such processers may contribute a secondarycomponent of observed high pressures and local variability in fluid pressures.

6. Modeling results indicate that overpressures in the Sacramento basin aremaintained by ongoing processes, except perhaps in areas of extremely lowpermeability.

In conclusion, we find it no surprise that overpressures are common to so manysedimentary basins around the world, especially in tectonically active areas. Multipleoverpressure mechanisms are probably active, and additive, in most overpressuredsedimentary basins. For the Sacramento basin, we agree with Berry (1973) that tectoniccompression seems appropriate as a hydrogeologically significant cause for overpressur-ing.

ACKNOWLEDGMENTSThis study was supported by a grant from the National Science Foundation

(Hydrologic Sciences Program, EAR-9526951) to Grant Garven. Under the auspices ofthe NSF grant, Brian McPherson was supported as a postdoctoral researcher while hecollaborated with Grant Garven at Johns Hopkins University. The authors are grateful toJohn Bredehoeft, who provided all of the observed overpressure data as well as insightfulideas and discussion. We are grateful to the editor and associate editors of the AmericanJournal of Science, as well as to Jeff Unruh and Kelin Wang for thorough, helpfulcomments on earlier versions of the manuscript. We also acknowledge Ray Pugsley forhis initial research and assistance with the pressure data.


REFERENCES

Almgren, A. A., 1978, Timing of Tertiary submarine canyons and marine cycles of deposition in the southernSacramento Valley, California, in Stanley, D. J., and Kelling, Gilbert, editors, Sedimentation in submarinecanyons, fans and trenches: Stroudsburg, Pennsylvania, Dowden, Hutchinson, and Ross, p. 276–291.

Argus, D. F., and Gordon, R. G., 1991, Current Sierra Nevada-North America motion from very long baselineinterferometry: implications for the kinematics of the Western United States: Geology, v. 19, p. 1085–1088.

Atwater, T., and Molnar, P., 1973, Relative motion of the Pacific and North American plates deduced fromsea-floor spreading in the Atlantic, Indian, and South Pacific oceans, in Proceedings of the conference ontectonic problems of the San Andreas Fault system: Stanford University Publications, Geological Sciences,v. 13, p. 136–148.

Barker, C., 1972, Aquathermal Pressuring: Role of Temperature in Development of Abnormal-PressureZones: American Association of Petroleum Geologists Bulletin, v. 56, p. 2068–2071.

Barker, C., and Horsfield, B., 1982, Mechanical versus thermal cause of abnormally high pore pressures inshales: discussion: American Association of Petroleum Geologists Bulletin, v. 66, p. 99–100.

Bear, J., 1972, Dynamics of Fluids in Porous Media: New York, Dover Publications, 764p.Belitz, K. R., and Heimes, F. J., 1990, Character and evolution of the ground-water flow system in the central

part of the western San Joaquin Valley, California: United States Geological Survey Water-Supply PaperW 2348, 28 p.

Berry, F. A. F., 1973, High fluid potentials in California Coast Ranges and their tectonic significance: AmericanAssociation of Petroleum Geologists Bulletin, v. 57, p. 1219–1249.

Bertoldi, G. L., Johnson, R. H., and Evenson, K. D., 1991, Ground Water in the Central Valley, California—A Summary Report: United States Geological Survey Professional Paper 1401-A, 44 p.

Bethke, C. M., 1985, A numerical model of compaction-driven groundwater flow and heat transfer and itsapplication to the paleohydrology of intracratonic sedimentary basins: Journal of Geophysical Research,v. 90, p. 6817–6828.

Bethke, C. M., and Corbet, T. F., 1988, Linear and nonlinear solutions for one-dimensional compaction flow insedimentary basins: Water Resources Research, v. 24, p. 461–467.

Bloch, R. B., Von Huene, R., Hart, P. E., and Wentworth, C. M., 1993, Style and magnitude of tectonicshortening normal to the San Andreas fault across the Pyramid Hills and Kettleman Hills South Dome,California: Geological Society of America Bulletin, v. 105, p. 464–478.

Bolt, B. A., 1979, Seismicity in the western United States: Geological Society of America Reviews inEngineering Geology, v. 4, p. 95–107.

Bradley, J. S., 1975, Abnormal formation pressure: American Association of Petroleum Geologists Bulletin,v. 59, p. 957–973.

Bredehoeft, J. D., and Norton, D. L., 1990, Mass and energy transport in deforming Earth’s crust, in The Roleof Fluids in Crustal Processes: Washington, D.C. National Academy of Sciences Press Studies inGeophysics, p. 27–41.

