A Middle Pleistocene Lacustrine
Delta in the Kern River
Depositional System: Structural
Control, Regional Stratigraphic
Context, and Impact on
Groundwater Quality
ABSTRACT
A 200- to 400-ft-thick coarsening-upward sequence of sediments was
mapped at depths of 300 to 700 fbgs under the western margin of the
Kern River alluvial fan in the southern San Joaquin Valley of California.
Sedimentary petrography and x-ray diffractometry of samples from wells
containing this unit indicate a predominantly Sierran source for the clastic
component of the sediments. The basal part of this unit is typically fine
grained (clays and silts) and grades upward into medium sands. The unit
thickens toward the Buena Vista Lake terminal basin and is interpreted to
have been deposited as a lacustrine delta as part of an alluvial fan-delta
system that prograded into its terminal basin. Relatively high amounts of
organic carbon and extremely low magnetic susceptibility suggest that the
unit was deposited under reducing geochemical conditions. Authigenic
euhedral and framboidal pyrite crystals indicate that reducing conditions
progressed to the sulfate reduction stage.
The coarsening-upward sequence is thickest in a rectangular, prismshaped trough bounded by the surface projection of normal faults that
were previously mapped in deeper sedimentary units. This observation
suggests that the deposition of this unit was at least in part structurally
controlled. Basinward, the unit laps onto or grades into a thick depocenter
clay layer that is tentatively correlated to the Corcoran Clay that was
deposited in a valley-wide lake prior to 600,000 years ago.
In the Kern Water Bank area, the thickest part of the coarseningupward unit lies in the same region as a set of wells with anomalously high
concentrations of groundwater arsenic. Relatively high concentrations
of easily-exchangeable arsenic were found in samples from a well
within the coarsening-upward unit. These observations suggest that,
in the reducing lacustrine setting of this unit, arsenic was locked up in
pyrite reservoirs. Dissolution textures observed in pyrite crystals under
a scanning electron microscope suggest a post-depositional transition
to more oxidizing conditions and the subsequent release of arsenic into
solution in groundwater and in easily-exchangeable sites on mineral
surfaces.
AUTHORS
Robert Negrini
Dirk Baron
Janice Gillespie
Robert Horton
Anne Draucker
Neil Durham
John Huff
Paul Philley
Carol Register
Department of Geology
California State University
Bakersfield, CA 93311
Jonathan Parker
Kern Water Bank Authority
5500 Ming Avenue, Suite 490
Bakersfield, CA 93309
Thomas Haslebacher
Kern County Water Agency
P.O. Box 58
Bakersfield, CA 93302
Pacific Section AAPG Publication MP48 95
INTRODUCTION
For the past million years or so the Central Valley of
California has received thousands of cubic miles of alluvial
sediments transported into the valley by several major
rivers, principally from the Sierra Nevada. These sediments
serve as important freshwater aquifers that are a source of
water for one of the world’s foremost agricultural centers,
for wildlife refuges, and for rapidly growing population
centers. Therefore, it is becoming increasingly important to
understand the details of this aquifer system, including the
stratigraphic and geochemical controls on water availability
and quality. The uppermost of these sediments were
deposited during the anomalously high amplitude climate
swings of the middle to late Pleistocene (e.g., Ruddiman,
2001; Zachos et al., 2001). Thus their study bears on the
paleoclimatology of western North America.
The current work addresses the above-described goals
while focusing on a major stratigraphic feature found within
the sediments of the Kern Water Bank (KWB) near the distal
edge of the Kern River alluvial fan system (Figure 1). This
feature, a combined unit consisting of a coarsening-upward
sequence and an underlying uniformly fine-grained unit,
appears to have been influenced both by local structures
and regional/global climate events. In turn, its lithological
features and associated environment of deposition appear
to play a significant role in local groundwater quality.
Figure 1. Location map of project area and surrounding features
discussed in the text. (a) Generalized geologic map after Page
(1986). (b) Well log coverage in project area. Logs with wells
shown as black filled circles. Sediment analyses were done on
samples from wells 23H and 24K (green filled circles). Magnetic
susceptibility was measured on samples from these wells and also
measured on samples from wells labeled with filled grey circles.
Kern Water Bank boundaries shown as red dashed lines.
96 Pacific Section AAPG Publication MP48
Figure 2. Late Tertiary stratigraphic column of southern
San Joaquin Valley (after Bartow, 1991). Sediments
of this study belong to the younger part of the Tulare
Formation.
BACKGROUND
The KWB is a large groundwater storage and
management facility with a capacity of ~1,000,000 acre-ft
located on the western margin of the modern Kern River
alluvial fan (Figure 1). The modern Kern River alluvial
fan system is located at the southernmost extreme of the
Central Valley (Figure 1). The alluvial fan covers an area of
approximately 800 mi2 (~2,000 km2). This system has been
the subject of many studies, mostly in the form of reports
associated with groundwater resources (Mendenhall et al.,
1908; 1916; Hoots et al., 1954; Dale et al., 1964; 1966;
Manning, 1967; Croft, 1972; Page, 1986; Williamson et al.,
1989; Bartow, 1991; PGA and Associates, 1991).
The late Pleistocene sediments of the modern Kern
River alluvial fan are traditionally assigned to either the
uppermost Tulare Formation or the Kern River Formation,
both principally units of nonmarine origin. This designation
is primarily dependent on proximity of the study area to
either the west or east side of the San Joaquin Valley (e.g.,
Plate 2 of Bartow, 1991). Because our study area is closer
to the west side of the valley and because recent work on
the Kern River Formation suggests that it is considerably
older than previously thought (Miller, 1999; Golob et al.,
2005; Baron et al., 2008), we assign the sediments of this
study to the Tulare Formation (Figure 2).
The lithology of the sediments of this study consists
of gravel, sand, silt, and clay with the proportion of fine
material increasing away from the mouth of the Kern
Canyon (Dale et al., 1966; Croft, 1972). Along the western
and southern edges of the fan these easterly-derived
sediments interfinger with relatively minor amounts of
sediments derived from the Coast Ranges and Transverse
Ranges, respectively (Croft, 1972).
