| |
Presented at the NABS Annual meeting, Pittsburgh, Pennsylvania, 2002
in Landscapes
FLOW CHARACTERIZATION FROM LANDSCAPE VARIABLES: AN ALTERNATIVE APPROACH FOR ESTIMATING NATURAL FLOW REGIMES.
L.C. Hinz Jr.1,2, M.J. Wiley1, P.W. Seelbach2, and J.D. Allan1. 1School of Natural Resources & Environment, University of Michigan, Ann Arbor MI 49109-1115, 2Institute for Fisheries Research, Michigan Department of Natural Resources, Ann Arbor, MI 48109-1084
Stream flow can be considered a master variable that influences lotic systems by regulating important physicochemical characteristics such as channel geomorphology, current velocity, habitat diversity, oxygen concentration, substratum stability, and water temperature. However, patterns of land and water use by humans have changed the pathways that water moves on the landscape and thereby altered natural flow regimes. These human-induced hydrologic perturbations can lead to novel conditions that no longer support native species or sustain healthy ecosystems. Traditional flow characterization methods typically require long-term gauge records under near homeostatic conditions. An alternative approach is to develop and apply predictive models using catchment characteristics. We found six major types of flow regimes illustrating tradeoffs between groundwater and runoff sources by determining flow exceedence curves for Michigan rivers. Existing data from gauged sites and landscape factor information (e.g., basin area, catchment slope, landcover, precipitation, soils, surficial geology) derived using a GIS were used to develop MLR models by adding landscape factors to basic hydraulic geometry equations. Preliminary models have explained 90% or more of the variation in specific flow exceedences. Further application of the "presettlement" landcover to these models allows for predicting "unaltered" hydrologic conditions or "natural flow regimes" for these rivers.
|
|
