Abstract
Remote sensing is one of the most important tools for determining water stress and related plant traits. In this study, it is aimed to estimate the corn yield depending on water stress by using the reflectance measurements performed at the leaf level via the hand-held spectroradiometer. For this purpose, in Canakkale, a field experiment consisting of six irrigation treatments (100%, 80%, 60%, 40%, 20% and 0%) was conducted in 2007 and 2008. The reflectance values were measured before each irrigation event and twelve different spectral indices were calculated from these measurements. Univariate and multivariate regression analyzes were performed between the cob yield obtained from the harvest and the spectral indices for each growth period of corn. As a result of the study, statistically significant models were obtained between the yield and spectral indices. In the univariate regression analysis, these relationships were further strengthened after the vegetative period. According to multivariate linear regression analysis; the highest coefficients of determination (R2) were 0.945, 0.905, 0.938 respectively according to vegetative, flowering and grain filling-harvest periods in 2007. They were found to be 0.703, 0.946, 0.914 respectively in 2008. Consequently, it can be said that the cob yield of sweet corn can be determined with high accuracy by using the spectral indices calculated at the leaf level with remote sensing.