Efficient Flood Detection through Hybrid Machine Learning and Metaheuristic Methods using Sentinel-1

Aug 5, 2026

Upload by Liam

Floods are among the most destructive natural disasters and demand precise, timely decision-making for effective management. Satellite remote sensing provides diverse imagery for monitoring and assessing flood impacts, enabling data-driven detection pipelines. This study uses Sentinel-1 SAR data with machine learning models Random Forest and HGBDT, paired with metaheuristic hyperparameter optimization via Harris Hawks Optimization and Ant Colony Optimization, and a pre-trained VGG-16 for deep feature extraction. Ensemble models are evaluated with statistical metrics, achieving validation accuracy above 95% and testing accuracy above 97%. The HGBDT-ACO model shows the best flood-pixel detection with the lowest error rate; HGBDT also offers favorable computational complexity over RF.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Introduction

Floods and the need for accurate monitoring

Role of satellite remote sensing (optical vs. radar)

Sentinel missions and SAR advantages

Existing flood detection approaches

Efficient Flood Detection through Hybrid Machine Learning and Metaheuristic Methods using Sentinel-1 Behnam Ebadati 1, Reza Attarzadeh 1, Mohammad Alikhani 2, Fahimeh Youssefi 3, Saied Pirasteh 3 1 Dept. of Geomatic Engineering, South Tehran Branch, Islamic Azad University, Tehran, [Iran-st_b_ebadati@azad.ac.ir](Iran-st_b_ebadati@azad.ac.ir), [reza.attarzadeh@iau.ac.ir](reza.attarzadeh@iau.ac.ir) 2 Geodesy and Geomatics Engineering Faculty, K. N. Toosi University of Technology, Tehran, [Iran-m.alikhani1@email.kntu.ac.ir](Iran-m.alikhani1@email.kntu.ac.ir) 3 Institute of Artificial Intelligence, Shaoxing University, 508 West Huancheng Road, Yuecheng District, Shaoxing, Zhejiang Province, Postal Code 312000, [China-youssefi@usx.edu.cn](China-youssefi@usx.edu.cn), [sapirasteh1@usx.edu.cn](sapirasteh1@usx.edu.cn)[ ](sapirasteh1@usx.edu.cn)Keywords: Flood Detection, Machine Learning, VGG-16, Hyperparameter Optimization, Metaheuristic Algorithms, Sentinel-1. Abstract Floods are considered among the most destructive natural disasters, requiring precise and timely management. Remote sensing, utilizing diverse satellite imagery data, enables effective monitoring and assessment of flood impacts. In this context, machine learning and deep learning methods, as effective and scalable approaches, can significantly enhance the accuracy of flood detection and management by analyzing remote sensing data, thereby playing a crucial role in mitigating flood-related risks. In this study, to flood detection using Sentinel-1 SAR data, machine learning algorithms, including Random Forest (RF) and Histogram-based Gradient Boosting Decision Tree, were employed, along with two metaheuristic algorithms, Harris Hawks Optimization (HHO) and Ant Colony Optimization (ACO), for hyperparameter optimization. Additionally, to enhance the models' ability to detect flooded pixels and improve overall performance accurately, a pre-trained VGG-16 Neural Network was used as a deep feature extractor. Finally, four ensemble flood detection models—RF-HHO, RF-ACO, HGBDT-HHO, and HGBDT-ACO—were implemented, and their performance was evaluated and compared based on statistical metrics. Based on the obtained results, all four ensemble flood detection models demonstrated excellent performance in the validation and testing phases. The overall accuracy of these models reached over 95% in the validation phase and exceeded 97% in the testing phase. However, the HGBDT-ACO model achieved the highest accuracy and the lowest error rate in detecting flood pixels, making it the best-performing model in this study. Generally, HGBDT models showed a relative advantage over RF models, as they required significantly less time for training while achieving comparable results. Therefore, they were efficient and performed better in terms of computational complexity. 1. Introduction Recently, the frequency of natural disasters worldwide has significantly increased (Munawar et al., 2022) . Among these, floods stand out as one of the most prevalent water-related calamities, directly or indirectly affecting approximately 23% of the global population, equivalent to 1.8 billion individuals (Amitrano et al., 2024) . The United Nations Office for Disaster Risk Reduction (UNDRR) reports that the occurrence of flood events has seen a remarkable rise across the globe over the past two decades (Sadiq et al., 2023) . Climate change, intense rainfall, snowmelt, glacier retreat, and dam breaches are among the primary triggers for flooding (Jeyaseelan, 2004) . Additionally, rapid urbanization and increased human activities pose an even more significant threat of floods to human communities (Alidoost and Arefi, 2018) . Floods can severely damage agricultural lands, residential areas, and transportation infrastructures such as roads and railways (Lamovec et al., 2013) . Despite the seemingly insurmountable challenge of preventing floods, effective management can reduce the risks and mitig

FAQs

Why is Sentinel-1 SAR useful for flood detection?

Sentinel-1 SAR can acquire images day and night and under adverse weather conditions, including dense cloud cover, making it more effective than purely optical approaches.

Which machine learning models are used in the study?

The study employs Random Forest and Histogram-based Gradient Boosting Decision Tree (HGBDT) for flood detection based on Sentinel-1 data and extracted features.

How are hyperparameters optimized in this work?

Two metaheuristic algorithms—Harris Hawks Optimization (HHO) and Ant Colony Optimization (ACO)—are used to perform hyperparameter optimization for the detection models.

Aug 5, 2026

Upload by Liam