Aller au contenu principal

Emergency Services: Coverage Analysis

Identify areas that lack adequate emergency service coverage to optimize facility placement and response planning.

Scenario Overview​

Goal: Find residential areas more than 5km from the nearest fire station to identify service coverage gaps.

Real-World Application:

  • Fire departments optimizing station placement
  • Emergency management planning response times
  • Urban planners evaluating service equity
  • Insurance companies assessing risk zones

Estimated Time: 12 minutes

Difficulty: ⭐⭐ Intermediate

Prerequisites​

Required Data​

  1. Fire Stations Layer (points)

    • Emergency service facility locations
    • Must include station names/IDs
    • Covers your study area

  2. Population Areas Layer (polygons)

    • Census blocks, neighborhoods, or postal zones
    • Population count attribute (optional but valuable)
    • Residential land use areas

  3. Optional: Road Network

    • For drive-time analysis (advanced)
    • Network topology for routing

Sample Data Sources​

Option 1: OpenStreetMap

# Use QGIS QuickOSM plugin

# For fire stations:
Key: "amenity", Value: "fire_station"

# For residential areas:
Key: "landuse", Value: "residential"
Key: "place", Value: "neighbourhood"

Option 2: Government Open Data

  • Municipal emergency services databases
  • Census boundary files with population
  • HIFLD (Homeland Infrastructure Foundation-Level Data)
  • Local GIS data portals

Backend Recommendation​

OGR - Best for this workflow:

  • Universal format compatibility (Shapefiles, GeoJSON, GeoPackage)
  • No complex setup required
  • Good for datasets <10,000 features
  • Works with any QGIS installation

Step-by-Step Instructions​

Step 1: Load and Prepare Data​

  1. Load layers into QGIS:

    • fire_stations.gpkg (or .shp, .geojson)
    • residential_areas.gpkg

  2. Verify CRS:

    Both layers must use same projected coordinate system
    Right-click → Properties → Information → CRS

    Recommended: Local UTM zone or state/national grid
    Example: EPSG:32633 (UTM Zone 33N)

  3. Inspect data:

    • Count fire stations: Should have at least 3-5 for meaningful analysis
    • Check residential areas: Look for population or household count attributes
    • Verify coverage: Fire stations should be distributed across study area
Finding Your UTM Zone

Use epsg.io and click on map to find appropriate UTM zone for your region.

Step 2: Create 5km Service Areas Around Fire Stations​

Using FilterMate:

  1. Open FilterMate, select fire_stations layer
  2. Enter expression: 
    -- Keep all fire stations
    1 = 1
  3. Enable Buffer operation:
    • Distance: 5000 meters
    • Type: Positive (expand)
    • Segments: 16 (for smooth circles)
  4. Apply Filter
  5. Export as fire_coverage_5km.gpkg

Result: Circular 5km buffers around each fire station (service coverage zones)

Step 3: Identify Under-Served Residential Areas (Inverse Query)​

This is the key step - finding areas NOT within 5km of any fire station:

Method 1: Using FilterMate (Recommended)

  1. Select residential_areas layer
  2. Choose OGR backend
  3. Enter expression:
-- Residential areas NOT intersecting fire coverage
NOT intersects(
  $geometry,
  aggregate(
    layer:='fire_coverage_5km',
    aggregate:='collect',
    expression:=$geometry
  )
)

Method 2: Using disjoint() predicate

-- Areas completely outside all coverage zones
disjoint(
  $geometry,
  aggregate('fire_coverage_5km', 'collect', $geometry)
)
  1. Click Apply Filter
  2. Review map - red/highlighted areas show coverage gaps

Step 4: Calculate Exact Distance to Nearest Station​

Add a field showing how far each under-served area is from nearest fire station:

  1. Open Attribute Table (F6) of filtered layer
  2. Open Field Calculator
  3. Create new field: 
    Field name: distance_to_nearest_station
    Type: Decimal (double)
    Precision: 2

    Expression:
    array_min(
      array_foreach(
        overlay_nearest('fire_stations', $geometry, limit:=5),
        distance(geometry(@element), $geometry)
      )
    ) / 1000  -- Convert meters to kilometers

