Incorporating Predictive Mapping to Advance Initial Soil Survey: An Example from Malheur County, Oregon

摘要

Digital soil mapping (DSM), including predictive mapping and GIS representation, is seen as the future for National Cooperative Soil Survey (NCSS) operations. Introduction of DSM technology to an initial soil survey, from inception through three seasons, is the focus of this article. We worked with the Malheur County, Southern Part Soil Survey (MCSPSS), providing predictive soil maps produced using machine learning (decision-tree analysis, DTA) in a series of iterations to reach a stable work flow with field soil scientists. Along the way, we took a close look at the interpersonal challenges involved in introducing a new technology for soil survey. Based on these experiences we make recommendations for incorporating DSM technology into the USDA-NRCS Major Land Resource Area (MLRA) Soil Survey Restructuring Plan.