Creating a 2x2 km Fishnet Grid Using CBS Data

Duration: 60-75 minutes

Learning Goals

By the end of this tutorial, you will be able to:

  • Install and use QGIS plugins (PDOK Services Plugin)
  • Access Dutch national geographic datasets through web services
  • Create regular grid structures (fishnet) with specific dimensions
  • Understand and work with coordinate reference systems (EPSG:28992)
  • Perform spatial clipping operations to match geographic boundaries
  • Calculate geometric attributes (area) using Field Calculator
  • Filter spatial data based on geometric criteria
  • Understand the importance of data quality in grid-based analysis
  • Prepare spatial frameworks for statistical data integration

Software and Data Required

Software: - QGIS 3.40 or higher - Internet connection for PDOK plugin and data access

Data sources: - PDOK Services Plugin (installed during tutorial) - Dutch national boundaries (Landgebied layer from PDOK)

Output files: - fishnet_2km.gpkg - Initial 2x2 km grid - fishnet_2km_clipped.gpkg - Grid clipped to national boundaries - fishnet_2km_filtered.gpkg - Final filtered grid (≥50% coverage)

Introduction

A fishnet grid is a regular rectangular mesh that divides a geographic area into uniform cells. This spatial framework is commonly used for:

  • Statistical data aggregation - Standardized units for census and survey data
  • Spatial sampling - Systematic selection of study locations
  • Pattern analysis - Regular grid for density and distribution studies
  • Data integration - Common framework for combining multiple datasets

In the Netherlands, Statistics Netherlands (CBS - Centraal Bureau voor de Statistiek) publishes data at various grid resolutions (100m, 500m, 1km). This tutorial creates a 2x2 km fishnet compatible with CBS statistical data, using the PDOK (Publieke Dienstverlening Op de Kaart) service for national boundaries.

Why 2x2 km? This resolution balances detail with computational efficiency, making it suitable for regional and national-scale analysis while remaining compatible with finer CBS grid data through aggregation.

Part 1: Setting Up PDOK Plugin and Loading Boundaries

Step 1.1: Install the PDOK Plugin

The PDOK Services Plugin provides easy access to Dutch national geographic datasets including administrative boundaries.

  1. Open QGIS 3.40

  2. Go to: Plugins → Manage and Install Plugins

  3. In the Plugins dialog, click on the All tab

  4. In the search box, type: PDOK

  5. Select PDOK Services Plugin from the results

  6. Click Install Plugin

  7. Wait for installation to complete, then click Close

Figure 1 - Installing the PDOK Services Plugin
What is PDOK?

PDOK (Publieke Dienstverlening Op de Kaart) is the Dutch national SDI (Spatial Data Infrastructure) that provides standardized access to key geographic datasets through web services. It includes administrative boundaries, topographic maps, cadastral data, and more.

Step 1.2: Load Dutch National Boundaries

Now we’ll use the PDOK plugin to load the Netherlands land area boundary.

  1. Set the project CRS to EPSG:28992 (Amersfoort / RD New)

    • Go to: Project → Properties → CRS
    • Search for: 28992
    • Select: EPSG:28992 - Amersfoort / RD New
    • Click OK

  2. Click on the PDOK Services Plugin icon in the toolbar (or go to Web → PDOK Services Plugin)

  3. In the PDOK Services panel, type Landgebied in the search bar

  4. Look for Typr: OGC API - Features

  5. Double-click to add it to your map, or right-click and select Add to map

  6. The national boundary should now appear in your Layers Panel and map canvas as Landgebied

  7. Verify the CRS: Check the bottom right corner of QGIS. It should show EPSG:28992

Figure 2 - PDOK panel with Landgebied layer selected
Understanding EPSG:28992 (RD New)

The Rijksdriehoeksmeting (RD) New coordinate system is the standard projected CRS for the Netherlands. It uses: - Units: Meters (making 2 km = 2000 units straightforward) - Origin: Amersfoort tower - Coverage: Optimized for the Netherlands territory - Applications: All Dutch official mapping and cadastral work

Always verify this CRS when working with Dutch spatial data!

Note: If you prefer to work with a specific region or province instead of the entire country, you can also select provincial boundaries or municipal boundaries from the PDOK plugin.

