Gezira Sulama Projesi'ndeki (Sudan) tarimsal ürünlerin izlenmesinde uzaktan algilama kullanimi

This thesis addresses agricultural management and food security challenges in Sudan's Gezira Irrigation Scheme (880,000 hectares). It uses advanced machine learning to enhance agricultural monitoring, focusing on wheat production and water productivity optimization, the study integrates multiple data sources, employing Support Vector Machine (SVM) and Object-Based Image Analysis (OBIA) for crop classification using Sentinel-2 imagery. Vari-ous models, including Random Forest and XGBoost, estimate yield and water productivity, Results show high accuracy in crop classification, with SVM slightly outperforming OBIA. Crop area estimation achieved a 2-3% error range compared to official records. The research reveals complex wheat cultivation dynamics, highlighting non-linear yield factors and sim-pler water productivity relationships, this work contributes to improved food security, farmer livelihoods, and sustainable water use, aligning with UN Sustainable Development Goals. The methodology has potential applications in similar global irrigation schemes.