Comparative Evaluation of Artificial Neural Network and Adaptive Neuro‑Fuzzy Inference System Models for Monthly Rainfall Prediction: A Case Study in Benghazi, Libya

Abstract

This study investigates the application of machine learning techniques, namely Artificial Neural Networks (ANNs) and Adaptive Neuro-Fuzzy Inference Systems (ANFISs), to monthly rainfall prediction in Benghazi (Benina region), Libya. Due to increasing rainfall variability under climate change, accurate forecasting is crucial for water resource management, infrastructure planning, and disaster mitigation in semi-arid regions. Historical meteorological data from 1970 to 2020, including rainfall, temperature, relative humidity, and wind speed, were obtained from the Benina Meteorological Station. Model performance was evaluated using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and the coefficient of determination (R²). The results demonstrate that both models capture rainfall patterns; however, the ANN exhibited superior prediction accuracy and generalization performance compared to ANFIS. Model accuracy was constrained by several factors, including the use of monthly data, missing records, and the exclusion of key climate variables such as sea surface temperature, cloud cover, and atmospheric pressure. The findings highlight the potential of machine learning in rainfall prediction while emphasizing the need for high-resolution, multi-variable datasets to improve reliability and support sustainable water and climate adaptation strategies in Benghazi and similar arid regions.