Are Rainfall Trends Associated with Deforestation in the Southern Amazon and Northern Cerrado?

In the southern Amazon and northern Cerrado, deforestation has the potential to affect rainfall patterns by increasing land surface albedo, decreasing leaf area index, decreasing land surface roughness, and decreasing vegetation rooting depth. These changes can affect evapotranspiration and energy fluxes, resulting in altered rainfall patterns. Understanding changes in rainfall patterns in this region is important not only for sustaining Amazon and Cerrado ecosystems, but also for rainfed agriculture, hydropower generation, and drinking water management. Previous studies suggest complex, scale-dependent interactions between forest cover and rainfall. For example, in simulations, the size and distribution of deforested patches has been found to affect rainfall quantity and spatiotemporal distribution. Here, we take an empirical approach, examining trends in rainfall intensity, frequency, temporal unevenness, and spatial coherence during 1982 - 2014, a period of intensive forest cover loss in the region. We then relate these rainfall metrics to percent forest cover. We use rainfall data from both the Climate Hazards Infrared Precipitation with Stations (CHIRPS) dataset and a gauged precipitation dataset. Forest cover data is from the MapBiomas dataset. All rainfall metrics are calculated for each season of each year and trends are examined using the Mann-Kendall trend test coupled with Sen's slope. Analyses reveal a slight decreasing trend in rainfall spatial coherence during austral winter and a slight increase during austral spring. Results from analyses of trends in rainfall intensity, frequency, and temporal unevenness are forthcoming. The association between trends in all rainfall metrics and deforestation will be examined using linear mixed effects models. Altogether, these analyses will quantify the relationship between deforestation in the Amazon-Cerrado transition zone and spatiotemporal trends in rainfall.