Result: Each residential area now shows distance to closest fire station

Step 5: Prioritize by Population at Risk​

If your residential layer has population data:

  1. Calculate total population in under-served areas:

    -- In expression filter or field calculator
    "population" > 0

  2. Sort by priority:

    Attribute Table → Click column header "population"
    → Sort descending

  3. Create priority categories:

    CASE
      WHEN "distance_to_nearest_station" > 10 THEN 'Critical (>10km)'
      WHEN "distance_to_nearest_station" > 7 THEN 'High Priority (7-10km)'
      WHEN "distance_to_nearest_station" > 5 THEN 'Medium Priority (5-7km)'
      ELSE 'Acceptable (<5km)'
    END

Step 6: Visualize Coverage Gaps​

Symbology Setup:

  1. Right-click residential_areas → Symbology
  2. Choose Graduated
  3. Value: distance_to_nearest_station
  4. Method: Natural Breaks (Jenks)
  5. Classes: 5
  6. Color ramp: Red (far) → Yellow → Green (close)
  7. Apply

Add Labels (optional):

Label with: concat("name", ' - ', round("distance_to_nearest_station", 1), ' km')
Size: Based on "population" (larger = more people affected)

Step 7: Export Results and Generate Report​

  1. Export under-served areas:

    FilterMate → Export Filtered Features
    Format: GeoPackage
    Filename: residential_areas_underserved.gpkg
    CRS: WGS84 (for sharing) or keep project CRS

  2. Generate summary statistics:

    Vector → Analysis Tools → Basic Statistics
    Input: residential_areas_underserved
    Field: population

  3. Create summary report (Python Console - optional):

    layer = iface.activeLayer()
    features = list(layer.getFeatures())

    total_areas = len(features)
    total_population = sum(f['population'] for f in features if f['population'])
    avg_distance = sum(f['distance_to_nearest_station'] for f in features) / total_areas
    max_distance = max(f['distance_to_nearest_station'] for f in features)

    print(f"=== Emergency Services Coverage Gap Analysis ===")
    print(f"Under-served residential areas: {total_areas}")
    print(f"Population affected: {total_population:,}")
    print(f"Average distance to nearest station: {avg_distance:.1f} km")
    print(f"Maximum distance: {max_distance:.1f} km")

Understanding the Results​

What the Filter Shows​

✅ Selected areas: Residential zones >5km from ANY fire station

❌ Excluded areas: Residential zones within 5km service radius

Interpreting Coverage Gaps​

Critical Gaps (>10km):

  • Response time likely exceeds national standards (e.g., NFPA 1710: 8 minutes)
  • High priority for new station placement
  • Consider temporary or volunteer stations
  • May need mutual aid agreements with neighboring jurisdictions

High Priority (7-10km):

  • Response time borderline acceptable
  • Should be addressed in next planning cycle
  • Consider mobile/seasonal stations
  • Evaluate road network quality (may be longer drive time)

Medium Priority (5-7km):

  • Technically under-served by strict standards
  • Low urgency if population density is low
  • Monitor for future growth
  • May be acceptable for rural areas

Validation Checks​

  1. Visual spot check: Use QGIS Measure tool to verify distances
  2. Edge cases: Areas just outside 5km may round differently
  3. Population accuracy: Verify sum matches known census totals
  4. Geometry validity: Check for slivers or invalid polygons

Best Practices​

Coverage Standards​

NFPA 1710 (USA) Recommendations:

  • Urban areas: 1.5 mile (2.4 km) travel distance
  • Rural areas: Up to 5 miles (8 km) acceptable
  • Response time goal: 8 minutes from call to arrival

Adjust threshold based on your region:

Urban areas:    2-3 km
Suburban areas: 5 km (as in this tutorial)
Rural areas:    8-10 km

Performance Optimization​

For large datasets:

  1. Simplify residential area geometry:

    Vector → Geometry → Simplify
    Tolerance: 50 meters (maintains coverage accuracy)

  2. Pre-filter to populated areas only:

    "population" > 0 OR "landuse" = 'residential'

