air, fire, water . . . Aristotles four basic elements of matter
come under a great deal of scrutiny in many of the environmental
and geologic projects at Lawrence Livermore. Some of this research
springs from issues of particular interest to California, such as
groundwater quality or earthquake prediction.
The Energy and Environment
Directorate, with the help of other groups across the Laboratory,
is a natural for contributing in this arena. For example, Robin
Newmark, a geophysicist in Energy and Environment, helped develop
thermal remediation technologies such as dynamic underground stripping
and hydrous pyrolysis/oxidation that are being used to clean up
contaminated groundwater sites at both government and privately
owned facilities across the country. These technologies have been
used at Lawrence Livermore and in Visalia, California, and are under
consideration by the U.S. and California Environmental Protection
agencies for use at other sites. She notes, We have a number
of small and mid-sized projects with a California perspective, and
others that have great potential for more interplay with the state
Dave Layton, division leader
for Health and Ecological Assessment, agrees. Id like
to see our research portfolio push further into California.
In one current project funded
by the California State Water Quality Control Board, researchers
from Energy and Environment, along with colleagues from the Chemistry
and Materials Science Directorate and the Environmental Protection
Departments Environmental Restoration Division, are using
isotopic tracers to examine the vulnerability of groundwater to
volatile organic compounds and other chemicals. Meanwhile, out on
dry land, researchers from the Biology and Biotechnology Research
Program and the Energy and Environment directorates are developing
a DNA-based measurement technology that will help them study the
environmental factors that allow the Valley Fever fungus to thrive
in Californias Central Valley. Another California-centric
example is a recent workshop hosted by the Laboratory and sponsored
by the Department of Energys Office of Fuels Development and
the Western States Petroleum Association. The workshop focused on
the increased use of ethanol and alkylates in automotive fuels in
California after a phaseout of the potentially toxic MTBE fuel additive.
Unique Laboratory capabilities
are also being used to help California with environmental issues.
The Atmospheric Release and Advisory Capability (ARAC) came to the
aid of the California Environmental Protection Agency three years
ago, using computer simulations to monitor smoke billowing from
a 30-acre tire fire near Tracy, California (see S&TR,
June 1999, Forewarnings
of Coming Hazards). The Environmental Restoration Division developed
GeoTracker, a public Web site (http://geotracker2.arsenaultlegg.com/),
which helps California regulators and the public evaluate the safety
of drinking water by identifying leaking underground fuel tanks
and their proximity to municipal water wells (see S&TR,
The Laboratory in
the News). A digital mapping or geographic information systems
capability that allows users to sift through layers of geographic
data (such as population densities, elevation data, and earthquake
faults) will support the California Energy Commission in siting
future power plants to alleviate the current California energy crisis.
Three projects with a subterranean
focustwo on groundwater and one on earthquakesprovide
examples of what the Laboratory is doing to address environmental
issues of importance to California and beyond. In a fourth, large-scale
efforta joint vision of Lawrence Livermore and the University
of California at Mercedenvironmental research related to the
California Central Valley and the Sierra Nevada would be integrated
into a supercomputer- and sensor-driven Virtual Valley,
with resources available to planners and public alike.
of a portion of South Lake Tahoe showing the area being modeled
to determine how MTBE moves in groundwater.
1992, many U.S. oil refineries have been adding MTBE (methyl tertiary-butyl
ether) to gasoline to reduce air pollution and fulfill a requirement
of the federal Clean Air Act. However, MTBE has turned out to have
certain drawbacks. Once it gets into the groundthrough spills
and leaksit can infiltrate groundwater supplies, giving drinking
water an unpleasant taste and odor. The taste is noticeable
even at very small quantitiesdown to 5 micrograms of MTBE
per liter of water, notes Laboratory hydrogeologist Steven
Carle. Some evidence also suggests that MTBE may be a human carcinogen.
These are no small issues.
Groundwater is a significant source of drinking water worldwide.
