Wednesday, November 4, 2009
Convention Center, Exhibit Hall BC, Second Floor
Abstract:
The truth of any model is at best partial, and hence relative; a model can only be “more-or less” true. Moreover, a model can seem to be true in some ways and far from it in other ways. In fact, what we are looking for in the application of digital data as it is applied to the process of soil survey is not “truth,” or “verity,” but validity. For example, when defining the validity of a soil taxon, we conform to accepted principles of sound pedologic classification. We discuss the usefulness of various tools used in digital soil mapping for providing variables to adequately model differences and composition of soils geospatially across a complex landscape ranging from arid to subalpine in southern Nevada. We stress the need for the use of field notes to ground truth the validity of the digital layers and to develop the soil-landscape concepts which represent the actual climatic variability reflected in elevation, surface orientation (slope and aspect), and shadows cast by topographic features to describe the high spatial and temporal heterogeneity in local energy and water balance. These features determine micro-environmental factors such as air and soil temperature regimes, evapotranspiration, snow melt patterns, soil moisture, and light available for photosynthesis. The resulting patterns then can be grouped into meaningful map units and scaled back to the landscape. These patterns can then be used to refine digital terrain models that provide information indispensable for land resources management and environmental assessments.