SRG · Physics
Storm Surge
GTSM hydrodynamic storm-tide return levels (COAST-RP) on a bare-earth elevation composite (GEDTM30 v1.2 + DeltaDTM v1.1 over Copernicus); tropical from the STORM 10,000-year synthetic catalog, extratropical from ERA5, de-duplicated at event level against IBTrACS · v1.0 · 2026-09
US Gulf Coast — the Katrina 2005 calibration coast
Coastal storm-surge screening from a hydrodynamic solve, so wind direction, fetch, basin geometry and tide-surge interaction are inside the numbers rather than approximated. Still-water only — no wave setup or runup — on a frozen 1980-2017 climatology with no sea-level rise, and flood defences are not modelled, so below-datum land reads as undefended. Screening for portfolio triage, not engineering design.
What ships
Surfaces — P(inundation) at 50 / 250 / 500-year horizons (500-year default) with a five-class cut, uncertainty and coverage. Returns — still-water inundation depth in metres at rp50/100/250/500, on the same return-period grid as Hurricane so wind and surge combine return-for-return.
Every release: 0–1 score · 5-class band · per-pixel uncertainty · input-coverage mask · STAC item · methodology doc.
Inputs
COAST-RP storm-tide return levels (Dullaart 2021), DeltaDTM v1.1 bare-earth elevation (Pronk 2024, from Copernicus WorldDEM-30), STORM synthetic cyclone catalog, ERA5, IBTrACS
Version history
What changed between releases — the same notes that ship with the data. Newest first.
Surfaces
v1.0
2026-09
Cut to 1.0 at this rebuild, not by relabelling v0.1. 1.0 means built on the current bare-earth elevation stack (GEDTM30 v1.2 + DeltaDTM v1.1 over Copernicus) with the region-scoped spike gate in the audit — not "the same data with a larger number". The rebuild is the reason the version moved.
- Coastal-fringe elevations corrected. The blend that joins land to sea floor was smearing bathymetry onto the first three pixels of land wherever the sea bed falls away steeply, inventing ground below sea level and therefore inventing depth on top of it. On the Puerto Rico coast the 500-year depth maximum fell from 10.3 m to 3.6 m against a regional 99th percentile of 0.75 m that did not move — the surrounding surface was always right; a handful of cells were not.
- Bare-earth terrain inland. GEDTM30 v1.2 now sits under the coastal DeltaDTM overlay, so inland coastal ground is no longer a surface model in which buildings and canopy read as terrain.
- Corrected terrain registration. The shared 90 m terrain features were displaced one pixel north-west relative to their own source; that is fixed, so terrain now aligns with land cover, buildings and soils.
- Per-cell elevation provenance ships beside each tile, recording which elevation model supplied every pixel.
- New audit gates: region-scoped spike detection with healthy controls, and a refusal to mosaic a partial product.
v0.1
2026-08
- Both storm populations, counted once. Tropical surge from the STORM 10,000-year synthetic cyclone catalog and extratropical surge (North Sea, nor'easters, Baltic, Mediterranean) from ERA5 — de-duplicated at the event level, so hybrid storms like Sandy are counted exactly once.
- Hydrodynamic source physics. Built on the GTSM global tide-and-surge solve (COAST-RP): wind direction — including negative surge on offshore-wind coasts — fetch, basin geometry and tide–surge interaction are inside the numbers, not approximated.
- Petrel coastal amplification. A tropical-branch enclosure amplification calibrated on Typhoon Haiyan (Tacloban) and Hurricane Katrina (Mississippi coast) closes the sub-grid gap where compact intense storms meet semi-enclosed shallow bays.
- Consistent with Petrel hurricane wind. Same underlying cyclone catalog — wind and surge losses can be combined without event-set mismatch.
- Surfaces: 50/250/500-yr inundation probability, 500-yr depth, 5-class exposure banding, uncertainty, coverage.
- Documented exclusions: wave setup, sea-level rise, flood defences (below-datum land reads as undefended), extratropical-transition surge.
Returns
v1.0
2026-09
Cut to 1.0 at this rebuild, not by relabelling v0.1. 1.0 means built on the current bare-earth elevation stack (GEDTM30 v1.2 + DeltaDTM v1.1 over Copernicus) with the region-scoped spike gate in the audit — not "the same data with a larger number". The rebuild is the reason the version moved.
- Coastal-fringe elevations corrected. The blend that joins land to sea floor was smearing bathymetry onto the first three pixels of land wherever the sea bed falls away steeply, inventing ground below sea level and therefore inventing depth on top of it. On the Puerto Rico coast the 500-year depth maximum fell from 10.3 m to 3.6 m against a regional 99th percentile of 0.75 m that did not move — the surrounding surface was always right; a handful of cells were not.
- Bare-earth terrain inland. GEDTM30 v1.2 now sits under the coastal DeltaDTM overlay, so inland coastal ground is no longer a surface model in which buildings and canopy read as terrain.
- Corrected terrain registration. The shared 90 m terrain features were displaced one pixel north-west relative to their own source; that is fixed, so terrain now aligns with land cover, buildings and soils.
- Per-cell elevation provenance ships beside each tile, recording which elevation model supplied every pixel.
- New audit gates: region-scoped spike detection with healthy controls, and a refusal to mosaic a partial product.
v0.1
2026-08
- Depth, not just probability. On low-lying deltas the probability of flooding within 500 years is effectively 1.0 everywhere — true, but it cannot tell a 2 m pixel from an 8 m one. This line carries the physical depth that drives damage.
- Consistent with Petrel hurricane wind. Both products are driven by the same STORM synthetic cyclone catalog, so wind and surge combine return-for-return without event-set mismatch.
- Both storm populations, counted once. Tropical surge from synthetic cyclones and extratropical surge (North Sea, nor'easters, Baltic, Mediterranean) from reanalysis, de-duplicated at the event level.
- Hydrodynamic source physics. Wind direction — including negative surge on offshore-wind coasts — fetch, basin geometry and tide–surge interaction come from a global tide-and-surge solve, not a parametric approximation.
- Elevation-bounded domain. The scored band follows terrain and hydrology rather than a fixed distance from shore, so it extends far up flat deltas and stops quickly on steep coasts.
- Documented exclusions: wave setup, sea-level rise, flood defences (below-datum land reads as undefended), compound rainfall/fluvial flooding, extratropical-transition surge.
Other hazards
LandslideAvalancheWildfireEarthquakeWindstormExtreme heatExtreme coldFrost / freezeLightningSoil erosionHurricaneDroughtVolcanic EruptionTsunamiHail