Utilizing Computer Simulations to Evaluate Ground-Water Flow Patterns within a Peatland System

Computer simulations provide a realistic constraint for conceptual models of wetland hydrology and the assessment of sparse hydrologic data collected within wetland systems. Three-dimensional computer models based on regional data (topography, climate data, and hydrography) have been created to assess ground-water monitoring data collected from Caribou Bog, a 2200 hectare multi-unit peatland located in central Maine. Computer models were developed using FiPy, a flexible PDE solver developed at the National Institute of Standards and Technology. Two suites of data have been collected from this peatland: 1) two years of data collected from 18 well clusters arranged along several transects through different sections of Caribou Bog intended to identify broad ground-water flow patterns and 2) one year of data collected along two short transects of well clusters located across an inferred esker imaged through geophysical surveys, intended to assess the role of biogenic gas on ground-water flow within the peatland. Our data indicates that ground-water within the peat flows in a fan shaped pattern that is easily emulated by computer simulations using regional data. This flow pattern is altered near the esker system, where ground water deviates from the regional pattern and converges toward the esker system (or associated pools). Field data and computer simulations indicate that the esker exerts a strong local control on ground-water flow, enhancing vertical flow, that is not evident in the broader regional data. While broad hydrologic function can be identified in our peatland system based on regional 'proxy' data constrained by computer simulation, details of Caribou Bog's hydrologic function require the collection of additional site specific data.