Bredehoeft, J. D., Wesley, J. B., and Fouch, T. D., 1994, Simulations of the origin of fluid pressure, fracturegeneration, and the movement of fluids in the Uinta basin, Utah: American Association of PetroleumGeologists Bulletin, v. 78, p. 1729–1747.

Burrus, J., Kuhfuss, A., Doligez, B., and Ungerer, P., 1992, Are numerical models useful in reconstructing themigration of hydrocarbons? A discussion based on the Northern Viking Graben, in England, W. A., andFleet, A. J., editors, Petroleum Migration: Geological Society of London Special Publication, v. 59,p. 89–109.

Cady, J. W., 1975, Magnetic and gravity anomalies in the Great Valley and western Sierra Nevadametamorphic belt, California: Geological Society of America Special Paper 168, 56 p.

Castillo, D. A., and Zoback, M. D., 1994, Systematic variations in stress state in the southern San JoaquinValley: Inferences based on well-bore data and contemporary seismicity: American Association ofPetroleum Geologists Bulletin, v. 78, p. 1257–1275.

Chapman, D. S., and Pollack, H. N., 1975, Global heat flow; a new look: Earth and Planetary Science Letters,v. 28, no. 1, p. 23–32.

Chapman, R. E., 1980, Mechanical versus thermal cause of abnormal high pore pressure in shales: AmericanAssociation of Petroleum Geologists Bulletin, v. 64, p. 2179–2183.

————— , 1982, Mechanical versus thermal cause of abnormal high pore pressure in shales: reply: AmericanAssociation of Petroleum Geologists Bulletin, v. 66, p. 101–102.

Daines, S. R., 1982, Aquathermal pressuring and geopressure evaluation: American Association of PetroleumGeologists Bulletin, v. 66, p. 931–939.

Davisson, M. L., and Criss, R. E., 1993, Stable isotope imaging of a dynamic groundwater system in thesouthwestern Sacramento Valley, California, USA: Journal of Hydrology, v. 144, p. 213–246.

Davisson, M. L., Presser, T. S., and Criss, R. E., 1994, Geochemistry of tectonically expelled fluids from thenorthern Coast Ranges, Rumsey Hills, California, USA: Geochemica et Cosmochimica Acta, v. 58,p. 1687–1699.

Deju, R. A., 1973, Global look at high pressures: American Association of Petroleum Geologists Bulletin, v. 57,no. 4, p. 775.

Dickinson, W. R., Hopson, C. A., Saleeby, J. B., Schweickert, R. A., Ingersoll, R. V., Pessagno, E. A., Jr.,Mattinson, J. M., Luyendyk, B. P., Beebe, W., Hull, D. M., Munoz, I. M., Blome, C. D., 1996, Alternateorigins of the Coast Range Ophiolite (California): introduction and implications: GSA Today, v. 6, no. 2,p. 1–10.

Dickinson, W. R., and Seely, D. R., 1979, Structure and stratigraphy of forearc regions: American Associationof Petroleum Geologists Bulletin, v. 63, p. 2–31.

Dickinson, W. R., and Snyder, W. S., 1979, Geometry of triple junctions related to San Andreas transform,Journal of Geophysical Research, v. 84, no. B2, p. 561–572.


Domenico, P. A., and Palciauskas, V. V., 1979, Thermal expansion of fluids and fracture initiation incompacting sediments: Geological Society of America Bulletin, v. 90, p. 953–979.

Domenico, P. A., and Schwartz, F. W., 1990, Physical and chemical hydrogeology: New York, John Wiley &Sons.

Dumitru, T. A., 1988, Subnormal geothermal gradients in the Great Valley forearc basin, California, duringFranciscan subduction; a fission track study: Tectonics, v. 7, n. 6, p. 1201–1221.

Fertl, W. H., Chapman, R. E., and Hotz, R. F., 1994, Studies in abnormal pressures: Developments inPetroleum Science, v. 38, 454 p.

Fox, K. F., Jr., 1983, Tectonic setting of late Miocene, Pliocene, and Pleistocene rocks in part of the CoastRanges north of San Francisco, California: United States Geological Survey Professional Paper 1239, 33 p.

Garcia, R., 1981, Depositional systems and their relation to gas accumulation in the Sacramento Valley,California: American Association of Petroleum Geologists Bulletin, v. 65, p. 653–673.

Garven, G., 1995, Continental-scale groundwater flow and geologic processes: Annual Reviews of Earth andPlanetary Science, v. 23, p. 89–117.

Garven, G., Ge, S., Person, M. A., and Sverjensky, D. A., 1993, Genesis of stratabound ore deposits in themidcontinent basins of North America: 1. Role of regional groundwater flow: American Journal ofScience, v. 293, p. 497–568.

