Publicações

On December 10th through 11th 2012, a strong convection activity over southern Brazil grew into a huge Mesoscale Convective System (MCS) accompanied by heavy precipitation, hailstorm, damaging winds and dangerous lightning. The Civil Defense Authority reported that over 700,000 people experienced power outages due to either strong winds or lightning strikes or even suffered from flash floods caused by high rain volumes. The Brazilian Lightning Location System (BrasilDAT) detected more than 90,000 cloud-to-ground (CG) lightning strikes during the MCS activity, and approximately 4,000 CG strokes presented large peak currents over 75 kA. The main objectives of this work were to achieve the most suitable configuration of the Weather Research and Forecasting (WRF) model, and to apply the Potential Lightning Region (PLR) tool, in order to forecast reliably this type of severe weather event with enough time in advance to adopt strategies that might minimize injuries and hazardous situations in the future. Firstly, the CG and IC lightning stroke rates, the peak currents distribution, and the IC/CG ratio were evaluated. Observed precipitation from a homogeneous network of surface meteorological stations was used as proxy data. Lightning and precipitation were compared during the MCS development in order to seek for correlations. As well known, current operational models cannot predict convective subgrid scale processes explicitly, due to their microscopic and discontinuous nature, and must do so via parameterization. In this case, a cumulus scheme should try to transport heat vertically, redistribute moisture, and reduce thermodynamic instability. Four cumulus parameterization schemes (CPS) in the WRFv3.3.1 model were investigated to proper simulate this MCS over southern Brazil. WRF grid points near meteorological stations with high accumulated precipitation values were chosen in order to assess the behavior of the simulated data against the observations. The Potential Lightning Region (PLR) is a tool that indicates the spatial distribution of lightning occurrence probabilities in a given area. PLR was developed from a combination of WRF output variables, and operationally tested over southeastern Brazil. The WRF convective parameterization selected from the previous analysis was used to generate new PLR maps over southern Brazil for the MCS. This case study served as a laboratory to expand PLR capabilities in the prediction of severe weather events.