METHODS
Sediment Analyses
Several sediment analyses were conducted mainly on
grab samples collected at 10 to 20 foot intervals from wells
T30S/R25E-23H and T30S/R25E-24K (Figure 1). These
wells were studied because they penetrated the stratigraphic
unit receiving primary focus in this study. Also, the wells
contained contrasting levels of groundwater arsenic
according to laboratory reports on file at the Kern Water
Bank Authority in Bakersfield, CA and Boockoff (2005).
Arsenic concentrations in well 23H were measured at 63
ppb in September, 2000, 55 ppb in April, 2003, and 30-33
ppb in February of 2005, whereas well 24K over the same
time period has consistently had concentrations that were
low enough (1.85 ppb) to be near detection limits.
Grain-size was determined by assigning sand,
silt, and clay percentages using the Wentworth scale and
standard percentage templates (e.g., Compton, 1985).
Mineralogy and petrography of sand- and silt-sized grains
were determined using petrographic microscopy augmented
by scanning electron microscopy and microprobe analysis.
Fifty-four thin sections were point counted (300 points per
section) using an automated point-counting stage. Because
the grab samples were generally unconsolidated, no
textural analysis was possible during thin-section analysis.
Mineralogy of fine-grain material in 32 samples was further
constrained using x-ray diffractometry.
Total organic carbon (TOC) content was determined
at 10-ft depth intervals using the loss-on-ignition method
(Dean, 1974). Samples were dried for two days at 45 ˚C in
a Fisher Isotemp Oven to expel water and then they were
exposed to 600 ˚C in a Lindberg Muffle furnace for 1 hour
to combust the organic carbon. The mass of the sample was
measured before and after each step.
Magnetic susceptibility (MS) was measured with a
Bartington MS2 susceptibility meter after samples were
placed in an MS2B bottle sensor. This was converted to
mass-specific MS by dividing through by the density of
each sample. Because MS can be measured relatively
rapidly, these data were acquired for a total of eight wells
(including wells 23H and 24K) along a transect parallel to
a radius of the Kern River alluvial fan.
Arsenic concentrations of sediment grab samples were
measured in a Perkin-Elmer ELAN 6100 ICP-MS after the
exchangeable fraction was isolated using the methods from
Tessier et al. (1979). This extraction isolates arsenic weakly
adsorbed to mineral surfaces that could be released into the
groundwater due to changes in pH and ionic strength, or
due to increases in the concentrations of competing ions
such as phosphate.
Pacific Section AAPG Publication MP48 97
Correlation and Mapping of Subsurface Units
The primary subsurface mapping data used in this study
consists of electric (short normal resistivity) logs gathered
from water and oil wells. Most of these logs are from the
Kern Water Bank. The database includes electric logs from
162 wells in the KWB and nearby groundwater projects,
covering an area of approximately 80 mi2 (~200 km2),
most of which penetrate to depths of several hundred feet
or more. These electric logs were digitized and then entered
into a project in Geographix™, a subsurface Geographic
Information System (GIS) software package commonly
used in the petroleum industry. Lithologic units were
defined based on grain-size distributions inferred from log
resistivity values, an assumption that is addressed below
in the Results section. These units were then correlated to
delineate surfaces corresponding to the bases and tops of
stratigraphic units. The resultant surfaces were contoured
as structure maps and the thicknesses of the units were
contoured as isochore maps.
RESULTS
Mineralogy and Petrography
Based on thin section analyses of the grab samples, the
sediments of the Kern Water Bank in the region of wells
23H and 24K were determined to be arkosic arenites and
wackes dominated by quartz, plagioclase, and K-feldspars
(orthoclase and microcline) (Figure 3). Plagioclase shows
evidence of dissolution at all depths. Coarser samples contain
abundant rock fragments; these are mainly microphaneritic
grains of granitic composition but these fragments include
quartzite, schist, chert, siltstone, and shale. Most finegrained lithic fragments consist of shale; these are much
more abundant in samples from well 23H than those from
well 24K (Figure 3). A few samples contain serpentinite
grains. Accessory minerals are dominated by biotite and
hornblende with lesser muscovite and opaques (including
ilmenite and framboidal pyrite composed of cubic and
octahedral crystals). Some samples contain epidote, zircon,
titanite, rutile, chlorite, calcite (or limestone), tremolite, and
/or phosphate. The fine-grained sediments are dominated
by smectite (R=0, I/S = 90% smectite) but include illite/
mica, kaolinite, and chlorite. Pyrite is common in fine-
Figure 3. Composition of the coarse fraction of Figure 4. (a) Typical pyrite (white) concentrations of shale
sands from Kern Water Bank wells 23H and 24K. Q = images inspected in this study. Pyrite represents ~1-2% of the
monocrystalline, polycrystalline and lithic quartz, F = total area shown in this view.
feldspars (includes lithic feldspars), L = fine-grain lithics
(mainly shale clasts). Plotted following the methodology
of Dickinson (1970); lithic quartz and lithic feldspars are
crystals in microphanerites and phenocrysts in volcanic
grains.
98 Pacific Section AAPG Publication MP48
d
e
Figure 4. (b-c) Closeup views of spherules of framboidal pyrite and octahedron-shaped authigenic pyrite crystals in shale
(polished thin section and grain mount). SEM backscattered-electron image. Well 23h, depth = 690 ft. (d) Microprobe-based
back-scattered electron micrograph and wavelength dispersive X-ray element maps of detrital grains from well 23H, depth =
690 ft. Pyrite is indicated by bright spots present in both the Fe and S maps (arrows). Arsenic is present in both pyrites. The
framboid on the left show exhibits a low-pyrite core and hig-pyrite rim while the mass at the right shows a spherical low-pyrite
core coated with a zone of variable but lower pyrite levels. (e) Microprobe-based wavelength dispersive spectrum of pyrite
in framboidal spherule showing La peak for arsenic distinctly above background. Depth = 690 ft. (f-g) Dissolution features in
framboidal pyrite (f) and octahdron-shaped authigenic pyrite crystals (g) in shale (grain mount). SEM backscattered-electron
image. Well 23h, depth = 690 ft.
