Physically-based modifications to the Sacramento Soil Moisture Accounting model. Part A: Modeling the effects of frozen ground on the runoff generation process

Abstract

Summary This paper presents the first of two physically-based modifications to a widely-used and well-validated hydrologic precipitation–runoff model. Here, we modify the Sacramento Soil Moisture Accounting (SAC-SMA) model to include a physically-based representation of the effects of freezing and thawing soil on the runoff generation process. This model is called the SAC-SMA Heat Transfer model (SAC-HT). The frozen ground physics are taken from the Noah land surface model which serves as the land surface component of several National Center for Environmental Prediction (NCEP) numerical weather prediction models. SAC-HT requires a boundary condition of the soil temperature at the bottom of the soil column (a climatic annual air temperature is typically used, and parameters derived from readily available soil texture data). A noteworthy feature of SAC-HT is that the frozen ground component needs no parameter calibration. SAC-HT was tested at 11 sites in the U.S. for soil temperature, one site in Russia for soil temperature and soil moisture, eight basins in the upper Midwest for the effects of frozen-ground on streamflow, and one location for frost depth. High correlation coefficients for simulated soil temperature at three depths at 11 stations were achieved. Multi-year simulations of soil moisture and soil temperature agreed very well at the Valdai, Russia test location. In eight basins affected by seasonally frozen soil in the upper Midwest, SAC-HT provided improved streamflow simulations compared to SAC-SMA when both models used a priori parameters. Further improvement was gained through calibration of the non-frozen ground a priori parameters. Frost depth computed by SAC-HT compared well with observed values in the Root River basin in Minnesota.