Community Heat Mapping

Early in the morning of July 31st, 2022, fifteen pairs of students and researchers from the Institut Teknologi Bandung mounted highly sensitive sensors on their bicycles and motorbikes and set out to capture ambient heat data across the City of Bandung, Indonesia. Their common purpose was to better understand how temperatures vary across their urban environment on a hot day. The volunteer data-collectors rode with their sensors through dense urban corridors, lush green campuses, rural farmland, and many other landscapes found across Bandung. As they navigated each terrain, their sensors gathered thousands of samples of air temperature and humidity across locations. They repeated their data collection routes again in the afternoon and evening.

Volunteers and researchers hypothesized that they would find places in Bandung to be hotter than others based on the common idea that materials of the built environment like concrete, asphalt and buildings trap and retain the sun’s radiation, whereas tree canopy and vegetative areas retain less heat and help to cool surrounding areas.

On the following weekend, the volunteer researchers furthered their investigation by visualizing the relationship between landscape and heat with a photo-mapping activity. Using thermal imagery cameras, they measured surface temperatures at various landscapes throughout Bandung while reflecting on the adaptation behaviors and infrastructure changes that might help to cool each location.

It is well known that the differences in temperature between urban neighborhoods can have significant and differential impacts on the health and wellbeing of local communities. The involvement and perspective of community members is key to better understanding this issue. By engaging community members in this data collection campaign we aim to advance the role Bandung residents play in determining effective solutions to climate change.

In this Story Map we share back the results from Heat Watch Bandung in two main components: first, the air temperatures measured by mobile traverses along with area-wide maps and analyses made from the data; and second, outputs of the photo-mapping exercise with thermal surface temperature photos and survey responses. Using these insightful datasets, planners and researchers can leverage the investigations of community members to better understand the issue of urban heat and advance equitable solutions in Bandung.

Heat Watch Bandung was conducted as a partnership between CAPA Strategies, The Institut Teknologi Bandung, and The World Bank.

Pictured below are data collection teams in action during the mobile campaign. As volunteers navigated their traverses across the city, their heat sensors measured temperature, relative humidity, and location with a unique measurement each second. Volunteers rode the same traverses at morning (6 to 7am), afternoon (1 to 2pm), and evening (5 to 6pm).

The maps below present the temperature measurements of each gathered data point. Using your mouse, you can zoom in and click on the points to see the temperature values. Pay close attention to the surrounding landscape – near a red hot point, what do you see? Are there more pavement and buildings nearby? Near a cool blue point, do you see more tree cover or parks?

Using the traverse data and recent satellite imagery, analysts at CAPA Strategies combined the datasets into machine-learning algorithm and generated air temperature models at morning, afternoon and evening. The models, shown below, depict modeled air temperature at a resolution of 10 meter by 10 meter pixel.

We can here investigate the relationship between heat and land cover by examining the underlying satellite imagery, and we can also see the shifts in temperatures across the area between the three time periods.

You can view all of the area-wide models along with the corresponding traverse data together in a  web-map  and also learn more about how these models are constructed in the Heat Watch Bandung  summary report .

To next examine the relationship between heat and land cover across the entire city, we can isolate two variables commonly associated with heat: tree canopy and impervious surface amount. Looking at each variable side-by-side with the heat data, we can visualize these relationships at varying scales across Bandung.

In the map slider below, the average temperature across morning, afternoon and evening is summarized by village. The canopy layer presents the percentage of that village that is covered by tree canopy. Sliding the arrows from left to right, can you see a relationship between areas with high temperature and tree canopy? How about for areas with low temperatures?

Below is a similar set of maps with impervious surface amounts by village. Considering that materials like asphalt and concrete are known to absorb and retain heat, can you see a pattern between heat and impervious surfaces here?

We can also notice a relationship between population density by village and heat.

The patterns between these variables of heat, land cover and population density uncover an important aspect of the impact of the built environment on heat and population exposure.

On the weekend following the mobile campaign (August 6th & 7th), volunteers investigated heat across Bandung’s urban environment using a different set of tools: FLIR thermal imagery cameras, and their personal intuitions about people, place, and heat. FLIR cameras provides a thermal snapshot and estimation of surface temperatures within nearby view.

Geared with the FLIR cameras and a survey application, volunteers teams spread across various landscapes of central Bandung to capture images of heat. They took photographs and recorded people’s coping strategies to heat as well as ideas on how to mitigate heat in each area.

Slide the arrow icon from left to right in the images below to compare the normal photo with its thermal equivalent. Below each photo are survey responses provided by the volunteers that visited the location.

Why is this place hot or cool? The area is very hot because it is a built-up area without any trees.

What are coping strategies and potential improvements here? Avoid locations that are exposed to direct sunlight, such as shelter and activities under the building canopy or inside buildings.

Why is this place hot or cool? Dense settlements, lack of trees, and zinc roof material makes the air hotter.

What are coping strategies and potential improvements here? Create a garden and increase the vegetation in this area. Also, change the roof of the house to tile.

Why is this place hot or cool? There is vegetation that casts a shadow.

How are people affected by heat? The heat is quite influential on people who are active around the location. Those affected by temperature conditions around the location area are tourists, pedestrians, and street vendors.

To view all of the results from the photo-mapping activity with photos and survey responses, navigate through the map and photo slide deck below.

You can also view the full set of FLIR survey responses in alternative  web-map  format.

The anecdotes provided by the FLIR photo-mapping exercise provide valuable insight into the lived experience of Bandung residents. Correlating these stories with the mobile heat data, models, and surrounding land-cover offers a robust evidence base and understanding into the issue of urban heat in Bandung. Using these data, policymakers can advance efforts towards mitigation strategies with targeted information and grounded knowledge.

The insights gathered through Heat Watch Bandung could not have been possible without the local volunteers & researchers at the Institut Teknologi Bandung -- thank you to all for your time and energy.

This activity was made possible through financial support from the Global Facility for Disaster Reduction and Recovery.

To provide broad access to the results, the summary report and datasets are available  here.