Evaluating the historical impacts of landscape transformation on hydrologic fluxes for environmental assessment and modeling

Alison Meredith Goss, Purdue University

Abstract

Land use/cover change can have a profound impact on the hydrologic cycle, with significant consequences for both water resources management and ecosystem services. However, a lack of concurrent land use/cover and water flux data for large watersheds over historical time scales has largely prevented scientists and water resource managers from being able to perform detailed hydrologic analyses spanning 50 or more years. To overcome the limitations provided by a lack of historical land use/cover data, a “backcast” method was developed to estimate historical land cover using proxies from U.S. Census data and agricultural statistics. This backcasting method was applied to the Muskegon River Watershed in southwestern Michigan. Land cover backcasts from 1910 to 1978 were then used as inputs into a macroscale water and energy balance model, the Variable Infiltration Capacity (VIC) model. Model results for simulations run using identical climate records but different land cover data suggest that widespread agricultural abandonment and afforestation have dramatically altered the seasonal water balance of the basin. Also, urbanization over the last century has changed the ratio of stream flow provided by overland flow versus near surface baseflow. However, despite strong patterns of change at sub-watershed scales, the model predicted no significant overall change in mean monthly or annual stream flow for the watershed as a whole. Thus water resources managers focusing solely on watershed-scale stream flow for impact assessment would not detect fundamental alterations in the water balance of the region. Future refinements of the method developed here to analyze the hydrologic dynamics of large watersheds will include modification of proxy datasets used to backcast land cover change and enhancement of the parameterization of vegetation and urban areas within the hydrologic model. Overall, this approach allows water resource managers to develop large-scale assessments of hydrologic changes resulting from land cover variations, using data that are available for most large watersheds in the U.S.

Degree

Ph.D.

Advisors

Harbor, Purdue University.

Subject Area

Physical geography|Hydrologic sciences|Environmental science|Geography

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS