Developing a logistic model for estimating the yield of autumn sugar beet in Khozestan province

Predicting crop yield is an important factor for efficient planning and management. In the present study, a simple mathematical model is developed for predicting the yield of autumn-sown sugar beet in Khuzestan region. In this model, the logistic function was used, in which the root yield and white sugar yield is defined as a function of irrigation, rainfall and evaporation from the pan. In order to evaluate the model, data of two-year trials conducted under five treatments of drip irrigation water based on percentage of water requirement (25, 50, 75, 100 and 125%) at Safiabad Agricultural Research and Education and Natural Resources Center, Dezful, were used. Results showed that the model was able to estimate root yield and white sugar yield with good accuracy so that the normalized root mean squares error (NRMSE) for estimating root yield in the calibration and validation stage were 9.4 and 13.0%, respectively. Also, the value of NRMSE for estimating white sugar yield in both calibration and validation years was less than 10% and equal to 6.8 and 9.8%, respectively. The model is able to have good accuracy during the growing season and estimate the values of root yield and sugar yield during different days after planting (DAP) with an error of less than 20%. Considering that the model presented in the current study is an experimental model, it is recommended to first recalibrate the model, and then use it in regions with different climatic conditions.