HR: 1330h
AN: H33B-05    [Abstracts]
TI: Exploring Linkages Between Land Use and Hydroecology Using Multivariate Analysis and Process-Based Models
AU: * Welty, N R
EM: weltynic@msu.edu
AF: Michigan State University, 206 Natural Science Building, East Lansing, MI 48824
AU: Hyndman, D W
EM: hyndman@msu.edu
AF: Michigan State University, 206 Natural Science Building, East Lansing, MI 48824
AB: It is well known that land uses influence water quality and ecosystem integrity but the linkage is often poorly understood. The causes of impaired water quality need to be understood to allow for educated land management decisions that will preserve our hydrologic resources and ecosystems for future generations. Frequently, these decisions deal with the mitigation of common stream stressors including low dissolved oxygen levels, increased temperature, and elevated nutrient concentrations. Urban and agricultural land uses are often blamed for harmful effects but it can be difficult to link land use with stream water quality. Statistical methods can infer land use-water quality linkages, but they provide little insight to underlying mechanisms. In contrast, process-based numerical models simulate the necessary hydro/bio/geo/chemical aspects of streams but often require extensive and/or idealized parameterization. This study utilizes a novel approach that combines detailed synoptic water quality data, land use data, multivariate techniques, and process-based numerical models to explore water quality-land use relationships. Principal component analysis identified land use-water quality signatures for 120 source areas ranging from 1 to 47 square km across Michigan. Correlations were established between water quality and land use, as well as with leaf area index, a MODIS data set. These associations were then explored using widely-used water flow and quality models, including QUAL2K and FEMWATER. The models were coupled to account for both surface water and groundwater, since Michigan streams are groundwater-dominated during the summer. This hybrid statistical/modeling approach has the advantage of simultaneously identifying stressor-response relations and mechanistically explaining these links.
DE: 1803 Anthropogenic effects
DE: 1871 Surface water quality
SC: Hydrology [H]
MN: 2005 Joint Assembly