Methodology

Flood

Peril Report looks up the FEMA National Flood Hazard Layer for the point that matches your address. The headline is the mapped flood zone, with the annual chance of flooding where FEMA publishes it. Show on Map overlays FEMA flood hazard zones on the address map and places a legend under the map. Street, satellite, and elevation views are available (OpenStreetMap streets, Esri satellite imagery, and USGS 3DEP lidar hillshade with highlighted NHD streams, rivers, and lakes, tinted from red at the lowest point in the original view to green 30 feet above it); a pin marks the looked-up point without an address popup.

Special Flood Hazard Areas (typically A and V zones) are the 1% annual-chance floodplain. Zone X includes areas outside that floodplain, including the 0.2% annual-chance (500-year) floodplain when FEMA maps it. A levee is noted only when FEMA’s zone subtype says the area is protected by a levee.

If the point is not in a mapped flood zone, the report shows the nearest mapped zone within 2 miles. Nearest flowing water is the closest creek, stream, or river on an NHDPlus flowline from the USGS National Hydrography Dataset, named from the EPA NHDPlus service. The same line says how many feet that closest point sits below the address, from USGS 3DEP ground elevation. The water value is the ground where the mapped line sits, not a flood stage. The 1% flow is the peak streamflow with a 1% annual exceedance probability at that point, from USGS StreamStats. The elevation graphic is the USGS 3DEP ground elevation from the address to 20 meters past that stream, on 1 meter ground near the channel. A satellite view of that same line is shown under it, on the same horizontal scale. Blue shading is the lowest 210 square feet of that cross section, the area that carries the 1% flow at about 4 feet per second.

Height Above Nearest Drainage (HAND) is the vertical distance from the address down to the stream it drains to. In the contiguous U.S. it is read from the National Flood Interoperability Experiment CONUS HAND rasters (Oak Ridge National Laboratory, hosted at the Texas Advanced Computing Center), computed from 10 m USGS 3DEP bare-earth terrain and NHDPlus medium-resolution streams. The cell containing the address is read directly from the HUC6 basin raster. Elsewhere, or if that read fails, it falls back to the Alaska Satellite Facility Global HAND dataset, computed from the 30 m Copernicus GLO-30 surface model, which includes tree canopy and buildings. HAND measures to the stream the address drains to, which can differ from the nearest flowing water.

The Flash Flooding figure is the 1% annual-chance (100-year) 1-hour rainfall intensity from NOAA Atlas 14, queried at the address from the Precipitation Frequency Data Server. Values are partial-duration series estimates in inches per hour. The figure is omitted in regions Atlas 14 does not cover, including much of the Pacific Northwest. The comparison says how much of the contiguous US has a lower value. It uses the area-weighted NOAA Atlas 14 grid of the same 1-hour figure. Alaska, Hawaii, and unpublished areas are left out.

Below it, the hydrologic soil group is the dominant-condition rating for the soil map unit at the address, from the USDA NRCS Soil Survey Geographic Database (SSURGO) through Soil Data Access. Groups A through D rank how quickly soil absorbs water after it is thoroughly wet, from A (sand and gravel, low runoff) to D (clay, high water table, or shallow bedrock, high runoff). Dual groups such as A/D give the drained and natural ratings for soils with a high water table.

Tropical Cyclones

Hurricane chances on Peril Report come from NOAA IBTrACS (International Best Track Archive for Climate Stewardship). Storms are counted from 1950 onward when the U.S. wind speed reaches hurricane force (64 knots, Category 1) or major-hurricane force (96 knots, Category 3).

A storm counts if its center passed within 50 miles of the Census county that contains the address. Unique storm IDs are used so one hurricane is not counted twice. The annual chance is a Poisson estimate from that history: 1 − exp(−count / years). The rating in that section is FEMA’s National Risk Index tropical-cyclone expected annual loss rating for the county, which is separate from the IBTrACS strike chances. The percentile bar is the tropical-cyclone expected annual loss rate compared with other U.S. counties. The text also prorates historic and modeled building-loss ratios over 30 years.

