funded by EPSCoR/NSF 

photograph taken by Mike Clair.

This is a multi-year, joint project among three universities, University of Idaho (UI), Boise State University (BSU), and Idaho State University (ISU).  The project is funded by the Experimental Program to Stimulate Competitive Research (EPSCoR) and the National Science Foundation (NSF).  The research team consists of professors from the three universities with Dr. John S. Oldow of UI as the team leader.  Dr. Paul Donaldson , Dr. John Bradford , Dr. William P. Clement , Dr. Mitchell Lyle , and Dr. Partha Routh are the research participants from BSU.  The research area is the Borax Lake Hydrothermal System (BLHS)  in southeastern Oregon.  The overall goal of the project is to study how physical and geochemical properties of the desert habitat affect the living organisms present, and in turn how the living organisms affect their own living environment.  The scientific fields of research being applied to this study are in thermophilic microbiology, bioinformatics, spatial modeling in ecology, biogeochemistry, hydrogeology, mineralogy, structural geology, and geophysics.

[return to top of page]

The BLHS is located in Alvord Basin which is part of what is known as the Northern Great Basin of the United States.   Part of what makes the Great Basin unique is the fact that surface water flows neither to the Pacific nor Atlantic Oceans.  Instead, the water flows to inland lakes and marshes or is evaporated. The geologic formations we see today were formed from Basin and Range extension.  For the past 20 million years, extensional geologic processes have been stretching this region in the east-west direction. Basin and Range geology consists of horsts and grabens.  Through extension, the Steens Mountains and Pueblo Mountains to the west of the Alvord Basin have been uplifted, while the Alvord Basin has settled downward. The same effect is seen on the east side of the valley with the Trout Creek Mountains being uplifted.  The fault on the west side of the basin is a northwest striking normal fault, while the fault on the east side is a southeast striking normal fault.  The Steens Mountains are mostly a west-tilted slab of Steen's basalt dating from the middle to late Miocene.  Below the Steen's basalt are volcanic and sedimentary layers that outcrop on the east side of the mountains which are dated in the early Miocene. Above the Steen's basalt are volcanic ash flows dating from 7 to 10 million years ago.  Basin fill consists of alluvial and lacustrine sediments resting on Miocene volcanic rocks that include tuffs, basalts, and rhyolites. Near the center of the basin extending north from Borax Lake, over 100 hot springs are linearly aligned. A large mid-basin basement high is aligned with the surface trend of the hot springs (Bradford et al., 2004). Bradford concluded that the most active fault system is near the center of the valley, which is also the location of the hot springs.  The linear trend of the hot springs indicates near surface faulting, even though no obvious fault scarps appear.  The left lateral offset in the trend of the springs may suggest rotational strain that is redirecting the upflow of the geothermal water.

[return to top of page ]

As part of this interdisciplinary project that aims to study the link between the physical characteristics of hydrothermal systems and the biota that occupy those systems, professors and students from BSU have begun a detailed geophysical characterization of the BLHS located near the center of the Alvord Basin. Basement rock is comprised of Miocene volcanic deposits overlain by up to 700 m of alluvium and lacustrine deposits. Previous studies, based on gravity data and surface mapping, suggest that the BLHS lies directly over a north/south trending fault (Cleary et. al, 1981). We are conducting seismic investigations on both a basin scale and a local high resolution scale. A seismic reflection profile collected across the entire basin will help us place the hydrothermal system in a larger geologic context, while the high resolution 3D surveys will be used for detailed imaging of fault structure and hydrothermal flow paths. A third seismic survey was conducted around one of the larger hot springs known as the Bridge Hot Spring (seen in the photo gallery). This survey will utilize seismic refraction tomography to image conduits near the surface around the hot spring.  Additionally, we acquired coincidental high resolution magnetic data, as well as regional magnetic data for a large portion of the Alvord Basin. Inversions of the magnetic data sets will be constrained by the interpretations of the seismic data. A 3D time-domain electromagnetics (TEM) survey and four east-west TEM profiles were also conducted to aid interpretation of the BLHS sub-surface structure. The seismic basin profile followed Powerline Road, while the 3D surveys TEM profiles were collected north of Borax Lake. Topographic map to show relative locations of study areas.

[return to top of page]

In the summer of 2002, the primary objectives were to verify that a fault zone is present beneath the Borax Lake hot springs and to conduct tests to determine acquisition parameters for a detailed 3D seismic investigation.  A 3.5 km section of the basin scale seismic profile was completed and noise tests were conducted at the 3D survey sight. The summer of 2003 followed up with completing the east and west sections of the seismic profile and conducting a 3D seismic survey. In the summer of 2004, we completed the 3D tomography survey.

