Heat Watch Cape Town
Community Heat Mapping in Cape Town, South Africa
Heat Watch Cape Town
Source: World Bank Climate Change Knowledge Portal
Due to the effects of climate change, extreme heat in South Africa poses significant risks across urban areas. The region will likely experience an increase in the number of hot days (defined by a maximum temperature over 35°C) and warm nights, extending the heat season and increasing heat-health risk to the population. With the compounding pace of urbanization and worsening temperature projections, South African cities will see continued pressure on public health, energy infrastructure and economic productivity. With one of the highest income inequality rates in the world coupled with a legacy of exclusion and insufficient employment opportunities, concerns of maintaining livelihoods with worsening levels of exposure to climate-induced threats is of eminent concern. Though it is a middle-income country, South Africa faces high poverty and inequality which contributes to its significant vulnerability to climate change impacts. High potential for disproportionate effects of these hazards at the local level will likely hinder mitigation and adaptation efforts.
In response to these growing challenges, a group of local climate researchers and community volunteers recently engaged in a multi-day field campaign across Cape Town to better identify the exact locations facing greater exposure to extreme heat. The campaign was part of a larger effort across South Africa in three cities total including Cape Town, Tshwane , and Buffalo City . Using a well-established participatory mapping process, "Heat Watch", the team generated tens of thousands of air measurements in a single day, helping to achieve two main objectives:
- Develop high-resolution descriptions of the distribution of air temperature, heat index, and air quality across Cape Town;
- Engage local communities to foster relationships and build heat awareness amongst researchers, public officials and residents.
Heat Watch Cape Town volunteers.
Results gathered by volunteers indeed indicate differences in air temperature, as wide as 16.3°C across the city at the same time, confirming a significant level of disparity in heat exposure and risk to health effects. Examining the built environment, hotter measurements were found to correspond with greater amounts of impervious surfaces and lesser amounts of tree canopy in surrounding areas.
Combining the heat measurements with satellite-derived landscape descriptions through geostatistical modeling, we generated high-resolution models of temperature and heat index across the city throughout the morning, afternoon and evening, as well as an average model of daytime distribution. These new heat descriptions improve on otherwise coarse surface temperature descriptions derived from satellites, which can tend to overestimate air temperatures and offer little access for use by local communities.
Lastly, as a complement to the air measurements, volunteers also conducted a photo-mapping exercise at various heat-relevant "places of interest" across each city. The resulting anecdotes and thermal photographs from Heat Watch volunteers help to describe personal experiences in specific locations and generate community-based ideas for cooling solutions.
Digital and thermal imagery camera side-by-side, Cape Town.
These new datasets provide robust sets of information for a variety of purposes towards strengthening heat resilience in South African cities. By engaging community members in this data collection campaign we aimed to advance the role residents play in directing effective solutions to climate change and bring a degree of "civic legitimacy" to the data products. Using this new evidence base and hazard description, planners and researchers can leverage the investigations of community members to better understand the issue of urban heat and advance equitable heat solutions in Cape Town and South Africa as a whole.
In this Story Map, we share how volunteers generated the datasets, tour a sample of the resulting maps, and overview a set of possibilities for applying these data towards heat mitigation and adaptation actions.
Heat Watch South Africa was conducted as a partnership between CAPA Strategies, Blessing Mancitshana, Charlton Ziervogel, Ntombovuyo Sibutha, and Khashifa Chilwan at the Community Organisation Resource Center (CORC) in Cape Town, and The World Bank.
Campaign volunteers in action, learning to use Heat Watch sensors.
Data Collection
On the morning of each mapping day, volunteers arose early to be in position for their morning data collection route, from 6:00am to 7:30am. Volunteers mounted their Heat Watch sensors to the windows of their vehicles, started their navigation software, and turn-by-turn gathered ambient measurements through residential neighborhoods, parks, industrial districts, rural areas, and the like.
As they drove, their sensors sampled the temperature and humidity of each area at a rate of one measurement per second, resulting in thousands of measurements per traverse. They repeated the same traverse paths at afternoon (14:00 to 15:00) and evening (18:00 to 19:00) in each city. On a similar hot day, they visited several "places of interest" with FLIR thermal cameras to visually investigate surface temperatures and record observations of the built environment and human behaviors. Paired in teams, volunteers recorded their experiences throughout the campaign with photo and video imagery.