Pacific Section AAPG Publication MP48 99
grained sediments from both wells, occurring as spherical
framboids and irregular masses of euhedral cubic and
octahedral crystals (Boockoff, 2005; Negrini et al., 2006).
Pyrite in fine-grained sediments from well 23H was found
to contain trace quantities of arsenic as high as 0.35 weight
percent (Figure 4). Clinoptilolite, gypsum, Fe-oxide,
pyroxenes, rare microfossils, garnets, and/or tourmaline
are also present in some fine-grained samples. Biotite is
most abundant in the shallowest samples, but otherwise no
trends with depth were observed in either fine- or coarsegrained samples.
Total Organic Carbon, Magnetic Susceptibility, and
Exchangeable As
Results for TOC, MS, and exchangeable As (in
from wells 23H and 24K. TOC ranges from 0 to 8%, MS
from 7X10-8 m3/kg to 80X10-8 m3/kg, and exchangeable
As from 1 to 15 ppm. TOC is highest and MS is lowest
within the middle depth intervals of these wells (~650-250
fbgs). In fact, these anomalously low MS values are within
a factor of 2 or so of the instrument sensitivity. Except
for a one-sample spike near the bottom of the well, the
concentration of exchangeable arsenic in well 23H is at its
highest values in the middle interval with maximum values
occurring between 400 and 600 ft below ground surface,
toward the base of the interval with high TOC and low
MS values. Exchangeable arsenic in the 24K well is also
elevated slightly in this interval, but not nearly as much as
it is in the 23H well.
The MS logs for all of the wells surveyed in the Kern Water
sediment) are shown in Figure 5a-b for the grab samples
Bank area are shown in Figure 5c. The MS signature of
Figures 5a and 5b. Total organic carbon content (TOC), magnetic
susceptibility (MS), and sediment arsenic concentrations from
grab samples of (a) Well 23H and (b) Well 24K. Black lines
on the arsenic plots correspond to total arsenic concentrations.
Grey lines correspond to arsenic in easily exhangeable sites on
mineral surfaces (e.g., clays). The latter is usually associated
with higher concentrations of groundwater arsenic.
100 Pacific Section AAPG Publication MP48
Figure 5c. Magnetic susceptibility (MS) logs from grab
samples of eight wells from the Kern Water Bank. Note
that wells 23H and 24K have extremely low MS values in
the middle depth ranges.
the 23H and 24K wells was not typical. For the six wells
from other regions of the KWB, MS decreased with
depth though never attained values as low as those found
in the middle depth intervals in wells 23H and 24K. If
MS characteristically varies in opposition with TOC as
suggested by the data in Figure 5a-b, then the TOC content
of the sediments in the region around wells 23H and 24K
is anomalously high, particularly in the depth range below
a few hundred fbgs.
Electric-Log Resistivity as an Indicator of Grain-Size
The formation fluid in the unconsolidated sediments
from the upper 1,000 ft in the Kern Water Bank is fresh
water with an average TDS value of ~220 ppm (e.g.,
Figure 6a. Grain-size of grab samples compared to electric log
resistivity for Well T30R25S23H. i. Sand/silt/clay percentage
diagram. Percentages were determined from grab samples
spaced at 10ft intervals. The symbol “gr*” indicates the
presence of gravels and very coarse sands. ii. Short-normal
electric log for well T30R25S23H. Resistivity was measured at
0.5 ft intervals. iii. Clay percentage of grab samples. iv. 10 ft
averages of electrical resistivity data centered on depths from
which sediment grab samples were taken. Dashed lines in iii
and iv are smoothed values using a depth window equal to
25% of dataset.
KWBA laboratory reports; Boockoff, 2005). Under these
conditions the bulk sediment resistivity should depend
principally upon clay content which, in turn, is closely
related to bulk grain size. The relationship between clay
content in bulk sediment and resistivity is due to the cation
exchange capacity of clays and the ease of current flow
through the corresponding diffuse layer of exchangeable
cations near grain boundaries (Keller and Frischknecht,
1970; Hallenburg, 1984; Ward, 1990). Thus resistivity
potentially can be used as a crude estimator of grain-size
distribution (e.g., Abu-Hassanein et al., 1996).
This presumption has precedent in the study region.
For example, resistivity logs were found to give the best
grain size distinction between fine- and coarse-grained
Figure 6b. Grain-size of grab samples compared to
electric log resistivity for well T30R25S24K.
Pacific Section AAPG Publication MP48 101
Figure 6c. Electrical resistivity as a function of %fines for well 23H. Resistivity plotted is electric
log resistivity from Figure 6.a.iv averaged for all grab samples with corresponding %fines value.
alluvial sediments in the Kern River oilfield that lies east
of the water bank near the apex of the Kern Fan (Miller,
1986). The Kern River oilfield is a good analog for the
Kern Water Bank in that the sediments in question are
unconsolidated, are within ~1500 feet of the surface, and
have relatively fresh (<650 ppm total dissolved solids) pore
waters (Coburn, 1996).
We tested this presumption for the sediments of the
Kern Water Bank by comparing grain-size with well log
resisitivity data (Figure 6). To a first order the proposed
inverse relationship between clay content and resistivity
holds as shown in Figure 6a.i-ii. The resisitivity is highest
(~100 ohm-m) when no clay is present (e.g., sand spikes
centered at 275 and 625 fbgs in well 23H) and is relatively
low when clay percentage is high (e.g., average resistivity
is ~20 ohm-m in the interval from ~600-300 fbgs). In order
to account for sampling interval differences between that
of the grab samples (10 ft) and logging tool measurements
(0.5 ft) in Figures 6a.i-ii and 6b.i-ii, the clay percentage
of the grab samples (Figures 6a.iii and 6b.iii) is plotted
alongside log resistivity (Figure 6a.iv and 6b.iv) after
the latter is averaged over 10ft intervals centered on the
depth of each corresponding grab sample. It’s evident from
these plots that long-wavelength resistivity changes are
associated with similar changes in clay percentage.
102 Pacific Section AAPG Publication MP48
Following the approach of Abu-Hassanein et al.
