Urban validation site for testing impervious surface models derived from remotely sensed imagery

Accurate quantification of impervious surfaces is a necessary input in a variety of urban applications including hydrologic and hydraulic models and landscape change. The purpose is to assess the performance and effectiveness of multiple sensor platforms for the delineation of impervious surface in an urban setting using multifarious classification strategies. Data acquired from airborne and satellite based sensors are used, along with a variety of classification and data fusion strategies, to gather a cost versus reliability measurement for each of the systems. A framework is presented to aid users in selecting the appropriate dataset and methodology for their specific situational needs. The impervious surface generation models are applied to remotely sensed data collected over the Springfield Urban Validation Site (UVS), an approximately 1-km N-S by 4-km E-W urban corridor within the City of Springfield, Missouri. The site is highly documented with respect to position and composition of structures and land covers and consists of varying aged residential developments, commercial, institutional, parks and open space, and light industrial land uses. The models and comparisons developed here can be reliably used to estimate the costs and spatial variability of different methods of impervious surface generation from various imagery inputs, aiding urban planners and managers in the assessment of the errors and biases of various impervious surface generation strategies