The single day for data collection in each city ("campaign day") was selected for representing typical conditions that produce urban heat island effects: high temperatures, low cloud cover, low wind, and no precipitation. In the results we expect to see structures such as buildings, roads, and other infrastructure absorbing and re-emitting the sun’s heat more than natural features such as forests and water bodies. The method of collecting many in-situ ambient measurements simultaneously across a study region provides a detailed description of how ambient heat varies across urban landscapes.
Typical surface and air temperature patterns across urban to rural environments (Source: EPA)
The following section presents maps featuring field data products generated for Cape Town. Initial data interpretations below provide an initial look into urban heat distribution in Cape Town. Further exploration can be done in the interactive web map. Products include:
(1) Temperatures measured along the data collection traverses ("Heat Data"); (2) Temperature models and comparisons with land-cover ("Heat Models"); (3) Thermal imagery photographs and survey responses ("Photo-Maps").
Heat Watch Sensor collecting temperature and relative humidity measurements during the Cape Town Campaign.
Heat Data
We will next examine the trends across the study area, though each neighborhood also displays variation that is worthy of examining. Following a successful campaign day, CAPA retrieved and processed the collected data into traverse point datasets and maps. The below summary statistics table briefly describes results from for Cape Town as well as the other two South African Heat Watch cities, Buffalo City and Tshwane alongside several key geographic attributes. Note that "Max Temperature" indicates the maximum temperature measured during any of the three traverse periods and "Max Difference" describes the widest simultaneous range of temperatures across the three time periods.
City | Study Size (Area) | Population (People) | Min Elevation (meters) | Max Elevation (meters) | Max Temperature (Measured) | Max Difference (Measured) | Campaign Day (2024) |
---|---|---|---|---|---|---|---|
Cape Town | 208 km 2 | 4,772,846 | 0 | 124 | 41.6°C (Afternoon) | 16.3°C (Afternoon) | February 28th, 2024 |
Tshwane | 208 km 2 | 4,040,315 | 1272 | 1574 | 34.1°C (Afternoon) | 6.1°C (Evening) | March 5th, 2024 |
Buffalo City | 165 km 2 | 975,225 | 0 | 634 | 39°C (Afternoon) | 17.4°C (Afternoon) | March 1th, 2024 |
Measurement summary by city.
Several patterns can be observed from these statistics.
- A simultaneous temperature difference of 16.3 °C was recorded in the afternoon in Cape Town, indicating that disparities may be more present during this time.
- While the elevation range is smaller for the Cape Town study area, the surrounding geographical features and distance to the coast may be a factor in heat distribution.
We will next examine the trends across Cape Town and examples of local variation.
Visualized here are temperature data collected from the routes (or "traverses") in Cape Town. Colors closer to red indicate warmer temperatures, and those closer to blue indicate cooler temperatures. You can click each point to view the measurements.
Examining the temperature data collected from the afternoon periods in Cape Town and the surrounding land cover reveals several emergent patterns.
Heat concentrates in the middle of the study area during the afternoon collection period.
In this highly urbanized region of the city, where development is dense, some of the hottest measurements are gathered.
Image taken by a Heat Watch volunteer.
View here Albert Road and New Market Street. These streets are surrounded by impervious surfaces with little vegetation.
Despite its proximity to Cape Town's urban center, several neighborhoods showcase the effect that urban greenspaces have on moderating urban heat.
Surrounding nearby Hartleyvale Stadium, vegetation appears to be contributing to a cooling effect over the course of the day. Temperatures recorded here during the afternoon collection period are lower than nearby Woodstock and District Six neighborhoods.
Photos taken by volunteers here affirm that this area contains both dense vegetation in park areas.
And lining streets.
Compared to the concentration of heat in Downtown Cape Town, traverses collected in the southern portion of the Cape Town study area appear much cooler.
Here we notice a clear difference in temperature that seems to be moderated by the adjacent green space.
For instance, during the afternoon collection period, cooler temperatures are reported near the Swartklip Rd Sports Field
And adjacent to Bongani.
