Eficiência do uso da água na fronteira agrícola do Matopiba via sensoriamento remoto e estatística multivariada.

Abstract

Several environmental changes have occurred over the past few decades due to the robust agricultural expansion undertaken to population growth. Agribusiness is in strong development in the Brazilian cerrado area. These changes affect the carbon coupling in vegetation. As an indication to quantify changes in the carbon-water balance of vegetation, WUE (water use efficiency) is a key gauge for comprehending these changes. This study used orbital sensors and multivariate statistics to evaluate the spatial and temporal behavior of multiscale WUE and its interaction with environmental variables at the MATOPIBA agricultural frontier. Gross Primary Productivity (GPP, MOD17A2H) and Evapotranspiration (ET, MOD16A) products were obtained by Moderate Resolution Imaging Spectroradiometer (MODIS) orbital sensor on the Aqua satellite. To understand the response of environmental influence in the WUE, it used the land use data (MAPBIOMAS), Land Surface Temperature (LST, MOD11A2), and Precipitation data from the Climate Hazard Group InfraRed Precipitation product with Station data (CHIRPS). Pearson's Correlation Analysis and Analysis of Major Components (PCA) were used to evaluate the degree of connection and interaction between the environmental factors and WUE. The WUE obtained the highest negative correlation with precipitation (-0.45) and the highest positive correlation with LST (0.60), showing that the WUE achieves the best performance when LST values are greater, and low precipitation levels. Elevation, LST, Precipitation, and WUE were subsequently classified as medium (0.7> KMO 0.8) based on the results of the KMO test, which indicates the data's quality. These data were then put to PCA, revealing that four principal components were sufficient to explain 65.77% of the variance. Regarding the spatial dimension, MATOPIBA's northwestern section exhibited the highest rainfall values. In contrast, the extreme west of Baiano exhibited the highest LST and WUE values. In addition, the data suggested that the ENOS phases (El Niño and La Niña) possibly affected the WUE to water supply and temperature. The region of extreme western Baiano stood out for receiving the highest WUE scores, verified by highest GPP (580 gC/m2), ET (3000 mm), and WUE (3.5 gC/mm.m2) values in agriculture areas. This result can be attributed to cultural treatment and irrigation installation, which contributed to the higher WUE values. The conclusion is that the WUE is highly sensitive to environmental changes resulting from shifts in climate variables and/or anthropogenic factors induced by agricultural growth in the MATOPIBA.