Landsat and Sentinel-2 images as a tool for the effective estimation of winter and spring cultivar growth and yield prediction in the Czech Republic

The influence of climate and topography on crop condition and yield estimates is most effectively monitored by non-invasive satellite imagery. This paper evaluates the efficiency of free-access Sentinel 2 and Landsat 5, 7 and 8 satellite images scanned by different sensors on wheat growth and yield prediction. Five winter and spring wheat cultivars were grown between 2005 and 2017 in a relatively small 11.5 ha field with a 6% slope. The normalized difference vegetation index was derived from the satellite images acquired for later growth phases of the wheat crops (Biologische Bundesanstalt, Bundessorenamt and Chemical industry 55 - 70) and then compared with the topography wetness index, crop yields and yield frequency maps. The results showed a better correlation of data obtained over one day (R-2 = 0.876) than data with a one-day delay (R-2 = 0.689) using the Sentinel 2 B8 band instead of the B8A band for the near-infrared part of electromagnetic spectrum in the normalized difference vegetation index calculation.