Monitoring Agricultural Drought Using Geographic Information Systems and Remote Sensing on the Primary Corn and Soybean Belt in the United State

The study aims to evaluate various remote sensing drought indices to assess those most fitting for monitoring agricultural drought. The objectives are (1) to assess and study the impact of drought effect on (corn and soybean) crop production by crop mapping information and GIS technology; (2) to use Geographical Weighted Regression (GWR) as a technical approach to evaluate the spatial relationships between precipitation vs. irrigated and non-irrigated corn and soybean yield, using a Nebraska county-level case study; (3) to assess agricultural drought indices derived from remote sensing (NDVI, NMDI, NDWI, and NDII6); (4) to develop an optimal approach for agricultural drought detection based on remote sensing measurements to determine the relationship between US county-level yields versus relatively common variables collected.;Extreme drought creates low corn and soybean production where irrigation systems are not implemented. This results in a lack of moisture in soil leading to dry land and stale crop yields. When precipitation and moisture is found across all states, corn and soybean production flourishes. For Kansas, Nebraska, and South Dakota, irrigation management methods assist in strong crop yields throughout SPI monthly averages. The data gathered on irrigation consisted of using drought indices gathered by the national agricultural statistics service website. For the SPI levels ranging between one-month and nine-months, Kansas and Nebraska performed the best out of all 12-states contained in the Midwestern primary Corn and Soybean Belt. The reasoning behind Kansas and Nebraska's results was due to a more efficient and sustainable irrigation system, where upon South Dakota lacked. South Dakota was leveled by strong correlations throughout all SPI periods for corn only. Kansas showed its strongest correlations for the two-month and three-month averages, for both corn and soybean.;Precipitation regression with irrigated and non-irrigated maize (corn) and soybean levels show yields as a function of precipitation. The GWR models predicted that yields were significantly better than OLS performances for maize (corn) and soybean. The OLS regression model when used showed a general trend of correlation between observed yields and long-term mean precipitation totals, with 84% and 63% of the variability in mean yield explained by the mean annual precipitation for the non-irrigated crops. The GWR technique performance in predicting yields was significantly better than OLS performances. For instance in the months of June, July, and August precipitations had greater impacts on maize (corn) yields than soybeans under non-irrigated conditions as a result of the greater sensitivity maize (corn) had to water stress.;SPI is capable of offering various time-scales enabling it to show initial warning signs of drought conditions and accompanying severity levels. SPI calculation techniques used for various locations are reflected upon the precipitation records acquired during those periods. Over the 3, 6, and 9-month periods, NDII6 performed the best out of all of the MODIS indices as shown in its results in monitoring vegetation moisture and drought detection. NDII6 performed the best due to its detection abilities. The 9-month SPI provides an indication of inter-seasonal precipitation patterns over medium timescale duration.;A new approach used is to average corn and soybean yields for all counties of the study area in comparison with average anomalies of the MODIS indices for the growing season between May through September from 2006-2012. There was a strong correlation between average corn yields versus MODIS NDII6 averages for these years with R2 equaling 0.62. That means NDII6 is the best indicator to show drought conditions and vegetation moisture monitoring. There was a weak correlation with R2 = 0.16 between averages of soybean yields and averages of precipitation. Irrigation and management systems, technological improvements from hybrids, producer management techniques, and other management practices have an impact on crop yield productions. (Abstract shortened by ProQuest.).