In the United States alone, 60 percent of the water used is groundwater.
In California, at least 10 public drinking water wells have been
closed because of the intrusion of MTBE.
With this as background,
Laboratory researchers Carle, Reed Maxwell, and Dave Layton developed
three-dimensional computer simulations of how MTBE moves underground.
As it turns out, MTBE behaves differently in groundwater from other
petroleum products such as benzene: it is highly soluble in water,
does not easily adsorb to soil, and does not degrade readily on
its own. The team chose South Lake Tahoe as the location to model.
MTBE could be a crucial issue for this area, Carle says.
The South Lake Tahoe area receives lots of recharge in the
spring from melting snow. The snow melts into the ground, causing
pressure to build up that flushes the groundwater and anything in
it quickly through the subsurface system."
four-part simulation shows an extensive MTBE plume increasing
the contamination of nearby wells over time.
and the team set about creating a realistic geologic model of the
area using information supplied by the South Lake Tahoe Public Utilities
District. The subsurface is complex, notes Carle, with the aquifer
systems spanning multiple geologic formations. Using TSIM, a geostatistical
simulation code developed by Carle, the researchers built a below-surface
model of the geology of the area as well as of the permeability
of the various layers and subterranean features.
We had to use a certain
amount of statistical interpolation, says Carle. Even
though we used information gained from core samples and geologic
studies of the area, we dont know, inch by inch, exactly whats
down there. The model of the areaa chunk of real estate
600 meters by 1,200 meterswas built up of 7.2 million wafer-thin
geologic cells, 5 by 10 by 0.2 meters. We ended
up with a problem consisting of millions of cells, with materials
that had permeabilities differing by seven orders of magnitude,
he says. It was the first time that both the geology and permeability
of this area had been simulated in such detail.
The team then placed four
municipal wells in the model at their proper locations and set up
a realistic model of a leaking underground tank: a constantly replenished
pool of MTBE (1.5 kilograms per day) just under the surface, dissolving
into groundwater over 300 days. Knowing the pumping rates of the
wells, the team was able to use ParFlow (a parallel finite-difference
numerical flow modeling code developed at the Laboratory) to calculate
a high-resolution flow field. The flow field indicates the direction
that water and other liquids flow below the surface, rather like
a topographic map allows determination of the direction of flow
on the surface.
Finally, the simulations
were fed into SLIM-Fast, a numerical particle-tracking code, to
trace the fate of MTBE. With this code, tracking MTBE particles
in the groundwater is like tracking a bunch of ping-pong balls in
a river, explains Carle. SLIM-Fast also permits splitting
of the MTBE particles, allowing researchers to track increasingly
dilute concentrations of MTBE. This was a first, notes
Carle. Most other codes cannot trace the ultimate fate of
MTBE because they cant accurately reach the low concentrations
that evolve over time. Using the three codes, the researchers
tracked the MTBE over a 30-year period to its final dilutions in
The team also generated multiple
scenarios using different geologic heterogeneities, all equally
based on the information available, all equally plausible. We
generated a cloud of predictions that helped us determine
the uncertainties of our final results, says Carle. The bottom
line: According to the simulations, the most likely path of flow
for MTBE leads directly to the water wells where it would show up
in concentrations of tens of micrograms per liter.
When the results were compared
to an actual situation, the researchers found they were definitely
in the ballpark. MTBE was found in seven South Lake Tahoe wells
in concentrations of 0.5 to 68 micrograms per liter. One interesting
result of the simulations indicates that the MTBE problem could
persist for decades. This was something we hadnt expected,
says Carle. Wed assumedas did others in the fieldthat
the MTBE moves along at the pace of groundwater. Instead, it appears
that, through mechanisms we dont fully understand yet, the
MTBE lingers and is released over time. It could be that the chemical
is slowing down in low-permeability zones. To explore this issue
and others requires more realistic simulations that can produce
MTBE breakthrough curves, showing concentration of MTBE in wells
over time out to 30 years.
