Methodology

How the 1–10 scores are assigned

Each hazard layer has its own breakpoints, fixed nationally rather than derived from the neighborhood around the address. A score is therefore comparable between properties and between states: it says how this location ranks against the whole country for that hazard, not how it ranks against its own street.

Scores roll up into categories by their worst member, so a single severe hazard is never averaged away by mild ones sharing its category. The report stops there. It does not reduce those categories to one grade for the property. Eighteen hazards measured in feet, degrees, decibels and chances-per-year do not average into a word, and the word would be the least informative thing on the page: a house is not "Moderate", it is exposed to particular things to particular degrees. What the report gives you instead is a ranking of the categories worth acting on at this address, and the evidence under each one.

Traffic and aircraft noise sits further out than that. It is read, scored and shown as its own category like everything else, but it is kept out of the ranked risks entirely, along with heating and cooling demand, because that list answers what can happen to the building and a loud or expensive address is not a more dangerous one. A property can read Very High for noise and have nothing in its risk list at all. There is a section on it further down.

On the report, categories are ordered by severity at that specific address, and those scoring Low or Very Low are collapsed into a drawer at the bottom. Nothing is removed. Every hazard remains on the page, and unscored hazards stay visible rather than being filed under "lower-rated", because a gap in the data is not the same as a low risk.

The underlying measurements are the same ones published by scores.degreeday.org, and the method for turning a measurement into a 1–10 score is shared with that site and enforced by tests rather than by convention. Wildfire and wind are graded differently here, on purpose.

Both used to be graded against the whole world. Parts of Africa and Australia burn every few years, and the Pacific sees stronger storms than the United States does, so those places set the top of the scale and every American address was pushed toward the bottom. It was not a small effect: Paradise, California, where a wildfire destroyed the town and killed 85 people in 2018, scored 1 out of 10.

For those two, this site grades American addresses against the United States instead. We read every US pixel of the source map and cut it into tenths, so each score step is a tenth of the country: a 5 is a middling American address, an 8 is worse than four fifths of it. The top score is held back for the worst 1%, so that a 10 still means something on a coast where a great many places are exposed. Every other hazard on this page is unchanged. Traffic and aircraft noise is graded neither way: it is scored on the decibel reading itself, for the reason set out below.

The consequence, plainly: for those two, the same address can read differently here than on the global site. Neither number is wrong. They answer different questions, one about the world and one about this country.

Site-context flags

Three checks shown beside the address are memberships, not scores: a property is either inside them or it is not, so they appear as labels rather than as a 1–10 value.

• Wildland–Urban Interface: read from a 100 m WUI mask. It marks where development meets or intermingles with wildland vegetation, which raises a structure's exposure beyond what the wildfire score alone conveys.
• Leveed area: from the USACE National Levee Database, named where the database names it. This one changes how to read the rest of the report: the flood scores here take no credit for the levee. They describe the water before any protection is accounted for. Levees can also be overtopped or fail, so some risk remains either way.
• Hurricane zone: within roughly 100 km (62 mi) of a historical tropical-cyclone track of Category 1 or stronger, from IBTrACS. It marks exposure to the storm type; the wind speed itself is scored separately under Wind.

Each is shown only where it applies, and none has a "no" state. The absence of a flag is not a clean bill of health. It means the check did not fire, which can also happen when a dataset cannot be read.

Methodology by hazard

Every hazard scored on a report, grouped as the report groups them. Each entry describes the dataset and the modeling behind that layer.

Flooding

Water reaching the property from rivers, the coast, or heavy rain.

River & Coastal Flood Water depth in a flood with a 1% chance in any year

Riverine flood depths are based on the JRC global river flood model, which combines ERA5-forced GloFAS hydrology with the two-dimensional LISFLOOD-FP hydraulic model, run on the MERIT-Hydro digital elevation model, to estimate inundation depth for the 100-year return period at approximately 90 m resolution. Coastal flooding is screened using a connectivity-based inundation model driven by the 100-year COAST-RP extreme water level, which combines ERA5-based extratropical storm surge, synthetic tropical-cyclone surge, and tides over the approximately 30 m DeltaDTM coastal terrain model. Mean dynamic topography is used to account for differences between the vertical reference frame of the coastal extreme water levels and the coastal terrain elevations, so that water levels and land elevations can be compared consistently. Areas are identified as potentially inundated where low-lying terrain is hydraulically connected to the ocean and the adjusted extreme water level exceeds the local ground elevation. The coastal screening includes a friction-based attenuation component to account for the reduction of inland water levels with distance and land-surface resistance, so the method is not a simple bathtub fill. This approach represents a static coastal inundation screen rather than a full dynamic hydrodynamic simulation. Wave setup, estimated as a fraction of the significant wave height, is added on wave-exposed open coasts identified by a fetch-based exposure classification and omitted in sheltered bays, estuaries, and fjords. Wave runup, overtopping, erosion, and local drainage interactions are not included. Both flood layers represent undefended hazard conditions; levees, seawalls, surge barriers, and other local flood-protection infrastructure are not explicitly modeled.

Why this matters: A depth above roughly 0.3 m (1 ft) at the building footprint usually means structural flood damage, not just a wet yard.