Part 2: Creating the Initial Fishnet Grid

Step 2.1: Generate the Grid

Now we’ll create the 2x2 km fishnet grid using QGIS’s built-in grid creation tool.

  1. Go to: Vector → Research Tools → Create Grid

  2. In the Create Grid dialog, configure the following settings:

    • Grid type: Select Rectangle (Polygon)
    • Grid extent: Click the dropdown arrow ... next to the extent field and select Calculate from Layer → choose your Landgebied layer
    • Horizontal spacing: 2000 (meters)
    • Vertical spacing: 2000 (meters)
    • Grid CRS: EPSG:28992 - Amersfoort / RD New

  3. For the output:

    • Click the … button next to Grid
    • Select Save to File
    • Specify location and name: fishnet_2km.gpkg

  4. Click Run

  5. Wait for processing to complete (this may take 30-60 seconds)

  6. Once finished, click Close

Figure 3 - Create Grid dialog configuration

Your fishnet grid should now appear on the map as a layer of 2x2 km squares covering the rectangular extent of the Netherlands.

Grid Spacing Units

The spacing values are in the units of your CRS. Since EPSG:28992 uses meters:

- 2000 = 2 kilometers

- 1000 = 1 kilometer

- 500 = 500 meters

For other grid sizes, simply adjust these values!

Challenge 1: Inspect Your Grid

Before proceeding to clip the grid, let’s examine what we’ve created. This exercise helps you understand the relationship between geographic extent and grid generation.

Task 1.1: Check the Layer Extent

  1. Right-click on your fishnet_2km layer and select Properties

  2. Go to the Information tab

  3. Look at the Extent section. Record the coordinates:

    • xMin: ___________
    • xMax: ___________
    • yMin: ___________
    • yMax: ___________

  4. Compare these to the Landgebied layer extent

Question: Are the extents identical? Why or why not?

Figure 4 - Layer information showing extent
Expected Extent Values

You should see coordinates approximately: - xMin: ~10425 - xMax: ~278026
- yMin: ~306846 - yMax: ~621876

These represent the rectangular bounding box of the Netherlands in RD New coordinates.

Task 1.2: Count the Grid Squares

  1. Right-click on your fishnet_2km layer and select Open Attribute Table

  2. Look at the top of the attribute table window - you’ll see text like “Total features: X”

  3. Record: How many grid squares did you create? ___________

Question: Based on the extent (approximately 267 km × 313 km), can you calculate the expected number of 2×2 km cells?

Calculation:

Width in cells = (xMax - xMin) / 2000 = (~280,000 - 13,000) / 2000 ≈ 134 cells
Height in cells = (yMax - yMin) / 2000 = (~619,000 - 306,000) / 2000 ≈ 157 cells
Expected total = 134 × 157 ≈ 21,000 cells

Figure 5 - Attribute table showing total features

Task 1.3: Visual Inspection

  1. Zoom to the full extent of both layers (right-click layer → Zoom to Layer)

  2. Observe the map:

    • Notice that the grid extends beyond the actual land area
    • The grid includes areas over the sea
    • It covers rectangular space including neighboring countries

  3. Toggle layer visibility (checkbox in Layers Panel) to compare:

    • Grid alone
    • Landgebied alone
    • Both together

Question: Why does the grid extend beyond the Netherlands boundary?

Understanding Grid Extent

The Create Grid tool generates a rectangular grid covering the full bounding box of your reference layer. This means:

  • Included: All complete 2×2 km cells within the bounding rectangle
  • Not considered: The actual shape of the boundary (coastlines, borders)
  • Result: Many cells over water, neighboring countries, and empty space

This is why we need clipping in the next step!

Part 3: Clipping the Grid to National Boundaries

Step 3.1: Perform the Clip Operation

The fishnet grid currently extends beyond the national boundary in a rectangular extent. To create a grid that matches the actual shape of the Netherlands, we need to clip it.

  1. Go to: Vector → Geoprocessing Tools → Clip

  2. In the Clip dialog:

    • Input layer: Select your fishnet_2km layer
    • Overlay layer: Select the Landgebied layer
    • Clipped: Click … and save as fishnet_2km_clipped.gpkg

  3. Click Run

  4. After processing completes, click Close

Figure 6 - Clip dialog configuration

You now have a fishnet grid that matches the Netherlands boundary. However, notice that many cells along the coastline and borders are only partially within the country.