To test the fault zone interpretation, and begin to build a large scale image of basin geometry, we acquired an 11 km seismic reflection profile perpendicular to the suspected fault zone. The profile consists of 30-fold CMP data acquired using a trailer mounted, 200 lb accelerated weight drop. Reflections are evident to depths of at least 1 km. Evidence for faulting is clear with the seismic image showing a complex normal fault zone bounded to the west by a structural high. Refraction analysis suggests that the structural high is basement rock, and that this basement high controls the piezometric surface in the area with higher pressure to the west.

[Click here to see a larger version.]

Due to the strong site dependence of shallow seismic investigations, a critical first step in any study is a noise test. This is comprised of a set of measurements designed to test the local earth response to various combinations of source/reciever pairs. At this site we fired seven sources into two linear receiver arrays - one with sixty 10 Hz geophones and the other with forty-eight 100 Hz geophones.  Receiver spacing within each array was 0.5 m. The seven sources tested were a .223 rifle, a .22 firing rod, a 25 lb weight drop, a 16 lb weight drop, a dead blow, a silver dollar tamper on wood, and a silver dollar tamper on rock.

All seven sources were fired into the arrays at relatively small offsets (up to 48 m). The conditions at the site were excellent.  Good reflections were observed for all sources, but there were some significant differences. We found that the 100 Hz geophone array had a significantly better high frequency response, while not compromising bandwidth. The .223 rifle produced about 2 orders of magnitude more energy than any of the other sources. Surprisingly, the 25 lb weight drop, dead blow, and silver dollar tamper on wood and rock all produced about the same amount of energy and similar bandwidth. The .22 firing rod produced the lowest energy, but had good bandwidth and high frequency content.

Three sources were fired into the arrays at large offsets - up to 120 m. These were the .223 rifle, the 25 lb weight drop and the 16 lb weight drop. Reflections are clearly evident to times of at least 300 ms, which corresponds to a depth of roughly 270 m. The rifle source provided the highest energy and clearest reflections. It also produced a large air blast which obscures reflection energy at near offset. The 25 lb weight drop produced significantly less energy and had lower frequency content. On the other hand, it produced lower amplitude coherent noise, and some reflection energy is more clearly observed in the near offset range.  Overall, the .223 rifle appears superior to all other sources tested.  For the actual 3D survey data collection, the rifle was modified to include a blast plate and silencing system to reduce scattered debris and coherent air wave noise. 

Data collection for the high resolution 3D seismic survey was conducted the last two weeks of July, 2003.  The survey area is with 30 channels each.  The receiver lines were oriented roughly in the east-west direction, perpendicular to the alignment of the hot springs.  The inline spacing of the receivers was 5 meters, while the crossline spacing was approximately 10 meters.  Shot lines had a crossline spacing of 5 meters and extended 75 meters past the east and west ends of each receiver line.  The inline shot spacing located approximately 400 meters to the north of Borax Lake.  The basic acquisition geometry was a grid of eight receiver lines was 5 meters for odd numbered shot lines and 2.5 meters for even numbered shot lines with the shot line farthest south being line number 1. Even numbered shot lines coincided with receiver lines.  The surveyed was rolled along 20 meters at a time in the crossline direction approximately every 360 shots or four shot lines.  The seismic source was a 7.62x39 mm caliber SKS rifle with a blast shield.  Two shots were stacked for each shot location with a 25 Hz low-cut filter applied to each shot record.

The 3D stacked cube of seismic data consists of 105 cross-lines and 95 in-lines and is truncated at 250 ms. A time slice of reflection amplitudes at 72 ms (~ 60 m depth) from the migrated stacked cube shows linear features that are parallel to the surface trend of the hot springs. These linear features are reflection discontinuities related to faulting that likely controls groundwater flow paths. In-line slices from various locations show a semi-continuous reflector at 150 ms (~ 160 m depth). The reflector is planar to the south with offsets as you go to the north and disappears on the east side throughout the cube. This reflector horizon likely represents the Miocene volcanic rocks associated with the mid-basin basement high described by Bradford. The small offsets are probably due to smaller faults associated with the rotation of strain. The basement reflection terminates on the east side of the survey area because of a normal fault striking approximately NNW (North Northwest) and dipping 60° ESE (East Southeast). The fault tracks through overlying sediment structures and lines up with the surface expression of the hot springs; furthermore, this major fault is likely one of the main sub-surface structures that controls preferential fluid flow paths. Note that the linear features seen in the time slice at 72 ms correspond to the approximate orientation and location of the fault shown in the in-line slices. In-line slices also show signs of other small faults and/or fractures, but more detailed processing is necessary before sound interpretations can be made. 

[Click here to see the cross-line slices and time slice.]

[return to top of page]

Total field magnetic measurements were collected throughout the Alvord Basin. In particular, magnetic measurements were collected along Powerline Road and around the BLHS. The 3D magnetic survey was collected using a Geometrics 858G cesium vapor magnetometer. We collected total field magnetic readings at 3425 stations with 10 m station spacing along the east-west direction and 30 m station spacing along the north-south direction. The regional total magnetic field was removed from the raw data and the residual shows a good correlation with the locations of the hot springs.

