Key points are not available for this paper at this time.
Process-based crop models can predict harvested yield by reproducing the effects of the environment on plant phenology and physiology. Accurate yield forecasts are essential to support strategic and tactical actions in public and private sectors. Applications span from detecting critical areas for food security issues to optimizing selling/buying prices of crop products in main producing regions, to informing farmers on the best agricultural management practices. Most crop models are point-based and must be integrated in a spatially explicit environment to provide the yield information in a target area at the desired spatial resolution. Remote sensing (RS) represents an invaluable resource to inform crop models with actual vegetation dynamics based on consistent and timely views of Earth's surface with time and space continuity. The main advantage of incorporating RS data into crop models is hence the representation of the missing spatial information and the reliable description of the crops health condition throughout the growing season. This study presents, an open-source tool developed within the Google Earth Engine environment to monitor crop growth and estimate crop yield. It is based on a generic model (SIMPLE) executed over large areas at run-time and is easily adapted to different crops by adjusting a few physiological parameters. SIMPLE algorithmic implementation uses ERA5-Land as weather source and derives the leaf area index (LAI, unitless) and the actual crop evapotranspiration (ETc, mm day-1) using data from the MODIS Normalized Difference Vegetation Index (NDVI). Results show that integrating RS data into the SIMPLE model allowed currently identifying the limits of the growing season and mapping seasonal crop phenology evolution in the Piedmont region. Abiotic stresses have been correctly spotted, and aboveground and yield of winter wheat and maize have aligned with reference data. Our findings have significant implications for improving yield estimations by identifying spatial patterns of crop growth productivity for summer and winter crops. This tool also shows potential for near-real-time monitoring of crop growth dynamics in response to abiotic stresses in sensitive phenological phases.
Crecco et al. (Mon,) studied this question.