Impact of soil moisture on the dominant modes of North American temperature variability

Abstract

North American temperature variability and its link with soil moisture is examined in observational/reanalysis datasets and in carefully designed model experiments. The model experiments consist of a control simulation, where soil moisture is allowed to interact (coupled) with the atmosphere at each model time step, and an uncoupled simulation, where soil moisture is prescribed with climatological annual cycle values. Empirical Orthogonal Function analysis reveals that the first three dominant modes of temperature variability, which span synoptic to inter-annual time scales, are very similar in both the observations and the coupled simulation, suggesting that the coupled model successfully reproduces the dominant spatial and temporal patterns of temperature variability. Comparison of the coupled and uncoupled simulations indicates that damping soil moisture variability leads to a clear and robust reduction in temperature variability associated with one of these leading modes, and modest changes to the other two. To explore the underlying feedback processes, the principal component time series of this mode is correlated with the circulation, radiation and turbulent fluxes. This analysis suggests that soil moisture modulates the turbulent fluxes and regional circulation patterns, which in turn modifies the net shortwave radiation pattern to affect the spatio-temporal variability of temperature. Considering the long memory of soil moisture anomalies and the above role it played in modulating one of the leading modes of North Americas temperature variability, the results presented here can have implications for improving sub-seasonal to seasonal temperature forecast skill over North America.