Análise do índice padronizado de precipitação (SPI) na caracterização das secas do semiárido brasileiro.

Abstract

The high spatial and temporal variability of rainfall in the semiarid Brazilian makes it difficult to forecast and monitoring of drought. In this context, this study aims to subdivide the Brazilian semiarid in sub-regions of precipitation by using the Cluster Analysis and use the Standardized Precipitation Index (SPI) on different time scales (1, 3, 6, 12 and 24 months) to monitor drought in these sub-regions. We used monthly data of rainfall with at least 20 years of record from 270 meteorological stations. The method of K-Means clustering identified four sub-regions of the homogeneous rainfall. IPP-24 showed the highest incidence of moderate and severe droughts in group 1. Already the extreme droughts were present in Bom Jesus-PI and Upanema-RN for the groups 1 and 4 respectively. IPP-12 identified in Group 2 greater amount of gauge rain that had extreme drought. The largest amount of rainfall stations with more cases of severe droughts were in Group 4. IPP-06 showed a higher amount of rainfall stations with cases of moderate drought in group 2. The gauge rain Itaíba-PE belonging to group 3 had the highest number of cases of moderate droughts identified by IPP-03. IPP-01 identified a greater number of cases of severe droughts in the Mucugê-BA city in the group 1, as well as the largest amount of gauge rain that had extreme drought. The study of droughts in the long term considering the IPP-06 and IPP-12 indicated that the gauge rain Curimatá-PI and San Miguel-RN had a longer duration of moderate and severe droughts, as well as IPP-24 indicated a greater moderate drought in the period 1980-87 for Ipupiara-BA.