This paper is aimed at developing the model of prediction of the sugarcane yield based on the satellite data and mathematical modeling. Remote sensing satellites have been monitoring agricultural crops with regard to growing, harvesting, and other periods. Satellite data have provided accurate information on the earth surface and then easily interpreted which crop is in good health and which is unhealthy, and vegetation indices can give more valuable information for the prediction of crop yield. In this regard, if farmers can estimate the yield before harvesting this is very helpful to the farmers and countries. In this study, ground truth data were collected by farmer's fields and validated with satellite indices and predicted yield. In this model sugarcane crop is first selected for the prediction of yield because 72% of crops consist of sugarcane and the duration of this crop is 8-12 months. During the survey, information was collected on crop yield, the demonstration plots were verified and the observed yield computed. Sentinel-2 data were selected for crop yield forecasting. This crop model used three vegetation indices (Normalized Difference Vegetation Index [NDVI], Enhanced Vegetation Index [EVI], and Green Chlorophyll Vegetation Index [GCVI]), which have been computed from sentinel-2 data using Raster Calculator. To correlate the sugarcane observed crop yield, NDVI, EVI, and GCVI values were computed by linear model. Sugarcane crop yield correlated strongly with the NDVI, EVI, and GSVI (NDVI: R-2 = 0.65, EVI: R-2 = 0.598 and GCVI: R-2 = 0.746). The GCVI index has a high correlation with observed yield using a linear model. Therefore, three linear correlation models have been developed by vegetation indices to determine which indices correlated best with the prediction yield. The observed yield data were compared with normalized vegetation index and other indices. The sugarcane yield is high compared with the observed yield.