The ensemble weather forecast approach represents an attempt to extend the predictability of the dynamic atmospheric models. A simple away to produce ensemble weather prediction is to run the weather model starting from a set of slightly perturbed initial conditions. The objective of the present study is to present a methodology to determine the minimum and maximum surface temperatures using the grid history of the ensemble forecasts over Brazil available at the Center for Weather Forecasting and Climate Studies (CPTEC). The results indicated that forecasts of minimum (maximum) temperatures from the CPTEC general circulation model exhibit systematically higher (lower) temperatures than observed. A statistical correction based on the bias of the fifteenth earlier forecasts was used to improve the ensemble predictions. Such corrections enhance the performance of the forecasts as measured by statistical indexes as rms, bias and anomaly correlation coefficient.
Minimum temperature; maximum temperature; ensemble weather forecasting
