Flood Risk Follows Valleys, Not Grids: Graph Neural Networks for Flash Flood Susceptibility Mapping in Himachal Pradesh with Conformal Uncertainty Quantification

• URL: http://arxiv.org/abs/2603.15681v1
• Date: Sun, 15 Mar 2026 02:54:24 GMT
• Title: Flood Risk Follows Valleys, Not Grids: Graph Neural Networks for Flash Flood Susceptibility Mapping in Himachal Pradesh with Conformal Uncertainty Quantification
• Abstract summary: Flash floods are the most destructive natural hazard in Himachal Pradesh, India, causing over 400 fatalities and $1.2 billion in losses in the 2023 monsoon season alone.<n>Existing risk maps treat every pixel independently, ignoring the basic fact that flooding upstream raises risk downstream.<n>We address this with a Graph Neural Network (GraphSAGE) trained on a watershed connectivity graph.
• Score: 0.20871483263418808
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Flash floods are the most destructive natural hazard in Himachal Pradesh (HP), India, causing over 400 fatalities and $1.2 billion in losses in the 2023 monsoon season alone. Existing risk maps treat every pixel independently, ignoring the basic fact that flooding upstream raises risk downstream. We address this with a Graph Neural Network (GraphSAGE) trained on a watershed connectivity graph (460 sub-watersheds, 1,700 directed edges), built from a six-year Sentinel-1 SAR flood inventory (2018-2023, 3,000 events) and 12 environmental variables at 30 m resolution. Four pixel-based ML models (RF, XGBoost, LightGBM, stacking ensemble) serve as baselines. All models are evaluated with leave-one-basin-out spatial cross-validation to avoid the 5-15% AUC inflation of random splits. Conformal prediction produces the first HP susceptibility maps with statistically guaranteed 90% coverage intervals. The GNN achieved AUC = 0.978 +/- 0.017, outperforming the best baseline (AUC = 0.881) and the published HP benchmark (AUC = 0.88). The +0.097 gain confirms that river connectivity carries predictive signal that pixel-based models miss. High-susceptibility zones overlap 1,457 km of highways (including 217 km of the Manali-Leh corridor), 2,759 bridges, and 4 major hydroelectric installations. Conformal intervals achieved 82.9% empirical coverage on the held-out 2023 test set; lower coverage in high-risk zones (45-59%) points to SAR label noise as a target for future work.

