V52C-01
Volatile Flux and Composition at Yellowstone Reflects a Gas-Charged Hydrothermal System Above a Basalt-Fueled Silicic Magma Reservoir
Recent papers have documented the immense volatile (CO2, S, Cl-, F-) flux from the Yellowstone Caldera. Carbon dioxide is estimated to escape diffusively through soils at a rate of 45,000 t d-1 (Werner and Brantley, 2003) compared with 137 t d-1 of Cl- released from hot springs into rivers (Hurwitz and others, 2007). These high volatile fluxes and the CO2/Cl- ratio of ~300 are inconsistent with simple degassing of a mid- or upper-crustal silicic intrusion. For example, if the estimated CO2-flux were supplied solely by a 104 km3 silicic-magma reservoir with 500 ppm dissolved CO2, the reservoir would be exhausted in ~1000 years, less than 0.1% of the longevity of the present Yellowstone volcanic field. Moreover, silicate melt inclusions in phenocrysts from erupted rhyolites contain abundant dissolved Cl- and F-, but minimal CO2 and S, the dominant effluents from the hydrothermal system. The carbon budget is explained best by dominant basalt degassing (~0.3 km3 a-1 of magma) in the lower and mid- crust, augmented by metamorphic devolatilization of limestone and other sediments, plus what can be sourced from overlying rhyolitic magma. The volatiles then pass into and through the near-surface hydrothermal system. The relative abundances of emitted CO2 and Cl- appear to require that the shallow subsurface beneath Yellowstone is gas-saturated down to >2 km. Fournier (1989) concluded that the diverse Yellowstone geothermal waters are ultimately derived from a deep parent fluid with 400 ppm Cl-. If Cl- and CO2 are emitted in proportion to their abundance in the hydrothermal system, then given the CO2/Cl- of 300, this parent fluid would contain 12 wt.% CO2 (5 mol%). Solubility constraints reveal that such fluid would be saturated with CO2-rich steam within the upper few kilometers. The presence of a compressible and expandable vapor phase has important implications for the origin and interpretation of ground-surface displacements at active calderas such as Yellowstone. Fournier RO (1989) Ann Rev Earth Planet Sci 17, 13-53. Hurwitz S, Lowenstern JB, Heasler H (2007) J Volcanol Geothermal Res 162, 149-171. Werner C, Brantley S, (2003) Geochem Geophys Geosystems 4 (7) 1061, doi:10.1029/2002GC000473.
V52C-02
Geothermal Monitoring in Yellowstone National Park
When the first exploring parties surveyed the Yellowstone region in the late 19th Century, it was the geologic wonders - geysers, hot springs, mudpots, fumaroles - that captured their imaginations. Because of these treasures, the U.S. Congress set aside and dedicated this land of "natural curiosities" as the world's first "public pleasuring ground". Protection of Yellowstone's unique geothermal features is a key mission of Yellowstone National Park as mandated by U. S. Congressional law. In response to that mandate, the Yellowstone National Park Geology Program developed a peer-reviewed, Geothermal Monitoring Plan in 2003. With partial Congressional funding of the Plan in 2005, implementation of a scientific monitoring effort began. Yellowstone's scientific geothermal monitoring effort includes the acquisition of time-temperature data using electronic data loggers, basic water quality data, chloride flux data, estimates of radiative heat flux using airborne, thermal infrared imagery, geothermal gas monitoring, and the monitoring of groundwater wells. Time- temperature data are acquired for geysers, hot springs, steam vents, wells, rivers, and the ground. Uses of the time-temperature data include public safety, calibrating airborne thermal infrared-imagery, monitoring selected thermal features for potential hydrothermal explosions, and determining the spatial and temporal changes in thermal areas. Since 2003, upgrades of Yellowstone's stream gaging network have improved the spatial and temporal precision of the chloride flux, water quality, and groundwater components of the Geothermal Monitoring Plan. All of these methods serve both for geothermal monitoring and volcano monitoring as part of the Yellowstone Volcano Observatory. A major component of the Geothermal Monitoring Plan is remote sensing of the Yellowstone volcano and its active hydrothermal areas at various scales. The National Center for Landscape Fire Analysis at the University of Montana and the USDA Fire Sciences Lab acquired visible and mid-infrared (3-5 micron) airborne imagery (night and day flights) for Norris Geyser Basin during October 2005 and October 2006. The Remote Sensing Services Laboratory at Utah State University also acquired visible and thermal infrared (8-12 micron) airborne imagery (also day and night flights) for the Upper Geyser Basin, Midway Geyser Basin and Lower Geyser Basin during 2005 and 2006. Montana State University collaborators are analyzing Landsat satellite imagery for park-wide estimates of radiant heat flux and change detection of active geothermal areas. Geothermal gas and groundwater well monitoring efforts were initiated in 2006. The geothermal gas monitoring instrumentation, developed with assistance from both the Yellowstone and Hawaiian Volcano Observatories, measures hydrogen sulfide, carbon dioxide and basic weather parameters. A specially constructed well adjacent to the Norris Geyser Basin measures water temperature, pH, electrical conductivity, and water level.
