This project studies the potential feedback between soil moisture and precipitation. A positive feedback would mean that enhanced soil moisture leads to enhanced precipitation, as a wetter surface provides a source of evaporation for subsequent rainfall. In that case, precipitation would lead to wetter soil, which would in turn lead to further precipitation, while a lack of precipitation could lead to a lack of soil moisture and hence further reductions in precipitation leading to prolonged drought. This study assesses the influence of atmospheric and land surface heterogeneity and spatial scale on soil moisture-precipitation coupling, principally at daily timescales, through two interconnected approaches: 1) diagnostic analysis of available observations, reanalysis products, and global climate model (GCM) output; and 2) conceptual modeling of land-atmosphere interactions using process-based idealized models expresed as analytic equations as well as atmospheric column models coupled to simple soil models (eg "bucket" models). A primary objective of this activity is the development of robust, physically-based metrics for quantifying and comparing intraday soil moisture/precipitation coupling in models and observations.
The issues addressed in this project have relevance to society as well as science, as drought duration and intensity have substantial consequences for agriculture. In addition, the project will support and train a postdoctoral researcher, thereby helping to support and train the next generation of scientists.
in this work we have investigated the feedback between soil mositure, cloudiness and precipitation. We have used observations (reanalysis), general circulation models and idealized models of differnet scales. The results of our research led to new insights on the pathway between soil moisture and precipitation. In particular we have shown that soil msoiture can only control the frequency of precipitation but not the intensity of the precipitation. We have also evaluated the tight coupling between preciptiation, mositure convergence and cloud radiation over the Amazon. This has emphasized that in humid regions the feedbacks are essentially nonlocal. We have also developed the first single column model (1D version of a cliamte model) that can correctly reproduce the diurnal cycle of precipitation over land with a peak in precipitation observed in the afternoon. We belive that this is a major breakthrough and we expect the model to be developed into the NASA GISS model. In addition the model is able to represent some important cumulus clouds (forced clouds) that were not present earlier in previous models and that may have an important contribution to the surface energy budget and the planetary albedo. Our results have been published in numerous journals: Nature Geosciences, Journal of the atmopsheric sciences, Journal of hydrometeorology, Journal of climate...
