Determination of Crop Coefficient of Sugar Beet by Proximal Sensing Method Using Digital Images

This study aimed to determine the crop coefficient of sugar beet using canopy cover extracted from digital images under different irrigation managements. The crop coefficient and canopy cover were directly measured by water balance and image processing methods, respectively, in 10 days intervals during the growing season. The crop coefficient of sugar beet in three irrigation managements with maximum allowable depletion (MAD) of 40%, 60%, and 80%, was estimated using its regression equation with canopy cover. This was modeled for potential conditions and then validated by using the average measurements in two years. The findings showed that the estimated crop coefficients were in good agreement with the observations in irrigation managements that had MAD of 40% and 60%. The coefficient of determination (R2), normalized Root Mean Square Error (nRMSE), and model efficiency (EF) were 0.95, 0.11 and 0.95, for 40% MAD, 0.9, 0.13 and 0.85 for 60% MAD, respectively. The results illustrate that the crop coefficient of sugar beet, within the moisture range between field capacity to a MAD of 60%, can be reliably estimated by this approach. The values of determination coefficient (R2), normalized Root Mean Square Error (nRMSE) and model efficiency (EF) decreased to 0.49, 0.37 and 0.63, respectively, for 80% MAD, indicating poor performance of the model under severe drought stress conditions. The proposed method has some advantages including easy and fast data collection, greater accuracy and lower cost, the ability to provide the desired number of images, and no need for meteorological data. Therefore, this can be applied to study the plant growth and crop coefficient variations during the growth period.