Technology Category
Computer Vision: Low Level & Physics-based VisionMachine Learning: Multimodal LearningPlanning, Routing, and Scheduling: Model-Based Reasoning
Data-driven Super-Resolution of Flood Inundation Maps using Synthetic Simulations
🤖 AI Summary
To address the limited temporal coverage of high-resolution flood inundation maps (FIMs), this paper proposes a data-driven super-resolution method that downscales daily low-resolution VIIRS water fraction maps (WFMs) to 30-m high-resolution FIMs. The method leverages high-fidelity hydrodynamic simulations generated via HEC-RAS to synthesize physically consistent, high-resolution flood training samples—enabling the construction of a deep convolutional neural network for single-image super-resolution. Crucially, the model exhibits strong generalization under zero-shot cross-regional transfer without fine-tuning. Experimental evaluation on real flood events in Iowa demonstrates significant improvements over conventional interpolation and non-learning baselines. Moreover, the approach shows reliable transferability to climatically and hydrologically similar regions. This work establishes a scalable, physics-informed data augmentation paradigm for high-frequency flood dynamics monitoring.
Technology Category
Application Category
📝 Abstract
The frequency of extreme flood events is increasing throughout the world. Daily, high-resolution (30m) Flood Inundation Maps (FIM) observed from space play a key role in informing mitigation and preparedness efforts to counter these extreme events. However, the temporal frequency of publicly available high-resolution FIMs, e.g., from Landsat, is at the order of two weeks thus limiting the effective monitoring of flood inundation dynamics. Conversely, global, low-resolution (~300m) Water Fraction Maps (WFM) are publicly available from NOAA VIIRS daily. Motivated by the recent successes of deep learning methods for single image super-resolution, we explore the effectiveness and limitations of similar data-driven approaches to downscaling low-resolution WFMs to high-resolution FIMs. To overcome the scarcity of high-resolution FIMs, we train our models with high-quality synthetic data obtained through physics-based simulations. We evaluate our models on real-world data from flood events in the state of Iowa. The study indicates that data-driven approaches exhibit superior reconstruction accuracy over non-data-driven alternatives and that the use of synthetic data is a viable proxy for training purposes. Additionally, we show that our trained models can exhibit superior zero-shot performance when transferred to regions with hydroclimatological similarity to the U.S. Midwest.
Problem
Research questions and friction points this paper is trying to address.
Enhancing flood map resolution
Using synthetic data for training
Improving flood monitoring frequency
Innovation
Methods, ideas, or system contributions that make the work stand out.
Deep learning for super-resolution
Synthetic data for model training
Zero-shot transfer to similar regions
🔎 Similar Papers
No similar papers found.