The magnetic profile along Powerline Road supports the seismic interpretation of a mid-basin basement high. The high frequency anomalies between 3 km and 5 km suggest a magnetic disturbance due to near subsurface structure; while the broad, low frequency anomalies from 1 km to 3km and 5 km to 10.5 km suggest magnetic disturbances due to deep subsurface structure. The magnetic profile orientation is opposite the basin seismic reflection profile, but correlates well with the interpretation.


Partha Routh, a researcher from Boise State, has been working on inverting the magnetic data for a magnetic susceptibility model. We inverted the magnetic data to determine 3D magnetic susceptibility distribution for the survey area. We developed a smooth model 3D inversion code with depth weighting (Li and Oldenburg, 1996) that can invert magnetic residual data. The predicted data from the model fit the observed data well given a noise assumption of 3 nT. The model from the inversion of the magnetic data shows tubular structures that start at a depth of 200 m and continue to the surface. These tubular structures are the low susceptibility zones that may represent the conduits for upflowing geothermal water. The zones of low susceptibility correlate with the location of the hot springs, suggesting the validity of the magnetic inversion. Our future goal is to combine the interpretation of the magnetic inversion with the structural and stratigraphic information obtained from the seismic.

 
Inverted magnetic susceptibility model at BLHS.  The blue line indicates the line of hot springs.  The perspective of the left image is looking toward the NE, while the right image is looking toward the East.

[return to top of page]

To improve imaging capabilities, a transient electromagnetic survey was performed in the same survey area as the 3D seismic survey and noise test.  The TEM method measures the conductivity of the subsurface as a function of depth by inducing an electromagnetic field into the subsurface.  This is done by passing a current through a square transmitter loop.  Once the current is terminated in the transmitter loop, the time derivative of the vertical magnetic field is measured with a square receiver loop.  At times nearest the current termination time, the receiver loop collects information about the near surface conductivity; while at later times, the receiver loop collects information about conductivity at a greater depth.  We plan to combine seismic and EM methods to construct a detailed 3D image of the geothermal system.  We conducted a 3D TEM survey to determine the resistivity structure of the linear spring trend. Part of our goal was to understand the overlap of the linear trend of the springs. Each transmitter loop was 20x20 m with nine 5x5 m receiver loops inside. The survey was designed to provide 3-D coverage of the area. The loops were moved so that the surface was covered continuously in the survey region. The survey consisted of three 200 m North-South lines and three 200 m approximately East-West lines.

In addition to seismic source testing during the summer of 2002, we acquired 21 TEM soundings, at 10 m intervals along the same transect as the source tests. We used a 20 m x 20 m transmitter loop and 5 m x 5 m receiver loop to collect the data.  The receiver loop was located in the center of the transmitter loop. Conductivity along the profile was estimated by applying a 1-D inversion algorithm to each sounding.  A high conductivity zone comes to the surface in the vicinity of the hot springs.  This is most likely the water saturated zone, and indicates that the groundwater is highly conductive.  The lower conductivity zone, near the center of the profile at a depth of 10 - 20 m, may be indicative of the underlying geologic structure. 

The 3D TEM data was collected in the last two weeks of July, 2003.  The survey covered an area 200 m x 50 m, approximately covering the offset area of the hot springs.  We used a 20 m x 20 m transmitter loop and 5 m x 5 m receiver loop.  Nine different soundings were performed in each transmitter loop, so the central 15 m x 15 m area of the transmitter loop was completely covered. The loops were moved up so that the next set of receiver loops were adjacent to the previous set of receiver loops. Along with the 3D survey, three 150 m long profiles roughly in the east-west direction overlapped the same area.  Relative to the 3D survey, they were collected at 10 m, 90 m, and 185 m from the southern end of the survey.  The profiles were collected with the same loop setup as the 3D survey.

Preliminary results show a high resistivity layer from the surface to a depth of 3 m. A low resistivity layer exists from 3 m down to 10 m. A layer of increased resistivity is present between 10 to 23 m depth. A possible geologic model for this section is a thin vadose zone overlaying a spring water saturated zone. The layer of increased resistivity likely represents a zone of low permeability.

[return to top of page]

This project is funded by the Experimental Program to Stimulate Competitive Research (EPSCoR) and National Science Foundation (NSF).  The NSF Research Infrastructure Improvement (RII) Award, EPS 0132626, is a $9 million award aimed to improve and expand existing strengths in three research focus areas - Nanoscale Materials for Electronics and Sensor Applications, Life at Interfaces and the Biocomplexity of Extreme Environments, and Neuro-Fuzzy Soft Computing via Silicon Structures.

---

photograph taken by Mike Clair