Estimation of corn coefficients with vegetation indices using multispectral camera and drone

Abstract

Optimum irrigation scheduling and new technologies are the key to the successful practice of modern agriculture and natural resources, such as water management. A three-year research project was conducted at Velestino, Magnesia, Greece. The aim was to study whether vegetation indices can be used to estimate the crop coefficients of corn in order to apply an intelligent method of irrigation using drones in the future. The normalised difference vegetation index (NDVI), the soil-adjusted vegetation index (SAVI), the renormalised difference vegetation index (RDVI) and a new index [difference infrared - green vegetation index (DIGVI)] were calculated using multispectral photos from a camera adapted to a drone. Three different methods were applied to calculate the crop coefficients: (i) the water balance and the FAO Penman-Monteith reference evapotranspiration, (ii) the climatic data, (iii) the vegetation indices. The irrigation dose covered 100% of the crop water needs according to the soil moisture measurements and the single crop coefficient values. The statistical analysis and the simple linear regression method showed that the corn crop coefficients can be estimated when these indices are used as independent variables.