Remotely-sensed vegetation metrics for real-time crop monitoring and yield estimation

This research examines the utility of remotely-sensed vegetation indices for real-time crop monitoring, yield estimation, and climatic impact diagnosis.;First, a Climate Impact Index is derived from MODIS Leaf Area Index. This index provides a quantitative framework for assessing the importance of the length of the growing-season in determining vulnerability of vegetation to climate variability. This index can identify sensitive regions which are particularly susceptible to agricultural failure arising from month-to-month climate variations.;Next, a time-varying version of this index, the Climate-Variability Impact Index (CVII), is derived to quantify the fractional changes in the overall annual growth gained or lost during a given month at a particular location. The growing-season integrated CVII can provide an estimate of the fractional change in overall growth during a given year. In turn these estimates can provide fine-scale and aggregated information on yield for various crops.;Trained using historical records of crop production, a statistical model based upon the CVII can produce homogeneous production forecasts (in which the model is trained and tested for a particular region), as well as heterogeneous forecasts (in which the model is trained in a particular region and applied to a different region). By examining the model prediction as a function of time, it is possible to determine when the in-season predictive capability plateaus and which months provide the greatest predictive capacity.;Case studies for Illinois and North and South Dakota demonstrate that the model is capable of quantitatively predicting changes in production before harvesting and pinpointing regions where agricultural failure is greatest. More importantly, the case studies highlight the need for explicit monitoring of vegetation growth when estimating yield because drought-monitoring indices such as the Standardized Precipitation Index can both overestimate and underestimate changes in vegetation in drought-stricken regions.;Overall, this research has successfully shown that when properly formulated, new remotely-sensed vegetation indices can be used to (1) characterize and monitor the impact of climate variability on vegetation activity; (2) perform near real-time drought monitoring and famine prediction at regional and global scale; and (3) provide earlier yield forecasts for crops in the middle of the growing season.