What is bias correction in climate data?
In short: Bias correction adjusts systematic differences between simulated climate data and observations. A correction is calibrated where historical model output overlaps with observations, then applied to other model periods. It can correct features such as the mean or distribution, but it cannot eliminate every model error or projection uncertainty.
Why climate data are bias-corrected
GCM values represent grid-cell climate and may differ systematically from measurements at a station because of spatial scale, model structure, parameterization, and terrain. Impact models often need data whose historical statistics more closely match the local record.
Common correction approaches
A change-factor method adjusts a baseline by a modeled change, while quantile-mapping methods correct more of the distribution. The selected method should be evaluated on historical data and applied consistently to the future simulation.
SD-GCM bias-corrects CMIP5 and CMIP6 climate series with Delta, Quantile Mapping, and EQM.