(1996), we further investigated the resisitivity/grainsize relationship in well 23H by calculating the average
resistivity corresponding to sets of grab samples with
identical percentages of fine grains and then plotting this
value against “percent fines” (Figure 6c). These data also
support a first order, inverse relationship between the
percentage of fine sediments and resistivity.
In summary, down to a depth of at least 1000 ft, welllog resistivity values in the Kern Water Bank area exhibit
a first-order relationship with the percentage of finegrained sediments. This relationship justifies the inference
of relative sediment grain-size from log resistivity values
in defining lithologic units from well logs. Based on our
results and the previous results of Coburn (1996), we
extend this inference to rest of the southern San Joaquin
Valley well logs included in this report.
Stratigraphic Correlation and Mapping
Figure 7 shows a representative resistivity log in which
large-scale sedimentary sequences have been identified. One
sequence, herein called the “Coarsening-Upward Sequence
2” (CUS2) was chosen for detailed study because its entire
depth interval was present in most of the wells of this study,
thus allowing its correlation and mapping throughout the
Figure 7. E-log from a well containing
the prograding fan delta unit (CUS2)
in the Kern Water Bank region. CUS2
was defined to include the interval
of uniformly low resistivities usually
found below the coarsening upward
sequence. To the right of the e-log
is a schematic diagram of a vertical
sequence typical of prograding fan
deltas in Lake Hazar, Turkey (Wescott
and Etheridge, 1990). The lake bottom
subunit occasionally contains smaller
scale fining upward sequences which
may be turbidite deposits.
Figure 8. Structure contour maps on the (a) top and (b) bottom of the CUS2 (Coarsening-Upward Sequence 2) unit. (c)
Isochore (aka, thickness) contour map of the CUS2 unit. Area mapped is the same as that shown in Figure 1b. Contour
labels are elevations in feet relative to mean sea level. Note that, in this figure, the contour fill colors for both structure maps
are tied into the same absolute elevation scale. Thus, “deep” elevations for the “Top of LsCus2” map are similar in color to
the “shallow” elevations for the “Bottom” map. Interstate Highway 5 (black line), the modern Kern River (dashed blue line)
and the outlines of the Kern Water Bank properties (red dashed lines) are shown as landmarks. The base of the CUS2 unit
shows a broad, low-elevation feature that roughly follows the modern course of the Kern River and then takes a sharp turn
southward toward the location of the modern Buena Vista Lake terminal basin (Figure 1). This observation plus the much
lower relief exhibited by the top of the unit and the increasing thickness basinward of the unit support the hypothesis that
the CUS2 unit is a sublacustrine delta deposit prograding SSW-ward into a more extensive, ancestral Buena Vista Lake
terminal basin.
Pacific Section AAPG Publication MP48 103
study area (Figure 8). The uppermost coarsening-upward
sequence was not mapped because its top often resides
above the dramatically changing water table of the Kern
Water Bank through time. Thus the amplitude of the
resistivity for this uppermost unit is strongly dependent on
the logging date.
The CUS2 unit was found in 60 wells that were
primarily located in the southwestern quarter of the study
area (Figure 8). Based on typical resistivity values and
complementary inspections of grab samples from wells,
this unit is uniformly fine-grained (clays and silts) at its
base and coarsens upward typically into medium-grained
sands. The coarsening-upward log signature of this unit
was particularly well defined in logs from wells located
along the east side and close to the southern margin
Vista Lake terminal basin (Figure 1). The structural map
on the top of this unit (Figure 8a) exhibits markedly less
relief. The thickness of the CUS2 unit abruptly increases
basinward to >300 ft as it nears the terminal basin as
indicated by the isochore map in Figure 8c. A cross section
oriented approximately north-south through the CUS2 unit
(Figure 10) demonstrates that the southward thickening
CUS2 unit shown in Figure 8c continues southward into
the Buena Vista terminal basin where its lower part grades
laterally into and overlaps a thick unit of uniformly low
resistivity (Figure 10).
of township T30S/R25E and adjacent to this region in
township T31S/R25E. This location is also characterized
by net sand percentages of less than 60% within the Zone 4
depth interval (300 to 150 fbsl) most commonly containing
the uniformly fine-grained, bottom half of CUS2 (Figure
9). The structural map on the base of this unit (Figure 8b)
depicts a trough-like low that projects into this same region
from the east as part of a larger paleochannel. This feature
reaches maximum depths at the south end of the area in
township T31S/R25E toward the modern site of the Buena
The CUS2 is a coarsening-upward unit correlated
across the Kern Water Bank region. It is best defined and
found with more regularity in the south-central portion of
the area toward the toe of the Kern River alluvial fan. At
this location, the base of the CUS2 unit consists mostly of
uniformly fine-grained silts and clays and the unit abruptly
thickens basinward (Figures 8 and 10). On the structure map,
the bottom of the unit also increases basinward (Figure 8a)
and the structural relief of the base of CUS2 is significantly
greater than that of its top (Figure 8b). Taken together
DISCUSSION
Depositional and Geochemical Environment of CUS2
Figure 9. Net sand map of Kern Water Bank area for depth interval of 300-150 fbsl (Zone 4).
104 Pacific Section AAPG Publication MP48
Figure 10. Cross-section from Kern Water Bank into the Buena Vista Lake terminal basin showing relationship between
CUS2 unit and BV lake clay unit. The latter may be correlated to the Corcoran Clay of Croft (1972).
these observations suggest that CUS2 was deposited
in a prograding alluvial fan-delta setting that evolved
basinward into a prograding lacustrine delta that includes
a distal clay/silt unit (e.g., Wescott and Ethridge, 1990).
Out in the Buena Vista Lake depocenter, the CUS2 unit
grades laterally into and overlaps a thick unit of uniformly
fine-grained resistivity that we interpret to be a coeval lakebottom clay (Figure 10), the expected basinward facies of
the fan-delta system (Wescott and Ethridge, 1990).
