# Malaria Atlas Project data platform

> MAP provides modelled malaria risk, burden and intervention estimates, tabular administrative summaries, and spatial data. The interactive platform is at https://data.malariaatlas.org/ and the main project website is https://malariaatlas.org/.

The interactive explorer requires JavaScript. The public HTTP services below can be used independently of the explorer. These notes describe existing services; they do not grant additional access rights or change data licences.

Tabular administrative estimates are available directly from the Trends API. Raster download and aggregation are not required when the published administrative summaries match the requested metric and geography.

Releases can have different available metrics and years. A higher release number does not imply that every metric from an earlier release is available. Select and report an explicit release; do not assume the backend default is the newest release. The example below uses release 202508 and is not a recommendation to use that release for every question.

## Licensing and attribution

- [MAP open access policy](https://malariaatlas.org/open-access-policy/): MAP maps are available under the Creative Commons Attribution 3.0 Unported licence. MAP asks users to provide the correct citation when using its maps.
- [Creative Commons Attribution 3.0 Unported (CC BY 3.0)](https://creativecommons.org/licenses/by/3.0/): When reusing licensed maps, give appropriate credit, retain supplied attribution details, link to the licence and identify adaptations. Follow the licence terms and the citation information supplied with the maps.

Survey records and routine case data collated by MAP are owned by their original sources. MAP makes them available when they come from public sources or permission has been granted, and requests citation of the original source details supplied with downloads.

The policy explicitly assigns CC BY 3.0 to MAP maps; it does not establish that every tabular or third-party dataset uses that licence. For other datasets, follow the licence and citation information supplied with the dataset. Preserve those details in derived outputs and cite the dataset used, not only the platform homepage.

## Citation

Follow any dataset-specific citation instructions, including references to research publications and original sources. For MAP-produced datasets without a supplied citation, use this suggested fallback:

> Malaria Atlas Project. [Dataset or layer title], release [version, if available]. [Specific source URL]. Accessed [date].

Use the actual layer or dataset title, release identifier and most specific source URL available. Omit unavailable details rather than inventing authors, dates, versions or DOIs. Preserve original-source citations for third-party data; this fallback does not replace them. Keep applicable licence and adaptation notices alongside the attribution, separate from the bibliographic reference.

## Tabular estimates

- [Trends release metadata](https://data.malariaatlas.org/map-platform-app-backend/api/v1/trends/metadata): Available releases, data end years and corresponding GeoServer/boundary references. Check metric availability for the selected release.
- [Admin-1 Pf case-count means, release 202508](https://data.malariaatlas.org/map-platform-app-backend/api/v1/trends/malaria/Pf_incidence-count/admin1/mean?version=202508): JSON-stat estimates for all available admin-1 regions and years in this release. Replace the final path segment `mean` with `lci` or `uci` for lower or upper bounds. Available geographic levels include `admin0`, `admin1` and `global`.
- [Admin-1 names for boundary version 2023_mg_5k](https://data.malariaatlas.org/admin-unit-metadata/admin-units/ADMIN1?mapAdmin=2023_mg_5k): JSON records containing `id`, `name`, `countryName`, `iso3` and parent fields. Use `geoServerRefs.mapAdminVersion` from the selected release metadata; the boundary version here is an example, not a default for all releases.

The admin-1 JSON-stat response has `ID` and `Year` dimensions. Read their category indexes and the dataset's dimension order/size to decode values. Values can be an array or an object with numeric-string keys representing flattened positions. For dimensions ordered as `ID`, `Year`, the position is `adminIndex * numberOfYears + yearIndex`. Join the `ID` dimension to the admin-name service's `id` field, treating both as strings. For example, Kano, Nigeria is `10312921` in boundary version `2023_mg_5k`. The admin-name catalogue can contain regions absent from a particular metric dataset.

The documented Trends endpoint returns the whole selected geographic level; it does not support country/name/year query filtering. Filter the returned data locally. The dashboard's CSV export is generated in the browser and is not a separate public CSV endpoint.

These are modelled estimates, not a count of individually reported surveillance records. Preserve species, metric, release, geography, year and uncertainty information. Prevalence, incidence rates and case counts are different quantities. Availability in one release or metric does not establish availability in another.

## Spatial data and background

- [WCS capabilities](https://data.malariaatlas.org/geoserver/wcs?service=WCS&version=2.0.1&request=GetCapabilities): XML catalogue of raster coverages. This response is large. WCS provides raster data; WMS map images are a different product.
- [MAP malariaAtlas R client](https://github.com/malaria-atlas-project/malariaAtlas): Existing client for discovering and retrieving MAP spatial data.
- [Platform background and acknowledgements](https://data.malariaatlas.org/about): Existing interactive information page.
- [Malaria Atlas Project](https://malariaatlas.org/): Project background and research.