Impact of soil moisture feedback and vegetation feedback on seasonal prediction of precipitation over North America

Soil moisture-vegetation-precipitation feedback tends to promote long soil moisture memory in some areas of the globe, which contributes to seasonal precipitation predictability. The impact of initial soil moisture anomalies on subsequent precipitation within North America is investigated by conducting the numerical experiments using the coupled CAM3-CLM3 model. Here the coupled CAM3-CLM3 model has been modified to include the predictive vegetation phenology scheme validated by the MODIS data.

First, soil moisture-precipitation feedback is studied using the coupled model with prescribed vegetation seasonal variation by turning off the predictive phenology scheme (“SM” experiments). Our sensitivity experiments indicate that the future precipitation varies depending on the timing, magnitude and spatial coverage of the soil moisture anomalies. For example, for the same magnitude of soil moisture anomalies, the larger the spatial coverage is, the larger the magnitude of future precipitation anomalies they cause, and the longer the anomalies will last. We then study the role of vegetation feedback on the impact of initial soil moisture on subsequent precipitation using the coupled model with the predictive vegetation phenology scheme (“SM-Veg” experiments). The role of vegetation will be quantified by the simulated climate difference between the “SM-Veg” experiments and “SM” experiments. In particular, contribution from vegetation feedback to the magnitude and the persistence of the initialization-induced precipitation anomalies will be determined.