ClimateAF - Current, historical and projected climate data for Africa

Other climate data | Species range maps | Other data

ClimateAF: historical and projected climate data for Africa

On this page you can download the ClimateAF software and approximately 3,000 climate grids at 30 arcsec (~1 km) and 2.5 arcmin (5 km) resolution for historical climate (1901-2020) in monthly, annual, decadal, and 30-year steps and projected future climate (2020s, 2050s, 2080s) based on CMIP6 multi-model projections.

The database includes monthly base variables (Tmin, Tmax, Tave & Prec) as well as economically or biologically relevant (bioclim) variables such as growing and chilling degree days, heating and cooling degree days, Hargrave's climate moisture deficit and reference evaporation, and seasonal variables.

The climate grids were developed with a deep neural network that uses geographic and atmospheric information to model local weather patterns at medium resolution (e.g. see examples below, with the inset showing precipitation induced by orographic lift on the windward side, and rain shadows on the leeward side of major mountain ranges: Rift Valley, Mt Kenya, Mt Kilimanjaro area).

Subsequently, the ClimateAF software downscales the grids to higher resolutions with a digital elevation model in conjunction with local environmental lapse-rates. The software can also provide scale-free point estimates of climate variables for user-provided latitude, longitude and elevation coordinates.

Data downloads at 30 arcsec (~1 km) and 2.5 arcmin (5 km) resolution are available from the tables below, but you may also visually explore some sample grids: click on a thumbnails below and then zoom in or out of different areas ( ), or view the RGB-GeoTIFFs in GIS.

Mean Annual Precipitation (View, GIS)	Mean Coldest Month Temp (View, GIS)	Precipitation Dec-Jan-Feb (View, GIS)	Climate Moisture Deficit (View, GIS)

Mean Annual Temperature (View, GIS)	Mean Warmest Month Temp (View, GIS)	Avg Min Temp Dec-Jan-Feb (View, GIS)	Reference Elevation Grid (View, GIS)

ClimateAF software download and tutorials

This program does not require installation. Download, unzip, and double click the executable file ClimateAF.exe. The program should run on all versions of Windows. If you receive the error message "COMCTL32.OCX missing", you have to install these libraryfiles. The program also runs on Linux, Unix and Mac systems with the free software Wine or MacPorts/Wine).

ClimateAF v1.10 (2024-10-10) - includes AR6/CMIP6 projections and CRU 4.05 historical data from 1901 to 2020.

Poster describing the methodology for the underlying climate interpolations.

Video tutorials on how to use the software to obtain climate data for sample locations, or to produce climate grids for a study area.

Gridded data at 2.5 arcmin (~5 km) & 30 arcsec (~1km) resolution

This dataset was created with the ClimateAF v1.10 software package for historical and projected 30-year normal periods. To generate additional projections for individual GCMs, you can use the "ClimateAF input file" with the "Grid generator R code", included in the "Readme" file below. The download packages of historical and projected climate data contain geoTIFF files compatible with most GIS applications in the standard WGS84 geographic (= EPSG:4326) projection. See the "Readme" file for further details and explanations:

2.5 arcmin (~5km) resolution download:

Help file	ClimateAF input file	Elevation, ID reference	Area covered
Usage, variables, grid generator:	2.5 arcmin CSV:	2.5 arcmin geoTIFF:	Shapefile:

2.5 arcmin Historical 30-year climate normal periods
1931-1960 Bioclim: , Monthly:	1941-1970 Bioclim: , Monthly:	1951-1980 Bioclim: , Monthly:
1961-1990 Bioclim: , Monthly:	1971-2000 Bioclim: , Monthly:	1981-2010 Bioclim: , Monthly:
1991-2020 Bioclim: , Monthly:	(as a representation of current climate use the SSP2-2020s* projection below)

Average ensemble scenarios¹	27 Bioclimatic variables	48 Monthly variables
SSP1 (+2.6 W/m²) - Sustainability focus²	2020s³: , 2050s: , 2080s:	2020s: , 2050s: , 2080s:
SSP2 (+4.5 W/m²) - Middle of the road	2020s*: , 2050s: , 2080s:	2020s*: , 2050s: , 2080s:
SSP3 (+7.0 W/m²) - Regional rivalry	2020s: , 2050s: , 2080s:	2020s: , 2050s: , 2080s:
SSP5 (+8.5 W/m²) - Fossil-fuel focus	2020s: , 2050s: , 2080s:	2020s: , 2050s: , 2080s:

