While a variety of model experiments and analyses of observations haveexplored the impact of soil moisture variation on climate, it is not yetclear how large or detectable soil moisture feedback is across spatial andtemporal scales. Here, we study the impact of dynamic versus climatologicalsoil moisture in the GISS GCM ModelE (with prescribed sea-surfacetemperatures) on the variance and on the spatial and temporal correlationscale of hydrologically relevant climate variables (evaporation,precipitation, temperature, cloud cover) over the land surface. We alsoconfirm that synoptic variations in soil moisture have a substantial impacton the mean climate state, because of the nonlinearity of the dependence ofevapotranspiration on soil moisture.We find that including dynamic soil moisture increases the interannualvariability of seasonal (summer and fall) and annual temperature,precipitation, and cloudiness. Dynamic soil moisture tends to decrease thecorrelation length scale of seasonal (warm-season) to annual land temperaturefluctuations and increase that of precipitation. Dynamic soil moistureincreases the persistence of temperature anomalies from spring to summer andfrom summer to fall, and makes the correlationbetween land precipitation and temperature fluctuations substantially more negative.Global observation sets that allow determination of the spacetime correlation of variables such as temperature, precipitation, and cloud cover could provide empirical measures of the strength of soil moisture feedback, given that the feedback strength varies widely among models.
展开▼
