This paper estimates an econometric model of the relationship between climate and Brazilian agricultural activity to simulate the impacts of climate change on land use and on agricultural profitability. Elsewhere, researchers are studying how human behavior is likely to influence land cover change and how various future land use scenarios will affect regional climate. This paper describes the feedback of climate change on human behavior, directly, by predicting how changes in temperature and precipitation will affect the share of land in various use categories, and indirectly, by simulating the effects on net agricultural profits. To do this, we employ município-level hedonic cross-sectional models that exploit the spatial co-variation across Brazil between climate and agricultural profits, and between climate and land use patterns, allowing for a rich complement of control variables. The database integrates socio-economic data from the agricultural and demographic censuses with geo-referenced edaphic and geographic data produced by Embrapa and IBGE, and temperature and precipitation data from the Climate Research Unit (CRU). Simulations of spatially differentiated climate scenarios are based upon the HadCM3 and four other general circulation models (GCMs). Findings show that predicted land use and agricultural profitability are quite sensitive to the climate scenario simulated, in particular in the Amazon Basin.
Science Theme: LC (Land Use and Land Cover Change)