Limiting Imperviousness to Maintain Ecological Quality: Are Threshold-Based Policies a Good Idea?

The effects of urbanization on flows and water quality within a watershed have been documented for some time. Further, many models have been constructed that are sensitive to land use and land use change as predictors to make estimates of the impacts of urbanization. Impervious cover is probably the most basic quantity that can be used to quantify the degree and extent of urbanization across a landscape. Despite the apparent simplicity of recognizing imperviousness from the ground, measuring this quantity with accuracy from commonly available GIS-based products is not trivial. The situation is further complicated if one takes the perspective of the hydrologist. What imperviousness value should be used in a given analysis method? Is all imperviousness created equally? Here we investigate the several methods for characterizing imperviousness from several perspectives: remote sensor measurement, inference from land use/land cover, and "direct measurement". Our findings indicate that although different methods for measuring imperviousness are highly correlated, they can differ dramatically in their magnitude and there may exist local or systematic biases in the reported values associated with any one method. Therefore, if imperviousness values of 10% or greater are a concern, they are sensitive to the method used. As a final step, we illustrate how different measures of imperviousness can introduce error into a common hydrologic model — a regression model for peak flow estimation that uses imperviousness as a predictor. Peak discharge estimates generated from one measure of imperviousness averaged only 68.7% of those from a different measure. The implication is that it is imperative that methods for calculating and reporting imperviousness be standardized.