On the Performance of the Eta Model Microphysics Parameterization

Abstract:

The needing of accurate prediction of heavy rain is highlighted by the occurrence of floods and other disasters, at this point, numerical modeling of the atmosphere is essential. At most of situations, predictions with high resolution are needed, like in case of rainfall in small basins, since variations due to topographic details become more important as grid spacing decreases. For small grid spacing, it is usual to dispense with convective parameterizations. In that case, the microphysics parameterization becomes responsible for the total generation of rain and comes to be more important in models. This research is based on high-resolution simulations of heavy rain events with the Eta model. Object-based evaluations showed that Eta model forecasts with 1 km of grid spacing have a trend to overestimate the rain. For the purpose of detecting error sources, an assessment of the behavior of Ferrier's parameterization was performed, by analyzing the contribution of each process to the tendency of the variables that are influenced by microphysical phenomena. At this stage, the role of ice-phase processes was especially evaluated. Based on the results of that step, changes in the established ratio between cloud ice and snow were tested and evaluated using an object-based method. It was possible to obtain appreciable modifications in forecasted precipitation accumulations.