I recently taught a 1-month long course on GIS Applications in Urban and Regional Planning. We explored how GIS can be applied to solve problems in 6 different thematic areas. In this post, I will outline different applications and show concrete examples of using open-datasets and open-source GIS software QGIS.
The full course material – including data packages and PDF handouts – is now available for free download. Download [qgis_urban_planning.zip] containing detailed step-by-step instructions and datasets. These materials are ideal for self-study and complement the QGIS tutorials on my website.
Here are the 6 thematic areas
- Land Use Planning and Management
- Crime Mapping and Analysis
- Solid Waste Management
- Urban Infrastructure and Utilities
- Urban Transportation
- Spatial Planning
1. Land Use Planning and Management
Land use refers to the human activities on land. The key topics in this category where GIS is critical are
- Land Use Mapping: Present spatial distribution of land cover, land use and zoning
- Land Use Planning: Determine the desired future development pattern in a given area
- Land Use Analysis : Identify land use patterns and distributions and developability of a proposed land use.
- Land Suitability Analysis: Determine optimal functional use of the land considering social, physical, spatial and economic factors
Here are some example GIS Applications that show how spatial analysis techniques can be applied for such tasks.
1.1 Creating a Zoning Map
Zoning designations define and regulate what kinds of uses are allowed on specific parcels of land and outline design and development requirements and guidelines. Here’s an example of using QGIS Symbology and Print Layout to create a zoning map from using the zoning and split-zoning data layers from The City of Cape Town Open Data Portal (ODP)
1.2 Identifying Informal Settlements
Land-Cover data is a crucial reference dataset that informs a wide variety of strategic planning activities. The South African National Land Cover dataset is an excellent resource that provides land cover rasters in 3 epochs: 1990 <-> 2014 <-> 2018 under an open data license. We can use this dataset for quantification of landscape change over a 25 year period.
Using simple raster algebra operation using QGIS Raster Calculator one can identify areas under informal settlements.
1.3 Mapping URBAN GROWTH PATTERNS
We can use land cover change datasets that shows class transitions between 2 epochs to identify all pixels that have transitioned from a non-urban class to an urban class. This can be done using the Reclassify by layer algorithm in QGIS to take the input multi-class raster and convert it to a 3-class raster consisting of non built-up, existing built-up and new built-up areas. The resulting layer can be styled to show the urban growth pattern between the 2 epochs.
1.4 Determining Landuse Buffer Zones
A useful spatial analysis techniques is to determine a buffer of restricted area around all parcels belonging to a specific land use type. Such analysis can be used to establish a corridor of restrictions around institutional land use for controlling noise pollution or heavy traffic.
Here’s an example that uses Land use shapefile for San Francisco provided by DataSF Open Data Portal and establishes a buffer zone of 100ft around each institutional landuse parcel using the Buffer tool in QGIS with
Mitre join style for rectangular-buffers.
2. Crime Mapping and Analysis
There are several policy, planning, governance and technological approaches to address urban crime dimensions and its impacts. The applications of GIS are primarily in the following areas
- Crime Classification and Mapping: Aggregation and classification of crime incidents
- Crime Hotspot and Density Analysis: Identify crime heatmaps and prepare crime density maps
- Surveillance and Infrastructure Mapping: Mapping and coverage analysis of CCTV cameras, lighting infrastructure etc.
- Crime Prevention: Applying multicriteria analysis techniques for allocation and optimization of resources
A few curated examples that demonstrate how GIS is critical for crime mapping, analysis and prevention strategies.
2.1 Mapping Crime Statistics
It is important to understand the pattern of crime across administrative regions. Here’s an application to take crime statistics data and createt a choropleth map showing distribution of crime rate across police station boundaries.
We take the Station Boundaries and Points shapefiles provided by South African Police Service(SAPS) and join it with burglary counts from Crime Statistic of Republic of South Africa using the Join Attributes by Table algorithm in QGIS. We then normalize the raw counts using population data raster from WorldPop using the Zonal Statistics algorithm. The result is a choropleth map of crime rate.
