Wednesday, 12 January 2005
"Why are the lower level cumulus clouds often better organized near the east bank of the Tapajos River?" - a mechanistic study
The understanding of local mesoscale atmospheric circulation is important to interpret the spatial and temporal variations in energy, water and co2 fluxes, and to perform the regional carbon integration. We have carried out extensive numerical studies using Colorado State University Regional Atmospheric Modeling System (RAMS) to investigate the mechanisms leading to co2 variability. One phenomenon occurring persistently in our study region is that the lower level cumulus clouds is often better organized near the east side of Tapajos River. This cloud band is frequently observed both by local people and in satellite imagery. Previous research on this feature of regional climate stressed the role of thermally-driven mesoscale circulations associated with the forest-river contrast in sensible heat flux (the "riverbreeze effect"). The persistence of this feature under strong trade-wind conditions suggests that another mechanism may be involved as well. We performed a very high-resolution numerical simulation for this area with 4-level nested grids and 1 km grid spacing for the finest grid. The simulation ran from 1 through 15 August 2001, which was concurrent with our Santarem mesoscale field campaign. The results demonstrated that it was the mechanical forcings, i.e., the topography, the roughness length, the "T" shape juxtaposition of Amazon and Tapajos Rivers, and the resulting horizontal and vertical wind shears, that played the most important roles in producing the lower level convergence zone. The thermal forcing initiated by differential heating between land and water is only secondary. We then carried out two more numerical experiments to test our theory. In the first one we turned off the trade wind, left the surface temperature difference between the river and the land unchanged. We found that the riverbreezes, as well as the lower level convergences, formed on both sides of the Tapajos River. In the second experiment, we turned off the thermal forcing, and left the mechanical forcing untouched. We found that the lower level cumulus cloud still formed near the east bank of the river. In addition, a suite of numerical experiments are also being carried out to look at the relative importance of topography, roughness length, and land cover types, in the generation of this lower level convergence zone, by isolating and combining these forcing factors.
Our paper will present the detailed analysis of the events we simulated and explain the mechanisms that make the east bank of Tapajos River a preferred area for the lower level convergence to form.