Related papers

• YOLO26-RD: An End-to-End Road Damage Detection Network With Learnable Contrast Enhancement and Edge-Guided Downsampling [2.274915755738124]
  This paper tests the premise that a stride-4 level would claim 75.3% of anchors, and complete misses rather than localisation errors dominate baseline failures.<n>Fifteen models were trained from scratch under one recipe, five scales each of YOLO26-RD and of matched YOLO26 and YOLOv12 families.<n>YOLO26-RD-l is the best of the fifteen at 0.809 mAP50 and 0.497 mAP50-95, improving on the YOLO26 reference by 0.031 and 0.030 and leading all six per class entries.
  arXiv  Detail & Related papers  (2026-08-16T12:36:12Z)
• Physics-Informed CNN-LSTM for Street-Scale Urban Flood Prediction: Reconciling Aggregate Accuracy and Street-Level Plausibility [4.789111524845274]
  Deep learning surrogate models trained with mean-squared-error loss produce statistically accurate but physically unconstrained flood predictions.<n>We develop a physics-informed training framework for CNN-LSTM models that predict urban flood depths at 15 min intervals over a 128x128 spatial grid.
  arXiv  Detail & Related papers  (2026-07-27T23:46:25Z)
• Land cover and flood type govern the detection limits of satellite-based flood mapping across diverse global flood events [0.3448975792603021]
  Floods are among the most destructive natural hazards, and their increasing frequency under climate change makes satellite-based inundation mapping essential for disaster response.<n> Geospatial foundation models pretrained on satellite offer transferability, but their operational reliability across diverse, unseen events remains uncharacterized.<n>Here we deploy Prithvi-EO-2.0 across 19 out-of-distribution flood events spanning six continents, eight climate zones, and six flood mechanisms, validating against two independent reference products.
  arXiv  Detail & Related papers  (2026-06-05T18:46:03Z)
• Paired-CSLiDAR: Height-Stratified Registration for Cross-Source Aerial-Ground LiDAR Pose Refinement [46.13526676100776]
  We introduce Paired-CSLiDAR, a cross-source aerial-ground LiDAR benchmark for single-scan pose refinement.<n>The benchmark contains 12,683 ground-aerial pairs across 6 evaluation sites.<n>We propose Residual-Guided Stratified Registration (RGSR), a training-free, geometry-only refinement pipeline.
  arXiv  Detail & Related papers  (2026-05-01T13:14:20Z)
• Ensemble Graph Neural Networks for Probabilistic Sea Surface Temperature Forecasting via Input Perturbations [41.99844472131922]
  This work investigates ensemble learning strategies for sea surface temperature (SST) forecasting using Graph Neural Networks (GNNs)<n>We adapt a GNN architecture to the Canary Islands region in the North Atlantic and implement a homogeneous ensemble approach inspired by bagging.<n>Several noise-based ensemble generation strategies are evaluated, including Gaussian noise, Perlin noise, and fractal Perlin noise.
  arXiv  Detail & Related papers  (2026-03-06T11:04:12Z)
• BadCLIP++: Stealthy and Persistent Backdoors in Multimodal Contrastive Learning [73.46118996284888]
  Research on backdoor attacks against multimodal contrastive learning models faces two key challenges: stealthiness and persistence.<n>We propose BadCLIP++, a unified framework that tackles both challenges.<n>For stealthiness, we introduce a semantic-fusion QR micro-trigger that embeds imperceptible patterns near task-relevant regions.<n>For persistence, we stabilize trigger embeddings via radius shrinkage and centroid alignment.
  arXiv  Detail & Related papers  (2026-02-19T08:31:16Z)
• Predictive Modeling of Flood-Prone Areas Using SAR and Environmental Variables [0.0]
  Flooding is one of the most destructive natural hazards worldwide, posing serious risks to ecosystems, infrastructure, and human livelihoods.<n>This study combines Synthetic Aperture Radar (SAR) imagery with environmental and hydrological data to model flood susceptibility in the River Nyando watershed, western Kenya.
  arXiv  Detail & Related papers  (2025-12-06T16:24:10Z)
• HydroGAT: Distributed Heterogeneous Graph Attention Transformer for Spatiotemporal Flood Prediction [7.643667076993224]
  Graph Networks (GNNs) propagate information exactly along the river network, which is ideal for learning hydrological routing.<n>We introduce a heterogenous basin graph where every land and river pixel is a node connected by physical hydrological and inter-catchment relationships.<n>We propose HydroGAT, atemporal network that adaptively learns local temporal importance and the most influential upstream locations.
  arXiv  Detail & Related papers  (2025-09-02T16:31:40Z)
• ZIP: Scalable Crowd Counting via Zero-Inflated Poisson Modeling [29.570935239024138]
  We introduce ZIP, a scalable crowd counting framework that models blockwise counts with a Zero-Inflated Poisson likelihood.<n>We provide a generalization analysis showing a tighter risk bound for ZIP than MSE-based losses.<n>Experiments on ShanghaiTech A & B, UCF-QNRF, and NWPU-Crowd demonstrate that ZIP consistently surpasses state-of-the-art methods across all model scales.
  arXiv  Detail & Related papers  (2025-06-24T19:06:22Z)
• Graph Transformer-Based Flood Susceptibility Mapping: Application to the French Riviera and Railway Infrastructure Under Climate Change [0.0]
  This study introduces the first application of a graph transformer (GT) architecture for flood susceptibility mapping to the flood-prone French Riviera.<n>GT incorporates watershed topology using Laplacian positional encoders (PEs) and attention mechanisms.<n> Dimensionality reduction on Laplacian PEs revealed partial clusters, indicating they could capture spatial information.
  arXiv  Detail & Related papers  (2025-03-31T14:04:04Z)
• Satellite-Surface-Area Machine-Learning Models for Reservoir Storage Estimation: Regime-Sensitive Evaluation and Operational Deployment at Loskop Dam, South Africa [0.0]
  Prediction errors were decomposed into two regimes: Low (250 x 106 cubic meters) and High (>250 x 106 cubic meters) storage regimes.<n>Ridge regression achieved the lowest cross validated RMSE (12.3 x 106 cubic meters), outperforming GB by 16% and RF by 7%.<n>We recommend (i) GB retrained daily for routine operations, (ii) Ridge for drought early warning, and (iii) the stacked blend for all weather dashboards.
  arXiv  Detail & Related papers  (2025-02-27T11:14:14Z)
• Accelerating Flood Warnings by 10 Hours: The Power of River Network Topology in AI-enhanced Flood Forecasting [30.106533909897102]
  Graph Neural Networks (GNNs) underutilize river network topology due to tree-like structures causing over-squashing from high node resistance distances.<n>This study introduces a reachability-based graph transformation to densify topological connections, reducing resistance distances.<n>Tests show transformed-GNNs outperform EA-LSTM in extreme flood prediction, achieving 24-h water level accuracy equivalent to EA-LSTM's 14-h forecasts.
  arXiv  Detail & Related papers  (2024-10-07T22:25:37Z)
• How to predict on-road air pollution based on street view images and machine learning: a quantitative analysis of the optimal strategy [23.420609727152097]
  On-road air pollution exhibits substantial variability over short distances due to emission sources, dilution, and physicochemical processes. Integrating mobile monitoring data with street view images holds promise for predicting local air pollution. To bridge this gap, we employed 314 taxis to monitor NO, NO2, PM2.5 and PM10 dynamically and sampled corresponding SVIs.
  arXiv  Detail & Related papers  (2024-09-19T02:21:23Z)
• UrbanSARFloods: Sentinel-1 SLC-Based Benchmark Dataset for Urban and Open-Area Flood Mapping [24.857739769719778]
  UrbanSARFloods is a dataset featuring pre-processed Sentinel-1 intensity data and interferometric coherence imagery acquired before and during flood events. It contains 8,879 $512times 512$ chips covering 807,500 $km2$ across 20 land cover classes and 5, spanning 18 flood events. We used UrbanSARFloods to benchmark existing state-of-the-art convolutional neural networks (CNNs) for segmenting open and urban flood areas.
  arXiv  Detail & Related papers  (2024-06-06T14:28:43Z)
• Automating global landslide detection with heterogeneous ensemble deep-learning classification [44.99833362998488]
  Landslides threaten infrastructure, including roads, railways, buildings, and human life. Hazard-based spatial planning and early warning systems are cost-effective strategies to reduce the risk to society from landslides. Deep learning models have recently been applied for landside mapping using medium- to high-resolution satellite images as input.
  arXiv  Detail & Related papers  (2023-09-12T10:56:16Z)
• HRank: Filter Pruning using High-Rank Feature Map [149.86903824840752]
  We propose a novel filter pruning method by exploring the High Rank of feature maps (HRank) Our HRank is inspired by the discovery that the average rank of multiple feature maps generated by a single filter is always the same. Based on HRank, we develop a method that is mathematically formulated to prune filters with low-rank feature maps.
  arXiv  Detail & Related papers  (2020-02-24T11:50:09Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.