The CIROH Flood Inundation Mapping Community Ecosystem
Five open-source tools for generating, simulating, evaluating, and benchmarking flood inundation maps — from operational forecast to validated science.
One pipeline, multiple interconnected tools
Every stage shares formats, benchmarks, and infrastructure — so a forecast map and a research model are judged on identical terms. Producing a flood map is half the problem; knowing whether to trust it and how to improve it is the other half.
FIMserv
Operational FIM generation using the NOAA OWP HAND-FIM forecasting framework — National Water Model streamflow in, inundation extent and depth rasters out, at national scale and low computational cost.
- Runs multiple HUC-8 watersheds simultaneously, at any temporal resolution of the input streamflow
- Accepts NWM retrospective and forecast flows, USGS gauges, or any user-supplied discharge
- Cloud-ready Python package — Jupyter, Google Colab, and CIROH 2i2C
pip install fimserve
FIMsim
A web application that removes the technical barrier to 2D flood simulation. Define a study area and FIMsim downloads every input, writes all model configuration files, and runs the simulation on cloud infrastructure — no GIS expertise or local software required.
- Four standalone input tools — DEM, LULC & Manning’s n, flowlines, and streamflow — each usable on its own or as part of a full run
- End-to-end pipelines for LISFLOOD-FP and TRITON, executed on cloud infrastructure
- Multi-AOI batch processing, editable roughness table, and map preview at every step
- Runs in the browser, or as a standalone desktop installer for macOS and Windows
FIMeval
A large-scale FIM evaluation framework that automatically compares predicted flood maps against benchmarks — replacing tedious, error-prone manual geospatial analysis.
- Automated preprocessing: reprojection, binary flood maps, permanent-water removal
- Standard skill metrics plus building-footprint inundation comparison
- Queries the FIMbench database directly for the right benchmark at any location
pip install fimeval
FIMbench
A curated benchmarking database of reference flood maps derived from remote sensing imagery, aerial observations, and high-fidelity hydrodynamic model outputs.
- Four quality-based tiers, plus a high-water-mark–derived category
- Hosted on AWS S3 with an open API for research pipelines and workflows
- Interactive map visualizer — filter by tier and date, inspect metadata, download rasters
FIMbox
A FIM sandbox for model development and analysis — where new ideas are prototyped, stress-tested, and refined before joining the operational pipeline.
- Isolated environment for experimenting with models, parameters, and data
- Analysis utilities for exploring FIM outputs and diagnostics
- Bridges research code to the operational generation–evaluation loop