Ge, S., and Garven, G., 1989, Tectonically induced transient groundwater flow in foreland basins, in Price, R.A., editor, Origin and Evolution of Sedimentary Basins and Their Energy and Mineral Resources:American Geophysical Union, Geophysics Monograph Series: v. 48, p. 145–157.

————— , 1992, Hydromechanical modeling of tectonically driven groundwater flow with application to theArkoma Foreland Basin: Journal of Geophysical Research, v. 97, p. 9119–9144.

————— , 1994, A theoretical model for thrust-induced deep groundwater expulsion with application to theCanadian Rocky Mountains: Journal of Geophysical Research, v. 99, p. 13851–13868.

Godfrey, N. J., Beaudoin, B. C., Klemperer, S. L., Mendocino Working Group, 1997, Ophiolitic basement ofthe Great Valley forearc basin, California, from seismic and gravity data: Implications for crustal growth atthe North American continental margin: Geological Society of America Bulletin, v. 108, p. 1536–1562.

Gordon, D. S., and Flemings, P. B., 1998, Generation of overpressure and compaction-driven fluid flow in aPlio-Pleistocene growth-faulted basin, Eugene Island 330, offshore Louisiana: Basin Research, v. 10,p. 177–196.

Gordon, R. G., and Argus, D. F., 1993, The San Andreas fault system in Central California as the boundarybetween the Pacific Plate and the Sierra Nevada-Great Valley microplate: kinematics from VLBI geodesy:EOS, transactions of American Geophysical Union, v. 74, p. 64.

Green, D. H., and Wang, H. F., 1990, Specific storage as a poroelastic coefficient: Water Resources Research,v. 26, p. 1631–1637.

Haggart, J. W., and Ward, P. D., 1984, Late Cretaceous (Santonian-Campanian) stratigraphy of the northernSacramento Valley, California: Geological Society of America Bulletin, v. 95, p. 618–627.

Harrison, W. J., and Summa, L. L., 1991, Paleohydrology of the Gulf of Mexico Basin: American Journal ofScience, v. 291, p. 109–176.

Hart, B. S., Flemings, P. B., and Deshpande, A., 1995, Porosity and Pressure: Role of Compaction Disequilib-rium in the Development of Geopressures in a Gulf Coast Pleistocene Basin: Geology, v. 23, p. 45–48.

Harwood, D. S., and Helley, E. J., 1987, Late Cenozoic tectonism of the Sacramento Valley, California, UnitedStates Geological Survey Professional Paper 1359, 46 p.

Hull, L. C., 1984, Geochemistry of Ground Water in the Sacramento Valley, California: United StatesGeological Survey Professional Paper 1401-B, 36 p.

Hunt, J. M., 1990, Generation and migration of petroleum from abnormally pressured fluid compartments:American Association of Petroleum Geologist Bulletin, v. 74, p. 1–12.

Ingersoll, R. V., 1982, Initiation and evolution of the Great Valley forearc basin of northern and centralCalifornia, USA, in Leggett, J. K., editor, Trench-forearc geology: sedimentation and tectonics on modernand ancient active plate margins: Geological Society of London Special Publication 10, p. 459–467.

Ingersoll, R. V., 1988, Development of the Cretaceous forearc basin of central California, in Graham, S. A.,editor, Studies of the geology of the San Joaquin Basin: Society of Economic Paleontologists andMineralogists, Pacific Section, v. 60, p. 141–155.

Ingersoll, R. V., Rich, E. I., and Dickinson, W. R., 1977, Field guide: Great Valley Sequence, SacramentoValley: Geological Society of America, Cordilleran Section, Annual Meeting, Guidebook no. 8.

International Formulation Committee, 1967, A formulation of the thermodynamic properties of ordinarywater substance: Dusseldorf, Germany, International Formulation Committee Secretariat.

Lee, I. K., editor, 1968, Soil Mechanics, Selected Topics: New York, Elsevier.Luo, X., and Vasseur, G., 1992, Contributions of compaction and aquathermal pressuring to geopressure and

the influence of environmental conditions: American Association of Petroleum Geologists Bulletin, v. 76,p. 1550–1559.

————— , 1995, Modelling of pore pressure evolution associated with sedimentation and uplift in sedimentarybasins: Basin Research, v. 7, p. 35–52.

Magara, K., 1975, Reevaluation of montmorillonite dehydration as cause of abnormal pressure and hydrocar-bon migration: American Association of Petroleum Geologists Bulletin, v. 59, p. 292–302.

