Data, sources and methods
Everything on this site comes from free, open data. Here is where each figure comes from, how we calculate it, what it cannot tell you, and how to reuse it.
Sources and licences
| Data | Provider | Used for | Licence | Updated |
|---|---|---|---|---|
| Weather forecast | Open-Meteo (ECMWF, DWD and others) Weather data by Open-Meteo.com | Rain, temperature, feels-like temperature and UV for the next three days | CC BY 4.0 | Every 30 minutes |
| Air quality forecast | Open-Meteo, Copernicus Atmosphere Monitoring Service (CAMS) Generated using Copernicus Atmosphere Monitoring Service information [2026] | Fine dust (PM2.5), coarse dust (PM10) and Saharan dust | CC BY 4.0 | Every hour |
| River discharge | Open-Meteo, Copernicus GloFAS Generated using Copernicus Emergency Management Service information [2026] | Densu river above Weija (Accra) | CC BY 4.0 | Every 6 hours |
| Active fires | NASA FIRMS (VIIRS S-NPP) | Fires detected in Ghana in the last 48 hours (map) | Free use, NASA data policy | Every 3 hours |
| District boundaries (MMDAs) | geoBoundaries, William & Mary geoLab | Districts and regions (boundaries as of 2019) | CC BY 4.0 | Static |
| Clinics, schools, markets, water points, streets | © OpenStreetMap contributors (via Geofabrik and Protomaps) | Facilities, distances and the street map | ODbL 1.0 | When the data is rebuilt |
| Population 2025 | WorldPop, University of Southampton | Residents per 100 m cell, district population, access figures | CC BY 4.0 | Release R2025A |
| Terrain (Copernicus DEM GLO-30) | European Space Agency, Copernicus Produced using Copernicus WorldDEM-30 © DLR e.V. 2010-2014 and © Airbus Defence and Space GmbH 2014-2018 provided under COPERNICUS by the European Union and ESA; all rights reserved | Low-lying areas | Copernicus DEM licence (free use) | Static |
| Surface water 1984–2021 | EC Joint Research Centre / Google, Global Surface Water Source: EC JRC/Google | Where water stood and how often | Free use with attribution | Version 1.4 |
| Rainfall (CHIRPS and CHPclim) | Climate Hazards Center, UC Santa Barbara | Monthly rain of the last 12 months against the long-term normal | Public domain (CC0) | Monthly |
| Surface temperature | Landsat 8 and 9, courtesy of the U.S. Geological Survey, via Microsoft Planetary Computer | Hotter and cooler parts of each city in the dry season | Public domain (USGS) | Dry seasons 2024/25 and 2025/26 |
| Land surface temperature (overview) and night lights | NASA EOSDIS GIBS (MODIS, VIIRS Black Marble) | Regional overview layers on the map | Free use, NASA data policy | Fixed dates, named on the map |
| Land cover 2021 | ESA WorldCover © ESA WorldCover project 2021 / Contains modified Copernicus Sentinel data (2021) processed by ESA WorldCover consortium | Trees, built-up land, cropland and other cover | CC BY 4.0 | Version 200 |
| Built-up area 1975–2030 | Global Human Settlement Layer, European Commission JRC GHS-BUILT-S R2023A, European Commission, Joint Research Centre (JRC) | Growth of the built-up area; 2030 is a projection | CC BY 4.0 | Release R2023A |
| Satellite image | Sentinel-2 cloudless 2024 by EOX IT Services (contains Copernicus Sentinel data) Sentinel-2 cloudless – https://s2maps.eu by EOX IT Services GmbH (Contains modified Copernicus Sentinel data 2024) | Background image on the map | CC BY-NC-SA 4.0 | 2024 |
How we rate conditions
Each city page rates rain, heat, air and — for Accra — the Densu river with fixed thresholds. Our reading of open forecast data, not an official warning. For warnings, follow the Ghana Meteorological Agency and NADMO.
- Rain, next 24 hours
- Rain expected from 10 mm. Heavy rain likely from 30 mm, or when a single hour brings 20 mm or more — enough to overflow open drains. Very heavy rain from 50 mm.
- Heat, highest feels-like temperature
- Feels-like values of 32–39 °C are ordinary in Ghana, so we rate from the US National Weather Service’s “danger” band: Very hot from 39 °C, Dangerous heat from 45 °C, Extreme heat from 51 °C.
