Applied GIS: Urban Planning, Disasters, and Health

Applied GIS is the use of geographic information systems to solve real-world problems; in urban planning, it supports land-use decisions and transport planning, in disaster management it maps hazards, exposure and response, and in public health it tracks disease and measures access to care. In brief: planners use GIS to find suitable sites and to model growth, emergency managers use it to map flood zones, plan evacuations and assess damage from satellite images, and health agencies use it to find clusters of disease and to see who lives too far from a clinic. The best-known early example is John Snow’s 1854 map of cholera deaths in London, which linked cases to a water pump.

GIS turns data into maps and numbers that support decisions, and it does so across many fields. This guide explains how GIS is used in planning, disaster risk management and health, shows what data and methods are involved, discusses the limits and ethical issues, compares the applications in a table, works through examples, outlines the history and corrects common mistakes.

How is GIS used in urban planning?

Urban planners work with large amounts of spatial information, and GIS lets them combine it. Typical uses are:

• Land use and zoning: maps of existing uses, parcels and zoning rules, and the analysis of the effect of changes, such as whether a new development would be within walking distance of a station.
• Site selection: weighted overlay to find the best location for a school, park, housing, clinic or industrial area, taking account of access, land cost, environmental constraints and demand.
• Transport planning: network analysis for bus routes, service areas and accessibility, and the modeling of traffic.
• Growth modeling and scenarios: predicting where cities will grow, from past change detected on satellite images, and testing policies such as green belts.
• Infrastructure management: maps of water pipes, sewers, power lines and roads, used for maintenance and planning.
• Equity and environmental analysis: mapping green space per person, air pollution, heat and access to services in different neighborhoods to identify disadvantaged areas.
• Public engagement and 3D city models: web maps and visualizations that let citizens see and comment on plans, and “digital twins” of cities.

How is GIS used in disaster management?

GIS supports all phases of the disaster management cycle.

• Mitigation and risk mapping: hazard maps (flood zones, earthquake shaking, landslide susceptibility, wildfire risk, tsunami inundation) are overlaid with maps of buildings, population and critical facilities to show exposure and vulnerability, and guide land-use planning and building rules.
• Preparedness: planning of evacuation routes and shelters, with network analysis of travel times, and early warning systems that use real-time data from sensors and satellites.
• Response: mapping the extent of the disaster and the damage, finding the quickest routes for emergency vehicles, locating people who need help and coordinating relief. Satellite and drone images are compared before and after the event. After the 2010 Haiti earthquake, thousands of volunteers around the world traced roads and buildings on satellite images for OpenStreetMap, which provided maps for rescue teams within days. Tools such as Ushahidi collect crowdsourced reports.
• Recovery: tracking the progress of reconstruction, and identifying safer sites for rebuilding.

How is GIS used in public health?

• Disease mapping and surveillance: plotting cases to find clusters, spread and risk factors. John Snow’s map of the 1854 Broad Street cholera outbreak in Soho, London, showed that deaths clustered around one water pump, and the removal of the pump handle is a famous moment in the history of epidemiology. In the Covid-19 pandemic, dashboards such as the one by Johns Hopkins University, launched in January 2020, showed the spread in near real time.
• Environmental health: linking disease to pollution, water quality, heat or the habitats of disease vectors such as mosquitoes (malaria, dengue).
• Access to health care: calculating travel time or distance to hospitals and clinics, finding “service deserts” and testing the effect of new facilities. Spatial accessibility measures combine the supply of services (beds, doctors) and the demand (population) within a catchment.
• Resource allocation and campaigns: planning vaccination, mobile clinics and the siting of services to reach the most vulnerable, and the tracking of progress with spatial data.
• Health inequalities: mapping the differences in life expectancy and disease between neighborhoods.

What are the limits and ethical issues?

Results depend on data quality: out-of-date or incomplete data, or data at a coarse scale, can mislead. Models rest on assumptions (an even population within an area, constant travel speeds) and have uncertainty. The modifiable areal unit problem means that results can change with the boundaries used. Privacy is a concern, especially in health and in the use of mobile phone and location data: even maps of cases can identify individuals if the scale is too fine, so data are often aggregated or masked. There are also issues of equity: groups missing from the data (informal settlements, for example) are invisible to the analysis. Capacity and access to technology and training differ widely, and decisions should combine GIS with local knowledge and participation.

Applications compared

Worked example: access, flood damage and response time

The values are hypothetical and illustrate typical analyses.

1. Access to health care: a region has 200,000 people. Network analysis shows that the area within a 15-minute drive of a clinic contains 156,000 of them. The coverage is 156,000 ÷ 200,000 = 78%, and 44,000 people (22%) are outside the service area.
2. Choosing a new clinic site: three candidate sites would bring 12,000, 18,000 and 9,000 of the uncovered people within 15 minutes. The best site covers 18,000, raising the coverage to (156,000 + 18,000) ÷ 200,000 = 87%.
3. Beds per population: a hospital with 600 beds serves a catchment of 150,000 people. The supply is 600 ÷ 150,000 × 1,000 = 4.0 beds per 1,000 people.
4. Flood exposure: a flood map shows three depth zones: more than 1 m (3.2 km²), 0.5 to 1 m (4.8 km²) and less than 0.5 m (6.0 km²), a total of 14.0 km². With 800 buildings per km², the number of exposed buildings is 14.0 × 800 = 11,200.
5. Expected damage: with damage ratios of 60%, 30% and 10% for the three zones, the damage equals 3.2 × 800 × 0.6 + 4.8 × 800 × 0.3 + 6.0 × 800 × 0.1 = 1,536 + 1,152 + 480 = 3,168 buildings’ worth of loss, which is 3,168 ÷ 11,200 = 28% of the exposed stock.
6. Response time: a fire station is 6 km along the road from an incident, and the speed of the fire engine is 40 km/h. The time is 6 ÷ 40 = 0.15 hours = 9 minutes. If a target is 8 minutes, the station fails to reach the incident in time, and a new station is needed or the route improved.

How did these applications develop?

John Snow’s work in 1854 and Florence Nightingale’s diagrams in the 1850s are early examples of spatial analysis in health. Planners used overlays by hand in the 1960s, notably Ian McHarg, and the first computer GIS were applied to land management in the 1960s and 1970s. Disaster mapping advanced with remote sensing and GPS in the 1990s, and the 2005 Hurricane Katrina and 2010 Haiti earthquake highlighted the use of web mapping and volunteered information. Public health GIS expanded with the availability of geocoded data, and the Covid-19 pandemic made maps and dashboards part of daily news. Today, real-time data from sensors and phones, drones, artificial intelligence and open platforms such as OpenStreetMap and Google Earth Engine are widening the use of GIS for planning, risk reduction and the Sustainable Development Goals.

Which mistakes are common?

Mistake 1: Using straight-line distance for access to services

People travel along roads and paths, and barriers such as rivers matter. Travel time on the network is a better measure of access.

Mistake 2: Mapping counts of cases instead of rates

A map of raw numbers shows where people live. Rates per 1,000 people show where the risk is higher.

Mistake 3: Forgetting uncertainty in hazard and damage maps

Flood and damage maps are model outputs with errors. They should be shown with ranges and checked against local knowledge.

Mistake 4: Ignoring privacy and fairness

Detailed location data can identify individuals, and data gaps may leave some communities out. Aggregate sensitive data and involve communities in the analysis.