The sediment analyses of the CUS2 unit are consistent
with the fan-delta hypothesis. First, the CUS2 sediments are
relatively high in TOC and low in MS (Figure 5) and this
low MS signature appears to be confined to the region where
the CUS2 unit is thickest (Figures 1, 5c, and 8). High TOC
and low MS is an association that is commonly observed in
reducing lacustrine environments due to the dissolution of
magnetite during digestion of organic matter by anaerobic
bacteria (Cohen, 2003; Evans and Heller, 2003). Second,
clusters of pyrite found in fine-grained sediments of well
23H have either euhedral and/or framboidal morphologies
that indicate an authigenic origin (Figure 4). All of these
observations are suggestive of the reducing conditions
expected of a lacustrine depositional environment as
opposed to the relatively oxidizing conditions expected for
an alluvial fan setting (Cohen, 2003).
During sediment diagenesis, bacterial decay of organic
carbon can quickly deplete pore waters of dissolved oxygen.
If this is not replaced by additional oxygenated groundwater
infiltrating from the surface, the system quickly turns
anoxic. Under these conditions, remaining organic material
becomes a food source for anaerobic bacteria. More
specifically, as oxygen is depleted as an electron acceptor,
a series of bacterially-mediated reactions use first Fe3+
from ferric oxides and oxyhydroxides and then S6+ from
dissolved sulfate as electron acceptors for the continued
breakdown of organic materials (e.g., Langmuir, 1997).
As a consequence, detrital iron oxides and hydroxides
dissolve and iron is released into groundwater as Fe2+.
For this reason, MS is often used to estimate the extent
of iron reduction (Evans and Heider, 2003). In the case of
the CUS2 sediments, the disappearance of the MS signal
indicates that most, if not all, of the available Fe-oxides and
hydroxides were consumed. When all of the iron has been
reduced, sulfate acts as the electron acceptor and reduced
sulfur from sulfate reduction reacts with dissolved Fe2+
to form authigenic pyrite, thus explaining the framboidal
(and euhedral) pyrite crystals found in well 23H (Figure 4;
Boockoff, 2005; Negrini et al., 2006).
Pacific Section AAPG Publication MP48 105
Figure 12. Anomalous groundwater arsenic concentrations and
surface projection of subsurface faults (Jones and Gillespie, 1997)
superimposed upon CUS2 isochore contour map.
Figure 11. “Lake Clyde,” the Pleistocene lake in which
the Corcoran Clay (and CUS2?) were deposited (from
Harden, 2004).
Regional Stratigraphic Context and Hypothesized Age of
CUS2
As discussed above, the CUS2 unit grades into and overlaps
the upper part of a depocenter clay unit in the Buena Vista
Lake basin. Clay layers found in the Buena Vista Lake
basin at similar depths have been previously correlated to
the Corcoron Clay (Croft, 1972; Page, 1986; Kiser et al.,
1988). The Corcoran Clay is a widespread lacustrine clay
deposited throughout most of the Central Valley of California
800 to 600 kyr ago (Frink and Hues, 1954; Croft, 1972;
Sarna-Wojcicki, 1995; Harden, 1999). The lower age of the
Corcoran Clay is constrained by the Matuyama/Brunhes
magnetic reversal (789 kyr) found just below the Corcoran
Clay in a core from near Wasco, CA, and the eruption of the
Bishop ash bed (774±2 kyr), which is found just above the
base of the clay bed in the same core (Davis, 1978; SarnaWojcicki, 1995; Sarna-Wojcicki et al., 2000). The upper
age of the Corcoran Clay is constrained by the eruption of
106 Pacific Section AAPG Publication MP48
the Lava Creek ash bed (639± 5 kyr), which is found just
below the top of the clay in the Wasco core (Davis, 1978;
Sarna Wojcicki, 1995; Lanphere et al., 2002). The depth of
the Buena Vista Lake clay associated with CUS2 (~7001000 ft) is consistent with that of the Corcoran clay 60
miles north in the Tulare Lake basin (Croft, 1972) and the
depth of coeval units found in the Owens Lake basin 100
miles to the northeast (Sarna-Wojcicki et al., 1997). Thus
its presumed correlation with the Corcoran Clay yields a
consistent sedimentation rate (~1-2 ft/1000 yr) with other
nearby basins in similar depositional settings.
Presuming that the depocenter clay in Figure 10
is indeed the Corcoran Clay, then this clay and the
stratigraphically-related CUS2 fan-delta deposit were
deposited in Lake Clyde (Figure 11), a lake that covered
most of the Central Valley of California over a time period
of more than 100 kyr during the middle Pleistocene (SarnaWojcicki, 1995; Harden, 1999). Notably, the Corcoran
Clay has not been found in the region of the Bakersfield
Arch, an active structural feature that is responsible for a
broad, uplifted, northeast-trending region in the vicinity of
Bakersfield (PGA and Associates, 1991; Fig. 1) upon which
the Kern River alluvial fan was deposited. It is likely that
the Bakersfield Arch region stood above lake-level during
most of this time and separated the Tulare and Buena Vista
subbasins as it does today.
Because CUS2 lies on or near the top of the Buena Vista
Lake clay, we hypothesize that the CUS2 unit was deposited
toward the end of the lake episode, just after the eruption
of the Lava Creek volcanic ash which is found near the
top of the Corcoran Clay near Wasco, CA (Sarna-Wojcicki,
1995). If this is the case, then the CUS2 deltaic lobe was
deposited into Lake Clyde during Marine Oxygen Isotope
of this sedimentary package is due to some combination
of the prograding nature of the deltaic lobe and possibly
the regression of Lake Clyde as it drained into the Pacific
Ocean through the newly opened Carquinez Straits.
Stage 16 (MIS 16). MIS16 marks the highest amplitude
glacial event in the Pleistocene Epoch (Reheis, 2002).
This extreme glacial episode was likely responsible for the
deflection of the northern jet stream storm tracks southward
into the latitudes currently occupied by the Central Valley
and the Great Basin deserts (Antevs, 1948; Negrini, 2002).
