Skip to content

The climate-projection dataset landscape

What this page covers

The major climate-projection and climate-baseline datasets you’ll encounter doing African adaptation work. For each: what it is, what it’s best for, what its limitations are, and where to get it. The headline comparison table; then per-dataset write-ups for the most-likely-relevant ones.

At-a-glance comparison

To be written. A single table with columns:

DatasetTypeResolutionCoverageBias-correctedAfrica-suitedBest for

Populated from datasets_inventory.md.

Datasets you’ll meet most often

Per-dataset write-ups (1-2 paragraphs each) for:

NEX-GDDP-CMIP6

The Atlas default. 18 models, 0.25°, BCSD bias correction.

CORDEX-Africa

Dynamical downscaling. CMIP5-era mature, CMIP6-era in progress.

ISIMIP3b

Impact-sector default. 5 models, 0.5°, ISIMIP3BASD bias adjustment.

CHELSA-CMIP6

Very-high-resolution (1 km) for ecology / agroecology.

ERA5 / ERA5-Land

The dominant historical-reanalysis baseline for Africa.

CHIRPS / CHIRTS-ERA5

UCSB observational gridded rainfall + temperature.

AR6 Interactive Atlas

IPCC’s pre-processed CMIP6 + CORDEX organised by AR6 regions.

World Bank Climate Knowledge Portal

The most-likely alternative tool the Atlas’s audience already uses.

Mentioned briefly

One-line each on: CMIP6 raw (ESGF), CMIP5 raw, CIL-GDPCIR, WorldClim Future, MERRA-2, JRA-55, GMFD, W5E5.

How to choose for your use case

To be written. Decision-tree framing:

  • National / admin-1 adaptation work with multiple variables → NEX-GDDP-CMIP6
  • Impact-sector modelling (crops / water / health) → ISIMIP3b
  • Very-fine-resolution ecological work → CHELSA-CMIP6
  • Physical-consistency-critical extreme events → CORDEX-Africa
  • Historical baseline + observed-change analysis → ERA5 + CHIRPS/CHIRTS
  • Operational reporting + government policy framing → AR6 Interactive Atlas + CCKP

Why the Atlas chose NEX-GDDP-CMIP6

To be written. Brief justification cross-linking to Atlas approach.