What Does Clipping Do?

The clip operation: 1. Takes each grid cell polygon 2. Intersects it with the boundary polygon 3. Keeps only the portion inside the boundary 4. Creates new “cookie-cutter” shapes along edges

Result: Cells completely inside remain square; edge cells become irregular polygons.

Step 3.2: Examine the Clipped Results

  1. Open the attribute table of fishnet_2km_clipped

  2. Record the new feature count: ___________

  3. Compare to Challenge 1:

    • Before clipping: 21.172 cells
    • After clipping: _________ cells
    • Reduction: _________ cells removed

  4. Zoom to coastal areas and observe:

    • Some cells are complete 2×2 km squares
    • Others are partial (cut by coastline)
    • Some are very small slivers

Figure 7 - Clipped grid showing partial cells along coast

Challenge 2: Filter Partial Grid Cells

After clipping, many grid cells are only partially covered by land area. These partial cells can skew your analysis, especially when working with CBS data that assumes full grid cells. Let’s identify and remove cells with less than 50% coverage.

Task 2.1: Calculate the Area of Each Grid Cell

  1. Open the attribute table of fishnet_2km_clipped

  2. Enable editing mode (click the pencil icon or press Ctrl+E)

  3. Open Field Calculator (click the abacus icon or press Ctrl+I)

  4. Configure the area field:

    • ☑ Check Create a new field
    • Output field name: area_km2
    • Output field type: Decimal number (real)
    • Output field length: 10
    • Precision: 2

  5. Enter this expression:

$area/1000000

This calculates the area in square meters (since your CRS is EPSG:28992 in meters).

  1. Click OK

  2. Save edits (click the save icon) and toggle off editing mode

Figure 8 - Field Calculator creating area field
Alternative Area Expressions
  • round($area, 2) - Round to 2 decimal places
  • $area / 10000 - Convert to hectares
  • area($geometry) - More explicit function (same result)

Task 2.2: Identify the Maximum Grid Cell Area

  1. In the attribute table, click the area_km2 column header to sort by area (click twice for descending order)

  2. Record the maximum area value: ___________ m²

Question: What is the expected area of a complete 2×2 km square?

Calculation:

Expected area = 2000m × 2000m = 4,000,000 m²
  1. Scroll through the table and observe:
    • Many cells have exactly 4,000,000 m² (or very close)
    • Coastal/border cells have smaller values
    • Some cells are very small (< 500,000 m²)

Figure 9 - Sorted attribute table showing area distribution

Task 2.3: Calculate Statistics on Area Distribution

  1. Go to: Vector → Analysis Tools → Basic Statistics for Fields

  2. Configure:

    • Input layer: fishnet_2km_clipped
    • Field to calculate statistics on: area_km2

  3. Click Run

  4. Review the output in the Results panel:

    • Count: Total number of cells
    • Min: Smallest cell area
    • Max: Should be ~4,000,000 m²
    • Mean: Average cell area
    • Median: Middle value

Question: What percentage of cells are complete squares? - If mean ≈ 4,000,000, most cells are complete - If mean < 3,500,000, many partial cells exist

Figure 10 - Basic statistics results

Task 2.4: Filter Out Small Partial Cells

Now we’ll remove cells with less than 50% of the maximum area (i.e., less than 2,000,000 m²).

  1. Right-click fishnet_2km_clipped layer → Filter

  2. In the Query Builder dialog, enter the following expression:

"area_km2" >= 2000000
  1. Click Test to verify the filter works
    • You should see a message like “Filter returned X features”
  2. Click OK to apply the filter

Figure 11 - Query Builder filter dialog
  1. Check your attribute table again:
    • Note the new feature count (filtered)
    • All visible cells now have area_km2 ≥ 2,000,000
Filter vs. Selection vs. New Layer

The filter is temporary - it only hides features, doesn’t delete them. To make it permanent: - Export to new layer (recommended - see next step) - Or delete filtered-out features (destructive)

Removing the filter (Properties → Source → Query Builder → Clear) restores all features.

