Modeling Wetland Communities Based on Groundwater Discharge as Predicted from Thermal Imagery and Geomorphology

Migratory birds in Western North America, especially those dependent on wetlands, are experiencing declines due to habitat loss and ecosystem degradation. An integrated understanding of the geohydrology and ecology of Western wetlands is needed to understand, predict, and manage linkages between geohydrology and aquatic ecosystems and the response of wetland habitat to climatic and management actions. The NRMSC has begun ecological research to understand the interrelated processes that drive wetland productivity in the Northern Rocky Mountains. One area of focus has been the Red Rock Lakes National Wildlife Refuge. The Refuge is critical for many avian species, from white-faced ibis to trumpeter swans; indeed, it supports the largest concentration of nesting swans in the U.S. portion of
the Rocky Mountain Subpopulation. The Refuge is also a key research site because of USGS collaboration with the U.S. Fish and Wildlife Service, which must optimize water management for wetland birds in the Refuge in relation to agricultural water rights, cattle grazing, and wilderness values.

Wetland birds, especially swans, need large tracts of submergent vegetation on which to feed and emergents in which to nest. Management of wetlands for habitat is hampered by lack of knowledge about ecological interrelationships. Ground water discharge is thought to be a large component of the water input in the Red Rock Lakes wetland complex, and is also likely to ameliorate the effects of drought in this system and in montane systems throughout the West. There is evidence that ground water discharge is vitally important to wetland conditions, and that it has a major role in governing the dynamic nature of those wetlands. Additionally, climate change may affect the magnitude, spatial extent, and seasonal timing of ground water discharge differently than surface water inputs to wetland systems.

Documenting ground water discharge in the field is laborious, especially at a landscape scale. More efficient methods are needed, particularly when working in extensive wilderness and remote areas. Finding an affordable means to remotely determine areas influenced by ground water would be of tremendous benefit. One helpful characteristic that should lend itself to a remote mapping approach is water temperature; ground water is relatively constant in temperature throughout the year, while surface water varies seasonally. Airborne and satellite thermal data have been used for mapping thermal features of large water bodies, and airborne thermal data have been used successfully to locate areas of ground water discharge in some aquatic settings. However, airborne data are expensive and require speculation of optimal acquisition windows in advance. Repeat-acquisition satellite data are more attractive from these standpoints, but we know of no studies that have tried to apply satellite data to map thermal characteristics in wetlands, particularly as they pertain to ground water influence. Several characteristics of the Refuge wetland complex lead us to anticipate a strong ground water signal and success with this approach: (1) seasonal temperature differences between ground water and surface water are large, (2) water depths are shallow, often less than 1m, (3) flux of ground water in the wetland complex is large and some areas of obvious ground water discharge have already been located, (4) the water table is within 1m of the land surface over large portions of the watershed when surface water is absent, and (5) known locations of geothermal discharges will provide an exceptionally strong signal that can be used to calibrate our interpretations of the satellite imagery.

From past field observations at the Red Rock Lakes Refuge, we have found that ground water temperature remains at about 10?C year-round, while the temperature of surface water fluctuates from 0-25?C. Ground water is colder than surface water in summer, is likely rich in calcium carbonate, and has higher specific conductivity than surface water. This translates to differences in the vegetation communities, and, hence, wildlife communities. Furthermore, the Centennial Valley, in which the wetlands are situated, is one of the most tectonically active areas in the Lower 48 states. A suspected fault may also contribute ground water from a deeper, warmer source and at a different scale. We have not yet been able to document ground water discharge associated with this fault in the wetlands, partly due to its remoteness. However, a warm spring is located along the strike of the fault in the upper watershed and should be clearly visible on satellite imagery. Studies on the role of earth surface dynamics as they relate to wetland formation and maintenance are in their infancy in the Centennial Valley. By investigating the near-surface geological processes, especially landform development and tectonics, we hope to isolate other abiotic parameters associated with ground water discharge and their biotic communities.

The set of linkages described here is the impetus for this study. We will test the hypothesis that wetland bird communities are associated with specific vegetative communities that, in turn, are associated with specific ground water discharge conditions that can be predicted on the basis of geomorphology and thermal characteristics. Our general approach will be to use existing field and remotely sensed data to build models that predict areas of ground water influence, then to test the models with new field data from 2005, 2006, and 2007. The first step in understanding such relationships is to determine specifically where groundwater discharge occurs in these montane wetlands. Therefore, we will determine if affordable remotely sensed data can distinguish differences in surface water temperatures within wetlands, and whether those differences relate to groundwater discharge. We seek to identify discharge sufficiently large in magnitude, areal extent, or velocity to impact the distribution of nearshore and within-lake vegetation. This will be done by relating thermal anomalies to areas of discharge, supported by hydrologic measurements and geomorphologic controls. Surface and subsurface geology will be evaluated in relation to any patterns of groundwater discharge detected; such geological investigation will include Late Quaternary alluvial and lacustrine deposits.

Return to Rick Sojda's Research Projects
Return to Lake and Wetland Ecosystems
Return to Science