Marks, S. M., and Lindh, A. G., 1978, Regional seismicity of the Sierran foothills in the vicinity of Oroville,California: Bulletin of the Seismological Society of America, v. 68, no. 4, p. 1103–1115.

McPherson, B. J. O. L., ms, 1996, A Three-Dimensional Model of the Geologic and Hydrodynamic History ofthe Unita Basin, Utah: Analysis of Overpressures and Oil Migration: Ph. D. Disseratation, University ofUtah, Salt Lake City, 119 p.


Melchioore, E. B., Criss, R. E., and Davission, M. L., 1999, Relationship between seismicity and subsurfacefluids, central Coast Ranges, California: Journal of Geophysical Research, v. 104, p. 921–939.

Miall, A. D., 1990, Principles of Sedimentary Basin Analysis, 2nd edition: New York, Springer-Verlag.Mount, V. S., and Suppe, J., 1987, State of stress near the San Andreas fault—implications for wrench tectonics:

Geology, v. 15, p. 1143–1146.Moxon, I. W., ms, 1990, Stratigraphic and structural architecture of the San Joaquin-Sacramento basin: Ph.D.

dissertation, Stanford University, California, 563 p.Moxon, I. W., and Graham, S. A., 1987, History and controls of subsidence in the Late Cretaceous-Tertiary

Great Valley forearc basin, California: Geology, v. 15, no. 7, p. 626–629.Neuzil, C. E., 1986, Groundwater flow in low-permeability environments: Water Resources Research, v. 22,

p. 1163–1195.————— , 1993, Low fluid pressure within the Pierre Shale: a transient response to erosion: Water Resources

Research, v. 29, p. 2007–2020.————— , 1995, Abnormal pressures as hydrodynamic phenomena: American Journal of Science, v. 295,

p. 742–786.Nilsen, T. H., and McKee, E. H., 1979, Paleogene paleogeography of the western United States, in Armentrout,

J. M., Cole, M. R., and TerBest, H., Jr., editors, Cenozoic paleogeography of the western United States:Society of Economic Paleontologists and Mineralogists, Pacific Section, Pacific Coast PaleogeographySymposium 3, Anaheim, California, p. 257–276.

Ord, A., and Oliver, H. S., 1997, Mechanical controls on fluid flow during regional metamorphism: somenumerical models: Journal of Metamorphic Geology, v. 15, p. 345–359.

Page, R. W., 1986, Geology of the fresh ground-water basin of the Central Valley, California, with texture mapsand sections: United States Geological Survey Professional Paper 1401-C, 54 p.

Person, M., Raffensperger, J. P., Ge, S., and Garven, G., 1996, Basin-scale hydrogeologic modeling: Reviews ofGeophysics, v. 34, p. 61–87.

Pruess, K., 1991, TOUGH2—a general-purpose numerical simulator for multiphase fluid and heat flow:Berkeley, California, Lawrence Berkeley Laboratory Report LBL-29400,

Redwine, L. E., ms, 1972, The Tertiary Princeton submarine valley system beneath the Sacramento Valley,California: Ph.D. dissertation, University of California, Los Angeles, 480 p.

Rice, J. R., and Cleary, M. P., 1976, Some basic stress diffusion solutions for fluid-saturated elastic porousmedia with compressible constituents: Reviews of Geophysics and Space Physics, v. 14, p. 227–241.

Sass, J. H., Blackwell, D. D., Chapman, D. S., Costain, J. K., Decker, E. R., Lawver, L. A., and Swanberg, C. A.,1981, Heat flow from the crust of the United States, in Touloukian, Y. S., Judd, W. R., and Roy, R. F.,editors: Physical Properties of Rocks and Minerals: New York, McGraw-Hill, p. 503–548.

Sass, J. H., Lachenbruch, A. H., Munroe, R. J., Greene, G. W., and Moses, T. H., Jr., 1971, Heat flow in thewestern United States: Journal of Geophysical Research, v. 76, p. 6376–6413.

Sass, J. H., and Morgan, P., 1988, Conductive heat flux in VC-1 and the thermal regime of Valles Caldera,Jemez Mountains, New Mexico: Journal of Geophysical Research, v. 93, p. 6027–6039.

Schneider, F., Burrus, J., and Wolf, S., 1993, Modelling overpressures by effective-stress/porosity relationshipsin low-permeability rocks: empirical artifice or physical reality?: Elsevier, Amsterdam, NorwegianPetroleum Society Special Publication 3, p. 333–341.

Screaton, E. J., Wuthrich, D. R., and Dreiss, S., 1990, Permeability, fluid pressures, and flow rates in theBarbados ridge complex: Journal of Geophysical Research, v. 95, p. 8997–9007.