   ¹) The ensemble projections are averages across 8 CMIP6 models (ACC, CNRM, EC, GFDL, GISS, MIR, MPI and MRI) that were chosen
     based on a variety of quality criteria and representativeness. For details on GCM selection, see Mahoney et al. (2022).
   ²) Average projected global warming by the 2080s for different Shared Socioeconomic Pathways (SSP) scenarios:
       SSP1-2.6: 1.5-2.0°C; SSP2-4.5: 2.5-3.0°C; SSP3-7.0: 3.5-4.0°C; SSP5-5.8: 4.0-5.0°C.
   ³) Projections for 30-year normal periods: 2020s: 2011-2040, 2050s: 2041-2070, 2080s: 2071-2100.
   *) SSP2-2020s is a good choice to represent current climate (midpoint of "middle of the road" climate normal estimate).

30 arcsec (~1km) resolution download:

Help file	ClimateAF input file	Elevation, ID reference	Area covered
Usage, variables, grid generator:	30 arcsec CSV:	30 arcsec geoTIFF:	Shapefile:

30 arcsec Historical 30-year climate normal periods
1931-1960 Bioclim: , Monthly:	1941-1970 Bioclim: , Monthly:	1951-1980 Bioclim: , Monthly:
1961-1990 Bioclim: , Monthly:	1971-2000 Bioclim: , Monthly:	1981-2010 Bioclim: , Monthly:
1991-2020 Bioclim: , Monthly:	(as a representation of current climate use the SSP2-2020s* projection below)

Average ensemble scenarios¹	27 Bioclimatic variables	48 Monthly variables
SSP1 (+2.6 W/m²) - Sustainability focus²	2020s³: , 2050s: , 2080s:	2020s: , 2050s: , 2080s:
SSP2 (+4.5 W/m²) - Middle of the road	2020s*: , 2050s: , 2080s:	2020s*: , 2050s: , 2080s:
SSP3 (+7.0 W/m²) - Regional rivalry	2020s: , 2050s: , 2080s:	2020s: , 2050s: , 2080s:
SSP5 (+8.5 W/m²) - Fossil-fuel focus	2020s: , 2050s: , 2080s:	2020s: , 2050s: , 2080s:

Scenario selection to quantify uncertainty for different regions of Africa

If you want to quantify uncertainty in future projections, you need to work with a selection of multiple, individual models. This is easily done with the provided software for locations of interest (see video tutorial), the pre-selected grids downloadable below, or it is also fairly straight-forward to generate custom grids with the help of some R code (included in the help file) if your data needs are different.

To help with the selection of a representative set of models for different regions of Africa (or for the entire continent), we used the Katsavounidis-Kuo-Zhang (KKZ) algorithm which selects an optimally representative set of future projections for different IPCC climate reference regions considering multiple climate variables (click on left panel dendrogram to visualize similarity among models):

IPCCregions

A common approach is to select a median, a pessimistic and an optimistic projection, i.e., a subset size of 3. For example, to assess uncertainty in West Africa (WAF) with a 3-model ensemble, you would choose the models: GISS (median), MPI (optimistic), and EC (pessimistic) from the table below (compare with the dendrogram above). Adding models (rows 4 to 8) will provide increasingly better representation of uncertainty in future predictions in multivariate space.

To include the sensitive "outlier" model UKES in the model selection process, use the lower portion of the table. If UKES is included, this scenario will almost always be picked second, as the most "pessimistic" model of global warming. UKES is not a very likely outcome, but if you work with a larger ensemble, you may include it as one possible outcome. For details on GCM selection process, see Mahoney et al. (2022)