Elevation Above High Tide Ground height above the local high-tide line

The Coastal Exposure Index identifies low-lying coastal land that falls below selected extreme high-water levels and is hydraulically connected to the ocean. It is intended as a screen for chronic tidal exposure and sea-level-rise susceptibility, not as a storm-surge, wave, or event-based flood simulation. Water levels are referenced to the mean higher-high-water tidal datum, using NOAA VDatum in the United States and FES2022 tides elsewhere. Outside the United States, mean dynamic topography is used to align the modeled tidal water levels with the land-elevation reference frame before comparing water levels with terrain. Elevation data come from lidar-based DEMs in the United States and the approximately 30 m global DeltaDTM coastal terrain model outside the United States. A connectivity check removes low-lying terrain that is isolated from the coast, helping distinguish potentially exposed coastal areas from inland depressions or disconnected basins. The index does not model groundwater emergence, drainage interactions, land subsidence, shoreline change, or flood defenses.

Why this matters: Negative or near-zero values indicate land already at or below high tide. Expect nuisance flooding and rising insurance costs.

FEMA Flood Zone Official FEMA designation

FEMA flood zones are based on FEMA’s National Flood Hazard Layer (NFHL), the official regulatory dataset used for U.S. floodplain management and flood insurance. The card reports the mapped flood designation at the selected location, such as Zone AE, including Special Flood Hazard Areas associated with the 1% annual-chance flood and areas of 0.2% annual-chance flood hazard. Coverage is limited to the United States and depends on digitized NFHL availability. FEMA maps change over time, and some areas may be unmapped, incomplete, or updated after the data snapshot used here.

Why this matters: Any zone beginning with A or V means a mortgage lender requires flood insurance. Zone X is not a guarantee of safety: roughly a quarter of federal flood claims come from outside the high-risk zones.

Wildfire

Chance of a wildfire burning at this location.

Wildfire Chance of burning in any given year

This model produces global annual burn-probability estimates at 10 arc-second (~300 m) resolution from a tree-based gradient-boosting regressor trained on observed burned area worldwide (2003–2023, filtered to wildfire). Predictors span fire weather (Canadian FWI System indicators), climate, fuels and land cover (LULC, aboveground biomass), terrain (elevation, slope, TPI), and human/ignition factors (population, roads, lightning); relationships are learned globally and past fire location is excluded as a predictor. Over the United States the delivered layer substitutes USFS FSim, and a post-processing step then “oozes” wildland burn probability into adjacent wildland–urban interface pixels to represent structure-fire exposure.

Why this matters: Insurers increasingly price, or decline, on this. Worth pairing with a look at the defensible space around the house and what the roof is made of.

Wind

Wind, tornado and straight-line windstorm exposure.

Wind Strongest gust in a severe storm

Extreme wind hazard is represented by terrain-aware 3-second peak-gust return levels at 30 arc-second (~1 km) resolution. The global wind model combines separate tropical-cyclone and non-tropical-cyclone wind estimates, then applies terrain and surface-roughness adjustments to better represent local exposure.

The non-tropical-cyclone component is derived from ERA5 pressure-level winds, with tropical-cyclone days removed using historical track data. Wind profiles are vertically interpolated to represent near-surface conditions, annual maxima are extracted, and extreme-value statistics are used to estimate return levels. These estimates are refined to 30 arc-second (~1 km) resolution using elevation and surface-roughness exposure adjustments.

The tropical-cyclone component is based on CHAZ, the Columbia HAZard model, a statistical–dynamical tropical-cyclone model that generates large synthetic tropical-cyclone catalogs from large-scale environmental conditions. CHAZ represents tropical-cyclone activity through separate genesis, track, and intensity components, allowing coastal wind return levels to be estimated from many physically plausible storms rather than the limited observed record alone. In this workflow, CHAZ tropical-cyclone wind return levels are converted from 1-minute sustained winds to 3-second gusts using the WMO offshore conversion factor and then combined with the non-tropical-cyclone wind branch after surface-roughness and terrain-speedup adjustments.

Why this matters: Above about 110 mph, roof tie-downs, garage doors and window protection dominate whether a house survives intact.

Tornado Chance of a strong tornado within 25 miles per year

Tornado hazard is represented by a U.S. climatology of significant tornado occurrence based on NOAA Storm Prediction Center records. The metric is the mean number of days per decade with an EF2 or stronger tornado reported within 25 miles (40 km) of the location over 1986–2015. The 25-mile search radius accounts for spatial uncertainty in tornado tracks and provides a smoother local climatology, while the EF2+ threshold focuses on events most likely to cause substantial damage. This layer represents historical tornado occurrence, not a forecast or event-based tornado simulation. Coverage is limited to the United States.

Why this matters: The 25-mile radius reflects how narrow tornado tracks are. It measures the neighborhood’s climate, not the odds for this specific roof.

Widespread Thunderstorm Winds Widespread windstorms per year

Derecho hazard is represented by a U.S. climatology of long-lived, widespread straight-line windstorms produced by organized thunderstorm complexes. The layer is based on a 4 km observational climatology for 2004–2021 that uses machine-learning bow-echo detection and storm tracking to identify derecho events. Event wind-damage footprints are summed, annualized, and smoothed to produce a derecho-frequency surface expressed as events per year. This layer captures organized convective wind hazard that is not represented in the Extreme Wind layer. Coverage is limited to the United States east of the Rocky Mountains.

Why this matters: The classic cause of "the whole street lost its trees and power for a week".

Hail

Large hail, which damages roofs, siding and vehicles.

