Combined Use of Landsat 8 and Sentinel 2A Imagery for Improved Sugarcane Yield Estimation in Wonji-Shoa, Ethiopia

In this study, a support vector regression (SVR) approach based on radial basis function was used for estimating sugarcane yield in the Wonji-Shoa plantation (Ethiopia) combining Landsat 8 (L8) and sentinel 2A (S2A) data. Vegetation Indices(VIs) involving visible, near-infrared, shortwave infrared bands were calculated from L8 S2A sensor observations, seasonal cumulative values computed period June to October 9th month November 10th of year 2016/17 2018/19 cropping seasons. Sugarcane predicted using SVR, Multilayer perceptron neural network (MLPNN), Multiple linear (MLR) methods. Then, tenfold cross-validation implemented performance evaluation. The results showed significant correlations between VIs during growing season. also revealed that estimation accuracy better combined (RMSE = 12.95 t/ha, MAE 10.14 t/ha) than data alone 14.71 12.18 t/ha). Comparing SVR with MLPNN MLR disclosed outperforms other two models terms prediction accuracy. Overall, study demonstrated successful application developing model it may provide guideline improving estimations area.