Bias-corrected data set of daily precipitation for a study region in Burkina Faso and parts of Ghana, Togo and Benin. The precipitation simulations of the CORDEX Africa ensemble have been bias-corrected with a geostatistically based Quantile-Mapping with dry day correction. The distribution parameters probability of rainfall and mean precipitation on wet days were Kriged to the ungauged locations from the observed parameters of point measurements (172 stations in total) for nine season (dry season NDJF and remaining eight months individually). Observed precipitation was modelled with the exponential distribution, RCM precipitation was modeled with a Kernel Density Estimation of the CDF. Future scenarios were bias-corrected with the Double-Quantile-Mapping approach presented in http://onlinelibrary.wiley.com/doi/10.1029/2010WR009689/abstract All models were stretched to the Gregorian calendar (original CORDEX-RCMs are available as 360d/a, 365d/a and 365.25d/a) to facilitate running impact models. - Historical period: 1950-2005 (exception: SMHI-models start 1951) - calibration period - Future period: 2006-2100 (exception: models driven by HadGEM2 end 2099) - validation period - Temporal resolution: daily - Spatial resolution: 0.44° This is a test version - please inform me about any potential problems you may have with the data via manuel.lorenz@geo.uni-augsburg.de

Bias-corrected data set of daily precipitation for a study region in Burkina Faso and parts of Ghana, Togo and Benin. The precipitation simulations of the CORDEX Africa ensemble have been bias-corrected with a geostatistically based Quantile-Mapping with dry day correction. The distribution parameters probability of rainfall and mean precipitation on wet days were Kriged to the ungauged locations from the observed parameters of point measurements (172 stations in total) for nine season (dry season NDJF and remaining eight months individually). Observed precipitation was modelled with the exponential distribution, RCM precipitation was modeled with a Kernel Density Estimation of the CDF. Future scenarios were bias-corrected with the Double-Quantile-Mapping approach presented in http://onlinelibrary.wiley.com/doi/10.1029/2010WR009689/abstract All models were stretched to the Gregorian calendar (original CORDEX-RCMs are available as 360d/a, 365d/a and 365.25d/a) to facilitate running impact models. - Historical period: 1950-2005 (exception: SMHI-models start 1951) - calibration period - Future period: 2006-2100 (exception: models driven by HadGEM2 end 2099) - validation period - Temporal resolution: daily - Spatial resolution: 0.44° This is a test version - please inform me about any potential problems you may have with the data via manuel.lorenz@geo.uni-augsburg.de

Bias-corrected data set of daily precipitation for a study region in Burkina Faso and parts of Ghana, Togo and Benin. The precipitation simulations of the CORDEX Africa ensemble have been bias-corrected with a geostatistically based Quantile-Mapping with dry day correction. The distribution parameters probability of rainfall and mean precipitation on wet days were Kriged to the ungauged locations from the observed parameters of point measurements (172 stations in total) for nine season (dry season NDJF and remaining eight months individually). Observed precipitation was modelled with the exponential distribution, RCM precipitation was modeled with a Kernel Density Estimation of the CDF. Future scenarios were bias-corrected with the Double-Quantile-Mapping approach presented in http://onlinelibrary.wiley.com/doi/10.1029/2010WR009689/abstract All models were stretched to the Gregorian calendar (original CORDEX-RCMs are available as 360d/a, 365d/a and 365.25d/a) to facilitate running impact models. - Historical period: 1950-2005 (exception: SMHI-models start 1951) - calibration period - Future period: 2006-2100 (exception: models driven by HadGEM2 end 2099) - validation period - Temporal resolution: daily - Spatial resolution: 0.44° This is a test version - please inform me about any potential problems you may have with the data via manuel.lorenz@geo.uni-augsburg.de

Bias-corrected data set of daily precipitation for a study region in Burkina Faso and parts of Ghana, Togo and Benin. The precipitation simulations of the CORDEX Africa ensemble have been bias-corrected with a geostatistically based Quantile-Mapping with dry day correction. The distribution parameters probability of rainfall and mean precipitation on wet days were Kriged to the ungauged locations from the observed parameters of point measurements (172 stations in total) for nine season (dry season NDJF and remaining eight months individually). Observed precipitation was modelled with the exponential distribution, RCM precipitation was modeled with a Kernel Density Estimation of the CDF. Future scenarios were bias-corrected with the Double-Quantile-Mapping approach presented in http://onlinelibrary.wiley.com/doi/10.1029/2010WR009689/abstract All models were stretched to the Gregorian calendar (original CORDEX-RCMs are available as 360d/a, 365d/a and 365.25d/a) to facilitate running impact models. - Historical period: 1950-2005 (exception: SMHI-models start 1951) - calibration period - Future period: 2006-2100 (exception: models driven by HadGEM2 end 2099) - validation period - Temporal resolution: daily - Spatial resolution: 0.44° This is a test version - please inform me about any potential problems you may have with the data via manuel.lorenz@geo.uni-augsburg.de

Raw field data from the WASCAL core research sites Dano, Bontioli, Nazinga, and Bolgatanga on ecosystem services in terms of vegetation utilization, degradation and conservation by inhabitants of nine villages located in different savanna formations.