V52C-03
Chlorofluorocarbon and Tritium Concentration in Springs in the Norris Geyser Basin Area: Constraining Circulation Rate, Mixing Processes, and Ground-Water Evolution of Locally Derived Ground-Water in Yellowstone National Park, USA.
Chlorofluorocarbon (CFC) and tritium concentrations were measured in cool to warm springs in the Norris area of Yellowstone National Park to constrain the scale and rate of circulation within the shallow, local ground-water system and investigate interactions between locally derived water and the geothermal system. Water from more than 65 springs in and surrounding the Norris Geyser Basin, varying in temperature from 3°C to 90°C, was sampled and analyzed for CFC-11, CFC-12, CFC-113, and tritium. Apparent CFC age was calculated by using noble gas recharge temperature and the elevation of the surrounding volcanic flows. Measured tritium content ranged from 1 to 14 tritium units and apparent CFC age ranged from 1965 to modern. Tritium values were corrected for decay during transport by using the apparent CFC age, yielding the CFC age corrected initial tritium content. Initial tritium concentrations were then compared to an interpolated tritium precipitation curve for Yellowstone National Park, that has been smoothed via transport in the subsurface over a characteristic length scale. CFC ratio analysis shows no evidence of large-scale mixing, or systematic contamination or degradation. CFC age-corrected initial tritium contents match the historical input curve within the uncertainty of the input function, further indicating lack of large-scale mixing, and viability of CFC ages. These results imply that a large volume of ground water in the Norris Geyser Basin area is relatively young and likely recharged within the park.
V52C-04
High Permeability of Yellowstone Quaternary-age Volcanics: Evidence From Streamflow Hydrograph Analysis and Ground-water Ages
Hydrograph analysis and ground-water age of water from low-temperature springs in the greater Yellowstone area provide evidence of high permeability at the basin scale and rapid ground-water circulation in Quaternary- age volcanic rocks. Ratios of annual maximum to minimum mean daily discharge, recession index, base-flow index, and mean annual basin recharge were calculated from streamflow data for the period of record at 32 U.S. Geological Survey gaging stations in the area. Ground-water age was determined from tritium and CFC analyses of water from 30 low-temperature springs discharging from Quaternary volcanics in the Norris area of Yellowstone National Park. The ratio of annual maximum to minimum mean daily discharge varied from 1.9 to 226. Streams with basins underlain by Quaternary-age rhyolites and tuffs had ratios ranging from 2 to 7, indicating that these streams were dominated by ground-water inflows. Streams in basins with high ratios (30- 226) generally were underlain by Precambrian crystalline, Paleozoic sedimentary, and Eocene volcanic rocks and streamflow was dominated by runoff. Recession and base-flow indices have similar characteristics and correlation with geologic units. Water from all springs sampled had measurable CFC and tritium concentrations and a substantial fraction of modern water, indicating rapid flow rates through Quaternary volcanics. Ground- water recharge calculated from base-flow hydrographs with RECESS indicates that 50 percent of precipitation becomes recharge, providing further evidence that Quaternary volcanics are highly permeable. Becasue of the high permeability of volcanic rocks, the ground-water flow system is dynamic and likely an important component in the surficial expression of the Yellowstone geothermal system.
V52C-05
Norris Geyser Basin: An example of Yellowstone National Park's Effort to Monitor Geothermal Resources Using Remote Sensing
Protection of Yellowstone's unique geothermal resources is the main focus of Yellowstone National Park's effort to map and scientifically monitor heat emitted from selected hydrothermal areas and the entire 2.2 million acres of Yellowstone National Park. In 2005, Yellowstone National Park geologists began collaborations with researchers from the University of Montana's National Center For Landscape Fire Analysis(UM-NCLFA), Utah State University's Remote Sensing Services Laboratory, Montana State University and the USDA Fire Sciences Lab to accomplish this task. A goal of the remote sensing component of Yellowstone's Geothermal Monitoring Plan is the estimation of radiant heat flux for Norris Geyser Basin, the Upper Geyser Basin, Midway Geyser Basin, the Lower Geyser Basin, the Mud Volcano area, Mammoth Hot Springs, Hot Spring Basin, the Sour Creek resurgent dome and the entire Park. Norris Geyser Basin is an example of a fracture-controlled hydrothermal basin. The nine hydrothermal sub- basins that compose Norris Geyser Basin include: Porcelain Basin, Steamboat-Echinus, Gray Lakes-Porkchop, the Gap, the West Gap, Reservoir-Upper Tantalus Creek, Lower Tantalus Creek, One Hundred Spring Plain, and Sulfur Dust. Two orthogonal sets of fractures (northeast and northwest; north-south and east-west) direct the flow of heat and water through the otherwise impermeable Lava Creek B tuff within these sub-basins. Airborne mid- infrared (3-5 micron) imagery acquired at night clearly shows the flow of heat and water along these fracture trends. These major fracture sets also partition Norris Geyser Basin into numerous blocks, potentially allowing independent movements among blocks and hydrothermal sub-basins. We estimated radiant heat flux for the Norris Geyser Basin using airborne mid-infrared (3-5 micron) daytime imagery acquired by a thermal imagery contractor on 9 October 2002 and daytime imagery acquired with a different sensor but similar mid-infrared bandpass on 12 October 2005. Because of geolocation issues between the two airborne datasets, UM-NCLFA researchers used "grid-based", basin-wide (3.2 km2) and hydrothermal sub-basin approaches for estimating radiant heat flux. In all hydrothermal sub-basins, the mean radiant heat flux was 3-10% greater in October 2002 than in October 2005 with basin-wide change from 2.1 x 109 W/m2 to 1.9 x 109 W/m2.
