LSTM Raw Data for Irrigation Forecasting in Mango (Mangifera indica L.), Chimichagua, Colombia

This dataset contains daily time-series of averaged soil matric potential and calculated reference evapotranspiration from a commercial mango (Mangifera indica L.) plantation in Chimichagua, Cesar, Colombia, for the period of December 21, 2022, to April 10, 2023. Although original sensor data were collected at 15-minute intervals, the provided dataset consists of daily values processed for validating predictive models based on LSTM neural networks for efficient irrigation management. The file includes both field-derived observed values and LSTM model predictions, enabling applications in precision agriculture, water resource management, and agricultural sustainability in tropical climates.本数据集收录了哥伦比亚塞萨尔省奇米查瓜一处商业芒果（Mangifera indica L.）种植园在2022年12月21日至2023年4月10日期间的平均土壤基质势与计算得到的参考蒸散量日度时间序列数据。尽管原始传感器数据以15分钟为采集间隔，但本数据集提供的是经处理后的日度数据，用于验证基于长短期记忆（Long Short-Term Memory, LSTM）神经网络的高效灌溉管理预测模型。该文件同时包含野外实测观测值与LSTM模型预测结果，可应用于热带气候下的精准农业、水资源管理及农业可持续发展领域。