Spatial Disaggregation Techniques for Evaluating and Visualizing Soil Map Unit Composition

The soil-landscape model is the fundamental basis of most soil surveys. Soil-landscape relationships are described and documented in soil map units and extrapolated across a region through soil maps. In many cases, soil map units contain more than one soil class (component); however, the spatial distribution of components within a map unit must be inferred from the soil-landscape model contained in the soil map unit description.

Soil map units can be disaggregated into individual components based on soil-landscape relationships documented during a soil survey. Disaggregating existing soil maps facilitates the introduction of more information about the distribution of soil components on the landscape without requiring additional field data collection. This process may allow us to visualize and evaluate soil-landscape relationships documented in the aggregate data and more precisely estimate map unit and component properties. In addition, the resulting soil component maps may enable visualization of component-level properties and interpretations.

We mined soil-landscape relationships identified in the Soil Survey of Denali National Park and Preserve Area manuscript and in the supporting point and aggregate databases in order to develop a soil component map for selected soil map units. Data describing soil-landscape relationships critical to the development of quantitative soil-landscape models for selected soil map units and components was extracted from the National Soil Information System database. Key soil forming factors were identified and assigned ancillary data layers as proxies, and a database of rules based on data extracted from the database was developed. The resulting models were extrapolated based on digital elevation model derivatives and land cover classifications using simple raster analysis methods in ArcGIS.