An investigation of land/atmosphere interactions: Soil moisture, heat fluxes, and atmospheric convection

The theme of this study is land/atmosphere interaction: the roles of soil moisture, surface heat fluxes, and atmospheric convection in the continental hydrologic cycle at a hierarchy of spatial scales. Three projects involving both observations and modeling were completed. Initial results from a fourth project that combines ideas explored in the completed projects are also described.; The first project quantified the contribution to regional and global tropical rainfall by convective system type, size, and intensity. A global database of tropical cloud clusters was constructed from microwave (85 GHz) satellite imagery. A small population of cloud clusters defined as mesoscale convective systems (MCS) both by size and intensity were found to contribute the majority of tropical rainfall. Continental regions such as Sub-Saharan Africa had the largest contribution from large, intense MCSs than oceanic regions.; The second and third projects involved a case study from the Southern Great Plains 97 Hydrology Experiment in southwestern Oklahoma during 9–16 July 1997. In the second project the effect of a simple initialization scheme was tested using an untuned land surface model. The most accurate simulation of soil moisture and heat fluxes occurred during the 48 hours following heavy rainfall. Over the course of a long dry down, the predicted evapotranspiration and thus the partitioning of surface energy was a function of moisture availability in the subsurface soil where roots were located. The quality of the simulation depended on the accuracy of the rainfall data, soil texture, vegetation density, and initial soil moisture in the sub-surface layers.; The third project considered the effect of small-scale (<1 km) variability of soil moisture on the mesoscale energy budget. Simulations incorporating small-scale variability had significantly more realistic latent heating patterns over the mesoscale than simulations with homogenous soil moisture. Adding information about the landscape either explicitly or statistically improved the area-average results for soil moisture and heat fluxes. Research on the sensitivity of mesoscale convective system development to land cover variability is ongoing. Initial results indicate that vegetation density can affect storm-average and storm-total hydrologic properties such as total rainfall.