Task 2.5: Make the Filter Permanent

To create a new layer with only the filtered cells:

  1. Right-click the filtered fishnet_2km_clipped layer → Export → Save Features As...

  2. Configure export:

    • Format: GeoPackage
    • File name: Click … and save as fishnet_2km_filtered.gpkg
    • ☑ Check Add saved file to map
    • CRS: EPSG:28992 (should be default)
    • Select fields to export: Keep all (or deselect unnecessary fields like fid)

  3. Click OK

  4. Clean up your workspace:

    • Remove the filtered layer: right-click fishnet_2km_clipped → Remove Layer
    • Keep the new fishnet_2km_filtered layer
    • Optionally remove the original unclipped grid

Figure 12 - Export layer dialog

Task 2.6: Compare Before and After

  1. Record final statistics:
Stage Number of Cells Notes
Initial grid (before clip) __________ Rectangular extent
After clipping __________ Matches NL shape
After filtering (≥50%) __________ Only substantial cells
Cells removed by filter __________ Difference
  1. Calculate the percentage removed:
Percentage removed = (Cells after clip - Cells after filter) / Cells after clip × 100%
                   = _________ %

Question: Is this what you expected? Why might this percentage be high or low?

Understanding the Filtering Decision

Why Remove Partial Cells?

Let’s explore the reasoning behind the 50% threshold with concrete examples.

1. Data Accuracy

CBS and other statistical data are typically reported for complete grid cells. Consider population data:

Example - Coastal Cell: - Complete 2×2 km cell: 4,000,000 m² land area - Partial cell (30% land): 1,200,000 m² land area

If CBS reports “500 residents” for the grid cell: - Complete cell: 500 residents / 4 km² = 125 residents/km² - Partial cell: 500 residents / 1.2 km² = 417 residents/km² ❌ Misleading!

The partial cell appears to have 3× higher density, but this is an artifact of incomplete coverage.

CBS Grid Data Assumptions

CBS typically: - Reports data for cells with substantial coverage - May exclude or flag partial cells - Aggregates from finer resolutions (100m, 500m) - Assumes the grid cell represents the statistical unit

Using partial cells violates these assumptions!

2. Spatial Analysis Integrity

When performing calculations like totals or averages across neighborhoods:

Scenario: Summing building counts per neighborhood

Cell Type Land Area Buildings True Density Calculated Density
Complete 4.0 km² 200 50/km² 50/km² ✓
75% partial 3.0 km² 150 50/km² 50/km² ✓
25% partial 1.0 km² 50 50/km² 50/km² ✓

Problem: The 25% partial cell contributes the same weight to neighborhood statistics as complete cells, but represents less area.

3. Consistent Comparison

Scenario: Comparing commercial area across municipalities

Municipality Complete Cells Partial Cells Total Cells Commercial Area
Rotterdam 450 120 570 2,500,000 m²
Den Haag 380 200 580 2,400,000 m²

If we include partial cells: - Rotterdam: Higher proportion of complete cells → more reliable data - Den Haag: Many partial cells → potentially inflated/deflated statistics

Fair comparison requires comparable spatial units!

4. Coastline and Border Effects

Coastal Example:

[Complete Grid Cell]     [Partial Cell - 20% land]
┌─────────────────┐     ┌─────────┐
│     LAND        │     │  LAND   │
│                 │     │         │
│                 │     ├─────────┘
│                 │     │  WATER
│                 │     │
└─────────────────┘     └─────────

The partial cell: - May lack CBS data (below reporting threshold) - Represents coastal geography, not inland patterns - Could bias analyses toward maritime characteristics

5. The 50% Threshold

Why specifically 50%?

Too Strict (e.g., 90%): - Removes many valid cells - Loses data coverage in interesting regions - May create gaps in your analysis

Too Lenient (e.g., 10%): - Keeps cells with minimal coverage - Introduces unreliable data points - Reduces analysis quality

50% Balance: - Cell represents “more land than not” - Likely to have CBS data reported - Maintains good coverage - Industry standard for grid analysis

Alternative Thresholds

Depending on your analysis, you might choose:

  • 75%: Conservative - very reliable cells only
  • 60%: Moderate - slight preference for complete cells
  • 40%: Lenient - accept cells that are “mostly there”

For CBS integration, 50-60% is recommended.

