Title: Classification based forecasting models using medium range hydro-meteorological data of the flood affected districts of Kerala

Authors: Sweta Shukla; T.V. Vijay Kumar; Vikas Mittal

Addresses: Special Centre for Disaster Research, Jawaharlal Nehru University, New Delhi, 110067, India ' 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

Abstract: Floods are one of the most recurring and damaging phenomena across the globe. India experiences frequent flood incidents due to its geographical setting, climatic conditions and poor drainage system. In recent past, uncontrolled and unplanned human interventions in flood plains have aggravated the flood impacts leading to increase in flood events in riverine and coastal plains. To minimise such occurrences, preparedness measures need to be adopted and implemented that would enable better forecasting of flood events. This paper aims to design flood forecasting models for the districts of Kerala where recurring floods are one of the most concerning issues. Five machine learning techniques are used for forecasting floods in Kerala using the hydro-meteorological flood dataset consisting of meteorological parameters. The experimental results show that amongst all models, Random Forest (RF) model has a comparatively better ability to forecast meteorological conditions that may lead to floods in Kerala.

Keywords: hazard; disaster risk reduction; flood forecasting; artificial intelligence; machine learning.

DOI: 10.1504/IJW.2026.153206

International Journal of Water, 2026 Vol.17 No.3, pp.245 - 269

Received: 12 Jun 2025
Accepted: 27 Jul 2025
Published online: 29 Apr 2026 *

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