Skip to content
Conference

Deep Learning Models for Multivariate Rainfall Forecasting Using Meteorological Time Series

Sep 2026 · International Conference on Communications and Information Technology · pp. 1-6 · 0 citations · 33 references

Abstract

Accurate rainfall forecasting is essential for disaster preparedness, water resource management, agricultural planning, and sustainable regional development, particularly in regions vulnerable to hydrometeorological hazards such as floods and extreme weather events. However, rainfall prediction remains a complex challenge due to the nonlinear, dynamic, and multivariate interactions of atmospheric processes. This study aims to evaluate the effectiveness of deep learning models for multivariate rainfall forecasting using meteorological time-series data. The research investigates three state-of-the-art architectures: Long Short-Term Memory, Bidirectional, LSTM, BILSTM, and Transformer, integrated within a multivariate framework that incorporates historical rainfall, temperature, air pressure, wind speed, wind direction, relative humidity, and solar radiation. The dataset comprises 10-minute interval observations recorded from 2023 to 2025 at the Jatiwangi Meteorological Station in Majalengka, Indonesia. Hyperparameter optimization was performed using the Hyperband algorithm to obtain the optimal model configurations, and predictive accuracy was assessed using RMSE and MAE. The experimental results demonstrate that the BILSTM model provides the most accurate predictions, achieving the lowest RMSE and MAE values among the evaluated models. These findings indicate that bidirectional temporal learning improves the model’s ability to capture sequential dependencies and complex atmospheric patterns. Overall, the study highlights the potential of deep learning–based multivariate forecasting approaches to enhance operational rainfall prediction and supports their integration into data-driven meteorological decision-making systems.

View source

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.