  3. Use spatial index (OGR creates automatically for GeoPackage)

  4. Backend selection guide:

    < 1,000 areas:    OGR (sufficient)
    1k - 50k:         Spatialite
    > 50k:            PostgreSQL

Real-World Adjustments​

Consider road network reality:

  • Straight-line 5km may be 8km by road
  • Mountains/rivers may block direct access
  • Use network analysis for drive-time instead (advanced)

Network Analysis Alternative (QGIS built-in):

Processing → Network Analysis → Service Area (from layer)
Input: fire_stations
Travel cost: 5000 meters OR 10 minutes
Creates drive-time polygons instead of circles

Data Quality Considerations​

  1. Fire station accuracy:

    • Verify stations are operational (not decommissioned)
    • Check if volunteer stations should have smaller radius
    • Consider specialized stations (airport, industrial)

  2. Residential area quality:

    • Remove parks, industrial zones misclassified as residential
    • Update with recent census data
    • Account for new developments

  3. CRS importance:

    • Distance calculations require projected CRS
    • Geographic (lat/lon) will give incorrect results
    • Always reproject if needed before analysis

Common Issues​

Issue 1: All residential areas selected (or none selected)​

Cause: CRS mismatch or buffer not created properly

Solution:

1. Check fire_coverage_5km layer exists and has features
2. Verify both layers in same CRS
3. Re-create buffers with correct distance unit (meters)
4. Check buffer layer name matches expression exactly

Issue 2: Distance calculation returns NULL or errors​

Cause: overlay_nearest() not finding fire stations layer

Solution:

1. Ensure fire_stations layer is loaded in project
2. Check layer name matches exactly (case-sensitive)
3. Alternative: Use aggregate() with minimum distance:

distance(
  $geometry,
  aggregate('fire_stations', 'collect', $geometry)
)

Issue 3: Results show unexpected patterns​

Cause: Data quality issues or projection problems

Troubleshooting:

1. Zoom to specific result and measure distance manually
2. Check for overlapping residential polygons
3. Verify fire_stations actually cover the area
4. Look for invalid geometries:
   Vector → Geometry Tools → Check Validity

Issue 4: Performance very slow​

Cause: Large geometries or complex residential areas

Solutions:

1. Simplify residential geometry (50-100m tolerance)
2. Create spatial index on both layers
3. Process by administrative districts separately
4. Use PostgreSQL backend for >10k features

Next Steps​

Advanced Techniques​

1. Multi-Station Coverage (areas served by ≥2 stations):

-- Count overlapping coverage zones
array_length(
  overlay_intersects('fire_coverage_5km', $geometry)
) >= 2

2. Priority Scoring (distance + population):

-- Higher score = higher priority for new station
("distance_to_nearest_station" - 5) * "population" / 1000

3. Optimal New Station Location:

1. Export under-served areas with population
2. Find centroid weighted by population:
   Processing → Vector Geometry → Centroids
3. Manual analysis: Place new station at highest-priority centroid

4. Response Time Modeling (advanced):

# Requires road network and routing
# Uses QGIS Network Analysis tools
# Models actual drive time vs. straight-line distance
# Accounts for road speed limits and turn restrictions

5. Temporal Analysis (future growth):

-- If you have population projection data
("population_2030" - "population_2024") / "population_2024" > 0.2
-- Areas expecting >20% growth

Further Learning​

Summary​

✅ You've learned:

  • Creating service area buffers around facilities
  • Inverse spatial filtering (NOT intersects)
  • Distance calculations to nearest feature
  • Population-weighted priority analysis
  • Exporting results for planning reports

✅ Key techniques:

  • NOT intersects() for coverage gap analysis
  • overlay_nearest() for distance calculations
  • aggregate() with spatial predicates
  • Priority scoring with attribute + spatial data

🎯 Real-world impact: This workflow helps emergency management agencies identify service gaps, optimize resource allocation, improve response times, and ensure equitable emergency service coverage across communities.

💡 Pro tip: Run this analysis annually with updated census data to track coverage changes as populations shift and adjust station placement accordingly.