V52C-06
Characterizing Thermal features in Norris Basin, Yellowstone National Park, Using Multi- spectral Remote Sensing Data and Dynamic Calibration Procedures
A thermal infrared remote sensing project was implemented to develop methods for identifying, classifying, and mapping thermal features. This study is directed at geothermal features, with the expectation that new protocols developed here will apply to the wildland fire thermal environment. Airborne multi-spectral digital imagery were acquired over the geothermally active Norris Basin region of Yellowstone National Park, USA. Two image acquisitions were flown, with one acquisition near solar noon and the other at night. Raw data from the five sensors were uncalibrated, so a vicarious calibration procedure was developed to compute reflectance for the visible and NIR bands using an independently calibrated hyperspectral dataset. Calibration of the thermal sensor band utilized a dynamic, in-scene calibration procedure that exploited natural, pseudo-invariant thermal reference targets instrumented with in situ kinetic temperature recorders. The calibrated reflectance and radiant temperature data from each acquisition were processed and analyzed to develop a suite of thermal attributes, including radiant temperatures, a daytime-nighttime temperature difference (DeltaT), albedo, an albedo derivative (one minus albedo), and apparent thermal inertia (ATI). The albedo terms were computed using a published weighed-average albedo algorithm based on ratios of the narrowband red and near-infrared (NIR) reflectances to total solar irradiance for the respective red and NIR bandpasses. The weighing factors for each band were the proportion of total solar irradiance incident on the surface within each segment represented by a respective bandpass. In the absence of verifiable "truth," a step-wise chain of unsupervised classification and multivariate analysis exercises was performed, drawing heavily on "fuzzy truth" to assess the quality, efficiency, and efficacy of classification procedures and results. A final classification synthesizes a "geothermal phenomenology" comprised of four analytical components: spectral, statistical, geographical/contextual, and feature space. The four-component phenomenology developed for this geothermal classification provides the "1st-of-its-kind" basis for field-based explorations as well as for image-based analyses utilizing other geospatial information. This study has demonstrated that a classification gradient provides more discriminating information than a ‘hard' classification.
V52C-07
Distribution of Hydrothermal Mineral Assemblages in the Sevenmile Hole Area, Grand Canyon of the Yellowstone River, Yellowstone National Park, Wyoming
Incision of the Grand Canyon of the Yellowstone River in Yellowstone National Park has exposed approximately 350 vertical meters of hydrothermally altered rhyolites. This older alteration formed in the shallow portion of a hydrothermal system that was most likely similar to the modern Yellowstone hydrothermal environment. Hydrothermal fluid circulation is related to the ongoing rhyolitic magmatism that produced the Yellowstone caldera at 640 ka. The rhyolitic magmatism and hydrothermal system are shallow expressions of deeper mantle- derived basalts. The older alteration is well exposed in the Sevenmile Hole area, near the northeastern margin of the caldera. Here, the alteration protolith is the high silica, low-18O, rhyolitic Tuff of Sulfur Creek. The tuff erupted at about 480 ka after resurgent doming associated with the third cycle collapse of the Yellowstone caldera. The tuff is a rheomorphically deformed densely welded agglutinate fallout ash that was deposited along the caldera wall. It contains phenocrysts of quartz, sodic plagioclase, and potassium feldspar. The tuff is exposed from the rim of the canyon, which is very close to the pre-alteration paleosurface, to the river bottom where it is covered by detrital sediments and actively forming hot spring deposits. Rocks exposed within the field area are pervasively hydrothermally altered. Mineral phases in approximately 90 samples were determined in the field using a Portable Infrared Mineral Analyser (PIMA). Subsequently, more precise mineral determinations were made using standard petrographic and powder XRD techniques. The alteration mineralogy consists of variable assemblages that include zones of kaolinite + opal; kaolinite + alunite with local dickite and typically high opal and/or quartz concentrations; highly silicified zones containing illite with or without smectite; and weakly silicified zones containing mostly illite. Minor (less than 1 percent) fine-grained disseminated pyrite is ubiquitous. The alteration is predominantly an advanced argillic, acid sulfate assemblage. Distribution