2.     Methodology
2.1. Dataset and sources
For this study, we used population, precipitation, and crop production sorghum, maize, and paddy rice. We downloaded population and crop production data from the Food and Agriculture Organization Statistics website (FAOSTAT). FAOSTAT is a platform dedicated to the United Nations’ mission to improve data collection and dissemination for socio-economic development and the fight against global hunger and malnutrition. We obtained precipitation data from the Global Precipitation Climatology Project (GPCP). GPCP data were from rain gauges, satellites, and sounding observations, which have been merged to estimate rainfall from 1979 through 2015. GPCP provides global coverage for climate variability studies and estimates of uncertainty due to random errors, and it is useful for validation of global precipitation in climate models (Pendergrass et al. 2016). We collected data for the years 1980 through 2011, a total of 32 observations. Table 1 summarizes the units and temporal resolutions of the variables in the dataset.