Monitoring Crop Status in the Continental United States Using the SMAP Level 4 Carbon Product

Abstract:

Accurate monitoring of crop condition is critical to detect anomalies that may threaten the economic viability of agriculture and to understand how crops respond to climatic variability. Retrievals of soil moisture and vegetation information from satellite-based remote sensing products offer an opportunity for continuous and affordable crop condition monitoring. This study compared weekly anomalies in accumulated gross primary production (GPP) from the SMAP Level-4 Carbon (L4C) product to anomalies calculated from a state-level weekly crop condition index (CCI) and also to crop yield anomalies calculated from county level yield data reported at the end of the season. We focused on barley, spring wheat, corn, and soybeans cultivated in the continental United States from 2000 to 2018. We found that significant (alpha <0.1) correlations between SMAP L4C GPP anomalies and both crop condition and yield anomalies were lower (r: 0.4 to 0.6) at the beginning of the season, but increased as crops developed and matured, and that the agreement was better in drier regions (r: 0.6 to 0.8). The L4C provides weekly GPP estimates, permitting the evaluation and tracking of anomalies in crop status at higher spatial detail than metrics based on the state level CCI or county level crop yields. In some areas, the correlations between GPP and county level yield were significant as soon as the crop had emerged, indicating the potential of the L4C to provide a prediction of yield anomalies at the start of the season. The potential of the L4C to predict yield anomalies increases as the season progresses, as correlations were significant weeks or months prior to harvest. We demonstrate that the L4C GPP product can be used operationally to monitor crop condition with the potential to become an important tool to inform decision making and research.