Bias-adjustment consists in “calibrating model simulations to ensure their statistical properties are similar to those of the corresponding observed values” (as from climat4impact.eu). Some authors claim that bias-adjustment techniques introduce another level of uncertainty making evaluation of projections uncertainty even harder (e.g., Ehret et al. 2012, Maraun et al., 2016). There are still differences of opinion regarding whether direct or bias-adjusted climate model simulations should be used in impact modeling and assessment. On the one hand, the use of direct climate model simulations ensures spatial and temporal consistency across variables, on the other hand the substantial biases of raw variables renders direct climate model simulations unrealistic and ultimately unsuitable for climate change impact modeling. While the climate modeling community continues to improve climate models, statistical bias adjustment is currently necessary  to make climate projections fit for purpose in impact modeling and assessment (Ficklin et al 2016, climate4impact.eu).

To generate the data sets on the climate data factory we use the Cumulative Distribution Function transform (CDF-t) method (Michelangeli et al., 2009, Vrac et al., 2016, Famien et al., 2017) we co-developed with academics. The CDF-t method assumes a reference period (i.e., 1979-2005 for CMIP5, 1989-2005 for CORDEX) over which gridded observations data are available and over which CDF-t is calibrated. We use gridded observations designed for model evaluation (i.e., from WFDEI for CMIP5, from MESAN for CORDEX). CDF-t performances are not sensitive to the model performance but to the variability and trend of the driving large-scale fields (reanalysis or GCMs/RCMs control runs) which can perform better or worse depending on the variable but also on the season (Vrac et al., 2012).

The CDF-t bias-adjustment method preserves long-term trend in climate models data. However, to represent a correct CDF under historical and/or present climate conditions does not guarantee to correctly represent the evolution of the CDF in a climate change context. In addition, CDF-t method is a univariate adjustment method which is applied location by location, and is not designed to reproduce multi-dimensional properties (e.g., variable covariance and spatial correlations). Development of a multivariate and spatial version of the CDF-t method is underway (Vrac 2018, in review). The aforementioned limitations are not specific to the CDF-t method but common to any univariate Quantile-Quantile method. 

For a complete description of the data processing you can also read our document “Technical note: bias adjusting climate model projections”.

-----------------------------------------------------------------------------------------------------------------------------

References

Ehret U. et al. (2012) Should we apply bias correction to global and regional climate
model data?, Hydrol. Earth Syst. Sci., 16, 3391–3404.

Famien, A. M., Janicot, S., Ochou, A. D., Vrac, M., Defrance, D., Sultan, B., and Noël, T.: A bias-corrected CMIP5 dataset for Africa using CDF-t method. A contribution to agricultural impact studies, Earth Syst. Dynam. Discuss., https://doi.org/10.5194/esd-2017-111, in review, 2017.

Ficklin D. et al (2016) The Influence of Climate Model Biases on Projections of Aridity and Drought, Journal of Climate, 1269. 

Maraun, D., 2016. Bias Correcting Climate Change Simulations - a Critical Review. Current Climate Change Reports 2, 211–220.

Michelangeli, P. A., Vrac, M., & Loukos, H. (2009). Probabilistic downscaling approaches: Application to wind cumulative distribution functions. Geophysical Research Letters, 36(11).

Vrac, M., P. Drobinski, A. Merlo, M. Herrmann, C. Lavaysse, L. Li, and S. Somot, 2012: Dynamical and statistical downscaling of the French Mediterranean climate: Uncertainty assessment. Nat. Hazards Earth Syst. Sci., 12, 2769–2784, doi:https://doi.org/10.5194/nhess-12-2769-2012. Crossref

Vrac, M., T. Noël, and R. Vautard (2016), Bias correction of precipitation through Singularity Stochastic Removal: Because occurrences matter. J. Geophys. Res. Atmos., 121, 5237–5258, doi:10.1002/2015JD024511.

Vrac, M.: Multivariate bias adjustment of high-dimensional climate simulations: The Rank Resampling for Distributions and Dependences (R2D2) Bias Correction, Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2017-747, in review, 2018.

Did this answer your question?