(Water) Bugs Life
Chemicals arent the
only undesirables in the groundwater supplies. Microbes, virusesbugs
in popular parlancealso find a welcome environment in subterranean
aquifers. Livermores Reed Maxwell worked with fellow researchers
from the Laboratory, the U.S. Geological Survey, and Drexel University
to better understand how these organisms are transported underground.
Specifically, the team applied simulation and modeling techniques
to see how the varied geology of riverbankssand, clay, cobbles,
gravel, and so onaffect the movement of colloids (submicroscopic
particles, such as viruses) dispersed in a different medium, such
as water. Using a site in Californias Orange County for their
simulations, the researchers examined how viruses could attach to
different geologic media and compared the arrival time at the wells
for both viruses and tracer chemicals often used to track groundwater
movement. They also compared their results with results of simulations
using models that assume a uniform, or homogeneous, type of material
throughouta riverbank composed of only sand, for instance.
Water management districts
such as Orange County are looking for better ways to manage and
obtain clean sources of drinking water, notes Maxwell. In
the process, researchers are examining technologies used over the
past century in other countries and finding that these methods are
actually quite effective in producing clean water. For instance,
a number of European countries obtain their water from the Rhine
River by placing a well on the riverbank and pulling in river water
through the subsurface of the bank. This riverbank filtration technique
filters out much of the microbial activity in the water. Wed
like to better understand how this process works, says Maxwell.
At what rate are these organisms filtered out? Do they stick
to some particular medium, and if so, what medium and for how long?
How does the transport of these colloids differ from that of tracer
These questions are of particular
interest to Orange County. Nearly 80 percent of its water supply
is composed of groundwater. The Orange County Water District is
currently monitoring surface water and groundwater for viruses,
but so far, it has detected none. Furthermore, Orange County is
now considering using tertiary-treated wastewater (wastewater treated
three times) to artificially recharge lakes, which will, over time,
recharge groundwater aquifers. The Environmental Protection Agency
has guidelines in place to ensure that such water is clean before
its used again.
Maxwell explains, There
are guidelines on such matters as the distances that wells must
be from the water sources, and the residence time that
treated water must remain underground. These guidelines are expected
to account for complicated natural processes that need to be better
understood. Both wastewater recharge and riverbank filtration are
potential pathways for microbes to contaminate drinking water sources,
so understanding how various geologic media affect movement of these
organisms is important.
To gain insight into how
viruses move in groundwater and potentially prevent virus problems
before they arise, Maxwell and his collaborators used a three-dimensional
computer model developed by Laboratory hydrologist Andrew Tompson
to simulate the movement of groundwater in an area used in a water
reclamation operation managed by the Orange County Water District.
simulation of the subsurface geology, the streamlines tracing
travel paths of groundwater, and the predicted mean age of the
water as it travels from recharge sources to wells P-4, PL-5,
wells were studied. Two (PL-5 and PL-10) derive much of their water
directly from the Santa Ana River, and one (P-4) receives recharge
water from Anaheim Lake and Warner Basin, which in turn receive
their water from
the river. Water arriving at P-4 was about 1.3 years old, water
at PL-5 was 0.5 year old, PL-10 water was 1.2 years old. Although
PL-3 is closer to the water source than PL-5, the same water actually
takes longer to travel the shorter distance because of the permeability
and complexity of the subsurface. This anomaly, Maxwell notes, would
not have appeared if the geologic model had been homogeneous.
The team took this three-dimensional
model and numerically built in the ways in which submicrometer-size
organisms attach and detach from the materials found beneath the
surface. From there, the team deconvolved the problem. That is,
they took the complicated three-dimensional flow field that simulates
the motion of groundwater, its tracers, and its hitchhiking organisms
and broke it down into a number of one-dimensional transport problems.
The problems were solved using Livermores massively parallel
computing systems. The results were then recombined to find out
what arrived first at the wells.