Thus it was responsible for the largest extents of pluvial
lakes in the Great Basin during the Pleistocene (Reheis,
2002) and, hence, most likely also resulted in extreme
pluvial conditions in the Central Valley of California,
conditions driven by anomalously high discharge of surface
water from Sierran streams (e.g., the Kern River). SarnaWojcicki (1995) has proposed that this event resulted in
an extreme highstand of Lake Clyde. The lake was then
able to overtop a spillover sill into the San Francisco Bay.
The highstand, coupled with the anomalously low sea
levels associated with the glacial maximum, resulted in
an increase in hydraulic head between Lake Clyde and the
Pacific Ocean. This, in turn, led to the incision of the modern
outlet of the Central Valley hydrologic system into the San
Francisco Bay through the Carquinez Straits, an event that
drained Lake Clyde soon thereafter. The Sarna-Wojcicki
hypothesis is supported by the age of the oldest sediments
in San Francisco Bay that exhibit the distinct mineralogy
of sediments sourced primarily from the Central Valley
(Sarna-Wojcicki, 1995; Harden, 1999). In this scenario, the
CUS2 unit was deposited during the high discharge event
associated with MIS 16 and the coarsening-upward nature
the bulk sediment (Figure 5).
The above association may be due to the reducing
geochemical environment more likely to be found in
lacustrine than alluvial fan settings. In this hypothesis,
arsenic is bound in authigenic pyrite formed when reducing
conditions are extreme enough and persist long enough for
the reduction of groundwater sulfate to occur after the iron
reduction phase is completed. Arsenic readily substitutes for
iron in pyrite (Savage et al. 2000). A later change to more
oxidizing geochemical conditions can potentially dissolve
the pyrite and release the arsenic into the groundwater. This
hypothesis is supported by preliminary SEM and microprobe
analyses reported by Boockoff et al. (2005) and Negrini et
al. (2006) wherein As-bearing, authigenic pyrite was found
in shale samples from well 23H (Figure 4). Furthermore,
dissolution features were observed in some of the pyrite
crystals supporting a change to more oxidizing conditions
and corresponding release of As to groundwater and/or to
easily exchangeable sites on mineral surfaces (Figure 4).
Association of CUS2 with Elevated Groundwater Arsenic
Concentrations
The areal distribution of the thickest part of the CUS2
deltaic lobe is remarkably coincident with an isolated,
contiguous region in the Kern Water Bank within which
groundwater arsenic concentrations are above 50 ppb
(Figure 12). The depth interval containing these deltaic
sediments also generally coincides with depth intervals
characterized by elevated concentrations of groundwater
arsenic and of arsenic in the easily exchangeable state in
Structural Control of CUS2
The maximum thickness of CUS2 occupies a welldefined 2-3 mile-wide rectangular prism in the part of the
mapping area closest to the Buena Vista Lake terminal
basin (Fig. 8c). This prism lies within the surface projection
of northeast-striking normal faults mapped by Jones and
Gillespie (1997) in underlying units. These faults define
a series of stair-stepped fault blocks that are down to the
Pacific Section AAPG Publication MP48 107
southeast relative to a horst mapped in the southwest
corner of township 30S25E (Figure 12). The relief on the
base of CUS2 in this region (~100-150 ft of relief at a
depth of 700 fbgs) is also approximately consistent with
that predicted by the 400-500 ft total throw across the fault
system of the older middle Mya unit in the Plio-Pleistocene
San Joaquin Formation at a depth of 2500-3000 fbgs (Jones
and Gillespie, 1997). From these observations we conclude
that the associated fault system was probably active at least
up to and throughout the time of deposition of CUS2 and
likely controlled site of deposition for this unit.
IMPLICATIONS AND FUTURE WORK
Given increasing population stress on water resources
over time, there is a growing need to better understand the
relationship between groundwater resources, groundwater
quality, and underlying geological controls. The work
presented here begins to address this need by demonstrating
probable links between specific environments of
deposition, structural setting and groundwater quality.
Much work remains to be done to test and refine these
nascent hypotheses. The prograding fan-delta model
requires testing with a higher density well log database that
should become available as new water wells are drilled and
logged. Efforts should be made to core selected intervals
from these wells so that age control can be obtained. Two
volcanic ash layers (the Lava Creek and Bishop tephra)
and a major magnetic polarity boundary (the Matuyama/
Brunhes) provide coring targets of the appropriate age
(~800 to 600 kyr) with which to test the geologic history
proposed herein. Additional SEM and microprobe work
is required on samples from both wells 23H and 24K to
test the proposed role of the formation and subsequent
dissolution of arsenic in the contamination of groundwaters
and to help determine why only one of these wells has high
concentrations of groundwater arsenic. Such refinements
would lead to an extension of this study to elsewhere on the
Kern River alluvial fan and the greater San Joaquin Vallley
which, in principal, should contain related stratigraphic
packages that are controlled by the same underlying
geological events and similarly affect groundwater quality
and resources in their local region.
108 Pacific Section AAPG Publication MP48
ACKNOWLEDGMENTS
Funding for this study was provided by the US
Department of Agriculture (USDA-CREES Grant #200101170) and by the California Department of Water
Resources (Assembly Bill 303 “Local Groundwater
Assistance” grants awarded to the Kern Water Bank
Authority on July of 2003 and 2004). Al Tanabe, Tom
Osborn, Elizabeth Powers, Lael Monrian, Cari Meyer,
assisted with sediment analyses. Jennifer Shives assisted
in mapping the Buena Vista Lake clay unit. Karen Blake
was invaluable with respect to setting up the Geographix™
project upon which the well correlations and mapping were
based. Conversations with Robert Crewdson and Brian
Hirst greatly improved this manuscript. Quantitative XRD
analyses were performed by K/T GeoServices, Inc., under
the direction of James P. Talbot, P.G. Licenses were made
available for the Geographix™ software under a grant to
the CSU Bakersfield GeoTechnology Training Center from
Landmark Graphics Corporation, Inc. of Houston, Texas.
The Hitachi Corporation donated SEM time. Sarah Roeske
assisted with the operation of the microprobe at UC Davis
Department of Geology facility.
REFERENCES
Abu-Hassanein, Z.S., C.H. Benson, L.R. Blotz, 1996.
