Search

The present study assesses the medium-range probabilistic forecast issued using an improved version of the Centre for Weather Forecasting and Climate Studies (CPTEC) Ensemble Prediction System (EPS). This version includes improvements made to the method used to select the fastest growing modes (the empirical orthogonal functions (EOF)-based method), and the bias correction. Results demonstrate that this improved CPTEC EPS surpasses the skills of the current operational version. An assessment was conducted using the framework of the THORPEX Interactive Grand Global Ensemble (TIGGE), comparing three configurations of the CPTEC EPS to the EPS from the National Centre for Environment Prediction (NCEP) and the Korean Meteorological Agency (KMA). The score adopted for metrics was the continuous ranked probability score (CRPS), calculated for absolute temperature at the 850 hPa level (T850), and sea level pressure (SLP). The best improvements were found for T850 over the Northern and Southern Hemispheres, with a general increase of 9 days in the useful forecast horizon. SLP augments are more modest, with useful performance extension of 2 days in the Northern Hemisphere and 3 days in the Southern Hemisphere. All the EPSs over the Southern Hemisphere showed a general decrease in performance when compared with their respective counterparts in the Northern Hemisphere, both for T850 and SLP. Over the Tropics, SLP is better predicted than T850 by most EPSs, but the NCEP EPS presents the best performance over this region with similar scores for the two variables. [ABSTRACT FROM AUTHOR]

Empirical conditional probabilities of extreme monthly precipitation averaged over three regions of the southern part of the La Plata basin were calculated as a function of three atmospheric indices: two of them represent low level moisture convergence and the third is related to the mid tropospheric circulation over southern South America. The first region includes southern Paraguay, northeastern Argentina, north of Uruguay and the western sector of southern Brazil, the second region stretches over part of southern Brazil, while the third region, surrounding the Plata Estuary extends over western Uruguay and eastern Argentina between 37 ° S and 32 ° S. Regional monthly precipitation was considered extreme when it exceeded its mean value plus one standard deviation. Conditional empirical probabilities of Months with Extreme Precipitation ( MEP) are more than twice and even four times higher than climatic probabilities when they are calculated under the restriction that the atmospheric indices were beyond certain extreme thresholds empirically defined. Combined conditional probability is even higher when the values of two or the three indices over their respective thresholds take place at the same time, but these cases are less frequent. The index that renders higher conditional probabilities is the one that accounts for the low level moisture convergence in a region that is basically to the west and outside of the three regions. This reflects the fact that most of the mesoconvective systems, which originate to the west of these regions and then move eastwards, produce heavy precipitation that lead to monthly accumulated extremes. Copyright © 2013 Royal Meteorological Society [ABSTRACT FROM AUTHOR]