Shi, Y., and Wang, C., 1986, Pore pressure generation in sedimentary basins: overloading versus aquathermal:Journal of Geophysical Research, v. 91, p. 2153–2162.

————— , 1988, Generation of high pore pressures in accretionary prisms: inferencers from the Barbadossubduction complex: Journal of Geophysical Research, v. 93, p. 8893–8910.

Silver, E. A., 1971, Tectonics of the Mendocino triple junction: Geological Society of American Bulletin, v. 82,p. 2965–2978.

Spencer, C. W., 1987, Hydrocarbon generation as a mechanism for overpressuring in Rocky MountainRegion: American Association of Petroleum Geologists Bulletin, v. 71, no. 4, p. 368–388.

Timko, D. J., and W. H. Fertl, 1971, Relationship between hydrocarbon accumulation and geopressure and itseconomic significance, Journal Petroleum Technology, v. 23, p. 923–933.

Unruh, J. R., Davisson, M. L., Criss, R. E., and Moores, E. M., 1992, Implications of perennial saline springs forabnormally high fluid pressures and active thrusting in western California: Geology, v. 20, p. 431–434.

Unruh, J. R., Loewen, B. A., and Moores, E. M., 1995, Progressive arcward contraction of a Mesozoic-Tertiaryfore-arc basin, southwestern Sacramento Valley, California: Geological Society of America Bulletin,v. 107, p. 38–53.

van der Kamp, G., and Gale, J. E., 1983, Theory of earth tide and barometric effects in porous formations withcompressible grains: Water Resources Research, v. 19, p. 538–544.

Villeges, M. E., Bachu, S., Ramon, J. C., and Underschultz, J. R., 1994, Flow of formation waters in theCretaceous-Miocene succession of the Llanos Basin, Colombia: American Association of PetroleumGeologists Bulletin, v. 78, no. 12, p. 1843–1862.

Wakabayashi, J., and Smith, D. L., 1994, Evaluation of recurrence intervals, characteristic earthquakes, and sliprates associated with thrusting along the Coast Range—Central Valley geomorphic boundary, California:Bulletin of the Seismological Society of America, v. 84, p. 1960–1970.

Walder, J., and Nur, A., 1984, Porosity-reduction and crustal pore pressure development: Journal ofGeophysical Research, v. 89, p. 11539–11548.

Waltham, D., and Hardy, S., 1995, The velocity description of deformation, Paper 1: theory: Marine andPetroleum Geology, v. 12, p. 153–163.


Wentworth, C. M., and Zoback, M. D., 1989, The style of later Cenozoic deformation at the eastern front of theCalifornia Coast Ranges: Tectonics, v. 8, p. 237–246.

Wieck, J., Person, M. A., and Strayer, L., 1995, A finite element method for simulating fault block motion andhydrothermal fluid flow within rifting basins: Water Resources Research, v. 31, p. 3241–3258.

Williams, H., and Curtis, G. H., 1977, The Sutter Buttes of California, a study of Plio-Pleistocene volcanism:Berkeley, University of California Publications in Geological Sciences, v. 116, 56 p.

Williams, T. A., 1993, Current views on the deposition and sequence stratigraphy of the Great Valley Group,Sacramento basin, in Graham, S. A., and Lowe, D. R., editors, Advances in the sedimentary geology of theGreat Valley Group, Sacramento Valley, California: Society of Economic Paleontologists and Mineralo-gists, Pacific Section, Book 73.

Williamson, A. K., Prudic, D. E., and Swain, L. A., 1989, Ground-water flow in the Central Valley, California:United States Geological Survey Professional Paper 1401-D, 127 p.

Wilson, A. M., Garven, G., and Boles, J. R., 1999, Paleohydrogeology of the San Joaquin Basin, California,Geological Society of America Bulletin, v. 111, no. 3, p. 432–449.

Wong, I. G., Ely, R. W., and Kollmann, A. C., 1988, Contemporary seismicity and tectonics of the northernand central Coast Ranges—Sierran Block Boundary Zone, California: Journal of Geophysical Research,v. 93, p. 7813–7833.

Zoback, M. D., Zoback, M. L., Mount, V., Suppe, J., Eaton, J. P., Healy, J. H., Oppenheimer, D., Reasenburg,P., Jones, L., Raleigh, C. B., Wong, I. G., Scotti, O., and Wentworth, C. M., 1987, New evidence on thestate of stress of the San Andreas fault: Science, v. 238, p. 1105–1111.

B. J. O. L. McPherson and G. Garven466

june 1999 hydrodynamics and overpressure mechanisms in the sacramento...

Documents