Welcome · 09:00–09:30

Urban heat risk monitoring

Every March–May some Bengaluru neighbourhoods bake while others stay cool. Today you use free satellite pictures and a little Python to find where, who lives there, and when to warn them.

About me

I'm Aman Bagrecha, a geospatial consultant; I write at amanbagrecha.com and co-run Let's Talk Spatial, Bengaluru's free community for maps and satellites.
Presenting at FOSS4G 2023
Presenting at FOSS4G 2023, Prizren, Kosovo
Let's Talk Spatial meetup group photo
Let's Talk Spatial #22, Takshashila Institution, May 2026

Where we are going

By 3 pm you will turn the air temperature forecast into a heat alert for each of the 369 wards. Tomorrow's forecast high is 29.5 °C, +2.5 °C from an average 15 October but below the 33 °C alert line, so no ward is warned. On this year's hottest afternoon, 23 April 2026 (35.2 °C air temperature, ERA5, very hot), the same rule would have put 111 on emergency and 258 on heat alert of the 369 wards.

Today

Time
09:00 Welcome get set up
09:30 1 · Where is it hot? surface temperature from satellites, day and night
10:30 break
10:45 2 · Why are some places cooler? trees, lakes and concrete vs surface temperature
11:30 3 · Who is at risk? surface temperature, people and shade per ward
12:30 lunch
13:30 4 · When should we warn? air temperature forecast → tomorrow's alert per ward
15:00 close
Change a number in a cell and everything that depends on it reruns; hover, click and drag the charts, nothing can break.

The data

What How it is measured When Squares
Wards GBA ward boundaries (OpenCity) Dec 2025 369 wards
Surface temperature, day Landsat 8 & 9: thermal camera, 705 km up 10:40 am, Mar–May 2022–26 (44 pictures) 30 m
Surface temperature, night MODIS on Terra and Aqua: thermal camera, 8-day averages ~10 pm and ~2:30 am, Mar–May 2022–26 (105 files) 1 km
Ground cover ESA WorldCover: Sentinel-1 radar and Sentinel-2 photos sorted into 11 classes 2021 10 m
Trees Meta & WRI tree heights: an AI model reading Maxar satellite photos photos mostly 2018–2020 ~1 m
People Census 2011 headcount of each ward, in the GBA ward file (next census: February 2027) 1 March 2011 369 wards
Air temperature, past ERA5 (ECMWF): station, balloon and satellite readings fed into a weather model each day's highest and lowest, 1991 to Sep 2026 ~25 km
Air temperature, forecast Open-Meteo forecast: ECMWF's global weather model the next 7 days 9 km

Get set up

Windows (PowerShell):
powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"   # then reopen PowerShell
irm https://github.com/amanbagrecha/urban-heat-bengaluru/archive/main.zip -OutFile workshop.zip
Expand-Archive workshop.zip -DestinationPath .
cd urban-heat-bengaluru-main
uv sync
uv run marimo edit notebooks
macOS / Linux:
curl -LsSf https://astral.sh/uv/install.sh | sh     # then open a new terminal
git clone https://github.com/amanbagrecha/urban-heat-bengaluru && cd urban-heat-bengaluru
uv sync
uv run marimo edit notebooks
Then check that everything says True:
from wardheat.setup_check import checks

checks()
check ok detail
marimo True 0.25.1
numpy True 2.5.3
pandas True 3.0.6
geopandas True 1.2.0
rasterio True 1.5.2
xarray True 2026.9.0
odc.stac True 0.5.3
pystac_client True 0.9.0
planetary_computer True 1.0.0
altair True 6.3.0
matplotlib True 3.11.2
pooch True v1.9.0
Planetary Computer STAC True HTTP 200 in 0.6s
Open-Meteo forecast True HTTP 200 in 0.7s
Workshop files (GitHub) True HTTP 200 in 0.5s
Made with open data and open-source Python · marimo · Altair