ABSTRACT
Global climate model (GCM) and regional climate model (RCM) outputs reportedly exhibit systematic biases relative to observations from different sources, such as errors in convection parameterization and unresolved fine-scale orography, unrealistic large-scale variability, and unpredictable internal variability different from observations. A number of bias correction methods have been developed for adjusting mean, variance, and higher moment of distributions as well as all quantiles. It has been known that precipitation values from GCMs or RCMs are more frequently generated than observations. This chapter discusses different approaches of correcting biases of systematic errors in GCM outputs and the necessity of bias correction and the occurrence adjustment for precipitation. It presents a bias correction method, quantile delta mapping (QDM), employing the ratios of the cumulative distribution functions of future and base periods. A. J. Cannon et al. suggested QDM to preserve the relative changes of model projections.