Similar results are provided from the morning and evening collection periods.
Full sets of the results are available at the following link: Heat Watch Cape Town Web-Map
Note that each web-map also features relative humidity measurements and heat index models. Heat index is a calculation of how hot it feels when humidity is factored in with temperature. On the particular days of mapping in all three cities, relative humidity levels measured too low to elevate heat index above temperature. However, historical data in these areas do indicate that relative humidity can occasionally be high enough on hot days to raise heat index and intensify the effect of heat on the human body. Heat Watch data may be helpful in understanding the relative distribution of humidity across these cities under more humid and hot conditions.
Heat Models
After processing the traverse points, we integrated high resolution remotely-sensed data from the Sentinel-2 satellite and correlated patterns of measured temperature with surrounding land cover in a geostatistical model. The resulting outputs are temperature models (10-meter resolution) that span each study area at morning, afternoon and evening, as well as an average map.
In the morning time, we see wide differences in temperature, with heat concentrated in the center city and cooling towards rural, vegetated areas.
The afternoon period indicates the cooling effect that canopy cover has across the urban area, as shown in the traverse data.
The vegetation and open space in the southern portion of the study area seems to have a cooling affect, in addition to the proximity of the coast.
Compared to the northern portion of the study area, in which high concentrations of impervious surfaces in industrial areas appear to be contributing to increased temperature.
Image taken by a Heat Watch Volunteer.
The Montague Gardens area provides an example of an industrial area's ability to raise the overall temperature in an adjacent residential area.
In the evening, heat is again most concentrated in the center city with natural and undeveloped areas cooling off more quickly.
During this period, several distinct urban heat islands form.
Such as Woodstock, District Six, and Salt River.
The Dunoon / Killarney Gardens Areas
And the Parow East / Hrdekraaltjie Neighborhoods.
The average model combines the three time periods with equal weight, providing a picture of daytime exposure.
As explored in the previous maps, we can track patterns between concentrations of heat and existing land cover variables like canopy cover and impervious surface presence as well as population density. Looking at each variable side-by-side with the heat data, we can visualize these relationships at varying scales across each city. The patterns that emerge reveal important considerations as these cities grow and change.
Canopy Cover Through the heat-mitigating effects of shade and evapotranspiration, we expect to see areas with greater amounts of tree canopy cover align with lower temperatures, and vice versa. From the regional perspective of Cape Town, we see most sub-places feature low amounts of tree canopy cover; the temperature gradient from north to south that may dominate the local effects of land cover at this scale. The graph to the right shows a comparison of the afternoon temperature and percent canopy of each sub-place boundary.
Canopy presence and average temperature model, Cape Town (Click to zoom; legends to right)
Impervious Surfaces Dense, built-up materials like buildings and pavement tend to concentrate and reemit the sun’s energy as heat throughout the day and night-time; we expect to see areas with greater amounts of impervious surfaces contribute to more significant heat islands. From the regional perspective of Cape Town, we see most sub-places feature high amounts of impervious cover; the temperature gradient from north to south that may dominate the local effects of land cover at this scale Examining each sub-area individually may help to isolate patterns between land cover and heat.
Impervious surface presence and average temperature model in Cape Town. (Click to zoom; legends to right)
Population Density Examining population density in regards to heat can help us determine the number of people who are exposed to various levels of heat. In this analysis, we summarized the number of people within each subplace with the resulting temperature from the afternoon map. The map below shows a mixed pattern between temperatures and population density; the scatter plot to the right displays how many people are exposed to five quintiles of the afternoon temperatures.
Population density against average temperature model in Cape Town. (Click to zoom; legends to right)
Photo-Maps
FLIR One thermal camera attached to smart phone.
To further capture the experience of heat in South African cities, volunteers also conducted a photo-mapping activity at various locations throughout the campaign day. Using thermal imagery cameras attached to their smartphones, volunteers visually investigated surface temperatures and noted physical and social qualities of the surrounding environment. They then recorded their observations at each site and responded to a series of prompts through the platform Survey123. With these data CAPA generated a web-map featuring the responses below.
The green icons here represent a survey response from a volunteer, with relevant photos attached. We will highlight a few example responses in the following slides.