of clay minerals may indicate a vertical temperature gradient related to depth below the paleosurface. Kaolinite and opal tend to occur along and just below the canyon rim. Montmorillonite (smectite) is the dominant clay mineral at intermediate depths or in areas suspected to be near zones of higher heat flow. Illite is the most common clay in the deepest exposures, and also at intermediate depths along a strongly silicified ridge that locally contains vuggy quartz alteration and hydrothermal breccias. The ridge is believed to be a local center of higher temperature fluid upwelling. This hydrothermal environment was most likely produced by deeper seated vapors that rose and mixed with shallow surface meteoric waters. A single sample of alunite yielded an 40Ar/39Ar age of 150 ka for the Sevenmile Hole altered area. It is unknown if the hydrothermal system in this area has been continuously or intermittently active since caldera collapse. The distribution of clay and other hydrothermal minerals preserved in the Grand Canyon walls suggest patterns in temperature and zonation that can be applied to the interpretation of the temporal evolution of the active hot spring systems in Yellowstone National Park.
V52C-08
Magma chamber rejuvenation associated with Yellowstone 516-70 ka post-caldera rhyolitic lava flows : results from crystal geochemistry
After Yellowstone caldera collapsed at 640 ka, it has been the focus of many voluminous rhyolite effusions. Over 900 km3 of magma have been erupted during relatively short periods, at 516 ka, 486-479 ka, 198 ka, and principally from 165 to 70 ka when the largest lava flows were emplaced. The last period itself consists of three episodes, from 165 to 147 ka, 117 to 102 ka, and 72 to 70 ka (Christiansen, 2001). The first-erupted lavas contain mainly plagioclase and sanidine. Quartz becomes increasingly abundant in younger flows, and clearly dominates over plagioclase in the 165 ka and younger rocks. Plagioclase is not present in the 117-102 ka and 72-70 ka lavas. Recent micro-analytical techniques have led to a better understanding of pre-eruptive conditions in magmatic systems by studying trace element concentrations and zoning in volcanic crystals. We analyzed plagioclase by electron microprobe and Ti in quartz by laser ablation-ICP-MS in order to assess changes in crystallization conditions from one period to the next, and also to reconstruct the pre-eruptive history of individual lava flows, by searching for potential dissolution, overgrowth or zoning features. Plagioclase in the lavas shifts progressively from a median value of An28 at 516 ka to An21 at ~ 160 ka. The plagioclase also contains decreasing amounts of K over time for a given level of Ca; for example, K2O at An22 declines from ~ 1.9 wt % in the oldest lavas to ~ 1.4 wt % in the 160 ka episode lavas, suggesting a reduced ability of plagioclase to incorporate K. This may translate into crystallization in a progressively cooler or shallower environment. Average Ti contents in quartz rims decrease from ~ 150- 200 ppm in the older lavas to ~ 100-125 ppm at ~ 160 ka and ~ 75-90 ppm at 70 ka. Zoned quartzes are found in most flows. They show higher Ti contents in the cores, with up to 240 ppm in the older lavas, 150 ppm at ~ 160 ka and 120 ppm at 70 ka. The temporal Ti decline in quartz rims and the normal zoning patterns for Ti in individual crystals are also both suggestive of formation in a cooling environment. Texturally, the plagioclase is commonly sieved, especially in the most primitive flows of the 516-479 ka period. The rims are typically thin, discontinuous and/or absent and do not appear to reflect overgrowths; analyses of sieved cores and fresh rims did not reveal any compositional zoning. For quartz, long narrow glass re-entrants commonly penetrate through crystals and are connected to the matrix glass. Major and trace element glass analyses did not reveal any noticeable compositional change between the re-entrants and the matrix. Sieved textures are commonly formed by dissolution resulting from heating, and we envisage a similar mechanism for the re-entrants in quartz. We hypothesize that discrete episodes of magma recharge occurred after caldera collapse, with further recharge episodes related to the 165 to 70 ka period when most of the lava was erupted. Neither quartz nor plagioclase bears an imprint of remobilization in a more mafic melt, so two possibilities present themselves. (1) The replenishing material never came into contact with the crystals which were erupted, and the crystals may have been merely heated causing embayments and partial dissolution. (2) The replenishing magma was high-silica rhyolite. Where present, rims on plagioclase are not well developed, suggestive of only a short interval of time from replenishment to eruption.