Analysis of Model-Calculated Soil Moisture over the United States (1931–1993) and Applications to Long-Range Temperature Forecasts

Abstract

A long time series of monthly soil moisture data during the period of 1931?1993 over the entire U.S. continent has been created with a one-layer soil moisture model. The model is based on the water budget in the soil and uses monthly temperature and monthly precipitation as input. The data are for 344 U.S. climate divisions during the period of 1931?1993. The main goals of this paper are 1) to improve our understanding of soil moisture and its effects on the atmosphere and 2) to apply the calculated soil moisture toward long-range temperature forecasts. In this study, the model parameters are estimated using observed precipitation, temperature, and runoff in Oklahoma (1960?1989) and applied to the entire United States. The comparison with the 8-yr (1984?1991) observed soil moisture in Illinois indicates that the model gives a reasonable simulation of soil moisture with both climatology and interannual variability. The analyses of the calculated soil moisture show that the climatological soil moisture is high in the east and low in the west (except the West Coast), which is determined by the climatological precipitation amounts. The annual cycle of soil moisture, however, is determined largely by evaporation. Anomalies in soil moisture are driven by precipitation anomalies, but their timescales are to first order determined by both climatological temperature (through evaporation) and climatological precipitation. The soil moisture anomaly persistence is higher where normal temperature and precipitation are low, which is the case in the west in summer. The spatial scale of soil moisture anomalies has been analyzed and found to be larger than that of precipitation but smaller than that of temperature. Authors found that generally in the U.S. evaporation anomalies are much smaller in magnitude than precipitation anomalies. Furthermore, observed and calculated soil moisture anomalies have a broad frequency distribution but not the strongly bimodal distribution indicative of water recycling. Compared to antecedent precipitation, soil moisture is a better predictor for future monthly temperature. Soil moisture can provide extra skill in predicting temperature in large areas of interior continent in summer, particularly at longer leads. The predictive skill of soil moisture is even higher when the predictand is daily maximum temperature instead of daily mean temperature.