Crop Growth Monitoring and Yield Prediction System Applying Copernicus Data for Poland & South Africa
Radosław Gurdak, Katarzyna Dąbrowska‐Zielińska, Zbigniew Bochenek, Marcin Kluczek, Maciej Bartold, Solomon W. Newete, George Chirima
Abstract
The study is carried out for agricultural areas in Poland and South Africa for the years 2019–2021. The aim of the project is to develop crop growth monitoring system using Copernicus satellite data (Sentinel-1,-2,-3) and low-resolution data from Terra MODIS satellites. At the first stage of the analysis Sentinel-2 indices, based on various combinations of spectral reflectance, were derived for particular fields within JECAM sites. The correlation coefficients between S-2 indices and LAI most often exceed 0.7 and precision of LAI determination varies from 92 to 81%, depending on year and crop type. NDII and DSWI proved to be most appropriate for LAI assessment, although usefulness of narrowband indices based on red-edge bands, like CLRE, MTCI, NDREI has been also found. Finally the prognosis model based on meteorological data with the periodical input of NDVI and LST for winter wheat yield prognosis has been applied. RMSE error of the model is 5.4 dt/ha.