Visual Comparison

Let’s see the difference filtering makes:

  1. Style your layers to visualize the difference:

    Before filtering (fishnet_2km_clipped):

    • Right-click → Properties → Symbology
    • Fill: Transparent
    • Stroke: Red, 0.4mm

    After filtering (fishnet_2km_filtered):

    • Fill: Light blue (50% transparent)
    • Stroke: Dark blue, 0.6mm

  2. Load both layers and compare:

    • Toggle visibility
    • Zoom to coastal areas
    • Observe which cells were removed

Figure 13 - Before and after filtering comparison

What should you see: - Small slivers along coast: Removed ✓ - Narrow peninsulas: Partial cells removed ✓ - Interior regions: All cells retained ✓ - Island cells: Depends on size (Wadden Islands may lose some)

Part 4: Adding Grid Cell Identifiers

To make the fishnet useful for analysis and joining with CBS data, we should add unique identifiers and coordinate information to each cell.

Step 4.1: Create Unique ID Field

  1. Open the attribute table of fishnet_2km_filtered

  2. Enable editing mode (click the pencil icon)

  3. Open Field Calculator

  4. Configure the ID field:

    • ☑ Create a new field
    • Output field name: grid_id
    • Output field type: Whole number (integer)
    • Output field length: 10

  5. Enter this expression:

$id

This assigns each feature its unique feature ID.

  1. Click OK

Figure 14 - Creating grid_id field

Step 4.2: Add Centroid Coordinates (optional)

CBS data often uses centroid coordinates as grid cell identifiers. Let’s add X and Y coordinates of each cell’s center.

Create center_x field:

  1. Open Field Calculator (still in editing mode)

  2. Configure:

    • ☑ Create a new field
    • Output field name: center_x
    • Output field type: Decimal number (real)
    • Output field length: 10
    • Precision: 2

  3. Expression:

x(centroid($geometry))
  1. Click OK

Create center_y field:

  1. Open Field Calculator again

  2. Configure:

    • ☑ Create a new field
    • Output field name: center_y
    • Output field type: Decimal number (real)
    • Output field length: 10
    • Precision: 2

  3. Expression:

y(centroid($geometry))

  1. Click OK

  2. Save edits and toggle off editing mode

Figure 15 - Attribute table with new identifier fields
Why Centroid Coordinates?

CBS grid data typically uses the centroid (center point) of each grid cell as the spatial identifier. This allows joining:

  • Your fishnet cell at centroid (125000, 487000)
  • CBS data row with X=125000, Y=487000
  • → Matched! Can transfer statistics to your grid

The coordinates are in RD New (meters), making them easy to work with.

Tips and Best Practices

CRS Consistency

Important

Always verify all layers use EPSG:28992 when working with Dutch data:

  • Check: Project → Properties → CRS
  • Verify: Bottom-right corner of QGIS interface
  • Fix: Right-click layer → Export → Reproject to EPSG:28992

The RD New system uses meters, making distance/area calculations intuitive (2 km = 2000 units).

Grid Resolution Selection

Grid Size Cell Area When to Use
100m × 100m 0.01 km² Detailed urban analysis, building-level patterns
500m × 500m 0.25 km² Neighborhood characteristics, walking distances
1km × 1km 1 km² District patterns, facility catchments
2km × 2km 4 km² Regional trends, city-wide comparisons
5km × 5km 25 km² Provincial analysis, macro patterns

Your 2 km fishnet can aggregate data from finer resolutions (100m, 500m, 1km) using spatial joins and zonal statistics.

Performance Considerations

For large areas, fishnet grids can contain thousands of cells:

Optimization strategies: - Work with subsets first (single province) - Use GeoPackage format (faster than Shapefile) - Simplify analysis - don’t keep unnecessary attributes - Use spatial indexes: Vector → Data Management Tools → Create Spatial Index - Filter before processing (remove cells over water if appropriate)

Data Quality Checks

After joining CBS data, verify quality:

  1. Check for NULL values:
    • Open attribute table
    • Sort by CBS columns
    • Investigate why some cells lack data
  2. Validate ranges:
    • Use Basic Statistics tool
    • Check min/max make sense
    • Look for outliers (data errors or real phenomena?)
  3. Compare totals:
    • Sum your aggregated values
    • Compare to CBS published totals
    • Should match within rounding error

Saving Your Work

Recommended file organization:

project_folder/
├── data/
│   ├── raw/
│   ├── boundaries/
│   │   └── landgebied.gpkg
│   └── grids/
│       ├── fishnet_2km.gpkg
│       ├── fishnet_2km_clipped.gpkg
│       └── fishnet_2km_filtered.gpkg
├── outputs/
└── project.qgz

Save formats: - GeoPackage (.gpkg): Recommended - modern, efficient, single-file - Shapefile (.shp): Legacy - field name limitations, multiple files - GeoJSON: Good for web mapping

Documentation

Keep notes about: - Data sources: Where did CBS data come from? Which version? - Processing steps: What filters/joins were applied? - Field definitions: What does each column represent? - CRS information: Always EPSG:28992 for Dutch data - Date/version: When was the analysis performed?

