Data sources and validation
This page lists all data sources and the validation checks. The model pages link here.
Validation checks
| Check | Method | Result |
|---|---|---|
| Hail frequency | Comparison with NOAA storm reports, and with held-out years that were not used to build the model. | See the gap note below. |
| Hurricane winds | Comparison with ASCE 7-22 design wind speeds at return periods of 100–700 years. | Modeled median gusts are 5–7% below the design wind speeds. |
| Traceability | Each hazard run keeps a record of the exact source data used. | — |
These checks cover selected inputs and comparisons. They do not, by themselves, validate every hazard, damage curve, revenue output or asset-level decision.
Data sources
| Input | Source | Period |
|---|---|---|
| Hail and tornado records | NOAA MRMS and Storm Prediction Center. | Tornado paths since 1996 |
| Wind standards and hurricane simulation | ASCE 7-22 and STORM. | — |
| Wildfire | USDA Forest Service FSim. | — |
| Flood | NOAA flood mapping, National Water Model, Atlas 14, SLOSH and USGS 3DEP. | — |
| Generation | Gridded solar-radiation and wind-speed records. | 2012–2024 |
| Prices | Nodal wholesale averages; negative-price hours. | 2010–2024; 2025 |
| Market value | Berkeley Lab regional solar and wind factors. | — |
| Site constraints | Federal habitat and flood-zone maps; terrain, soil, ownership and NREL transmission-cost estimates. | — |
| Climate factors | CMIP6-based projections from a specialist climate-data provider. | — |
| Operating cost | NREL Annual Technology Baseline. | 2024 |
| Weather alerts | National Weather Service. | Live |
Version
Applies to methodology version 1.0, September 2026.