Observation-Based Algorithm Development For Subsurface Hydrology In Northern Temperate Wetlands

This study investigates wetland subsurface hydrology, as well as biogeochemistry - which is strongly influenced by water and temperature dynamics - as these interactions are expected to be highly significant, yet remain poorly represented in current ecosystem and climate models. Northern wetlands have received widespread public attention due to steadily increasing summer mean global temperatures, extreme precipitation events and higher rates of natural greenhouse gas emissions, as well as the significant impacts on them due to human activities. The goal of my graduate research has been to improve quantification of the role of subsurface hydrology in northern wetlands by using a macroscale hydrological model, the Variable Infiltration Capacity (VIC) model. The existing VIC model was modified to better represent the effect of surface and subsurface water storage in managed wetlands. An improved water table depth calculation, based on a drained to equilibrium assumption, was incorporated into a new subsurface drainage algorithm. The spatial variability of water table depth across landscape positions has been represented using a topographic index approach. By incorporating a water table gradient into the VIC grid cell, subsurface-surface water exchange within the wetland can also be represented, dependent on land surface class. This algorithm was developed and evaluated using data at scales ranging from field to small watershed, which included a small wetland at the Agronomy Center for Research and Education (ACRE), the long-term drainage experiment at the Davis-Purdue Agricultural Center (DPAC), and a cooperators mint farm in Pulaski, Indiana. The improved model has been used at larger scales - from large watersheds to regional scale - to better understand the subsurface hydrology affected by drainage practices throughout the poorly-drained Midwest agricultural regions. Recent concern regarding high rates of soil organic matter decomposition due to artificial drainage enhancements motivated an integrated field and modeling experiment to quantify the influence of water management on cultivated organic soils in the Kankakee River basin, a flat outwash plain covered with relatively deep, poorly drained soil with high organic matter content. Methane and carbon dioxide emissions were simulated by using soil temperature, water table position and net primary production generated from the VIC model and evaluated using CO2 flux measurements, water table height and soil moisture measurements. The model simulations do support the high rates of subsidence previously reported for these high organic matter soils, but most of the subsidence took place soon after the introduction of agricultural drainage. Another case study evaluated the role of anthropogenic modifications to drainage conditions and wetland extent on streamflow in the upper Wabash River basin. An initial test case demonstrated that a depressional wetland perched on the Tipton Till Plain tends to recharge soil moisture in riparian areas by late summer, reducing the volume of baseflow downstream. When scaled up to the upper Wabash River basin , the study demonstrated that wetlands provided more temporal surface water storage and served to reduce peak flows. Subsurface drainage increased the high flow, mean flow, and Richard-Baker flashiness Index (RBI), and reduced the low flow and flow distribution. Stream network density analysis showed that simulations with lower drainage density (representing historic, natural conditions) had relatively lower high flow and smaller RBI. These results provide evidence that although drainage creates more pore space in the soil profile - reducing surface runoff - it also creates more flow paths, allowing water to travel to the watershed outlet more quickly