Evaluation of the EPIC Model to Predict Soil Moisture and Temperature Regimes.

NRCS soil scientists have historically lacked measured soil moisture and temperature data to assign soil moisture and soil temperature regimes for taxonomic classification and therefore, have relied upon the use of non-validated models.  Enhanced technology in recent years has made available measured daily soil moisture and temperature data by depth.  The Soil Climate Analysis Network (SCAN) is a network of soil and atmospheric climate stations designed to monitor soils that would be sensitive to slight changes in climate.  Long-term SCAN data collected at sites that represented map unit components with a large extent (i.e., benchmark soils) were used to evaluate soil temperature and moisture predictions from simulations with a modified version of the Environmental Policy Integrated Climate (EPIC) model.  The modified EPIC also included a module to assign the soil moisture and soil temperature regimes defined in the second edition of Soil Taxonomy.  Measured soil moisture and temperature data from SCAN soil climate stations located in Arizona, Georgia, Kentucky, New Hampshire, North Dakota, and Washington were compared to soil moisture and temperature data simulated by the modified EPIC model.  In addition, predicted soil moisture and temperature regimes classified by the EPIC module were compared to the current soil taxonomic classifications from each of the SCAN stations evaluated.  Normal years for the period 1995 to 2010 were evaluated.  The modified EPIC model reasonably demonstrated the ability to automate the identification of soil moisture and temperature regimes as currently defined in Soil Taxonomy for these study sites and can be used to predict soil climate regimes in other locations.