A tracer chemical was first
to break through into the simulated wells and displayed the highest
concentrations at the wells over time. The viruses broke through
last and showed the lowest concentrations. What that told
us is that the well is most vulnerable to a tracer . . . not to
microbes. says Maxwell. Tracer chemicals, in other words,
are poor predictors of viral arrival times in wells according to
this particular model. These results show the importance of using
a model that accurately reproduces the geology of the given area.
There are, he adds, many
more questions regarding microbial transport in groundwater. Its
pretty clear that different filtration processes dominate for different
types of soil. Less permeable sedimentstightly packed sand,
for instanceare good at filtering out microbes. We still have
a ways to go in understanding the physics of viral transport in
San Francisco Bay Area is located within the PacificNorth
American plate boundary. The focus of the Bay Area Paleoseismic
Experiment has been to develop past earthquake chronologies
on the seven major faults in the area: the San Gregorio, San
Andreas, Hayward, Calaveras, Rodgers Creek, ConcordGreen
Valley, and Greenville faults.
photomosaic trench log, a vertical cross section of the San
Andreas Fault at Mill Canyon near Watsonville, California, shows
evidence of three past earthquakesthe San Francisco earthquake
of 1906 and two previous events labeled MC-2 and MC-3. Each
of these earthquakes produced characteristic downward-tapered
infilled fissures. (This study was led by Tom Fumal at the U.S.
Search of Ancient Earthquakes
What could be more California
than earthquakes? The San Francisco Bay Area has the highest density
of active faults and the highest rate of seismic moment release
per square kilometer of any urban area in the U.S. In 1998, in the
hope of providing a framework for more precise forecasts of future
large and damaging quakes, the Center for Accelerator Mass Spectrometry
(CAMS) at Livermore joined a multi-institutional effort, led by
the U.S. Geological Survey, in a region-wide cooperative project
called the Bay Area Paleoseismic Experiment, or BAPEX.
BAPEXs goal is to develop
a 2,000-year chronology of large earthquakes in the Bay Area and
look for patterns in the timing, locations, and magnitudes of prehistoric
shakers. Unlike other paleoearthquake projects, which usually limit
their focus to the history of a single fault, BAPEX targets essentially
an entire plate boundary. This regionthe 45-kilometer-wide
area between the San Gregorio and Greenville fault zones from west
to eastrepresents most of the plate boundary between the Pacific
and North American plates. As part of that effort, geologists such
as Livermores Gordon Seitz are focusing on the areas
seven major fault systemsthe San Andreas, San Gregorio, Hayward,
Rodgers Creek, Calaveras, Concord Green Valley, and Greenville.
Seitz explains, When
we find a promising site, we excavate down as far as 9 meters and
examine the geologic layers or strata. Although most earthquakes
are triggered 10 to 15 kilometers below the surface, large earthquakes
cause ground rupture, which is recorded in these near-surface layers.
In cross-cuts through the strata of a fault zone tens of meters
in width, the researchers look for evidence of former ground-rupturing
deformation such as fault scarps at the surface or buried beneath
it, fissure fills, upward-terminating faults, folds, and sand volcanoes.
When geologists find such layers, they look for organic material
within the sediments that can be carbon dated at CAMS.
Not much material is required
for dating. About one-thousandth of a graman amount equal
to a couple grains of dirtwill do. The age of the material
is determined by the amount of carbon-14, or radioactive carbon,
in the sample. The Livermore accelerator is extremely sensitive,
able to find one carbon-14 isotope among a quadrillion other carbon
atoms. (For more information on CAMS, see S&TR, July/August
Research Benefits from Counting Small.)
The CAMS team can produce
results in daysa big improvement over performance of other
laboratories that typically can take months to return results to
field geologists. We call it real-time dating, says
Seitz. With CAMS as part of the process, geologists can quickly
determine the age of key layers and move on to the next step in
their fieldwork. What used to take several seasons can now
be accomplished in one season, he adds. This quick turnaround
is particularly important when the cross-cut or trench is located
in an urban area and cannot be left open for more than a few days.
