Date of Award
Doctor of Philosophy
Timely monitoring and prediction of the trajectory of crop development provides scientific information to agronomists and climate scientists. For example, estimates of when crops reach certain growth stages can help estimate the length of the growing season
or the timing of cessation of crop transpiration. The Soil Moisture and Ocean Salinity (SMOS) and Soil Moisture Active Passive (SMAP) satellites collect data at high spatial and temporal resolution compared to traditional remote sensing devices. In regions with intensive agriculture, such as Iowa, their measurements track the progression of crops through their growth stages.
Using data products from these two satellites, we develop three modeling approaches to describe crop growth signatures over a collection of spatially referenced satellite footprints in Iowa. We first propose a state space model, followed by a nonlinear parametric hierarchical model, and lastly a functional modeling approach. In addition to describing the entire seasonal growth curve, we estimate, and provide uncertainty quantification, for the timing of when corn reaches its milk (R3) growth stage. Lastly, we develop forecasting methods to predict the timing of R3 mid-season when only partial data is available.
Lewis-Beck, Colin, "Modeling crop phenology using remotely sensed data" (2018). Graduate Theses and Dissertations. 17239.