Hurricane storm surge comes from the National Hurricane Center’s National Storm Surge Risk Maps, the same public maps at https://www.nhc.noaa.gov/nationalsurge/. Those maps take SLOSH Maximum of Maximums (MOM) water levels at high tide and convert them to inundation above ground on a ~10 m DEM. Peril Report lists the NHC depth class for each hurricane category (1–5) that inundates the address, labeled as water depth from a worst-case storm of that category. Depths use NHC’s published bins (< 3 ft, 3-6 ft, 6-9 ft, and > 9 ft). A category is omitted where that storm does not flood the location. Leveed areas are shown as such. If the maps show no inundation but historic hurricane chance is at least 0.1%, the same cited box states that the location is not expected to be exposed, based on storm, tide, and elevation modelling. NHC does not cover the Pacific Northwest or Alaska.

For Gulf and East Coast counties that Colorado State University covers, Peril Report also shows that year’s Tropical Cyclone Impact forecast: the probability of a hurricane and of a major hurricane tracking within 50 miles. Those seasonal odds are adjusted from 1880–2020 climatology using CSU’s latest Atlantic forecast. They are omitted when CSU does not publish a forecast for the county.

Tropical cyclone is the generic name for a warm-core storm with a closed circulation. A tropical depression has winds below 39 mph (34 kt), with rain and flooding the main threats. At 39–73 mph (34–63 kt) it is a tropical storm and is given a name. At 74 mph (64 kt) or higher the same storm is called a hurricane in the Atlantic and Northeast Pacific, a typhoon in the Northwest Pacific, and a cyclone (or severe tropical cyclone) in the South Pacific and Indian Ocean. Super typhoon is a regional term for the strongest Northwest Pacific storms. A subtropical cyclone is a hybrid of tropical and extratropical storms; it can still bring damaging wind, rain, and surge, and may later become fully tropical.

Forecasting

The Forecasting page shows live tropical and climate outlooks alongside the address lookup. The National Hurricane Center 7-day Atlantic Graphical Tropical Weather Outlook is the official short-range product. Current Atlantic tropical cyclone public advisories from NHC are listed beside that map; click an advisory for the full bulletin. Google Weather Lab’s public 1000-member WeatherNext Cyclones file is paired to official observed storms, so it omits invests and genesis. Forecasting prefers a locally run WeatherNext Cyclones Mini ensemble (1°, Direct Tracker with cyclogenesis) when that grid has been uploaded; otherwise it uses the public 1000-member file, then falls back to the 64-member WeatherNext 3 file. Probability is the share of ensemble members with a track point inside 200 km of a 0.5° Atlantic cell at any lead in the run. Mini is coarser than the operational 0.25° 1000-member model and is experimental, not an official warning. Recent Weather Lab cycles are subject to Weather Lab’s terms of use. ECMWF’s AIFS ENS (50 members plus a control) is shown on the same 200 km / 0.5° grid, using that model’s tropical-cyclone trajectory files, which include storms that form during the forecast. AIFS is experimental, not an official warning; ECMWF open data is CC BY 4.0.

El Niño–Southern Oscillation status uses NOAA Climate Prediction Center’s Oceanic Niño Index and the latest ENSO Diagnostic Discussion. Historical hurricane tracks on Forecasting come from NOAA IBTrACS. Each storm is a simplified track whose stretches are colored by U.S. 1-minute sustained wind on the Saffir–Simpson scale at that time and place (Category 1 at 64 kt through Category 5 at 137 kt). All hurricanes in the archive are shown at once and can be filtered by peak category or year, or found by name. Choosing a search match hides the other tracks until Show all or a click on empty map. Search matches whose name was retired after that storm are shown in bold, using the WMO/NHC Atlantic and eastern/central Pacific retired-name lists. Clicking a track or choosing a search match shows that location’s category plus the storm’s name, dates, basin, and peak wind.

Wildfires

Peril Report searches mapped wildfire perimeters within 20 miles of the address. Historic fires come from the NIFC Interagency Fire Perimeter History layer. Recent fires, including current-year perimeters, come from WFIGS Interagency Fire Perimeters. Prescribed burns are left out.

If the point sits inside a perimeter, the report says the location burned over and, when more than one unique fire is mapped there, how many times. Otherwise it names the closest mapped fire, when it burned, and its size. Beyond 20 miles, it only says the closest mapped wildfire is more than 20 miles away.