Electrical resistivity of compacted clays, Journal of
Geotechnical Engineering, v. 122, p. 397-406.
Antevs, E., 1948. Climatic changes and pre-white man:
Bulletin University of
Utah, v. 38, p. 168-191.
Baron D., Negrini R.M, Golob E.M.*, Miller D., SarnaWojcicki A, Fleck R., Hacker B., Erendi A., 2008.
Geochemical correlation and 40Ar/39Ar dating of the
Kern River Ash and related tephra: Implications for the
stratigraphy of petroleum-bearing formations in the San
Joaquin Valley, California, Quaternary International, v.
178, pp. 246-260.
Bartow, J.A., 1991. The Cenozoic evolution of the San
Joaquin Valley, California, U.S. Geol. Surv. Prof. Paper
1501, 40 pp.
Boockoff, L.A., 2005. Arsenic in Groundwater from Water
Bank Well 30S/25E-23H01, San Jaoquin Valley, California,
M.S. Thesis, Department of Geology, C.S.U. Bakersfield.
Boockoff, L., D. Baron, R. Horton, R. Negrini, J. Parker,
2005. Arsenic in sediments and groundwater from a well in
the southern San Joaquin Valley, California. GSA Abstr. w/
Progr., v. 37, p. 376.
Coburn, M.G., 1996. The hydrogeology of the Round
Mountain, Santa Margarita, Chanac and Kern River
Formations, Kern River Field, California: Unpublished
M.S. thesis, California State University Bakersfield, 156
p.
Cohen, A.S., 2003, Paleolimnology, Oxford University
Press.
Compton, R. R., 1985. Geology in the Field. New York:
John Wiley and Sons.
Croft, M.G., 1972. Subsurface geology of the Late Tertiary
and Quaternary water-bearing deposits of the southern part
of the San Joaquin Valley, California, U.S. Geol. Surv.
Water-Supply Paper 1999H, 29 p.
Dale, R.H., J.J. French, H.D. Wilson, Jr., 1964, The story
of ground water in the San Joaquin Valley, California, U.S.
Geol. Surv. Circ. 459.
Dale R.H., J.J. French, G.V. Gordon, 1966, Ground-water
geology and hydrology of the Kern River alluvial fan area,
California, U.S. Geol. Surv. Water Res. Div. Open-File
Report.
Davis, P., J. Smith, G.J. Kukla, N.D. Opdyke, 1977.
Paleomagnetic study at a nuclear power plant near
Bakersfield, California, Quaternary Research, v. 7, pp. 380397.
Dean, W.E., Jr., 1974, Determination of carbonate and
organic matter in calcareous sediments and sedimentary
rocks by loss on ignition: comparison with other methods:
Journal of Sedimentary Petrology, 44: 242-248
Dickinson, W.E., 1970. Interpreting detrital modes Of
graywacke and arkose. J. Sediment. Petrol., v. 40, p. 695707.
Evans, M.E. and F. Heller, Environmental Magnetism :
Principles and Applications of Enviromagnetics, Academic
Press, San Diego, CA., 382 p.
Frink, J.W., H.A. Kues, 1954. Corcoran clay - A Pleistocene
lacustrine deposit in the San Joaquin Valley, California,
Am. Assoc. Petr. Geol. Bull., v. 38, pp. 2357-2371.
Golob, L., D. Baron, R. Negrini, A. Sarna-Wojcicki, 2005.
Trace-element analysis of four Neogene volcanic ashes:
implications for the stratigraphy of petroleum-bearing
formation in the San Joaquin Valley, CA, Geol. Soc. Am.
abstr. w/ prog., v. 37, p. 181.
Hallenburg, J.K., 1984. Geophysical Logging for Mineral
and Engineering Applications, Pennwell Publishing Co.,
254 p.
Harden, D., 2004. California Geology, Prentice Hall, 576
p.
Hoots, H.W., T.L. Bear, W.D. Kleinpell, 1954. Geological
summary of the San Joaquin Valley, California in Jahns,
R.H., ed., Geology of Southern California, Cal. Div. Mines
Geol. Bull. v. 170, pp. 113-129.
Jones, C.J., and J.M. Gillespie, 1997, Listric normal faults
in the Miocene-Pliocene section at North and South Coles
Levee Fields: a response to Pleistocene growth of the
Elk Hills anticline, in Geologic Evolution of the Western
Cordillera: Perspectives from Undergraduate Research;
Cooper, J. D. and Girty, G.H., eds., Pacific Section SEPM,
p. 49-56.
Keller, G.V., F.C. Frischknecht, 1970. Electrical Methods
in Geophysical Prospecting, Pergamon Press, New York,
NY, 519 p.
Kiser, S.C., E.J. Greenwood, and L.M. Bazeley, 1988.
Lithofacies of the Pleistocene to Recent sediments of
western Kern County, California, in J.W. Randall, R.L.
Countryman, eds., Selected Papers - San Joaquin Geological
Society, Vol. 7, pp.14-21, 370 p.
Langmuir, D., 1977. Aqueous Environmental Geochemistry.
Prentice Hall.
Lanphere, M.A., Champion, D.E., Christiansen, R.L., Izett,
G.A., Obradovich, J.D., 2002. Revised ages for tuffs of
the Yellowstone Plateau volcanic field: assignment of the
Huckleberry Ridge Tuff to a new geomagnetic polarity
event. Geological Society of America Bulletin 114, 559568.
Manning, J.C., 1967, Report on ground water hydrology in
the southern San Joaquin Valley, Journ. Am. Water Works
Assoc., v. 59, pp. 1513-1526.
Mendenhall, W.C., 1908. Preliminary report on the ground
waters of the San Joaquin Valley, California, U.S. Geol.
Surv. Water-Supp. Paper, 222, 52 p.
Pacific Section AAPG Publication MP48 109
Mendenhall W.C., R.B. Dole, H. Stabler, 1916. Ground
water in the San Joaquin Valley, California, , U.S. Geol.
Surv. Water-Supp. Paper, 398, 310 p.
Miller, D.D., 1999, Sequence stratigraphy and controls
on deposition of the upper Cenozoic Tulare Formation,
San Joaquin Valley, California [Ph.D. thesis]: Stanford,
California, Stanford University, 179 p.