And with urban development decreasing the number of good sites,
accomplishing more per season has become increasingly important.
historical record of past Bay Area earthquakes suggests a pattern
of intense earthquake activity for some 50 years prior to the
great 1906 San Francisco earthquake. That pattern ends just
after the 1906 quake. Using the results of paleoseismology research,
scientists can extend the record into the past to see if this
pattern of earthquake activity is a persistent feature of the
BAPEX geologists have excavated
more than 28 Bay Area paleoseismic sites and determined over 900
radiocarbon dates at CAMS. In the process, theyve developed
earthquake histories covering several thousand years for individual
faults. To do a rigorous statistical analysis on the earthquake
patterns, Seitz estimates that they need more event records covering
longer periods of time. Fourteen is the most Ive seen
recorded on any one site, he notes.
When a comparison of timing
and magnitudes is made between historic northern California
earthquakes (1850 to present) and prehistoric quakes,
a pattern emerges: Leading up to the 1906 quake, the magnitude and
frequency of earthquakes increased; after 1906, activity shut off
for more than 50 years. Is the Bay Area at the start of another
cycle, with the1989 Loma Prieta quake being the first major quake?
We cant say for
certain, says Seitz. Historically, the Bay Area has
not experienced one complete earthquake cycle. But by the best estimate
of the experts, theres a 70-percent chance in the next 30
years that the Bay Area will see a quake of magnitude 6.7 or greater
quite possibly centered in a heavily populated urban area. BAPEX
should cast some additional light on these forecasts by providing
a more complete picture of earthquake patterns over time and space.
Seitz, along with Graham
Kent of the Scripps Institution of Oceanography at the University
of California at San Diego, is extending this research by taking
a close look at paleoearthquakes under water. Previously,
the research of paleoearthquakes has focused on dry land,
he explains, because underwater trenching is not yet possible.
But in light of several major technological advanceshigh-resolution
seismic imaging, accelerator mass spectrometrys ability to
provide carbon-14 analysis of small samples, and detailed bathymetry
mappingwe wondered how fault investigations could be done
Seitz and Kent turned to
a new high-resolution seismic technique called CHIRP to help with
underwater paleoseismic studies. Like a sonar system, CHIRP bounces
sound waves off submerged structures. However, unlike sonar, the
CHIRP system can image sediment layers beneath the lakebed at unprecedented
resolutions. With CHIRP, we can image layers as thin as 20
centimeters, he says. Using Lake Tahoe as a test bed, the
team created seismic profiles at the bottom of the lake with imaging
depths of as much as 50 meters.
The use of CHIRP technology
in seismic hazard assessment is new, notes Seitz. Skills learned
at Lake Tahoe will not only help solve many of the outstanding local
questions, but in the future, they could also be used to understand
a subset of neotectonic problems that are hidden underwater.
Says Seitz, Our approach
of acoustic trenching combined with carbon dating of
strategically located sediment cores will, we hope, allow future
studies of many active fault systems that have been largely ignored,
mostly because of water coverage. Being able to image the tectonically
deformed sediments under water and having a way to determine their
Offshore faults along the
California coastline and the parts of faults that extend offshore
are examples of largely ignored fault systems that would benefit
from underwater peleoearthquake research. International locales
such as the Marmara Sea adjacent to quake-threatened Istanbul would
also benefit from this technological advance. For Seitz, Lake
Tahoe is an ideal place to develop these techniques. The logistics
are easier on this lake than at sea, and ship costs are 5 to 10
times less expensive.
seismic profile at unprecedented submeter resolution across
the North TahoeStateline Fault at a water depth of about
500 meters. The unique high resolution of this profile is achieved
the source through a range of frequencies (500 hertz to 15 kilohertz),
a technique developed largely
by Neal Driscoll, now at the Scripps Institution in San Diego.
