| description abstract | The U.S. Great Plains experienced a number of multiyear droughts during the last century, most notably the droughts of the 1930s and 1950s. This study examines the causes of such droughts using ensembles of long-term (1930?2000) simulations carried out with the NASA Seasonal-to-Interannual Prediction Project (NSIPP-1) atmospheric general circulation model (AGCM) forced with observed sea surface temperatures (SSTs). The results show that the model produces long-term (multiyear) variations in precipitation in the Great Plains region (30°?50°N, 95°?105°W) that are similar to those observed. A correlative analysis suggests that the ensemble-mean low-frequency (time scales longer than about 6 yr) rainfall variations in the Great Plains are linked to a pan-Pacific pattern of SST variability that is the leading empirical orthogonal function (EOF) in the low-frequency SST data. The link between the SST and the Great Plains precipitation is confirmed in idealized AGCM simulations, in which the model is forced by the two polarities of the pan-Pacific SST pattern. The idealized simulations further show that it is primarily the tropical part of the SST anomalies that influences the Great Plains. As such, the Great Plains tend to have above-normal precipitation when the tropical Pacific SSTs are above normal, while there is a tendency for drought when the tropical SSTs are cold. The upper-tropospheric response to the pan-Pacific SST EOF shows a global-scale pattern with a strong wave response in the Pacific and a substantial zonally symmetric component in which U.S. Great Plains pluvial (drought) conditions are associated with reduced (enhanced) heights throughout the extratropics. The potential predictability of rainfall in the Great Plains associated with SSTs is rather modest, with about one-third of the total low-frequency rainfall variance being forced by SST anomalies. Further idealized experiments with climatological SST suggest that the remaining low-frequency variance in the Great Plains precipitation is the result of interactions with soil moisture. In particular, simulations with soil moisture feedback show a fivefold increase in the variance in annual Great Plains precipitation compared with simulations in which the soil feedback is excluded. In addition to increasing variance, the interactions with the soil introduce a year-to-year memory in the hydrological cycle. The impact of soil memory is consistent with a red noise process, in which the deep soil is forced by white noise and damped with a time scale of about 1.5 yr. As such, the role of low-frequency SST variability is to introduce a bias to the net forcing on the soil moisture that drives the random process preferentially to either wet or dry conditions. | |
