Using Satellite Data to Estimate Corn Emergence in Kansas.

Broad scale emergence information is critical to evaluate current crop status at different scales and perspectives: farmers, policy makers, and insurance companies. Conventional methods to quantify emergence are labor-intensive, costly, and only summarize information at agriculture district level. Satellite data can assist large scale monitoring. The objective of the project is to implement a procedure to detect early season emerge. Two AGs (Agriculture Districts) were selected due to the different emergence time window. The procedure integrates the Conterminous United States fields layer (Roy and Yan, 2016) and the PROBA-V 5 days 100m TOC products. The temporal and spectral differential response of corn is utilized to identify the target fields. Steps included: 1) Select crop fields from CONUS layer. 2) Pre-season and early season NDVI thresholding to detect early emergence. 3) Mid-season SWIR evaluation to differentiate between delayed corn and others crops. A confusion matrix procedure using ground-truth data is implemented to evaluate the performance of the proposed workflow.