Title: Ensemble learning based flood forecasting models for the northern districts of Bihar

Authors: Vikas Mittal; T.V. Vijay Kumar; Aayush Goel

Addresses: School of Computer and Systems Sciences, Jawaharlal Nehru University, New Delhi, 110067, India ' School of Computer and Systems Sciences, Jawaharlal Nehru University, New Delhi, 110067, India ' Indian Institute of Management Mumbai, Powai, Mumbai, Maharashtra, 400087, India

Abstract: During past few decades, global warming and climate change have engendered certain change in relationships between various environmental parameters. There has been a significant increase in the number of occurrences of various natural hazards, such as floods, which is being observed globally. In recent years, floods have become the most regularly occurring natural hazard in India that has resulted in continuous and significant loss to lives and property. Continuously changing weather patterns make forecasting of such hazards increasingly difficult. In order to capture the changing dynamics of key weather parameters to improve forecasting, many flood forecasting models that use machine learning techniques, have been proposed in the literature. In this paper, flood forecasting models, using ensemble learning techniques, have been proposed that seek to enhance the flood forecasting capability of the existing machine learning based flood forecasting models. Experimental results have shown that these proposed models have performed better than the existing flood forecasting models using machine learning techniques on key performance metrics such as accuracy, precision, recall, F-measure and AUC-ROC.

Keywords: global warming; natural hazards; disaster; floods; forecasting; ensemble learning.

DOI: 10.1504/IJW.2025.148164

International Journal of Water, 2025 Vol.17 No.1, pp.13 - 36

Received: 08 Sep 2024
Accepted: 16 Mar 2025
Published online: 27 Aug 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article