This helps maintain reproducibility and data quality.

Common Issues and Solutions

Issue 1: PDOK Plugin Not Showing Layers

Symptoms: - PDOK panel is empty - Cannot load Landgebied layer

Causes & Solutions:

Cause Solution
No internet connection Check network, try again
Plugin not activated Plugins → Manage Plugins → check PDOK is enabled
PDOK service temporarily down Try again later, check https://www.pdok.nl/status
Firewall blocking Configure firewall to allow QGIS web services

Issue 2: Grid Cells Are Not Square

Symptoms: - Cells appear rectangular or distorted - Grid doesn’t align with background maps

Diagnosis: 1. Check bottom-right corner of QGIS 2. Look at coordinate values (should be 5-6 digits in meters)

Solution: - If showing lat/lon coordinates: 1. Project → Properties → CRS 2. Search for: 28992 3. Select: EPSG:28992 - Amersfoort / RD New 4. Click OK 5. Right-click layers → Set Layer CRS → EPSG:28992

  • If layers have wrong CRS:
    1. Right-click layer → Export → Save Features As
    2. Target CRS: EPSG:28992
    3. Replace old layer

Figure 21 - CRS mismatch visualization

Issue 3: Clip Operation Takes Too Long

Symptoms: - Clipping has been running for 5+ minutes - QGIS appears frozen

Solutions:

  1. Try different algorithm:
    • In Clip dialog, click dropdown next to Run
    • Try: GDAL instead of QGIS (or vice versa)
  2. Simplify boundary first:
Vector → Geometry Tools → Simplify
Tolerance: 10 meters
Use simplified Landgebied for clipping
  1. Work in batches:
    • Clip by province instead of whole country
    • Merge results afterward
  2. Check processing settings:
    • Processing → Options → General
    • Increase thread count if you have multi-core CPU

Issue 4: Area Values Are Wrong

Symptoms: - Areas are tiny (< 1000) or huge (billions) - Area doesn’t equal expected 4,000,000 m²

Diagnosis:

Area Value Likely CRS Units Problem
~0.004 Kilometers ✓ Correct!
~4,000,000 Meters Unit mismatch
~0.015 Degrees Wrong CRS (lat/lon)
~4,000,000,000,000 Centimeters or error Calculation error

Solution: 1. Verify layer CRS is EPSG:28992 2. Recalculate area with correct CRS 3. Convert units if needed:

Area in hectares: $area / 10000
Area in km²: $area / 1000000

Issue 5: CBS Data Won’t Join Properly

Symptoms: - All fishnet cells show NULL for CBS data - Only partial cells have CBS values

Possible causes:

  1. Coordinate mismatch:

    • Check: Are CBS coordinates in same CRS as fishnet?
    • Fix: Reproject CBS layer to EPSG:28992

  2. Resolution mismatch:

    • Check: Is CBS data 100m but you’re joining on exact 2km centroids?
    • Fix: Use spatial join within grid cells, not coordinate match

  3. Field type mismatch:

    • Check: Are join fields both numeric? Both text?
    • Fix: Use Field Calculator to standardize:
    Cast to integer: to_int("field_name")
Cast to string: to_string("field_name")
Round coordinates: round("center_x", 0)

  4. Precision issues:

    • Check: Coordinates like 125000.00001 vs 125000
    • Fix: Round before joining:
    round("center_x", -2)  -- Round to nearest 100

Issue 6: Function Names Appear Inconsistent

Symptoms: - Some tools have different names - Cannot find mentioned menu items

Cause: QGIS version or language differences

Solution: - English vs Dutch interface: - Settings → Options → General → Override system locale - Choose English for consistency with tutorial

  • Version differences:
    • This tutorial assumes QGIS 3.40
    • Older versions may have tools in different locations
    • Use Processing Toolbox search (Ctrl+Alt+T) to find tools by name

Extension Activities

Ready to go further? Try these advanced exercises:

Activity 1: Multi-Resolution Grid System

Create a nested grid system at multiple resolutions:

  1. Generate 1km × 1km grid using same workflow
  2. Generate 5km × 5km grid
  3. Style them with different colors
  4. Create spatial relationships (which 1km cells are within which 5km cells?)

