Skip to main content

Data and Methods

Overview

This page documents the data sources, processing methods, and analytical approaches used in the Safer Parks project. Our analysis combines multiple geospatial datasets to provide comprehensive insights into perception of park safety in the Bradford Local Authority District area.

Project Background

The Safer Parks project develops geospatial analysis tools and methodologies to assess park safety, focusing on factors that contribute to the perceived and actual safety of public green spaces. The project provides both analytical tools and data visualizations to support evidence-based decision making for park management and safety improvements.

Scaling Gender-Responsive Park Safety Insights with Open GIS by Fran Pontin

Data Sources

Primary Datasets

Parks and Green Spaces Data

Ordnance Survey Open Greenspace
  • Source: Ordnance Survey
  • License: Open Government Licence v3.0
  • Description: Comprehensive dataset of publicly accessible green spaces including parks, recreation grounds, playing fields, and other green areas
  • Processing:
    • Filtered to Bradford District boundaries
    • Classified by green space type (parks, playing fields, recreation grounds, etc.)
    • Merged with additional attribute data from local authority sources
  • Fields Used: Green space boundaries, names, types, access information
OpenStreetMap (OSM) Parks Data
  • Source: OpenStreetMap
  • License: Open Database License (ODbL)
  • Description: Community-contributed geographic data including parks, recreational areas, and associated facilities
  • Processing:
    • Extracted using Overpass API queries
    • Filtered for leisure=park, recreation_ground, and related tags
    • Quality checked and validated against OS data
  • Fields Used: Park boundaries, names, amenity types, recreational facilities
Bradford District Parks Information
  • Source: Bradford Council Parks Services and Bradford District Parks website
  • License: Local Authority Open Data (terms vary)
  • Description: Detailed information about park facilities, opening hours, accessibility features, and amenities
  • Processing:
    • Manual data collection and verification
    • Geocoded park locations where coordinates not available
    • Standardized facility categories and attributes
  • Fields Used: Park facilities, opening hours, accessibility information, contact details

Administrative Boundaries

Local Authority Districts
  • Source: ONS Geography Portal
  • License: Open Government Licence v3.0
  • Description: Official administrative boundaries for Bradford District
  • Processing: Clipped and simplified for web visualization
  • Usage: Geographic filtering and contextual mapping

Ward Boundaries - Source: ONS Geography Portal - License: Open Government Licence v3.0 - Description: Electoral ward boundaries within Bradford District - Processing: Simplified geometry for performance optimization - Usage: Local area analysis and demographic contextualization

Transportation and Access

Public Rights of Way
  • Source: Ordnance Survey Open Roads and Bradford Council
  • License: Open Government Licence v3.0
  • Description: Public footpaths, bridleways, and cycle paths
  • Processing:
    • Filtered for pedestrian and cycle access routes
    • Intersection analysis with park boundaries
    • Network analysis for accessibility assessment
  • Usage: Park accessibility and connectivity analysis
Park Entrance Points
  • Source: Calculated from OSM data and geometric analysis
  • License: Derived from ODbL-licensed OSM data
  • Description: Estimated park entrance locations based on path intersections and OSM tags
  • Processing:
    • Geometric intersection of public paths with park boundaries
    • Validation against OSM entrance tags where available
    • Manual verification for major parks
  • Usage: Access point mapping and pedestrian route analysis

Points of Interest and “Eyes on the Park”

Commercial and Social Facilities
  • Source: OpenStreetMap
  • License: Open Database License (ODbL)
  • Description: Businesses and facilities that provide natural surveillance of park areas
  • Categories Included:
    • Bars and pubs (amenity=bar, amenity=pub)
    • Social facilities (amenity=social_facility)
    • Newsagents and convenience stores
    • Restaurants and cafes near park areas
  • Processing:
    • Extracted using Overpass API queries
    • Buffered around park boundaries to identify relevant facilities
    • Categorized by business type and operating hours where available
  • Usage: Natural surveillance and activity level assessment
Residential Areas
  • Source: Ordnance Survey Open Map Local and OSM
  • License: OGL v3.0 (OS) and ODbL (OSM)
  • Description: Residential land use areas providing potential overlooking of parks
  • Processing: Land use classification and proximity analysis to parks
  • Usage: Natural surveillance assessment

Crime Data

West Yorkshire Police Crime Statistics
  • Source: Police.uk API and West Yorkshire Police
  • License: Open Government Licence v3.0
  • Description: Recorded crime incidents with spatial and temporal attributes
  • Processing Methods:
    • [Processing details to be completed when crime data methodology is finalized]
    • Spatial aggregation using H3 hexagonal grid system
    • Temporal analysis and trend identification
  • Categories: All crime types
  • Usage: Crime density mapping and risk assessment

Note: Detailed crime data processing methodology will be documented separately as this component is processed independently.

