Prediksi Cuaca Berdasarkan Variasi miliVolt Xylem Lannea coromandelica Menggunakan Model Artificial Neural Network Backpropagation

The rate of fluid flow in tree xylem generates an electrical potential difference (mV), which serves as a physiological indicator for monitoring plant conditions and predicting weather. This study aimed to develop a regression model based on Artificial Neural Network Backpropagation (ANN-BP) to estimate weather parameters from mV data of Lannea coromandelica. Electrical potential data were collected continuously for seven days using xylem-mounted sensors and synchronized with actual weather data, including air temperature, relative humidity, and light intensity. ANN-BP models employing three training algorithms (traingdx, traincgb, and traingd) were compared using mean squared error (MSE) as the evaluation metric. The traincgb algorithm achieved the best performance with an MSE of 3.29 × 10??. These findings demonstrate that variations in xylem electrical potential can reliably predict weather conditions in real time, supporting the development of an energy-efficient, biologically based weather monitoring system for precision agriculture and climate change mitigation.