Is the soil frozen or not? An algorithm using weather records

Abstract

Frozen soil water is important in hydrologic events because it reduces water infiltration. The presence of soil ice can be predicted reasonably well from detailed knowledge of the soil and microclimatic variables, but this type of information is generally unavailable. Consequently, the purpose of this study was to start with fundamental relations and see how well frozen soil conditions could be identified from daily weather station records of maximum‐minimum temperatures, solar radiation, and snowfall. Two relations were developed, one based on the soil‐atmosphere energy budget and the other on the heat flux across the soil surface layer. Conceptually, the two equations may be used together to give daily snowmelt as well as soil thawing and freezing rates, but in practice, the snowmelt prediction is probably not yet accurate enough for most practical applications. The simpler equation, describing the heat flux in the soil surface, does not require solar radiation input, yet it gave fair predictions of frozen soil on five diverse sites studied in the Palouse region of eastern Washington. Both approaches require only a single constant that accounts for individual site conditions such as slope, aspect, cover, and soil properties.