GENERALIZATION OF RAINFALL MODELS UNDER DIFFERENT CLIMATIC REGIMES IN NORTHEASTERN BRAZIL

Resumo

Climate change has caused significant alterations in rainfall variability, increasing the occurrence of extreme events, both dry and with intense rainfall, which directly impacts water availability and agricultural production. In the municipality of Garanhuns, located in the Agreste region of Pernambuco, evidence of changes in rainfall patterns reinforces the need for approaches that allow for the individual analysis of rainfall regimes. In this context, the present research aimed to analyze the generalizability of rainfall models under different climatic conditions in Garanhuns-PE. For this purpose, information referring to the period from 1993 to 2022 was used, including rainfall (mm) and meteorological variables such as air temperature (°C), atmospheric pressure (hPa), relative humidity (%), and wind speed (m/s), selected based on their influence on each rainfall regime. The years were classified according to quantiles as Very Dry, Dry, Normal, Rainy, and Very Rainy. The results indicated good performance in all adjusted models, with the best fit in extreme precipitation regimes, particularly the very rainy regime, which showed R¹ (0,90)=99,98%. Furthermore, the adjustments were subjected to different climatic regimes in order to analyze the generalization capacity; however, for all situations, the adjusted model retains a high explanatory capacity only for the rainfall regime corresponding to its individual quantile.