Displaying statistics linked with spatial data in a map is very powerful. We explore the DataPlotly plugin to create a time-series chart of burglaries.
2.2 Crime Hotspot Identification and Mapping
There are 3 primary techniques for mapping crime incidents.
- Point mapping: Individual incidents are mapped with 1 point for each incident. Techniques such as using ‘Feature Blending’ mode in QGIS can help show areas of high number of crimes and create a dot map like this.
- Binning: Aggregate the individual incidents over a rectanglular or hexagonal grid to show areas with high crime. This can be achieved in QGIS with the Create Grid and Count Points in Polygon algorithms. One needs to be aware that this technique suffers from the Modifiable Area Unit Problem (MAUP) and care needs to be taken to test various configurations before drawing conclusions from the results.
- Heatmap / Kernel Density Estimation: Hotspot mapping is the most widely used technique to identify areas of concentrated crime. This is the most robust technique for identifying crime hotspots. A fine grid is generated over the point distribution. A moving window (i.e. Kernel) of a specified radius visits each cell in the grid and calculates weights for each point within the kernel’s radius. The final value of the grid is determined by summing values from all points
Mapping Crime: Principle and Practice: Research Report and Mapping Crime: Understanding Hot Spots: Special Report from National Institute of Justice, U.S. Department of Justice are very references for practitioners working in this domain. Though a bit dated, the material covers the topic in depth and gives practical advice that is still relevant.
We take a dataset of bicycle robberies in Greater London area and identify
high-crime-density areas. QGIS can create such heatmaps easily using the built-in Heatmap renderer for point layers.
We will also see how we can explore the ‘temporal’ dimension of the dataset and see how the hotspots change month-to-month. This is achieved in QGIS using the Temporal Controller that can display time slices and create animations.
3. Solid Waste Management
The applications of GIS in Solid Waste Management can be categorized into following
- Infrastructure: Visualizing and analyzing spatial spread of waste management infrastructure and capacity (i.e. bins, landfills, recycling centers etc.)
- Service Delivery: Identifying underserved areas and optimizing waste lifting cycles
- Resource Mobilization: Identifying area potential and budget allocation for waste management
Here are some applications that show how GIS can be applied to solve waste management challenges
3.1 Mapping Waste Disposal Volumes
We take a spreadsheet of waste entering city disposal facilities and a shapefile of landfill sites from the City of Cape Town Open Data Portal. We create a proportional-symbol map showing amount of waste processed at each landfill using Data-defined Overrides in QGIS. The result is a beautiful and informative data visualization that also uses Data-defined Size Legend feature to show the distribution of waste across different sites.
3.2 Service Area Analysis
Continuing to work with waste infrastructure data for the city of Cape Town, we will take locations of waste collection facilities and determine which areas of the city are within 15-minutes of driving time. This will allow us to determine opportunities for improving the service delivery with potential new locations. The service area computation is done via the excellent ORS Tools plugin for QGIS using the Isochrones From Layer algorithm. We get polygons representing actual drive times along the road network – instead of more commonly used circular buffer zones.
3.3 Location-Allocation Analysis
Location-allocation analysis is used locate a set of new facilities such that the travel cost from facilities to demand-areas is minimized and assigns the closest facility to each demand point. The ‘travel cost’ can the total distance or travel time. We take the Refuse Collection Beats polygon centroids as Demand Points and locations of Waste Drop-off Facilities as Facility Locations and evaluate 2 potential waste drop-off facilities to determine which facility is the optimal location to minimize the overall travel cost for citizens. The network analysis is done via the Distance Matrix algorithm provided by QNEAT3 plugin for QGIS.
4. Urban Infrastructure and Utilities
Cities are generally faced with infrastructure and service delivery challenges and GIS can be applied to effectively solve them. The applications primarily fall into the following categories:
- Asset Management: Cataloguing, operation and maintenance of existing infrastructure such as water supply network, sewerage and storm water drainage, street lighting, and telecom
- Service Delivery: Identifying gaps and planning for new infrastructure
4.1 Field Data Collection
Mapping infrastructure and associated attributes is a key step in asset management. We use the QGIS based open-source mobile field data collection app Mergin Maps to design a form to survey streetlights.