Nearby fuel treatments are searched separately, within 2 miles: completed USDA Forest Service FACTS Hazardous Fuel Treatment Reduction polygons (thinning, mastication, pile or broadcast burning, and similar work) and NIFC or WFIGS prescribed-fire perimeters. Those operations can reduce local fuels. They are not mixed into the closest-wildfire sentence and are not the same as the WRC Community Wildfire Risk Reduction Zone.

The rating is FEMA’s National Risk Index wildfire expected annual loss rating for the county. The percentile bar is the wildfire expected annual loss rate compared with other U.S. counties. The text also prorates historic and modeled building-loss ratios over 30 years. Red-flag warning counts from the past 10 years are included when that record is available, on the same line as the closest mapped wildfire.

Risk to Potential Structures comes from the USDA Forest Service Wildfire Risk to Communities 30-meter raster (USFS WRC Risk to Potential Structures). It combines wildfire likelihood and intensity with generalized consequences to a home at that pixel: the relative risk to a house if one existed there. Annual burn probability at the pixel comes from the WRC Burn Probability raster (Wildfire Likelihood) and is shown as the chance that location burns in any given year. Classes are Low, Medium, High, and Very High using national pixel percentiles from the WRC methods (below the 40th, 70th, and 90th percentiles). The Community Wildfire Risk Reduction Zone (USFS WRC CWiRRZ 4-class raster) is the exposure type at that pixel: minimal, indirect (embers and home-to-home), direct (adjacent vegetation plus embers), or wildfire transmission. The underlying spatial datasets are cataloged in the Forest Service Research Data Archive as RDS-2020-0016-2 (https://doi.org/10.2737/RDS-2020-0016-2). The full recommended citation appears at the bottom of each address report.

For California ZIP codes, the report also shows the CAL FIRE Fire Hazard Severity Zone at the point (CAL FIRE OSFM FHSZ maps). State Responsibility Area zones are the adopted April 2024 map. Local Responsibility Area zones use CAL FIRE’s 2025 recommended classes (Moderate, High, Very High). Points outside those polygons are reported as not in a mapped zone.

Severe Convective Storms

Hail

Hail on Peril Report uses FEMA’s National Risk Index for the county. The headline is the expected annual loss rating (for example, Relatively Low), not the composite hail risk rating that also folds in social vulnerability and community resilience.

The percentile bar is the expected annual loss rate compared with other U.S. counties. The text also reports how often hail occurs in the NRI record and prorates historic and modeled building-loss ratios over 30 years.

Tornadoes

Tornado chances use NOAA Storm Prediction Center SVRGIS tornado paths from 1950 through 2024 (the same historical layer as FEMA’s Resilience Analysis and Planning Tool). Peril Report counts tracks that intersect the Census county containing the address, for any tornado (EF0 or above) and for strong tornadoes (EF3 or above). The annual chance is a Poisson estimate from that county history: 1 − exp(−count / years). The rating in the section is FEMA’s National Risk Index tornado expected annual loss rating, which is separate from those track-based chances. The percentile bar is the tornado expected annual loss rate compared with other U.S. counties. The text also prorates historic and modeled building-loss ratios over 30 years.

Wind

The Wind section shows a 3-second gust at 25 feet above ground for Exposure Category C, based on the ASCE 7-22 basic wind speed from the ASCE 7 Hazard Tool (100-year mean recurrence interval, Section 26.5.2) scaled with the Exposure C K_z coefficient in Section 26.10. The cited box adds a Beaufort force and description below hurricane strength, or a plain-language Saffir–Simpson damage description at 74 mph and above (no category label). Those comparisons use conventional wind definitions and are not a formal conversion of the ASCE gust. This is not a forecast or a historic wind report.

Earthquake

Earthquake results are FEMA National Risk Index values for the county, not a site-specific shaking model. The headline is the expected annual loss rating (for example, Relatively Moderate), not the composite earthquake risk rating that also folds in social vulnerability and community resilience. Frequency comes from the NRI annualized frequency field. Event counts, when present, are from the NRI record, which FEMA treats as a 100-year earthquake history for this index.

Peak ground acceleration comes from the USGS National Seismic Hazard Model at the address, for a 2% chance of exceedance in 50 years on site class BC (Vs30 = 760 m/s). Instrumental intensity and potential damage follow the USGS ShakeMap PGA ranges (Wald et al. 1999).