Climate-Driven Crop Yield Forecasting: A Review of Machine Learning, Deep Learning, and Remote Sensing Approaches

Effective crop yield prediction is also crucial in food security, agricultural management, and reducing the effects of climate variability. Over the recent years, data analytics on climate, coupled with artificial intelligence tools, have been enhanced to a great extent prediction abilities of crop yield models. The paper is a review of the position of crop yield prediction methods that utilize climate variables, deep learning, machine learning, and remote sensing methods. The review conducts a systematic study of both classical statistical approaches and current regression-based machine learning models, as well as classification models, such as neural networks, gradient boosting, support vector machines, and random forests. More so, long short-term memory networks and convolutional neural networks are covered under the deep learning algorithm. The research paper identifies the frequently employed climate variables of temperature, rainfall, soil moisture, humidity, and the multi-source data available. The most important issues, such as data heterogeneity, time uncertainty, climate extremes, and generalization of models across regions, are highly discussed. Lastly, the paper also provides future research directions with a focus on multi-modal data fusion, explainable AI, and scalable forecasting frameworks in order to make them robust and relevant in the real world.