The southern region of Brazil, especially the region of Campanha, has experienced periods of drought in the last decade, with greater severity and frequency than normal, seriously affecting agricultural production and therefore the economy. Monitoring of the drought-affected area and the drought's quantification are important for the management of this type of natural disaster. This study aimed to analyze the severity of drought using the number of days without rain (NDSC), the rate of normal percentage (IPN) and the standardized precipitation index (SPI), from the town of Bagé/RS/Brazil, representing the Campanha region of the state. The calculation of the SPI was based on monthly rainfall data, using the Gamma distribution. Adjusting this probability function to study data was confirmed by using the Chi-square test. The results indicated that both the NDSC and IPN were able to detect extreme drought and severe events, which contributed to the decrease in soybean yield in Bagé. The years of drought occurrence, classified as severe and extreme, coincide with the years of occurrence of the phenomenon called "La Niña". Out of 1164 months analyzed considering the values of the SPI, 6.8% (79) were classified as severe and extreme drought events.
monthly precipitation; Glycine max; water deficit; Gamma distribution; SPI