Data Processing Methods

Geospatial Processing Pipeline

The project uses a comprehensive Python-based processing pipeline built on the safer_parks library. Key processing steps include:

1. Data Acquisition and Standardization

  • Automated download and caching of external datasets
  • Projection standardization to British National Grid (EPSG:27700)
  • Data format conversion to GeoJSON for web compatibility

2. Geographic Subsetting

# Example using safer_parks library functions
from safer_parks import subset_to_LAD

# Subset data to Bradford District
bradford_parks = subset_to_LAD(
    LAD_gdf=local_authority_boundaries,
    LAD_column_name='LAD_name',
    LAD_name='Bradford',
    data_to_subset=parks_data
)

3. Geometry Processing and Validation

  • Boundary Merging: Adjacent or overlapping park polygons are merged using merge_touching_or_intersecting_polygons() function
  • Topology Validation: Geometry validation and repair for invalid polygons
  • Simplification: Coordinate reduction for web performance while maintaining accuracy

4. Attribute Data Enhancement

  • Data Cleaning: Standardization of text fields and removal of duplicates using clean_and_deduplicate() function
  • Facility Categorization: Standardized classification of park amenities and facilities
  • Accessibility Assessment: Analysis of accessibility features and ratings

5. Spatial Analysis Operations

  • Buffer Analysis: Creation of analysis zones around parks
  • Intersection Analysis: Calculation of park-facility relationships
  • Network Analysis: Assessment of connectivity and access routes
  • Point-in-Polygon: Assignment of crime incidents and facilities to park catchments

Quality Assurance

Data Validation Procedures

  1. Geometric Validation: Automated topology checking and repair
  2. Attribute Validation: Range checking and categorical validation
  3. Cross-Dataset Consistency: Comparison between OS and OSM sources
  4. Manual Verification: Spot-checking of key locations and attributes

Error Handling and Documentation

  • Comprehensive logging of processing steps and data quality issues
  • Version control for all data processing scripts
  • Documentation of known limitations and caveats

Analysis Methods

Park Safety Assessment Framework

Multi-Criteria Analysis

The dashboard employs a multi-criteria approach considering: - Physical Safety: Crime statistics and incident patterns - Perceived Safety: Environmental design factors and visibility - Accessibility: Physical access and inclusive design features - Activity Levels: Presence of facilities and natural surveillance

Spatial Analysis Techniques

  • Hot Spot Analysis: Identification of crime concentration areas using density estimation
  • Accessibility Modeling: Network analysis for pedestrian access routes
  • Visibility Analysis: Assessment of sight lines and natural surveillance opportunities
  • Buffer Analysis: Creation of influence zones for facilities and safety features

Statistical Methods

Aggregation Techniques

  • H3 Hexagonal Binning: Spatial aggregation of crime data using Uber’s H3 system for consistent spatial units
  • Temporal Aggregation: Monthly and seasonal crime pattern analysis
  • Categorical Aggregation: Grouping of facilities and amenities by safety relevance

Density Estimation

  • Kernel Density Estimation: Smooth interpolation of point-based crime data
  • Adaptive Bandwidth: Variable smoothing parameters based on local data density
  • Edge Correction: Adjustment for boundary effects in density calculations

Technical Implementation

Processing Environment

  • Language: Python 3.8+
  • Key Libraries:
    • GeoPandas for spatial data manipulation
    • Shapely for geometric operations
    • Folium/Leaflet for web mapping
    • H3-py for hexagonal spatial indexing

Web Visualization

  • Mapping: Leaflet.js with custom styling and interaction
  • Data Format: GeoJSON for client-side rendering
  • Performance Optimization: Geometry simplification and data tiling
  • Accessibility: WCAG 2.1 compliant interface design

Version Control and Reproducibility

  • Code Repository: GitHub - Safer-Parks/safer-parks
  • Data Versioning: Systematic tracking of dataset versions and processing dates
  • Documentation: Comprehensive README and inline code documentation
  • Testing: Automated testing suite for data processing functions

Data Limitations and Considerations

Temporal Considerations

  • Data Currency: Most datasets updated annually or quarterly
  • Temporal Alignment: Not all datasets cover identical time periods
  • Seasonal Variation: Crime patterns may vary seasonally but current analysis uses annual aggregates

Spatial Limitations

  • Boundary Precision: Some park boundaries may not reflect current ground conditions
  • Access Point Accuracy: Calculated entrance points are estimates and may not reflect all actual access routes
  • Private vs Public: Some green spaces may have restricted access not captured in the data

Data Quality

  • OSM Completeness: OpenStreetMap data quality varies by area and contributor activity
  • Crime Reporting: Recorded crime represents reported incidents only
  • Attribution Accuracy: Some facility and amenity information may be outdated

Future Development

Planned Enhancements

  • Integration of additional safety indicators (lighting, CCTV coverage)
  • Temporal analysis capabilities for trend identification
  • Enhanced accessibility assessment tools
  • Community feedback integration mechanisms

Data Expansion

  • Extension to additional local authority areas
  • Integration of demographic and socioeconomic indicators
  • Real-time or near-real-time crime data feeds
  • Community-contributed safety assessment data

Contact and Contributions

For questions about the data sources, processing methods, or to contribute to the project:

  • Project Repository: https://github.com/Safer-Parks/safer-parks
  • Issues and Suggestions: Use the GitHub Issues system for bug reports and feature requests
  • Data Contributions: Guidelines for contributing additional datasets or corrections available in the repository

Citation

When using data or methods from this project, please cite:

[Citation format to be determined based on project publication status]

Last updated: October 2025