Miller, D., R. Negrini, M. McQuire, C. Huggins,*M.
Minner, B. Hacker, A. M. Sarna-Wojcicki, C. Meyer, R.
Fleck, S. Reid, 1998. New upper age constraint on the Kern
River Formation, in, S.A. Reid, ed. Outcrops of the Eastern
San Joaquin Basin, San Joaquin Geological Society.
Miller, G.E., 1986. Sedimentology, depositional
environment and reservoir characteristics of the Kern River
Formation, southeastern San Joaquin Basin: Unpublished
M.S. thesis, Stanford University, 80 p.
Negrini, R.M., 2002, Pluvial lake sizes in the northwestern
Great Basin throughout the Quaternary Period. In: Eds. D.
Currey, D. Madsen, and R. Hershler, Great Basin Aquatics
Systems History. Smithsonian Press, pp. 17-59.
Negrini, R., D. Baron, R. Horton, J. Gillespie, P. Wigand,
M. Palacios-Fest, L. Boockoff, M. Casterline, E. Powers, R.
Nunneley, K. O’Sullivan, 2006. Section 2. CSU Bakersfield
contributions, in Hydrogeology of the Kern Fan Element
and Implications on Local Groundwater Management of
Kern Water Bank, final report submitted to CA Dept. Water
Resources for AB303 Local Groundwater Assistance grant,
Kern Water Bank Authority, Bakersfield, CA.
PGA (Pacific Geotechnial Associates, Inc., 1991. Study
of the regional geologic structure related to ground water
aquifers in the southern San Joaquin Valley groundwater
basin, Kern County, CA, submitted to Kern County Water
Agency, Bakersfield, CA, 48 p., 11 plates.
Page, R.W., 1986. Geology of the fresh ground-water basin
of the central valley, California, with texture maps and
sections, U.S. Geol. Surv. Prof. Paper 1401C, 54 p.
Reheis, M.C., A.M. Sarna-Wojcicki, R.L. Reynolds,
C.A. Repenning, and M.D. Mifflin, 2002. Pliocene to
middle Pleistocene lakes in the Great Basin: Ages and
Connections, In Hershler, R., D.B. Madsen, D.R. Currey,
(eds.), Great Basin Aquatic Systems History, Smithsonian
Contributions to the Earth Sciences, No. 3, Smithsonian
Press, Washington, D.C.
110 Pacific Section AAPG Publication MP48
Sarna-Wojcicki, A.M., 1995. Age, areal extent, and
paleoclimatic effects of “Lake Clyde,” a mid-Pleistocene
lake that formed the Corcoran Clay, Great Valley,
California. Glacial History of the Sierra Nevada, California:
a Symposium in Memorial to Clyde Wahrhaftig.
Sarna-Wojcicki, A.M., C.E. Meyer, E. Wan, 1997. Age and
correlation of tephra layers, position of the MatuyamaBrunhes chron boundary, and effects of the Bishop ash
eruption on Owens Lake, as determined from drill hole
OL-92, Southeast California, In Smith, G.I., J.L. Bischoff,
(eds.), An 800,000 Yr Paleoclimatic record from Core OL92, Owens Lake, Southeastern California, Special Paper
317, Geological Society of America, 160 p.
Sarna-Wojcicki, A.M., Pringle, M.S., and Wijbrans, Jan,
2000, New 40Ar/39Ar age of the Bishop Tuff from multiple
sites and sediment rate calibration for the MatuyamaBrunhes boundary. Journal of Geophysical Research, v.
105, B9, p. 21,431-21,443.
Savage, K.S., T.N. Tingle, P.A. O’Day, G.A. Waychunas,
D.K. Bird, 2000. Arsenic speciation in pyrite and secondary
weathering phases, Mother Lode Gold District, Tuolumne
County, California, Applied Geochemistry, v. 15., p. 12191244.
Selley, R.C., 2000. Applied Sedimentology, Academic
Press, San Diego, CA., 523 p.
Steel, R., S. Maehle, H. Nilsen, S.L. Roe, and A. Spinnangr,
1980. Coarsening-upward cycles in the alluvium of
Hornelen Basin (Devonian), Norway:
sedimentary
response to tectonic events. Geol. Soc. Am. Bull., v. 88, p.
1124-1134.
Swartz R.J., Thyne G.D., and Gillespie J.M., 1996.
Dissolved arsenic in the Kern Fan, San Joaquin Valley,
California:
Naturally occurring or anthropogenic?
Environmental Geosciences, v. 3, pp. 143-153.
Tessier, A., Cambell, P., and M. Bisson, 1979. Sequential
extraction procedure for the speciation of particulate trace
metals, Analy. Chem., v. 51., p. 844-851.
Ward, S., 1990. Resistivity and induced-polarization
methods, In Ward, S., Ed., Geotechnical and Environmental
Geophysics, v. 1, p. 147-190.
Welch A.H, Lico, M.S., and Hughes J.L., 1988. Arsenic in
ground water of the Western United States. Ground Water,
v. 26, pp. 333-347.
Welch A.H., Westjohn D.B., Helsel D.R., and Wanty
R.B., 2000. Arsenic in ground water of the United States:
occurrence and geochemistry. Ground Water, v. 38, pp.
589-604.
Wescott, W.A., F.G. Ethridge, 1990. Fan deltas: Alluvial
fans in coastal settings, in Rachocki, AH., M. Church, Eds.,
Alluvial Fans: A Field Approach, John Wiley and Sons,
New York, 391 p.
Wilson, R.I., 1993. Paleoenvironmental Analysis of the
Shallow Sediments within the Kern Water Bank Project
Area of the Southern San Joaquin Valley, Western Kern
County, California: Implications for Local Hydrogeology,
M.S. Thesis, CSU Fresno.
Zachos, J., M. Pagani, L. Sloan, E. Thomas, K. Billups,
2001. Trends, rhythms, and aberrations in global climate
65 Ma to present, Science, v. 292, p. 686-693.
Pacific Section AAPG Publication MP48 111