FM-LC: A Hierarchical Framework for Urban Flood Mapping by Land Cover Identification Models

2025-07-26

AI Summary

Urban flood mapping in arid regions faces challenges including low spectral contrast, high surface heterogeneity, and rapid hydrological dynamics, limiting the accuracy of conventional remote sensing approaches. To address these issues, this paper proposes FM-LC, a hierarchical framework leveraging multi-temporal PlanetScope imagery. It employs a three-stage strategy: (1) multi-class U-Net for initial land cover and water segmentation; (2) lightweight binary expert models tailored to confusion-prone classes for misclassification correction; and (3) Bayesian neighborhood smoothing to enhance boundary continuity. Evaluated on the 2024 Dubai flash flood event, FM-LC achieves a 29% average F1-score improvement over a single-stage U-Net baseline across all land-cover classes. Flood delineations exhibit sharper boundaries and significantly improved spatial coherence. The framework delivers scalable, high spatiotemporal resolution flood monitoring—enabling timely emergency response and resilience-oriented urban planning in arid-zone cities.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Remote Sensing / Geospatial AI
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Urban flooding in arid regions poses severe risks to infrastructure and communities. Accurate, fine-scale mapping of flood extents and recovery trajectories is therefore essential for improving emergency response and resilience planning. However, arid environments often exhibit limited spectral contrast between water and adjacent surfaces, rapid hydrological dynamics, and highly heterogeneous urban land covers, which challenge traditional flood-mapping approaches. High-resolution, daily PlanetScope imagery provides the temporal and spatial detail needed. In this work, we introduce FM-LC, a hierarchical framework for Flood Mapping by Land Cover identification, for this challenging task. Through a three-stage process, it first uses an initial multi-class U-Net to segment imagery into water, vegetation, built area, and bare ground classes. We identify that this method has confusion between spectrally similar categories (e.g., water vs. vegetation). Second, by early checking, the class with the major misclassified area is flagged, and a lightweight binary expert segmentation model is trained to distinguish the flagged class from the rest. Third, a Bayesian smoothing step refines boundaries and removes spurious noise by leveraging nearby pixel information. We validate the framework on the April 2024 Dubai storm event, using pre- and post-rainfall PlanetScope composites. Experimental results demonstrate average F1-score improvements of up to 29% across all land-cover classes and notably sharper flood delineations, significantly outperforming conventional single-stage U-Net baselines.

Problem

• Accurate flood mapping in arid urban areas with spectral challenges
• Distinguishing spectrally similar land covers like water and vegetation
• Improving flood delineation precision using hierarchical segmentation and smoothing

Innovation

• Hierarchical U-Net for multi-class land cover segmentation
• Lightweight binary expert model for misclassification correction
• Bayesian smoothing for boundary refinement and noise removal