Geocoding Layers

Understand data layers used in geocoding operations.

Overview

Geocoding is the process of enriching a location exposure with geolocation data that precisely identifies the global coordinates of a site.

Geocoding is a prerequisite for all catastrophe modelling processes. A location must be enriched with precise geolocation data before it can be modelled. The global coordinates define a spatial reference system that can be recognized by a model.

Data layer versions

A geocode data layer is a geographic database that contains location data: street addresses, counties, postal codes, and global coordinates. During geocoding, a geocoding engine uses the available location data to "look up" information about a location exposure in the geocode layer. The geocoding engine assigns a latitude-longitude coordinate pair to the location, validates or standardizes address elements (e.g. postal codes, cities), and appends or "backfills" missing location information.

The Geohaz API and Bulk Geohaz API expose operations that enable client applications to geocode locations. These operations require that the request specifies the data layer name, model type, and model version.

The table identifies geocoding layers supported by the Geohaz API and Bulk Geohaz API:

Data LayerversionengineTypeBulk Geohaz engineType
geocode18.0RL0
geocode18.1HD1
geocode18.1RL0
geocode21.0HD1
geocode21.0RL0
geocode22.0HD1
geocode22.0RL0
geocode23.0RL or HD0 or 1
geocode24.0RL or HD0 or 1
geocode25.0RL or HD0 or 1

Every request to geocode an exposure must specify three parameters: the data layer (geocode), the version, the engine type, which corresponds to the engine used to model the account, or portfolio exposure.

📷

Note

Valid values for the engineType parameter differ between the Create Geohaz Job and Create Bulk Geohaz Job operations.

  • The RL (or 0) engine type supports geocoding of exposures modeled using detailed loss models (DLM).
  • The HD (or 1) engine type supports geocoding of exposures modeled using High Definition models (HD).