The AI Feature API allows you to assess the impact of wind via the Wind Vulnerability and Wind Risk scores. The following sections provide details on the calculation of these scores and guidance on their interpretation.
Wind Vulnerability Score
The effect of the impact of wind is derived via the Wind Vulnerability Score, which is a 1-5 score that differentiates between buildings based on their relative susceptibility to resulting in a wind claim if exposed to wind events. A score of 1 represents the most vulnerable to wind damage, and a score of 5 represents the most resilient. The score is based on roof characteristics, structural conditions, and surrounding buildings.
The Wind Vulnerability Score is calculated in real time using a machine learning model trained on actual wind damage outcomes. For this reason, it is omitted from the API response by default and comes with a latency cost when requested on the API.
Quick info
Available for properties in: | USA, AU*, NZ* |
|---|
Available in APIs | AI Features API |
Available in Applications | Betterview |
Model Version | B.0 |
* Data is available in Australia and New Zealand where it assesses vulnerability to tropical cyclones (the regional term for hurricanes). However, the model was trained on US hurricane data
Wind Risk Score (claim predictor)
In addition to the vulnerability score, the API also provides a Wind Risk Score that combines vulnerability with FEMA-based hazard frequency from the FEMA National Risk Index model. This risk score, also known as a "claim predictor", measures the likelihood of experiencing a wind claim, where 1 is most likely to file a claim and 5 is least likely.
Quick info
Available for properties in: | USA |
|---|
Available in APIs | AI Features API |
Available in Applications | Betterview |
Model Version | B.0 |
Vulnerability vs Risk scores
It is important to distinguish between the Vulnerability and the Risk scores.
Vulnerability score
- Building-specific susceptibility to wind damage
- Based on building characteristics only (roof condition, materials, surroundings)
- Independent of location and hazard frequency
- Two identical buildings in different locations have the same vulnerability score
Risk score (Claim Predictor)
- Combines vulnerability with geographic hazard data (FEMA frequency)
- Predicts likelihood of experiencing a wind claim
- Location-dependent
- Two identical buildings in different locations can have different risk scores
Formula
Overall Risk = Hazard × Vulnerability
The vulnerability score provides the susceptibility component, while the risk score combines both susceptibility and hazard to predict claim likelihood.
Model input features
The Wind Vulnerability Score is calculated using a machine learning model trained on actual wind damage outcomes. The following features are used as inputs to the model.
Roof Characteristics
pitchedRoofPresent: Boolean, 1 if pitched roof detected, 0 otherwise. Pitched roofs (gable, hip) can have different wind resistance compared to flat roofs. metalPresent: Boolean, 1 if metal roof detected, 0 otherwise. Metal roofs typically have different wind performance characteristics than shingle roofs
Roof Condition Indicators
logStaining: Log-transformed roof staining ratio. Staining can indicate aging, water damage, or deterioration.compromisedShinglesPresent: Boolean, 1 if missing or damaged shingles detected. Missing or damaged shingles indicate compromised shingle roofing system that can increase wind vulnerability.logCompromisedStructure: Log-transformed structural damage indicators. Includes missing tiles/shingles, active ponding, rusting, structural damage, worn shingles. These factors collectively indicate structural weakness or deterioration.
Building & Surroundings
logFootprintArea: Log-transformed building footprint area. Larger buildings can have different wind exposure and damage patterns.zone2RoofPresent: Boolean, 1 if other roofs detected in defensible space zone 2: 5-30 ft. Nearby structures in the 5-30 ft zone can affect wind patterns and damage.zone2or3RoofPresent: Boolean, 1 if other roofs detected in defensible space zones 2 or 3: 5-100 ft). Structures in the wider 5-100 ft area indicate building density and potential wind effects.
The modelInputFeatures object in the response provides transparency into which specific factors influenced the vulnerability score for each building.
Geographic Applicability
The Wind Vulnerability Score is theoretically applicable to any building exposed to risk of wind damage. The model was trained on data from U.S. regions and validated on actual wind claim outcomes.
The FEMA-based Wind Risk Score utilizes the FEMA National Risk Index for hazard frequency data, which provides coverage across the United States. For properties outside the United States, the vulnerability score can still be calculated, but the risk score will not be available due to omission of FEMA hazard data.
Peril Applicability
The Wind Vulnerability Score correlates with wind damage from non-hurricane wind events. It applies to:
- Thunderstorm winds
- Tornadoes
- Severe weather wind events
- Straight-line winds
- Other non-tropical wind perils
It does not apply to:
- Hurricane wind damage (use Hurricane Vulnerability Score instead)
- Flood damage
- Hail damage (use Hail Scores instead)
- Other non-wind perils
The score is not expected or designed to perform optimally on hurricane wind damage claims. For tropical cyclone wind risk, use the Hurricane Vulnerability Score.
Regulatory Approvals
The Wind Vulnerability Score and associated rate factors are filed with U.S. regulatory authorities and approved for use in insurance rating in specific states. For detailed information on regulatory approvals, SERFF filing numbers, and approved rate factors, contact your Nearmap account representative.
Parcel Mode behavior
For information about the AI Feature API’s behavior when parcelMode=true is included in the request. For more information, see Parcel Mode.