Large Hail Chance of 2-inch-plus hail per year

Large-hail hazard is represented by a global climatology of very large hail occurrence, defined as hailstones at least 5 cm (2 in) in diameter. The layer is based on the AR-CHaMo statistical hail model applied to ERA5 reanalysis for 1950–2023. AR-CHaMo estimates hail occurrence as the product of thunderstorm probability and the conditional probability of very large hail given a storm, using convective-environment predictors such as instability, vertical wind shear, and moisture. The model is calibrated against hail reports from Europe, the United States, and Australia. Because ERA5 does not explicitly simulate hailstones, occurrence is inferred from atmospheric environments favorable for very large hail rather than observed or modeled hail impacts at a specific site.

Why this matters: Hail is the single largest driver of roof-replacement claims in the central US, and a common reason for high deductibles.

Heat

Extreme summer temperature and humidity.

Extreme Heat Hottest day in a typical decade

Extreme heat is represented by the annual maximum dry-bulb air temperature at a specified annual-chance level. Quality-controlled NOAA GHCN station records are fit with a Generalized Extreme Value (GEV) distribution in a nonstationary Bayesian framework, allowing distribution parameters to vary with global mean temperature. The resulting station-level estimates are regionalized to a 30 arc-second (~1 km) global grid using machine-learning methods trained on high-resolution climatological predictors derived from reanalysis, terrain-aware downscaling, and global land-surface climatologies, together with terrain and climate-zone variables. This layer represents extreme outdoor air-temperature hazard, not indoor heat exposure or heat stress, which also depend on humidity, radiation, wind, building characteristics, acclimatization, and exposure duration.

Why this matters: Sets how much cooling the house needs, and in older housing stock, whether it stays habitable through a heat wave at all.

Heat + Humidity Stress Worst combination of heat and humidity

Human heat stress is represented by the annual maximum wet-bulb temperature at the 10% annual-chance level. Wet-bulb temperature combines air temperature and humidity into a single thermodynamic measure related to the body’s ability to cool through evaporation. It is computed from quality-controlled station air-temperature and humidity records. Extremes are fit with a Generalized Extreme Value (GEV) distribution in a non-stationary Bayesian framework, allowing distribution parameters to vary with global mean temperature, and are then regionalized to a 30 arc-second (~1 km) global grid using the same machine-learning framework applied to extreme heat and cold. This layer represents large-scale outdoor heat-humidity hazard; it does not account for direct solar radiation, wind speed, clothing, activity level, indoor conditions, acclimatization, or exposure duration.

Why this matters: Where this runs high, air conditioning is a safety system rather than a comfort, and a summer power cut becomes a health problem rather than an inconvenience.

Cold

Extreme winter temperature.

Extreme Cold Coldest night in a typical decade

Cold stress is represented by the annual minimum dry-bulb air temperature at the 10% annual-chance level, with lower temperatures indicating greater hazard. Quality-controlled station minimum-temperature records are fit with a Generalized Extreme Value (GEV) distribution in a nonstationary Bayesian framework, allowing distribution parameters to vary with global mean temperature. The resulting station-level estimates are regionalized to a 30 arc-second (~1 km) global grid using the same machine-learning framework applied to the heat metrics, with gridded predictors representing local climate, elevation, terrain, geographic location, and climate-zone context. This layer represents outdoor cold-temperature hazard; actual impacts depend on asset design, exposure duration, wind, snow/ice conditions, building characteristics, and operational tolerances.

Why this matters: Sets the bar for pipe-freeze protection, insulation, and back-up heat.

Earthquake

Ground movement: earthquakes and landslides.

Earthquake Ground shaking in a rare, severe earthquake

Seismic hazard is represented by peak ground acceleration (PGA) with a 10% probability of exceedance in 50 years, equivalent to an approximately 475-year return period or about a 0.2% annual chance. Values represent firm-rock conditions and are provided on an approximately 1 km global raster. The layer combines the best available regional probabilistic seismic-hazard models, including USGS, ESHM20, GEM, and other regional sources, with GSHAP used as background coverage where higher-resolution regional models are unavailable. Where models overlap, the maximum PGA value is used. The layer does not include local soil amplification, liquefaction, earthquake-triggered landslides, tsunami hazard, or other secondary effects. Boundaries or discontinuities may appear where independently developed regional models meet.

Why this matters: Above about 0.2 g, foundation bolting and cripple-wall bracing are the retrofits that matter most for older houses.

Landslide How prone this terrain is to slipping

Landslide hazard is represented by a relative susceptibility index that reflects terrain conditions and rainfall-triggering potential. Outside the United States, the layer uses a global rainfall-driven landslide model that combines ERA5 daily rainfall with five susceptibility classes derived from factors such as slope, lithology, vegetation, and soil moisture to estimate landslide-initiation potential at approximately 90 m resolution. Within the United States, the layer uses USGS slope–relief susceptibility models trained on more than 600,000 documented landslides and 10 m elevation data. The outputs represent relative susceptibility and screening-level landslide potential, not event-specific probability, runout extent, depth, velocity, or site-scale slope stability.

Why this matters: On sloped ground, elevated values are the point at which a geotechnical report stops being optional. An engineer reads the slope itself rather than the map of it.

Air Quality

How often the air here is unhealthy to breathe.

