The Use of the Normalized Difference Vegetation Index to Analyze the Influence of Vegetation Cover Changes on the Streamflow in the Manhuaçu River Basin, Brazil

Abstract

The intensification of anthropogenic activities in river basins has significantly altered river streamflow. Thus, the aim of this work is to evaluate the influence of temporal and spatial changes in vegetation cover on the streamflow of the Manhuacu River, Brazil using the normalized difference vegetation index (NDVI). The NDVI values were obtained from Landsat images for 1986 to 2014 and processed on the Google Earth Engine (GEE). To characterize the hydrologic behavior of the basin, we used data from eight streamflow gauge stations and 18 rain gauge stations, and we conducted Mann-Kendall and Pettitt tests to evaluate the stationarity hypothesis. To verify the relationship between the vegetation cover changes and the streamflow, we used multiple regression models. In the analyzed period, we observed a slight increase in the vegetation cover and a reduction in pasture areas. In most cases, the annual low flow exhibited a decreasing tendency because of an increase in vegetation area. We obtained satisfactory adjustments of the multiple regression models by examining the correlations among the change in vegetation cover, rainfall and evapotranspiration data, which resulted in adjusted $$ {\mathrm{R}}_{\mathrm{a}}^2 $$ values ranging from 0.59 to 0.96. The NDVI values were a good indicator of vegetation cover dynamics in the Manhuacu River basin, which influenced the nonstationary streamflow behavior.