The National Cooperative Soil Survey has completed initial mapping in Pennsylvania, and is entering an update cycle. Soil survey update projects increasingly rely on geospatial technologies to investigate and define complex soil-landscape relationships. We propose a holistic integration of geospatial tools, landscape metrics, and elevation, geology and climate datasets to build rule structures for field mapping in the soil survey update of Potter County, Pennsylvania. The county is characterized by the Late-Wisconsinan glacial border, the deeply dissected valleys of the Allegheny Plateau, and a periglacial transition zone between the two regions. Our research goals are to: 1) capture existing knowledge and landscape relationships from the original Potter County Soil Survey (Goodman et.al., 1958); 2) develop landscape metrics to define geomorphologic and hydrologic characteristics of the respective physiographic regions; 3) model soil climate regimes across highly orographic and dissected landscapes; and 4) apply these metrics to the development of rule structures to predict map unit occurrences for the Nolo, Cattaraugus and Germania soil series. Germania soils were not extensively mapped, but they are associated with distinctive landforms and have significance to agriculture and cultural resources. Cattaraugus soils were mapped extensively, both within and beyond the glacial boundary, and therefore need a new landscape model to predict their occurrence or re-correlation with appropriate soil series. Nolo soils have limited spatial extent, but these soils are associated with distinct high elevation depressional landforms and unique (relict) plant communities. Soil temperature and moisture regime surfaces were generated with ArcGIS using output data from the Newhall Simulation Model (NSM, Van Wambeke, 1992). Thirty-year precipitation and temperature normals were derived from 84 National Weather Service (NWS) cooperative stations in a 100 km region around Potter County. Based on the population of NWS stations, mean annual soil temperature (MAST) can be predicted in a terrain regression as:

MAST = 137 – (1.97 * LONG) + (0.038 * LAT) – (0.00248 * ELEV)

where latitude and longitude are in decimal degrees (geographic coordinate system, NAD83) and elevation is expressed in feet. Our modeling indicates that frigid and possibly cryic temperature regimes are more extensive than previously mapped. Similarly, soils of the higher elevations in the unglaciated part of Potter County have climate regimes which are influenced by orographic effects and near-perudic conditions. Landscape metrics such as curvature, slope gradient, and slope aspect were derived from 10m DEMs in ArcGIS. Additional products were generated from these layers, including slope shape (Ruhe, 1975; Wysocki, 2000), flow direction, flow accumulation, and topographic wetness index (TWI; Beven and Kirby, 1979) will be used to develop rule structures for legend construction and the prediction of soil map unit occurrences. The digital terrain models function as a preliminary soil survey update process and will test approaches to support field mapping activities.