Intelligent Prediction Method for Phenological Period of Cold-Region Rice Based on Meteorological Data and Deep Learning

This study proposes an intelligent prediction method for the phenological period of cold-region rice based on meteorological data and deep learning. The method aims to enhance the accuracy and stability of phenological period prediction while overcoming the cost and scalability limitations of traditional approaches. By collecting daily meteorological data during the rice growth cycle, including maximum temperature, minimum temperature, average daily temperature, daily precipitation, sunshine duration, humidity, and wind speed, a time-series model is constructed. The Seq2Seq architecture, integrating Bi-LSTM and CNN, is employed to predict key phenological stages of rice. Additionally, a Bayesian neural network is utilized for uncertainty estimation, providing confidence intervals for the predictions. Experimental results demonstrate that the proposed method outperforms traditional approaches in both prediction accuracy and computational efficiency, making it adaptable to precision agricultural management across different years and regions. This work provides technical support for the advancement of intelligent agriculture.