The thickening of sediments near the
fault on the downthrown right side block is characteristic of
sedimentation after an earthquake.
the Virtual Valley
Suppose one could take the
Laboratorys environmental research capabilities and projects,
join them to the research capabilities of a leading university,
and focus them on a particular region. This is the premise of the
Virtual Valley, a joint vision of Lawrence Livermore and the University
of California at Merced, UCs newest campus, which is in the
planning and development stage.
The Laboratorys activities
to implement such a Virtual Valley concept would be pursued in partnership
with those of UC Merceds Sierra Nevada Research Institute.
The institute has been chartered to focus on the challenges surrounding
the rapid development and transformation of Californias Central
Valley and Sierra Nevada region.
The task facing Virtual Valley
designers and planners is to provide a comprehensive environmental
simulation and observation system focused on regional issues. Issues
the Virtual Valley might tackle include wildfire management and
prediction, the effects of urban development, air quality and water
resources management, earthquake prediction, and groundwater management
and cleanup. The Virtual Valley will have the information, tools,
and computational power needed to explore all these topics and more.
The Virtual Valley would
tie together supercomputers, sensor networks, geographic information
systems, field measurement sites, historical data sets, wireless
communications simulation and modeling, advanced visualization systems,
and Internet access. Its data sets and computing power would be
available to students, educators, scientists, planners, residents,
and the public-at-large.
Users would be able to combine
diverse areas of environmental research together, look for commonalties,
determine causes and effects, and work from a common platform. The
effects of this research have the potential of reaching beyond Californias
borders. For example, other areas in the United States have groundwater
management issues. Other countries must deal with the specter of
devastating earthquakes. The Virtual Valleyand all of the
research that could feed into itprovides us with a way to
benefit the world, by focusing on the valleys and mountains out
Bay Areas Paleoseismic Experiment (BAPEX), Center for Mass Accelerator
Spectrometry (CAMS), CHIRP, groundwater management, hydrogeology,
MTBE, Orange County, paleoearthquakes, ParFlow, San Andreas Fault,
SLIM-Fast, South Lake Tahoe, TSIM, University of California at Merced,
information contact Steve Carle (925) 423-5039 (email@example.com),
Reed Maxwell (925) 422-7436
(firstname.lastname@example.org), or Gordon
Seitz (925) 423-8469 (email@example.com).
(left) holds a B.S. from the University of Miami (1992) and
an M.S. from the University of California at Los Angeles (1994)
in mechanical engineering, and a Ph.D. from the University of
California at Berkeley (1998) in civil and environmental engineering.
He joined the Laboratory as a postdoctoral fellow in 1998 and
became a physicist in the Geosciences and Environmental Technologies
Division in 2000. He specializes in the study of radionuclide
transport at the Nevada Test Site and in environmental risk
assessment and management.
received his B.S. and M.S. in engineering geoscience from the
University of California at Berkeley in 1986 and 1987, respectively,
and his Ph.D. in hydrologic science from the University of California
at Davis in 1996. He came to Livermore as a postdoctoral fellow
in 1997. In 2000, he joined the Geosciences and Environmental
Technologies Division as a physicist. His research focuses on
the development of geostatistical methods, the hydrogeologic
modeling of groundwater flow and contaminant transport, and
the integration of diverse data sets.
SEITZ (right) holds a B.S. in geology
from San Diego State University (1983) and a Ph.D. in geological
sciences from the University of Oregon at Eugene (1999). He
joined the Laboratory as a postdoctoral fellow in 1999 to work
at the Center for Accelerator Mass Spectrometry on the Bay Area
Paleoseismic Experiment. His research interests include improving
scientific understanding of past earthquake chronologies and
interpreting patterns of fault behavior in space and time based
on accelerator mass spectrometry carbon-14 dating, with special
emphasis on the San Andreas Fault.