Uzaktan algılama (UA) ve coğrafi bilgi sistemi (CBS) teknolojisi kullanılarak mısır bitkisinin veriminin tahmin edilmesi

In the 21st century, precision agriculture is becoming more and more common, increasing crop yield and reducing costs in agriculture. Advanced geographic technology (Geographic Information System-GIS, Remote Sensing-RS technology, Global Navigation Satellite System-GNSS) played an important role in determining input variables of precision agriculture system. In particular, remote sensing technology with higher-end image processing techniques (spatial, spectral, temporal, and radiometric resolution) enhanced the accuracy of the crop yield models. This thesis has two main purposes: the first one presented an overview of the capabilities of remote sensing technology in maize yield forecast from its beginning time up to now through the many specific studies. It also reviewed a variety of vegetation indices (VIs) derived from remote sensing data that had a high correlation with corn yield measured. Furthermore, the literature proved that the remarkable development of remote sensing technology at each different platforms have improved significantly the accuracy of maize yield prediction before harvest. The other goal determined the correlation of the vegetation indices (NDVI, EVI, SAVI, WDRVI, GNDVI) derived from Landsat-8 and Sentinel-2 images and building the maize yield predicting models for the study area based on the Bayesian Model Averaging (BMA) method. This research determined 2 optimal models which were validated with RMSE = 9.38 t/ha for Landsat-8 and RMSE = 12.33 t/ha for Sentinel-2. The QGIS and R software were used throughout the process of spatial analysis and statistical data analysis. In the near future, remote sensing and open-source software will definitely become an absolutely necessary component of precision agriculture.