Use case: Hierarchical aggregation from detailed to regional scale

Activity 2: Hexagonal Grid Alternative

Repeat the tutorial using hexagons instead of squares:

  1. In Create Grid dialog, choose Hexagon instead of Rectangle
  2. Horizontal spacing: 2000m (this is the side length)
  3. Compare hexagon vs square grids:
    • Edge effects
    • Number of cells
    • Area calculations

Question: Why might hexagons be preferred for some analyses?

Figure 22 - Hexagonal grid example

Activity 3: Custom Regional Grid

Create a fishnet for a specific province or municipality:

  1. Load provincial boundaries from PDOK instead of Landgebied
  2. Select your province of interest (e.g., Zuid-Holland)
  3. Export selection as new layer
  4. Use this as extent for grid creation
  5. Compare grid statistics between provinces

Summary

What You’ve Learned

By completing this tutorial, you have:

✓ Installed and configured the PDOK Services Plugin ✓ Loaded national boundaries from Dutch web services (Landgebied) ✓ Created a regular grid with specific dimensions (2km × 2km) ✓ Understood coordinate systems and the importance of EPSG:28992 ✓ Performed spatial clipping to match grid to geographic boundaries ✓ Calculated geometric attributes (area) using Field Calculator ✓ Applied data quality filtering to remove partial cells (< 50% coverage) ✓ Added unique identifiers and coordinate fields for data integration ✓ Prepared a spatial framework ready for CBS statistical data ✓ Visualized and styled vector data appropriately

Key Concepts Reinforced

Concept Application in Tutorial
Coordinate Reference Systems EPSG:28992 (RD New) for Dutch data
Spatial Resolution 2km × 2km cells as statistical units
Geoprocessing Workflows Multi-step process: Create → Clip → Filter
Spatial Extent Bounding box vs. actual boundary shape
Data Quality Filtering partial cells for reliable analysis
Attribute Calculations Field Calculator for area, IDs, coordinates
Data Integration Preparing framework for joining statistical data

Fishnet Grid Applications

Your completed fishnet grid can be used for:

  • Statistical data aggregation - Joining CBS population, income, energy data
  • Spatial sampling - Systematic sample point or area selection
  • Heat mapping - Choropleth visualization of density, rates, totals
  • Change detection - Temporal comparison of grid cell statistics
  • Accessibility analysis - Distance calculations from cell centroids
  • Pattern analysis - Spatial autocorrelation, clustering, hot spots

Files You Should Have

Verify your outputs folder contains:

  • ☑ fishnet_2km.gpkg - Initial full rectangular grid
  • ☑ fishnet_2km_clipped.gpkg - Grid clipped to NL boundaries
  • ☑ fishnet_2km_filtered.gpkg - Final grid with ≥50% coverage cells
  • ☑ QGIS project file (.qgz) with all layers configured

Next Steps

Immediate applications: 1. Download CBS grid data from https://opendata.cbs.nl 2. Join statistics to your fishnet 3. Create choropleth maps showing spatial patterns 4. Calculate neighborhood or municipal statistics

Further learning: - Explore other PDOK datasets (topography, cadastre, infrastructure) - Learn spatial statistics (Moran’s I, Getis-Ord Gi*) - Automate workflows with Python and QGIS processing - Create web maps with qgis2web or Leaflet

Additional Resources

QGIS Documentation

  • Vector analysis tools: https://docs.qgis.org/latest/en/docs/user_manual/processing_algs/qgis/vectoranalysis.html
  • Field calculator functions: https://docs.qgis.org/latest/en/docs/user_manual/working_with_vector/expression.html
  • Geoprocessing tools: https://docs.qgis.org/latest/en/docs/user_manual/processing_algs/qgis/vectorgeometry.html
Back to top