Image taken by a Heat Watch volunteer at this location.
Location Description: "Informal settlement they use zink for houses. Close by there is Informal rank, democracy phoenix. Also its crowded."
Does this location feel warmer or cooler than the rest of your area? "Much warmer."
Do people seem to be affected by heat in this location? How so? "They are sweating and covering their faces."
What are potential strategies or improvements for cooling here, if needed? "Changing building material and crowding."
Slide between digital and thermal imagery taken at this location.
Image taken by a Heat Watch volunteer at this location.
Location Description: "Houses are close to each other. This place is also close to a bus road."
Does this location feel warmer or cooler than the rest of your area? "Average."
Do people seem to be affected by heat in this location? How so? "Yes they are affected, people are wearing caps and look tired."
What are potential strategies or improvements for cooling here, if needed? "None needed."
Slide between digital and thermal imagery taken at this location.
Location Description: "Mixed use area, gas stations, residential and commercial purposes. Very busy area with semi built up structures for both residential and commercial. The area is surface all over and drainage facilities."
Does this location feel warmer or cooler than the rest of your area? "Average."
Do people seem to be affected by heat in this location? How so? "Yes. Most people observed in the area are mainly under shed and all have sun hats."
What are potential strategies or improvements for cooling here, if needed? "This is a commercial area hence not easy to add more things infrastructure not even growing more trees."
Slide between digital and thermal imagery taken at this location.
Based on the images and responses captured during the photo-mapping exercise, we note several initial findings:
- Based on FLIR images and survey responses, many high density areas in Cape Town display very little tree cover and small, non-shade providing vegetation if any.
- Residents that reported cooler temperatures in their yards are surrounded by abundant tree cover.
- Crowding and insufficient building material was reported in areas with dense informal settlements toward the northern portion of the study area.
- Frequent suggestions to improve conditions include:
- Planting trees and vegetation.
- Changing building material / infrastructure upgrades.
- On very hot days like campaign day, survey respondents reported few people outdoors, and those that did exhibited discomfort i.e. sweating, covering faces, looking fatigued.
Next Steps
The results of Heat Watch Cape Town provide new insights into the increasing risk extreme heat poses to the city and surrounding areas through descriptions of ambient heat and relationships with the built environment and human experience. As snapshots-in-time, they describe present conditions of heat in these cities and serve as a basis of evidence for taking action on heat.
The primary takeaway from these activities is that on a given hot day, heat is not evenly distributed within Cape Town, expecially in the afternoon: in the morning, a temperature differential of 11°C was measured; in the afternoon, 16.3 °C, and in the evening, 13.8°C. Depending on where people live, work, travel and recreate in these cities, they may experience disparate levels of exposure to heat, and face higher risk of heat-related health conditions.
Together these distributional descriptions identify areas and locations across Cape Town that are experiencing disparate levels of urban heat. Broad physical interventions like increasing shade cover and decreasing the amount of heat-retaining materials like concrete and asphalt are likely to mitigate the effects of urban heat. Integrating the results with social vulnerability information will help to identify the populations most in need of awareness building and social safety net solutions.
As the work on heat progresses in Cape Town, collaborative efforts such as Heat Watch that involve community, local researchers, and supporting partners will be essential for effectively managing risk and enhancing resilience to climate change. As a first step to continuing the impact of this project, the CAPA and World Bank team recently met with the Cape Town to review these campaign products and consider direction applications of the maps towards municipal climate adaptation efforts. Potential directions included further research on heat and the transportation sector, integrating high-resolution information into early warning and heat advisory systems, increasing awareness of the relationship between heat and trees, and protecting existing trees during development. In addition, examining variations at smaller, neighborhood scales will help to apply the Heat Watch data more locally.
Acknowledgements
Heat Watch South Africa was conducted by CAPA Strategies in partnership with the World Bank Group and local implementation partner “Community Organisation Resource Center (CORC)”. We would like to especially thank local campaign organizers Charlton Ziervogel, Blessing Mancitshana, Ntombovuyo Sibutha and the entire CORC team. The insights gathered through Heat Watch Buffalo City would not have been possible without the local volunteers -- thank you to all for your time and energy.
To provide broad access to the results, the summary report and datasets are available here .