Air Quality Unhealthy-air days a year, with the worst days counted more

A long-term air-quality health rating on a 1–10 scale, published from US Environmental Protection Agency monitor records: the AirNow combined AQI grid, with AQS county and monitor data and a preliminary AirNow gap-fill used for Hawaii. The period is 2017 to 2025, and 2012 to 2025 for Hawaii, which qualifies only where observations exist in every year and cover at least 80% of days. The rating counts days above AQI 100, weighted by severity: AQI 101–150 counts once, 151–200 twice, 201–300 four times, and 301 or above eight times. Those weighted days are mapped onto the scale as 1 + 9 × log2(1 + weighted days) / log2(129), clipped to 1–10, so a rating of 10 corresponds to about 128 weighted days a year and a rating of 1 to none. Because the rating is a published score rather than a raw measurement, this site does not rescale it; the number shown on the card is the weighted day count recovered by inverting that formula. The grid is about 2.5 km across. Alaska, Puerto Rico and the US Virgin Islands are outside the source domain, and parts of the interior have no monitor near enough to rate at all.

Why this matters: This is the hazard on the page that acts on the people rather than the building, every year, without a storm to announce it. It matters most for children, older people and anyone with asthma or a heart condition, and it is the one a good filter and a well-sealed house can genuinely reduce.

Heating and Cooling Costs

How much heating and cooling this location needs.

Cooling Demand How much cooling this climate needs

Cooling demand is represented by annual Cooling Degree Days (CDD), the cumulative amount by which daily mean temperature exceeds the 18.3 °C (65 °F) balance point. CDD is computed from a 30 arc-second (~1 km) daily temperature climatology derived from reanalysis using terrain-aware statistical downscaling for 2004–2023. The metric approximates temperature-driven cooling energy demand, but does not account for humidity, solar gains, wind, building characteristics, occupancy, equipment loads, or cooling-system efficiency.

Why this matters: This is how much cooling the climate asks for over a whole year, not how hot it gets on the worst day. Extreme Heat answers that. Compare the estimate with your recent utility bills; if you are considering the property, ask for the previous 12 months.

Heating Demand How much heating this climate needs

Heating demand is represented by annual Heating Degree Days (HDD), the cumulative amount by which daily mean temperature falls below the 18.3 °C (65 °F) balance point. HDD is computed from a 30 arc-second (~1 km) daily temperature climatology derived from reanalysis using terrain-aware statistical downscaling for 2004–2023. The metric approximates temperature-driven heating energy demand, but does not account for wind, solar gains, building characteristics, occupancy, internal heat gains, heating-system efficiency, fuel type, or behavioral responses.

Why this matters: Compare this estimate with your recent utility bills. If you are considering the property, ask for the previous 12 months. This measures the climate, not the building’s efficiency.

Noise

How loud it is outside: background sound, road traffic and aircraft.

Outdoor Noise Average outdoor sound level in the daytime

Approximate daytime outdoor sound level (A-weighted decibels) at 30 m resolution across the contiguous United States, built as an energy sum of three layers. The first is a modelled ambient background: a gradient-boosted regressor over 90 predictors trained on 592 measured sites, which supplies the level of ground with no mapped transport source near it. The second is the Bureau of Transportation Statistics 2022 national transportation-noise surface, a 24-hour equivalent level for road and civil aviation. The third is military aviation: published AICUZ and ALUCP contours where an airfield has them (41 airfields) and a modelled convex-decay estimate elsewhere (122 airfields), partitioned so the two never double-count. THE THREE ARE NOT THE SAME METRIC. A daytime median background, a 24-hour equivalent level and a day-night contour with its night penalty removed are summed as sound energy, so the result is an approximate daytime exposure level rather than any standard index, and it should not be compared directly against a regulatory threshold stated in DNL or Lden. The 1-10 scale is cut against the World Health Organization's 1999 daytime guidance for outdoor living areas, which is stated in the same family of measurement: 50 dB for moderate annoyance and 55 dB for serious annoyance, where the High rating begins. Isolated and compact high cells are filtered before publication, and cells above 85 dBA are held to that level unless an aviation source is mapped within about a kilometre. The two modelled components carry roughly 5 decibels of held-out error, wider than the width of a single band, so a reading is best understood as the neighbourhood around an address rather than the address itself. Coverage is the contiguous United States; Alaska, Hawaii and the territories are not mapped. Sources other than transport, aviation and the modelled background — neighbours, industry, nightlife, construction — are not represented at all.

Why this matters: Noise is the complaint people make about a house they already bought. It is also the one figure on this page you can check yourself: stand outside at rush hour, and again after dark.

Large hail reported near here

Under the Hail card is the record: the days somebody near this address measured hail two inches across or bigger, with how big it was and how close the nearest report came. Of the four history blocks on this report it is the one most likely to describe something the owner has already paid for, because hail is what takes roofs off across the middle of the country.

The source is NOAA's Storm Prediction Center severe weather database, 1955 to last season, the same record behind the tornado block. Two inches is the Hail card's own threshold, so the list is the card's hazard itemised rather than a different question; it is also roughly the size at which stones start breaking roofs instead of stripping leaves, and it takes the record from 402,454 reports down to 28,360.

One row per day, not per report. A single hailstorm is reported over and over along its path: Dallas has 81 reports within 15 miles spread across 51 separate days. Each day is collapsed to one line carrying the largest stone reported nearby, how many reports there were, and how close the nearest one came. The largest stone and the nearest report are often different reports, so the row answers both questions rather than taking both numbers from whichever one it saw first.

