Evaluation of the SACZ index as a prognostic tool based on GFS forecasts

Abstract

The South Atlantic Convergence Zone (SACZ) is an atmospheric phenomenon typical of summertime where a band of nebulosity causes intense or persistent rainfall in many regions of Brazil. SACZ episodes can be responsible for many natural disasters. Besides, the impacts of rainfall on water availability and consequently on the energy sector are extensive. The main objective of this study was to investigate the implementation of the SACZ index as an objective forecasting tool using input data from the Global Forecast System (GFS) model. Initially, we compared the index with the SACZ events identified by the Center for Weather Forecasting and Climate Studies (CPTEC ) from 2017 to 2021. Results showed that the index represented all events identified SACZs by CPTEC. Finally, we used data from the GFS 0.25 Degree from 2017 to 2021 to calculate Accuracy, Probability of detection, and False alarm ratio to evaluate the SACZ index as a prediction tool. Three thresholds are defined for the binary classification of a possible SACZ event. Results showed that above the most sensitive threshold (h1), within 10 days in advance, the sign of a possible SACZ can be detected. For the intermediate threshold (h2), a forecast of 96 h can detect a sign. For the most specific threshold (h3), the index can detect the event within 72 h in advance with a probability of detection of almost 90%. The SACZ index proved to be an efficient tool for detecting the dynamics of the phenomenon and can be used to assist operationally and in decision-making.