Open Road Risk
  • Home
  • Start Here
  • Project
    • Project overview
    • Current model status
    • AI-assisted development
  • Literature
    • Literature overview
    • Literature evidence register
    • AI-assisted literature review
    • Literature-pipeline alignment
    • Crash frequency models
    • Exposure and traffic volume
    • Spatial methods and network risk
    • Junctions and conflict structure
    • Severity modelling
    • Validation and metrics
    • Transferability and open data limits
  • Data Sources
    • Overview
    • STATS19 Collisions
    • OS Open Roads
    • AADF Traffic Counts
    • WebTRIS Sensors
    • Network Model GDB
    • OS Terrain 50 (grade)
    • Deprivation (IoD 2025)
  • Methodology
    • Methodology Overview
    • Joining the Datasets
    • Feature Engineering
    • Empirical Bayes Shrinkage
  • Exploratory Data Analysis
    • Collision EDA
    • Collision-Exposure Behaviour
    • Vehicle Mix Analysis
    • Road Curvature
    • Months and Days of Week
    • Traffic Volume EDA
    • OSM Coverage
  • Models
    • Modelling Approach
    • Stage 1a: Traffic Volume
    • Stage 1b: Time-Zone Profiles
    • Stage 2: Collision Risk Model
    • Facility Family Split
    • Model Inventory
  • Investigations
    • Investigations overview
    • KSI atlas diagnostic
    • Staffordshire data quality
    • Temporal descriptors evaluation
    • AADF counted-only filter
    • Rank stability harness
    • Zero-calibration diagnostic
  • Outputs
    • Key figures
    • Top-risk map
    • QGIS GeoPackage (Kaggle)
  • Tools
    • ukgeo — UK Geocoder
  • Future Work

On this page

  • Sources
  • Features used
  • Join and coverage notes
  • Role in the pipeline

GB Deprivation Context

Within-country deprivation deciles from England IoD, Wales WIMD, and Scotland SIMD used as contextual predictors.

Open Road Risk uses small-area deprivation as a contextual predictor, not as a road attribute and not as a causal explanation. For GB coverage the pipeline keeps the three national indices separate, then exposes shared within-country decile columns to the model.

Sources

Country Source Geography Year Notes
England Indices of Deprivation (IoD) LSOA 2021 2025 MHCLG File 7 deciles
Wales Welsh Index of Multiple Deprivation (WIMD) LSOA 2011 2019 Ranks converted to within-Wales deciles
Scotland Scottish Index of Multiple Deprivation (SIMD) Data Zone 2011 2020 Overall decile from SIMD; income and employment ranks converted to within-Scotland deciles

The processed polygon layer is data/processed/context/deprivation_areas_gb.parquet. Link-level assignment is by road-link centroid within the deprivation polygon, with a short nearest fallback for boundary-edge cases. Assignment provenance is written to data/provenance/deprivation_gb_provenance.json.

Features used

Feature Meaning
deprivation_country Country/index used for the assigned area
deprivation_source IoD 2025, WIMD 2019, or SIMD 2020v2
deprivation_area_code LSOA/Data Zone code assigned to the road link
deprivation_area_type LSOA2021, LSOA2011, or DZ2011
overall_decile_within_country Overall deprivation decile within the source country
income_decile_within_country Income-domain decile within the source country
employment_decile_within_country Employment-domain decile within the source country

Deciles run from 1 (most deprived) to 10 (least deprived), but only within the source country/index. A Scottish decile 1, Welsh decile 1, and English decile 1 are not a single GB-wide most-deprived tenth.

Join and coverage notes

  • England uses IoD 2025 LSOA 2021 codes and the England/Wales OA 2021 boundary file dissolved to LSOA 2021.
  • Wales uses WIMD 2019 attributes and Welsh LSOA 2011 polygons.
  • Scotland uses SIMD 2020v2 attributes and 2011 Data Zone polygons.
  • deprivation_assignment_method records whether a link was assigned by polygon containment, nearest fallback, or left unmatched.

Role in the pipeline

These features sit alongside population density, rural/urban context, terrain grade, and road-network features in Stage 2. The first GB pass deliberately uses only country dummies plus overall, income, and employment deciles. Source-specific domains such as crime, housing, community safety, or access to services are left out until there is a deliberate modelling reason to add them.


Next: How the sources are joined

Open Road Risk

 

Built with Quarto