Subset size	IPCC climate reference region											Africa
Subset size	ARP	CAF	ESAF	MDG	MED	NEAF	SAH	SEAF	WAF	WCA	WSAF	Africa
Excluding UKESM1-0-LL
1	GISS	MRI	GFDL	MRI	CNRM	GISS	CNRM	GFDL	GISS	MRI	GFDL	GISS
2	EC	MPI	CNRM	ACC	EC	MIR	MPI	ACC	MPI	EC	CNRM	MPI
3	MRI	ACC	MPI	MIR	MPI	ACC	EC	MIR	EC	MPI	MIR	ACC
4	MPI	MIR	MIR	CNRM	GISS	GFDL	MRI	MPI	ACC	MIR	GISS	MIR
5	ACC	EC	GISS	GISS	MRI	MRI	ACC	EC	MIR	ACC	EC	MRI
6	CNRM	GFDL	EC	MPI	ACC	EC	GISS	MRI	CNRM	GFDL	MPI	EC
7	GFDL	CNRM	ACC	EC	MIR	MPI	MIR	CNRM	GFDL	CNRM	MRI	CNRM
8	MIR	GISS	MRI	GFDL	GFDL	CNRM	GFDL	GISS	MRI	GISS	ACC	GFDL
Including UKESM1-0-LL
1	CNRM	GISS	EC	CNRM	CNRM	GISS	CNRM	CNRM	CNRM	CNRM	EC	CNRM
2	UKES	UKES	UKES	UKES	UKES	UKES	UKES	UKES	MPI	UKES	UKES	UKES
3	EC	MPI	MPI	MPI	EC	MIR	MPI	MPI	UKES	EC	MIR	MPI
4	MPI	EC	MIR	MIR	MPI	ACC	EC	MIR	EC	MPI	CNRM	GFDL
5	MRI	MIR	CNRM	GISS	GISS	GFDL	MRI	EC	GFDL	GFDL	GISS	ACC
6	ACC	ACC	GISS	ACC	MRI	MRI	ACC	GISS	MIR	MIR	MRI	EC
7	GISS	GFDL	MRI	MRI	ACC	EC	GISS	MRI	GISS	ACC	GFDL	MIR
8	MIR	CNRM	ACC	EC	MIR	MPI	MIR	GFDL	ACC	MRI	MPI	GISS
9	GFDL	MRI	GFDL	GFDL	GFDL	CNRM	GFDL	ACC	MRI	GISS	ACC	MRI

   ¹) IPCC region abbreviations are: ARP, Arabian-Peninsula; CAF, Central-Africa; ESAF, East-Southern Africa; MDG, Madagascar MED, Mediterranean;
       NEAF, North-East-Africa; SAH, Sahara; SEAF, South-East-Africa; WAF, Western-Africa; WCA, West Central Asia and; WSAF, West-South-Africa.
   ²) Model abbreviations are: ACC, ACCESS-ESM1-5; CNRM, CNRM-ESM2-1; EC, EC-Earth3; GFDL, GFDL-ESM4; GISS, GISS-E2-1-G; MIR, MIROC6;
       MPI, MPI-ESM1-2-HR; MRI, MRI-ESM2-0; and UKES, UKESM1-0-LL.

Below, you can find a table with individual scenario downloads for uncertainty analysis. This is not the full set of SSPs at high resolution, as this is not typically necessary for a meaningful uncertainty analysis. The selection below will be suitable to quantify medium term (2050s) and long term (2080s) uncertainty in projections, as well as the dependence of outcomes on two SSPs.

Individual GCM downloads for uncertainty analysis (2.5 arcmin)

AOGCM	SSP2 (+4.5 W/m²) - Middle of the road		SSP5 (+8.5 W/m²) - Fossil-fuel focus
AOGCM	27 Bioclimate variables	48 Monthly variables	27 Bioclimate variables	48 Monthly variables
ACCESS-ESM1-5 (ACC)	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:
CNRM-ESM2-1 (CNRM)	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:
EC-Earth3 (EC)	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:
GFDL-ESM4 (GFDL)	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:
GISS-E2-1-G (GISS)	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:
MIROC6 (MIR)	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:
MPI-ESM1-2-HR (MPI)	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:
MRI-ESM2-0 (MRI)	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:
UKESM1-0-LL (UKES)	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:	2050s: , 2080s:

References

Note that the ClimateAF package has not undergone peer-review in a journal yet. In the interim, reference usage like this: "Climate data has been generated with the ClimateAF v1.10 softwarepackage, available at http://tinyurl.com/ClimateAF, based on methodology similar to Mahoney et al. (2022) and Wang et al. (2016)."

Mahoney, C.R., Wang, T., Hamann, A., Cannon, A.J. 2022. A CMIP6 ensemble for downscaled monthly climate normals over North America. International Journal of Climatology 42: 5871-5891.

Wang, T., Hamann, A. Spittlehouse, D.L. and Carroll, C. 2016. Locally downscaled and spatially customizable climate data for historical and future periods for North America. PLoS One 11: e0156720.

Hamann, A. and Wang, T., Spittlehouse, D.L., and Murdock, T.Q. 2013. A comprehensive, high-resolution database of historical and projected climate surfaces for western North America. Bulletin of the American Meteorological Society 94: 1307–1309.