- Air, average of the next 24 hours
- US EPA air quality breakpoints (2024), as used by IQAir. Fine dust (PM2.5): moderate from 9.1, unhealthy for sensitive groups from 35.5, unhealthy from 55.5 µg/m³. Coarse dust (PM10): from 55, 155 and 255 µg/m³. The worse of the two decides. When most of the dust is mineral dust, we name the Harmattan.
- River
- The highest forecast flow of the coming week against the same river cell over the last two years: higher than usual above the level reached on one day in ten, very high above one day in a hundred.
Methods
- Which districts belong to a city
- A district (MMDA) counts when it lies in the city’s region and either contains the city centre or lies mostly inside the built-up metropolitan area.
- Access to clinics, schools and markets
- For every 100 m cell with residents (WorldPop 2025) we measure the straight-line distance to the nearest facility mapped in OpenStreetMap, then add up the residents within 2 km and 5 km. Real walking routes are longer.
- Low-lying areas
- A 30 m cell counts as low-lying when it lies at least 5 m below the average height of its surroundings within about 500 m — valley bottoms and hollows where rainwater runs together.
- Surface heat
- Landsat 8 and 9 images from two dry seasons (December–March), with clouds and their shadows removed pixel by pixel. For each place we take the median temperature of all clear images, then compare it with the city’s median.
- Rain normals
- Long-term monthly averages (CHPclim) against the last twelve months (CHIRPS), averaged over each city’s area.
Every figure is produced by a scripted pipeline that can be re-run from the raw data, so results can be checked and updated.
Limits of the data
- Flood & Rain
- Low-lying areas come from a 30 m surface model that includes buildings and trees; they show where water tends to collect, not a flood forecast. Surface water is what satellites saw from 1984 to 2021.
- Heat & Air
- Heat maps show surface temperature from Landsat at about 10:30 on clear dry-season days, at 100 m detail — the ground and roofs, not the air you feel. Air quality is a model forecast (Copernicus CAMS), not a measurement.
- Services & Access
- Facilities come from OpenStreetMap, which volunteers keep up to date; anything not mapped is not counted. Water points are mapped so incompletely outside Accra that we show them on the map only. Distances are straight lines, not travel routes.
- Environment
- Built-up area is a satellite estimate (GHSL); 2030 is a projection. Land cover is from 2021 at 10 m.
- District boundaries
- The boundaries date from 2019. Districts created or split since then, such as parts of the Accra Metropolis, may not match today’s assemblies.
Services, API and downloads
Use these in QGIS, ArcGIS or your own code. All layers are in the workspace ghana.
- WMS
https://ghana.applied-geoinformatics.org/geoserver/ghana/wms?service=WMS&request=GetCapabilities - WFS
https://ghana.applied-geoinformatics.org/geoserver/ghana/wfs?service=WFS&request=GetCapabilities - WMTS
https://ghana.applied-geoinformatics.org/geoserver/ghana/gwc/service/wmts?request=GetCapabilities - API
https://ghana.applied-geoinformatics.org/api/cities·/api/cities/accra/live·/api/cities/accra/indicators - Fires
https://ghana.applied-geoinformatics.org/api/layers/fires.geojson
Downloads by city
- Accra Districts (CSV) Districts (GeoJSON) Facilities (GeoJSON)
- Kumasi Districts (CSV) Districts (GeoJSON) Facilities (GeoJSON)
- Tamale Districts (CSV) Districts (GeoJSON) Facilities (GeoJSON)
- Sekondi-Takoradi Districts (CSV) Districts (GeoJSON) Facilities (GeoJSON)
- Cape Coast Districts (CSV) Districts (GeoJSON) Facilities (GeoJSON)
- Wa Districts (CSV) Districts (GeoJSON) Facilities (GeoJSON)
Photos
- Accra: Office towers in Airport City, Accra, at dusk by Synth85, CC BY-SA 4.0, via Wikimedia Commons.
- Kumasi: Busy street with taxis and arcaded shops in central Kumasi by ZSM, CC BY-SA 3.0, via Wikimedia Commons.
- Tamale: Row of shops beside a large shade tree on a dusty street in Tamale by Matti Blume, CC BY-SA 4.0, via Wikimedia Commons.
- Sekondi-Takoradi: Old two-storey houses along a street in Sekondi by Kwameghana, CC BY-SA 4.0, via Wikimedia Commons.
- Cape Coast: Rooftops of Cape Coast seen from Fort Victoria, with the sea behind by Adam Jones, CC BY-SA 2.0, via Wikimedia Commons.
- Wa: Main street in Wa with shops, cars and motorbikes under a clear sky by Romain Tevels, CC BY-SA 3.0, via Wikimedia Commons.