ABSTRACT
GIS serve as useful extensions to spatial environmental models providing them with hierarchical data storage and preprocessing facilities. On the other hand the model can significantly extend the analytical power of a GIS by adding functionality in dynamic data analysis. Based on the Patuxent watershed (Maryland) case study, we illustrate how a process based ecological model and GIS (GRASS and ARC/INFO) can be mutually complementary for environmental analysis over a variety of spatial scales and resolutions. When rescaling from smaller to larger cell sizes, issues regarding the loss of information by averaging should be addressed. Different GIS filters are used to customize the method of grid averaging. In addition to the GIS spatial analysis of error associated with the rescaling operations we are focusing on the sensitivity of the model to these changes. The 1-Degree DEM (3- by 3-arc-second data spacing) and digital elevation data in 3.75' quad series format, that have a ground resolution of 4 feet per pixel, are used in this study in conjuncture with the hydrologic module of the Patuxent Landscape Model. In most cases the "geographic" error caused by averaging seems to be more significant than the "simulation" error, produced by the model. However in certain threshold conditions the "simulation" error is crucial.