The North American Summer Monsoon (NAM) affects summertime hydrologic cycle in a large part of United States and Mexico. Predicting regional hydrologic cycle associated with the NAM is important for assessing water resources, occurrence of weather-related natural disasters, and agricultural production in the region.

An analysis of century-long rainfall data shows that the interseasonal and continental-scale variations of precipitation anomalies associated with interannual variations of the NAM are notably different between the two periods of 1901-1962 and 1963-1994. In Arizona-New Mexico (AZ-NM), wet (dry) summers are preceded by dry (wet) winters only after 1962. A summer rainfall pattern of wetter (drier) AZ-NM, and drier (wetter) Great Plains (GP), is observed only after 1962, as well. The strongest contrast in summer rainfall anomalies appears between AZ-NM and the SEUS for 1901-1962, while it appears between AZ-NM and the GP for 1963-1994. For the period after 1962, the rainfall variability obtained here show the same features found in earlier studies. The hydrologic cycle associated with the NAM is affected mainly by two factors: global circulation and the regional-scale responses to the large-scale circulation. The long-term variations of rainfall characteristics found here suggest that the large-scale circulation that drives the NAM is affected by low frequency variability of the EarthÕs climate system. Causes of the changes in the rainfall characteristics around 1960 are not clear due to a lack of large-scale data before 1950. Instead, a coupled GCM experiment is being designed to investigate the long-term variability that is related to the observed transition of the NAM rainfall characteristics.

Concerning the role of regional-scale feedbacks in shaping the summertime hydrologic cycle of the SWUS, the effects of soil moisture content (SMC) on the summer rainfall has been investigated using a regional climate model, the interactively-coupled MAS/SPS. Results of seasonal simulations show that the differences in the simulated summer rainfall and evaporation are generally positively correlated with the differences in the SMC at the beginning of summer in the SWUS. The effects of the initial SMC differences last throughout the summer, in part via the positive feedback between the SMC and rainfall. These results suggest that improving the accuracy of the initial SMC is crucial for improving extended range forecasts for the summer season.