Avaliação da relação seca/produtividade agrícola considerando cenários de mudanças climáticas

• Autor(a) principal: Santos, Roziane Sobreira dos
• Data de Publicação: 2008
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/5223
• topic: Mudanças climáticas; Índice de seca; Produtividade agrícola; Climate changes; Drought index; Agricultural productivity; CNPQ::CIENCIAS AGRARIAS::AGRONOMIA::AGROMETEOROLOGIA

This study investigates the relationship between agricultural yield and drought in the state of Minas Gerais. Climate changes have warned to possible increases in meteorological extremes worldwide, as well as have uncertainties on how the climatic events might change the environment and affect world agricultural production in the future. Meteorological data from 1973 to 2004 and data projected by the ECHAM5/MPI-OM model for 2008 to 2020 based upon the scenario A1B. During the period from 1973 to 2004, corn productivity data obtained from the IBGE was used and future productivity estimates were performed using the Agro-ecological Zone (AEZ) methodology. The Palmer drought severity index (PDSI) and the Z index were used in a regression model which considered productivity as a dependant variable and the indices as independent variables. The performance of these models was verified by the following statistical coefficients: determination coefficient (r2), root mean squared error (RMSE), mean absolute error (MAE) and the Willmott index of agreement (d). The results demonstrated variation in the observed productivity (IBGE data) were poorly explained by the PDSI and Z index, as all model predictions for the different regions presented agreement values less than 0.5 with the observed values as well as extremely low determination coefficient values. Scenario A1B illustrated Willmot agreement values ranging from 0.48 to 0.90 as well as slightly more significant r2 values. However, productivity estimated by the AEZ presented significant loss for the harvests of 2008/2009, 2009/2010, 2014/2015 and 2018/2019 due to the lack of water in the Triângulo/Alto Paranaíba, Central Mineira and Jequitinhonha regions. The drought index predictions show that when working alone, these indices have not proven to be efficient to evaluate variations in productivity. Further, since the AEZ methodology only considers water limitations and one of the great uncertainties in climate models is precipitation, these results do not spare the need for further studies with more complete productivity simulation models, evaluating with different emission scenarios and more adequate agricultural drought indices.