The thing to understand about this record, and it matters more here than anywhere else on the page. These are reports, not measurements of where hail fell. A stone is in the database because a person saw it, sized it against a coin or a ball, and told the weather service. So the record is dense where people live and thin where they do not, and it thickens steeply through time as reporting improved: across all sizes, 2,214 reports in the 1950s against 137,255 in the 2000s, which is a sixtyfold change in reporting rather than in weather. Screening to two inches flattens most of that, because a stone that size gets noticed whoever is standing under it, but nothing removes the population effect. Read it in one direction only: hail listed here did fall, and hail absent from a rural or an old record may simply have had nobody to report it.

Distances are approximate, and rounded to whole miles for a reason. A report is filed against a place, not a point: every report before 2000 is recorded to two decimal places of latitude and longitude or coarser, which is about three quarters of a mile on its own, and many are simply geocoded to the nearest town. One coordinate in this record, Fort Worth's city point, carries fourteen separate reports. So the block rounds, says "about", and never claims hail fell on the house itself — which is the one thing the fire, hurricane and tornado blocks can each say and this one cannot.

Sizes are given both ways, as inches and as the object the weather service asks observers to compare against — hen egg, tennis ball, baseball, grapefruit, softball. That is not decoration: the recorded sizes cluster on those objects precisely because that is what people measure against.

Where nothing qualifies, the block does not appear, the same as the other three.

Damaging thunderstorm winds reported near here

The Wind card scores a modelled peak gust and the hurricane record above covers tropical systems. Neither describes the thing that actually takes roofs off across the interior of the country, which is a summer thunderstorm downburst. Kansas City has had sixty days of 75 mph gusts reported inside fifteen miles and not one hurricane. This block is that record, from the third file in the same Storm Prediction Center database as the tornadoes and the hail, 1955 to last season.

The threshold is 75 mph, the National Weather Service's "significant severe" level and about where thunderstorm wind stops taking branches down and starts taking roofs off. The plain severe criterion, 58 mph, returns 437,803 reports and would bury the reader in ordinary weather; this returns 37,335.

Most of these speeds are estimates, and each row says which. This is the thing to understand about the wind record, more than any other on the page. Of the reports at this strength, only 11% were measured by an instrument. About half are explicitly flagged as estimated, usually inferred from what the wind broke, and estimates made that way are known to run high; the rest carry no flag at all. Rather than average an anemometer reading and somebody's judgement into one number, the provenance travels with every report to the table: each row says whether its own speed was measured, estimated, or filed with no source stated.

As with hail, this is one row per day rather than per report, because a squall line is reported many times along its path; distances are approximate and rounded to whole miles, because a report is filed against a place rather than a point; and the same reporting bias applies, 519 reports at this strength in the 1950s against 10,576 in the 2010s, which is a change in reporting rather than in weather. Where nothing qualifies, the block does not appear.

Tornadoes that have touched down near here

The Tornado card scores a climatology: the mean number of days per decade with an EF2 or stronger tornado within 25 miles. That is a rate, and a rate is not a memory. Under the card is the same hazard itemised — every significant tornado that has touched down within those same 25 miles since 1950, with its rating, its path, and how close it came. The radius matches the card's on purpose, so the list is the card's number spelled out rather than a second, differently drawn answer.

It is not a second score, and the same warning applies here as to the fire and hurricane records, only harder: tornadoes do not repeat on a schedule and do not avoid places they have already hit. A street that took an F4 in 1974 is neither owed another nor spent.

The source is NOAA's Storm Prediction Center severe weather database, the national tornado record, which begins in 1950.

Only EF2 and stronger are listed. Three reasons, and they agree: it is the threshold the card's own metric uses, it is the meteorological definition of a "significant" tornado, and without it the list would mislead. The database holds 73,458 tornadoes and 58,508 of them are EF0 or EF1 — and the number of those recorded rose steeply through the 1990s because spotter networks, radar and storm chasing improved, not because more tornadoes happened. A list including them would show a trend that is not there. Tornadoes the survey never rated are not listed either, for the same reason a storm with no measured intensity is absent from the hurricane block. And where nothing qualifies, the block does not appear at all: no significant tornado within 25 miles in seventy-five years describes a place tornadoes do not reach, and the Tornado card above already says so with its score.

Distance is to the centre of the path, and width decides whether it hit. A tornado is not a line. The 2011 Tuscaloosa tornado was a mile and a half wide, so a house a quarter of a mile off its centre line was inside it. Rather than pick a round number for "close enough", the report compares the distance against half the path width the survey actually recorded, and says the tornado crossed this location only when the address falls inside it. About 30% of significant tornadoes have no recorded end point, only a touchdown; for those the distance is to that point, which is all anyone measured.

One wrinkle in the ratings. The Fujita scale was replaced by the Enhanced Fujita scale in February 2007, and the database stores both in one column. Each tornado is shown on the scale that was in force when it happened, so a 1974 tornado reads F5 and a 2011 one EF5. They are not quite the same measurement: the EF scale ties wind estimates to specific damage indicators and generally produces slightly lower wind numbers for the same damage.

Earthquakes that have shaken here

Under the Earthquake card is the record: the earthquakes whose shaking actually reached this address, with how big each one was, how far off, and how hard the ground is estimated to have moved here. The card above it is a modelled peak ground acceleration at a 1-in-500-year return period, on firm rock, in units of gravity. It is the least checkable number on this report, and this block is what a reader can check it against.