This gridded precipitation set is a test data set (BETA VERSION). The daily precipitation data set was created using a spatial interpolation technique using ordinary kriging. The gridded data set is based on 410 daily observed precipitation time series from stations located in Ghana, Benin, Togo, Burkina Faso and neighboring countries. They have been compiled from global (GHCN-daily), regional (e.g. AMMA-SYNOP, GLOWA-Volta) and national archives from the meteorological services of Burkina Faso, Ghana and Benin. The temporal coverage of the daily time series ranges from 1940 to 2013. The quality of the spatial interpolation was compared to various interpolation techniques (nearest neighbor, inverse distance and external drift kriging).

Bias-corrected data set of daily precipitation for a study region in Burkina Faso and parts of Ghana, Togo and Benin. The precipitation simulations of the CORDEX Africa ensemble have been bias-corrected with a geostatistically based Quantile-Mapping with dry day correction. The distribution parameters probability of rainfall and mean precipitation on wet days were Kriged to the ungauged locations from the observed parameters of point measurements (172 stations in total) for nine season (dry season NDJF and remaining eight months individually). Observed precipitation was modelled with the exponential distribution, RCM precipitation was modeled with a Kernel Density Estimation of the CDF. Future scenarios were bias-corrected with the Double-Quantile-Mapping approach presented in http://onlinelibrary.wiley.com/doi/10.1029/2010WR009689/abstract All models were stretched to the Gregorian calendar (original CORDEX-RCMs are available as 360d/a, 365d/a and 365.25d/a) to facilitate running impact models. - Historical period: 1950-2005 (exception: SMHI-models start 1951) - calibration period - Future period: 2006-2100 (exception: models driven by HadGEM2 end 2099) - validation period - Temporal resolution: daily - Spatial resolution: 0.44° This is a test version - please inform me about any potential problems you may have with the data via manuel.lorenz@geo.uni-augsburg.de

Raw field data from the WASCAL core research sites Dano, Bontioli, Nazinga, and Bolgatanga on vegetation composition, dendrometric parameters of the woody vegetation within the sampling plots, and observed basic ecological parameters of the sampling sites located in different savanna formations.

Bias-corrected data set of daily precipitation for a study region in Burkina Faso and parts of Ghana, Togo and Benin. The precipitation simulations of the CORDEX Africa ensemble have been bias-corrected with a geostatistically based Quantile-Mapping with dry day correction. The distribution parameters probability of rainfall and mean precipitation on wet days were Kriged to the ungauged locations from the observed parameters of point measurements (172 stations in total) for nine season (dry season NDJF and remaining eight months individually). Observed precipitation was modelled with the exponential distribution, RCM precipitation was modeled with a Kernel Density Estimation of the CDF. Future scenarios were bias-corrected with the Double-Quantile-Mapping approach presented in http://onlinelibrary.wiley.com/doi/10.1029/2010WR009689/abstract All models were stretched to the Gregorian calendar (original CORDEX-RCMs are available as 360d/a, 365d/a and 365.25d/a) to facilitate running impact models. - Historical period: 1950-2005 (exception: SMHI-models start 1951) - calibration period - Future period: 2006-2100 (exception: models driven by HadGEM2 end 2099) - validation period - Temporal resolution: daily - Spatial resolution: 0.44° This is a test version - please inform me about any potential problems you may have with the data via manuel.lorenz@geo.uni-augsburg.de

Bias-corrected data set of daily precipitation for a study region in Burkina Faso and parts of Ghana, Togo and Benin. The precipitation simulations of the CORDEX Africa ensemble have been bias-corrected with a geostatistically based Quantile-Mapping with dry day correction. The distribution parameters probability of rainfall and mean precipitation on wet days were Kriged to the ungauged locations from the observed parameters of point measurements (172 stations in total) for nine season (dry season NDJF and remaining eight months individually). Observed precipitation was modelled with the exponential distribution, RCM precipitation was modeled with a Kernel Density Estimation of the CDF. Future scenarios were bias-corrected with the Double-Quantile-Mapping approach presented in http://onlinelibrary.wiley.com/doi/10.1029/2010WR009689/abstract All models were stretched to the Gregorian calendar (original CORDEX-RCMs are available as 360d/a, 365d/a and 365.25d/a) to facilitate running impact models. - Historical period: 1950-2005 (exception: SMHI-models start 1951) - calibration period - Future period: 2006-2100 (exception: models driven by HadGEM2 end 2099) - validation period - Temporal resolution: daily - Spatial resolution: 0.44° This is a test version - please inform me about any potential problems you may have with the data via manuel.lorenz@geo.uni-augsburg.de