GeoAI for Flood Inundation Mapping
The broader research theme that frames the problem, methods, and linked outputs.
Read more
Back to research
Publication · GeoAI · Flood Inundation Mapping
Enhancement of low-fidelity flood inundation mapping through surrogate modeling
This paper focuses on how surrogate modeling can strengthen lower-fidelity flood inundation products by learning spatial patterns from simulations and geospatial predictors. It sits at the center of the GeoAI research theme and connects directly to operational evaluation workflows.
Overview
The manuscript examines how surrogate models can augment low-fidelity hydraulic outputs while preserving spatially meaningful flood patterns. The framing is research-forward but practical: improve runtime, improve map quality, and preserve compatibility with evaluation and decision-support workflows.
If you want to paste the full abstract later, this panel is the right place for it.
Why This Output Matters
- Connects machine learning directly to flood-inundation mapping quality.
- Supports the broader GeoAI research area with a concrete paper-level contribution.
- Links naturally to benchmarking and evaluation tools such as FIMeval.
- Provides a publication anchor for the conference and software narratives.