The source is the USGS ComCat catalogue, filtered to events the USGS classes as earthquakes — the catalogue also carries quarry blasts, mine collapses and several hundred nuclear tests, and none of those belong in a house's shaking history. Magnitude 5.0 and up. Below that the record stops describing earthquakes and starts describing seismometers: there are 20,305 U.S. events at magnitude 4.5 and up against 7,346 at 5.0, and nearly all of the difference is instruments installed since 1970.

An earthquake has to pass two screens to be listed. It must have happened within 60 miles of the address, the same radius the hurricane block uses and the widest on this report, which is the right end of the range for the hazard that travels furthest. And the ground under the address must be estimated to have reached intensity V on the Modified Mercalli scale: felt by nearly everyone, some windows and dishes broken, unstable objects overturned. That is the first rung at which an earthquake stops being a curiosity and becomes an event in the life of a building.

The list runs most recent first, unlike the other five, which run nearest first. These events span four centuries and the useful reading of an earthquake record is chronological: when was this address last shaken, and how often. Ten are shown, so on a long record the oldest fall off — and because the biggest earthquake is often also the oldest, the sentence above the table names the strongest separately, whether or not the table still has room for it.

What the 60 mile screen leaves out, because it is not nothing. Earthquakes are less local than the other hazards on this report: a magnitude 5 is barely felt at forty miles, while the 1811–12 New Madrid earthquakes were felt in Boston, eight hundred miles away. Measured against the catalogue, a 60 mile screen removes the 1700 Cascadia earthquake from Seattle and Portland and the 1964 Prince William Sound earthquake from Anchorage, and it empties the block altogether for Washington DC and Honolulu, each of which has felt a damaging earthquake centred about twenty miles beyond the line. It is close-run the other way too: the 1811 New Madrid earthquake sits 59 miles from Memphis and stays. This block answers "what has happened near here", which is the question the rest of the report answers, and not "what has this address ever felt".

The shaking figure is estimated, and it is the only modelled number in any of the six history blocks. It comes from Atkinson and Wald (2007), the intensity prediction equation the USGS itself uses, applied with the event's magnitude, depth and distance. Two coefficient sets exist and the choice matters: seismic waves travel far better through the old, cold crust east of the Rockies than through the fractured crust of the west, which is why New Madrid was felt in Boston and a Californian earthquake of the same size is a local event. Addresses east of longitude 105°W get the stable-continent set; everywhere else, including Alaska, Hawaii and the territories, gets the active-crust set.

What that estimate is worth. Checked against events with known outcomes, it puts the 1964 Prince William Sound earthquake at intensity VIII in Anchorage, which is what Anchorage got; the 2011 Mineral earthquake at V in Washington DC, which is what the USGS collected there from residents; and the 1906 earthquake off the top of the scale in San Francisco. It is still a point estimate on firm ground. It carries no site soils, and soft ground, fill or a basin can add a whole intensity unit; it also tends to run low very close to a large rupture. Every row that shows one says estimated beside it. The magnitude, the date and the distance on the same row are catalogue facts.

The record has no single starting year, which is why the sentence above the table does not claim one. It is instrumental from about 1900 and anecdotal before that, and how far back the anecdotes reach depends entirely on who was present to write them down: 1568 in New Mexico, 1638 in New England, the 1850s across much of the west, and later than that in Alaska. So the deep record is complete for large earthquakes in settled places and empty everywhere else, and a short list from a thinly settled area is not evidence of a quiet past. Distances are given to the nearest mile, which is as much as epicentres from before about 1930 support, and magnitude 5.0 is the floor: below it the catalogue is measuring instrument coverage rather than ground motion.

How the history lists are cut

Six blocks on this report list what has actually happened near an address rather than what a model expects: fires under Wildfire, hurricanes, thunderstorm winds and tornadoes under Wind, hail under Hail, and earthquakes under Earthquake. They share a shape, and these rules apply to all of them.

• Ten shown, best first. Five of the blocks put the nearest event first. The earthquake block puts the most recent first instead, because an earthquake record is read chronologically. Ten is a length somebody reads; the counts behind it often are not. The limit is on the display only — everything found is counted, the sentence above each table reports the true total, and the line under it says how many more there are.
• They are evidence, never a score. Nothing in these lists moves the 1–10 rating or the overall risk. A place that has burned, flooded or been hit is not thereby more or less likely to be hit again; that judgement belongs to the cards above, which are modelled.
• An empty record shows nothing at all. Where no qualifying event is on the record, the block does not appear rather than printing a sentence about a hazard that does not reach the address.
• Each list has a size threshold, and each one is stated below. They exist because small events are both far more numerous and far less consistently recorded, especially further back, so an unscreened list would show a trend in reporting and read as a trend in weather.
• Absence is weaker evidence than presence. An event on these lists happened. A short list means nobody recorded one, which in a rural area or an early decade is not the same thing as nothing having happened.

Hurricanes and typhoons that have passed near here

Under the wind cards is the track record: every hurricane whose centre passed within 60 miles of the address since 1920, with how close it came, how strong it was at that moment, and when. It is there for the same reason the fire record is. The Wind card scores a modelled peak gust, which a reader cannot check against anything they know; "Andrew passed 3.5 miles from here as a Category 5 in August 1992" is a fact about a real afternoon.

It is not a second score. The record never touches the 1–10 number or the overall risk. A coast that took three hurricanes in the 1960s and none since is not therefore safe, and forty quiet years is not a forecast.