Each participant does a survey of streetlights around a street block using the mobile app and then sync the field data to QGIS desktop. Loading the Dark matter basemap by CartoDB and applying a shapeburst fill symbology can be used effectively to visualize un-lit spots in the neighborhood.
4.2 Multicriteria Overlay Analysis
Multi-criteria overlay analysis is the process of the selecting areas on the basis of a variety of location attributes. We will apply geoprocessing techniques on multiple vector layers and find areas to build new parking facilities that can help reduce bicycle thefts.
Using Cycling Infrastructure data from Transport for London and London Individual crime and anti-social behavior (ASB) incidents from Police.uk we apply multi-criterial GIS analysis techniques to propose a new bicycle parking facility in the Westminster borough of London to reduce thefts. We use the following criteria
- The proposed area must be in a bicycle theft hotspot
- The proposed area must be within 50meters from a bicycle route
- The proposed area must NOT be within 50 meters from an existing bicycle parking
Using the vector layers representing each criteria, this analysis is carried out using the Buffer, Intersection, and Difference tools from the QGIS Processing Toolbox.
5. Urban Transportation
Spatial data and spatial analysis are key to effective transportation planning. Urban planners can apply GIS for
- Walkability: Determining conducive urban spaces for walkability at a city/ neighbourhood level.
- Mobility: Accessibility and coverage of various modes of transport
- Traffic: Analysis of traffic hotspots and congestion patterns
- Transit Oriented Development (TOD): Integration of land use, population and public transit network for site suitability and route planning.
- Smart Transportation: Analyzing transportation network for suitability of smart sensors, identifying locations for EV charging stations etc.
5.1 Analyzing Metro Rail Accessibility
When planning for transit-oriented development (TOD), a useful criteria is accessibility to public transport. We can apply spatial analysis techniques to determine what percentage of population in a given city lives in close proximity to metro stations. We first query OpenStreetMap database via the QuickOSM plugin in QGIS to get the location of functioning metro stations in the city of Bangalore, India. Then we calculate a 1km buffer and use the Zonal Statistics algorithm from QGIS on a population grid from WorldPop. The result of our analysis shows what percentage of the city’s population has easy access to the metro rail system.
6. Spatial Planning
Spatial planning is an interdisciplinary activity that takes a structured approach in terms of targeted areas interventions by looking at spatial patterns, trends, growth, integration, economic activities, infrastructure and limitations across the city. The key areas for GIS applications are
- Urban Space: Analyze patterns of land use (green spaces, informal settlements) across different ward/regions.
- Biodiversity/Ecology: Identify the eco-sensitive areas, represent overall biodiversity plans in terms of regional ecosystems (biomes, vegetation, wetlands, rivers, ground water, marine, others)
- Spatial Economy: Analyze the spatial pattern of the city in terms of economic hot spots and economically backwards areas requiring investments. Identify the potential urban and rural economic drivers.
- Integrated Planning/Spatial Development Framework: Identify urban growth corridors
- Smart City: Identify the potential areas for smart eco-development zones, walkable zones, high internet connectivity etc.
6.1 Analyzing Street Connectivityfor Walkability
Taking inspiration from the Johannesburg Spatial Planning Framework 2040, we replicate the walkability analysis using OpenStreetMap data and QGIS. We use several tools from the QGIS Processing Toolbox – Line Intersections, Dissolve, Multipart to Singleparts, and Delete Duplicate Geometries. By chaining these tools in a GIS workflow, we build a model to compute the density of street intersections. Through this example, participants learn how to chain various tools to build a typical GIS workflow.
Once we have the results, we design a data visualization using advanced cartography and labelling tools in QGIS including Label Callouts – which allow for flexible label placement.
Hope this post gives you an idea of the possibilities of spatial analysis and visualization in the field of urban planning. Open data and Open-source tools like QGIS allow anyone to adopt these technologies in making our cities better using a data-driven approach.
Special thanks to Aurobindo Ogra whose expertise and guidance were critical in the development of these lab exercises.