Yield estimation by satellite data

One prominent application of remote sensing is monitoring the surfaces covered by vegetation, the biosphere itself, including both the actual state and any changes on various timescales. Within the scope of this, both the estimation of the growing area of the main agricultural species and the acquisition of basic data for yield estimation has considerable economic importance. Our group has developed a yield estimation method for the main hungarian crops. A new index (GYURRI - General Yield Unification Robust Reference Index) was evolved, which has a very good correlation withe the yield of the crops. The figure below shows the connection between the GYURRI index and the yield of the winter wheat in Hungary in the period of 2003-2014.


Yield estimates are made shortly after the crop has been harvested, whereas forecasts are performed several weeks before the entire crop has been harvested. The figure below shows the forecasted and estimated yields of ten crops (winter wheat, whinter barley, spring barley, maize, sugar beet, sunflower, maize for silage, potatoes, alfalfa and peas) vs. the official yields in Hungary between 1996 and 2000.



In our most recent study, we tested our robust yield estimation procedure over a 21-year (2000-2020) MODIS data set. In the process, we examined the 4 plants with the largest sown area in Hungary (wheat, maize, sunflower, rapeseed) at 9 cloud cover thresholds, applying 5 different curve fits to the measured points. 16 vegetation indices were tested in the studies. The obtained results were corrected with meteorological (temperature, precipitation, radiation) and soil moisture data averaged over different periods.
Clivk here for details.