Remote sensing applications for crop type mapping and crop yield prediction for digital agriculture

This thesis addresses important topics in agricultural modelling research. Chapter 1 describes the importance of land productivity and the pressure on the agricultural sector to provide food. In chapter 2, a summer crop type mapping model has been developed to map major cotton fields in-season in the Murray Darling Basin (MDB) in Australia. In chapter 3, a robust crop classification model has been designed to classify two major crops (cereals and canola) in the MDB in Australia. chapter 4 focused on exploring changes in prediction quality with changes in the spatial resolution of predictors and the predictions. More specifically, this study investigated whether inputs should be resampled prior to modelling, or the modelling implemented first with the aggregation of predictions happening as a final step. In chapter 5, a new vegetation index is proposed that exploits the three red-edge bands provided by the Sentinel-2 satellite to capture changes in the transition region between the photosynthetically affected region (red region) and the Near-Infrared region (NIR region) affected by cell structure and leaf layers. Chapter 6 was conducted to test the potential of integration of two mechanistic-type model products (biomass and soil moisture) in the DDMs models. Chapter 7 was dedicated to discussing each technique used in this thesis and the outcomes of each technique, and the relationships between these outcomes. This thesis addressed the topics and questioned asked at the beginning of this research and the outcomes are listed in each chapter