Estimating rice yield using VSM model and satellite images in Guilan province

Estimating crop yield before harvesting is necessary for any agricultural management scheme. Accurate prediction of crop yield, indeed, reveals the impact of different agricultural inputs on the final yield. One of the new methods for such prediction is the use of so-called growth models. In spite of considerable progresses made for developing these models, their calibration and validation in large scales needs plenty of direct field inspections and sampling. For this reason, the use of satellite images and remote sensing tool is recently considered by different investigators. The objective of this study was to evaluate the possibility of using VSM model and satellite images for prediction of rice yield in Guilan province. Consequently, the MODIS and Landsat-8 satellite images were first obtained. A number of 20 rice farms were selected and the required samples and information including the grain yield were collected from these fields. By using the information of sampling points, the data were then extended to the entire study area by means of interpolation, using ArcGIS10.2 software. The grain rice yield was estimated using the satellite images and VSM model. To evaluate the performance of model, the actual crop yield was compared with those estimated by VSM model that has linear relation with high correlation (>80%) and coefficient of determination of 65%. The obtained results indicated a reasonable prediction of rice yield by VSM model. Furthermore, a strong (>80%) and significant relationship was obtained between the incorporated input parameters and the actual crop yield and satellite imagery has the suitable potential to estimate rice yield as input of VSM model. Considering the obtained results, it is possible to predict rice yield with this model and by calibratinge the satellite data in the studied area the yield for next years can be easily estimated. Consequently, there is no need for large field inspections as well as for spending the related extra costs.