Beyond Total Impervious Area: A New GIS Framework for Characterizing Urban Basins in Water Resources Management Applications incorporating Hydrological Connectivity

The percentage of total impervious area (TIA) is a popular proxy for the level of urbanization, adopted in many applications ranging from water quality assessments in developed watersheds to regional modeling for flood prediction in ungauged basins. However, TIA cannot satisfactorily capture important interactions between land development and its impacts on runoff patterns and peak flows, such as the effects of the spatial distribution of impervious patches, or the distinction between directly and indirectly connected impervious areas. In other words, TIA cannot incorporate information on hydrologic connectivity.  However, during a storm, these differences may have major implications on the surface runoff volumes that are contributed to the stream network from the impervious portions of a watershed, as well as their travel times, ultimately leading to large variability in the hydrologic response. E.g., the occurrence of pervious areas along the runoff paths from impervious patches to the stream may significantly decrease water volumes from those patches, attenuating both their impacts on direct runoff and the risks of stream contamination from localized pollution sources. Many recent strategies for flood mitigation at the local scale (also known as best management practices, or BMPs) exploit the concept of impervious-area disconnection to reduce peak-flow volumes via marginal landscape changes.

Although several other urbanization descriptors have been proposed in the literature, there is no agreement yet on alternative indices that could replace the traditional TIA in hydrological applications, so it is still predominantly used. One reason may be that these alternative measures may be difficult to derive for a given case-study basin. Some require the topology of the watershed’s stormwater drainage network, which is rarely available, especially in the case of large-scale studies. Other methods analyze patterns in concurrent flow and precipitation series, attempting to implicitly determine the proportion of directly connected impervious area from runoff coefficients, under the assumption that it is this component of the basin’s surface that governs its hydrologic response when smaller storms occur.  But this approach comes with major uncertainties related to the potentially variable contributions from pervious areas, depending on their antecedent soil moisture conditions.

We propose a new GIS framework for deriving connectivity-based urbanization measures using the digital elevation model, land-use, and soil maps of a watershed. We analyze its correlation to other, established urbanization measures, and test its predictive power in regionalization approaches. Our new index can aid urban water management on many fronts, including the assessment of alternative candidate BMPs on the overall connectivity of a watershed, enhancing the accuracy of regional models for prediction in ungauged basins (PUBs), and the analysis of the relationships between urbanization and water quality. The proposed methodology uses easily available datasets and can be implemented using Google Earth Engine and other open-source software, thus ensuring broad applicability irrespective of the study scale, as well as consistent analyses across different regions.