Soil moisture-vegetation-precipitation feedback at the seasonal time scale over North America
The soil moisture-vegetation-precipitation feedback at the seasonal time scale can play a critical role in the North American seasonal climate. To investigate such feedback mechanisms, we first incorporate into the land model (CLM) the vegetation phenology scheme, which can simulate the seasonal variation of vegetation in response to hydrometorological conditions. The modified land model is validated against the remotely-sensed MODIS LAI observations. Using the coupled land-atmosphere model CAM-CLM that includes the modified land model, a series of ensemble simulations are performed to examine how initial soil moisture anomalies influence the subse¬quent precipitation at the seasonal time scale. The simulations are performed with different initial soil moisture conditions and with different vegetation treatments. When vegetation feedback is excluded, initial wet/dry soil moisture anomalies in the model tend to cause increase/decrease of precipitation in general, leading to a pos¬itive soil moisture-precipitation feedback. Precipitation responses are found to vary with the characteristics of soil moisture anomalies, including their timing, magni¬tude/direction, spatial coverage and vertical depth. For example, during summer the decrease of precipitation due to dry soil moisture anomalies tends to persist longer than the increases of precipitation due to wet soil moisture anomalies. With vegeta- tion feedback included, LAI increases/decreases as a result of the initial wet/dry soil moisture anomalies. This increase/decrease of LAI tends to increase/decrease evapo¬transpiration and decrease/increase albedo, and thus further increase/decrease precip¬itation (i.e., positive vegetation feedback). On the other hand, the increase/decrease of LAI tends to increase/decrease water consumption by vegetation and decrease/increase soil moisture, and thus decrease/increase precipitation (i.e., negative vegetation feed-back). Whether positive feedback or negative feedback is dominant varies with the characteristics of soil moisture anomalies. As a result of initial wet summer soil mois¬ture anomalies, for instance, the increase of LAI leads to reduce soil moisture, but still enhance the precipitation response through vegetation's impact on evapotran¬spiration and albedo. Furthermore, additional ensemble simulations are performed after modifying CAM-CLM to reduce its dry biases. Results from the modified model suggest that modeling studies on the soil moisture-vegetation-precipitation feedback must consider the dependence of results on model climate and model parameteriza¬tion.