The source is IBTrACS, NOAA's international best-track archive: every forecast agency's post-season reanalysis of every tropical cyclone, merged into one record. Unlike the flood and fire layers there is no second source to reconcile, because this is the one everybody else reconciles to.

The strength shown is the strength here, not the storm's headline. A hurricane's intensity at your address is often far below what it reached at sea or at landfall two states away, and reporting the peak next to a distance from this house would overstate what happened, most of all for the addresses furthest from the damage. So the number in the table is the intensity as the centre passed, interpolated along the track. By the same rule, a storm that had already weakened below hurricane force by the time it got here is not listed: it did not pass as a hurricane, whatever it had been.

Only hurricanes and typhoons are listed. Tropical storms outnumber them roughly two to one — New Orleans has had 27 pass within 60 miles against 11 hurricanes — and listing them buried the storms people actually remember under a list of weather. The threshold is Saffir-Simpson Category 1, which is sustained winds of 74 mph. One consequence is worth stating plainly: a storm the archive never graded cannot clear that bar either, so it is absent. That costs nothing in the Atlantic, the Gulf, Puerto Rico or Hawaii, where every position since 1920 carries a measured wind, but it thins the western Pacific record before about 1940, and the note under the table says so.

The block does not appear at all where nothing has passed. An address with no hurricane on its record in a century is not in a hurricane zone, and a paragraph saying so would be about a hazard that does not reach it.

Why 60 miles, when the centre is a point. A hurricane is not its eye. Hurricane-force winds routinely reach 25 to 50 miles from the centre and tropical-storm-force winds two or three times that, so a track that missed by 60 miles was still a storm that happened to this address. Distances are to the storm's centre; best-track positions are filed to a tenth of a degree, so treat anything under a mile as "the eye came over here" rather than as a survey.

In Guam and the Northern Marianas these storms are called typhoons, and the report uses that word there. The name follows the ocean rather than the storm, so the wording is chosen from the address, not from where the storm formed. Hawaii keeps "hurricane", which is what the central Pacific uses.

Why the record starts in 1920. IBTrACS reaches back to 1842 in the Atlantic, but the early tracks are positions reconstructed from ship reports and landfall accounts, with no aircraft and no satellite, and their intensities are estimates of estimates. Even from 1920 the coverage is uneven: the Atlantic has a measured intensity for essentially every position, because the Hurricane Center reanalysed the whole record, while the western Pacific — which is Guam and the Northern Marianas — has no recorded intensity at all before about 1940. An unmeasured storm is not a weak one, but it is not a proven typhoon either, so those tracks sit outside a list headed "typhoons" rather than inside it carrying a blank.

Fires that have burned near here

Under the Wildfire card is a list of the large fires that have actually burned near the address, with the date, the size, and how far the fire came from the pin. It is there because the score above it is a modelled annual probability, and no reader can check one of those against anything they know. The record can be checked. A rating of 6 beside "this location burned in the Eaton Fire, January 2025" is a different sentence from a rating of 6 beside "nothing over 500 acres has burned within 25 miles since 1984", and the second is as much of an answer as the first.

It is not a second score. History is evidence, not a rating, and it is deliberately kept out of the 1–10 number and out of the overall risk. A place that has burned twice is not therefore twice as likely to burn again, and a clean record is not safety: in much of the West, the fuel that has not burned in forty years is exactly the fuel that carries the next fire.

Two sources, because neither covers the whole period. MTBS, the USGS and Forest Service burn-severity record, is the authoritative mapped archive back to 1984, but a fire is only added once its severity has been assessed from post-fire imagery, which runs a year or two behind. NIFC's interagency perimeters are published while a fire is still burning, and fill that gap. Where both hold the same fire, MTBS's assessed perimeter is the one kept.

Only fires of 500 acres and larger appear, and controlled burns are excluded. The floor is not arbitrary: it is MTBS's own mapping threshold, which is 1,000 acres in the West and 500 in the East, so nothing smaller is on the national record to show. That makes a quiet entry mean something specific. It means no large fire has burned nearby on the national record; it does not mean nothing has burned, and a fire that took houses off one street without reaching 500 acres will not be here. Prescribed burns are dropped because a controlled burn is fuel treatment, and listing one would read as "this area burns", which is close to the opposite of what it means.

A fire marked type not recorded was mapped without an ignition type, so it may have been a controlled burn rather than a wildfire. Those are kept rather than dropped because they are 4,311 of the records and 2,075 of them predate 2000: removing them would put a hole in a third of the early record.

Where nothing qualifies, the block does not appear at all rather than printing a sentence about a hazard that does not reach the address. The Wildfire card above still carries the score, which is where a low reading belongs.

Distance is measured to the edge of the burn, not to the middle of it, and a distance of zero means the perimeter encloses the pin: that spot is inside a mapped burn scar. Perimeters are simplified to about 50 m to keep the lookup fast, so treat a distance as accurate to roughly a twentieth of a mile, and a property right on a fire's edge as borderline rather than decided. The search runs to 25 miles, which is wide enough to describe the fire regime around an address rather than one lot.

Air quality

This is the only hazard on the report that acts on the people rather than the building, every year, with no storm to announce it. It is a health risk, so unlike heating costs or traffic noise it does count toward the biggest risks at an address, and it can appear in that list on its own.

