Evaluation of the CPTEC/AGCM wind forecasts during the hurricane Catarina occurrence

• Main Authors: A. F. Santos, A. M. Mendonça, J. P. Bonatti, J. G. Z. de Mattos, P. Y. Kubota, S. R. Freitas, M. A. F. Silva Dias, E. Ramirez, R. Camayo
• Format: Article in Journal/Newspaper
• Language: English
• Published: Copernicus Publications 2008
• Subjects: Science; Q; Geology; QE1-996.5; Dynamic and structural geology; QE500-639.5; Eta; South Atlantic Ocean
• Online Access: https://doaj.org/article/6bdde569618343c097ef5ddbc4f83e49

In March 2004 occurred the first hurricane registered at South Atlantic Ocean. The system named Catarina begun as an extratropical cyclone and remained quasi-stationary some days over the South Atlantic Ocean. The system displaced westward, acquiring characteristics of a hurricane and hit the Brazilian State of Santa Catarina (SC) between the 27 and the 28 March, causing destruction and deaths. The objective of this paper is to evaluate the Center for Weather Prediction and Climate Studies, Atmospheric Global Circulation Model (CPTEC/AGCM) forecast performance of some synoptic patterns associated with Catarina. The surface wind and reduced Sea Level Pressure (SLP) were examined. Moreover, the implementation of 10-m wind forecast (V10m) was evaluated. This variable was not available in the CPTEC/AGCM during the Catarina occurrence and in this study it was compared with the wind at first sigma-level of the AGCM. The CPTEC-Eta reanalyses were used to comparisons. According to reanalyses, more intense winds were observed in northeast, south and southwest edges of the cyclone. The system was not predicted by the CPTEC/AGCM forecasts longer than 24 h, then the analyses were carried out only for 24 h forecasts. In general, the first sigma-level wind forecasts underestimated the wind magnitude and the cyclone intensity. However, the Catarina formation and its displacement southeastward between the 20 and the 21 March were well represented by the model. The CPTEC/AGCM presents deficiencies to predict the system intensity, but in short-range forecasts it was possible to predict the system formation and its atypical trajectory. The wind results from the new implementation did not exhibit better performance compared with the wind at first sigma-level. These results will be better investigated in the future.