Progress and Challenges Related to Digital Soil Mapping in the United States.

Many of the scientific questions regarding land use and resource management require consistent and reliable information on soils and soil variability. Applications of soil information are becoming increasingly larger in extent, covering multi-county, multi-state, and multi-national areas in order to address issues that require regional to continental to global response. Digital soil mapping (DSM) approaches seek to maximize information delivery from existing soils databases and develop new geospatial soil data products. These DSM products are expected to have enhanced spatial detail, represent depth variation, and include information on uncertainty. To respond to these challenges, soil survey activities in the US must capitalize on DSM approaches to improve existing soil data products and develop complementary new data products. Harmonization and disaggregation techniques can be developed and applied to existing soil data (often referred to as legacy data), including maps, pedon data (characterization data and morphological descriptions), and point measurement data. From these data, pedotransfer and taxotransfer functions can be used to develop maps of soil properties. Development of new soil data products can occur through incorporation of new environmental covariates (e.g., gamma radiometrics) and implementation of novel pedometric models. Further, DSM has put much emphasis on mapping of stable soil properties. Approaches are needed that explicitly address the importance of the temporal component in soil variability and soil mapping and produce estimates of dynamic soil properties. Temporal databases, such as remotely-sensed imagery from Earth-orbiting satellites, are expected to have useful applications for DSM. Other important issues are validation of digital soil maps, estimation of uncertainty, scale, and resolution. With improved soil information products, the emphasis can shift from digital soil mapping to digital soil assessment.