The score is not ours. Every other number on this page is a measurement we convert to a 1–10 score. This one arrives already scored, published from US Environmental Protection Agency monitor records for 2017 to 2025, and 2012 to 2025 in Hawaii. There was no scale for us to cut and we did not invent one; rounding the published rating is the whole of the scoring.

What it counts is days the air went above the EPA's unhealthy line, weighted by how bad they were: a day that is unhealthy only for sensitive groups counts once, a fully unhealthy day twice, a very unhealthy day four times, and a hazardous day eight. The number on the card is that weighted count, recovered from the formula the publisher documents, because "about 15 days a year" is a fact a reader can hold and "a rating of 6.15" is not. It is a long-term average of what monitors actually recorded, not a forecast and not today's reading.

Two limits worth knowing. The grid is about 2.5 km across, so this describes the air over a neighborhood rather than over one lot. And it is built from monitors, so where there is no monitor near enough there is no rating: Alaska, Puerto Rico and the US Virgin Islands are outside the source entirely, and parts of the interior plains and desert have real gaps. Those addresses are shown as having no monitor nearby, never as having clean air. About 3% of US properties fall in that state.

Where the country lands, counted across all 170 million parcels rather than sampled: roughly a quarter come out Very Low and just under half Low, a quarter Moderate, about one in twenty-seven High, and one in eighty Very High. The worst readings in the country are in California's San Joaquin Valley, where Bakersfield rates 9 on about 64 weighted unhealthy days a year; Honolulu rates 1.

Traffic and aircraft noise

This one is not a hazard, and it is left out of the ranked risks. Nothing about a loud street can damage a building. It is on the report because it is one of the first things anyone asks about a house, and one of the few things a listing will never tell you.

The number is the average outdoor sound level across a full day, in decibels, from road traffic and from civil and military aviation. It covers the contiguous United States. Alaska, Hawaii and the territories are not mapped, and an address in one of them is shown as not applicable rather than as quiet.

It is a neighborhood reading, not a doorstep one. The grid is about a kilometer across and deliberately smoothed, and the people who publish it say plainly that it is meant for regional screening rather than for judging a single site. It will tell you whether this part of town is quiet or loud. It will not tell you which side of the house faces the road, whether a fence, a hill or a row of trees sits in between, or how well the windows hold sound out. Your own ears on a visit are the better instrument, which is why that is the step the card recommends.

Only transportation is counted. Neighbors, bars, schools, construction, industry, dogs and lawn equipment are not in this map at all, so a low score is not a promise of quiet.

The 1–10 scale here is set by the decibel reading itself, not by how this address ranks against others. Every other hazard on this report is ranked, because nobody has a feel for what a 0.4% annual chance of wildfire means and a ranking gives it one. Decibels are the opposite case. People already have a sense of them, and a ranking fights that sense: 62 dBA is an ordinary busy-street reading, and grading it Very High because most American homes happen to be quieter is true and useless. This page said otherwise until we looked at what the labels were doing to real readings, and the scale was rebuilt.

The bands follow published residential guidance instead. Below about 52 dBA is Very Low, the level under which the World Health Organization's 2018 review finds road traffic has little effect. High begins around 64, because US federal housing and aviation guidance treats roughly 65 as normally unacceptable for new homes and the federal highway criterion for residences is 67. Very High starts at 70, approaching the level that same housing guidance calls outright unacceptable. Those guidelines are written in metrics that add a penalty for night-time noise, while this layer is a plain 24-hour average, so lining them up is approximate; the mismatch errs toward calling a place quieter rather than louder.

Where that puts the country: we scored every parcel in the national parcel record, all 170 million of them, rather than a sample. About one in six comes out Very Low, three in ten Low, four in ten Moderate, one in seven High, and roughly one in five hundred Very High. Counting only the parcels that record a year built, the closest the data comes to "a building stands here", it shifts louder still: nearly half are Moderate and one in six is High. So a Moderate here is genuinely the middle of the country, and a High is a real minority rather than a rarity.

Property outlines and imagery

The lot line drawn on the aerial comes from a national parcel snapshot held by this site, with live county and state assessor services as a fallback where the snapshot cannot answer. Parcel data is a record of the legal lot, not a survey: boundaries carry the source assessor's accuracy, and a geocoded point can land on a driveway, a lot centroid, or the wrong side of a building. Owner names and mailing addresses are never shown.

Aerial imagery is Esri World Imagery. Street-level photography is Google Street View, and the report opens on it instead of the aerial when tree canopy covers the roof, measured from the aerial itself, so a house hidden under a mature canopy still gets a usable photo.

Data, time periods and uncertainty

These layers describe current average hazard conditions from the best recent data available. The exact period differs by hazard, because each relies on different sources, observation windows and methods: some use recent climate records, some use current observations, and some estimate rare events from many years of data.

Using recent data does not assume a static climate. It reflects that recent observations remain the strongest starting point for describing hazard conditions today. Climate change affects how often hazards occur, how severe they are and where, but not uniformly across hazards, places or periods, and that remains an active area of research.

Many values are modeled estimates rather than direct measurements, describing how often a rare event is expected or a long-term average. They are uncertain because observations are limited, extreme events are rare, models make assumptions, and national datasets cannot capture every local detail. That uncertainty grows the further ahead you look, so these scores should not be read as predictions of distant future conditions. Long-term decisions deserve analysis specific to the hazard and to the decision being made.

What this is not