Prediction of potato high-yield zones of a field: Bivariate frequency ratio technique

Abstract

Bivariate frequency ratio (BFR) technique was employed to determine high-yield zones in a 30 ha potato (Solanum tuberosum L.) field located in Wadi-Ad-Dawasir, Saudi Arabia. BFR was performed by inputting selected yield tendency factors (YTFs) and potato actual yield (YA). The YTFs were NDVIderived from Sentinel-2 images, soil electrical conductivity, nitrogen, pH and texture. The obtained yield tendency map (YP) was assessed against YA using the area under the curve metric. Although low accuracy (41-58%) was observed with the individual YTFs, high-yield zones were determined with an accuracy of 90% using the cumulative response of YTFs.