Comments
Systematic Bias in Simulated Monthly Mean Temperature and Precipitation in Regional Climate Models
A new paper has been recently published in GRL entitled: ‘On the need for bias correction of regional climate change projections of temperature and precipitation,’ by Christensen et al
The Abstract states:
Within the framework of the European project ENSEMBLES (ensembles-based predictions of climate changes and their impacts) we explore the systematic bias in simulated monthly mean temperature and precipitation for an ensemble of thirteen regional climate models (RCMs). The models have been forced with the European Centre for Medium Range Weather Forecasting Reanalysis (ERA40) and are compared to a new high resolution gridded observational data set. We find that each model has a distinct systematic bias relating both temperature and precipitation bias to the observed mean. By excluding the twenty-five percent warmest and wettest months, respectively, we find that a derived second-order fit from the remaining months can be used to estimate the values of the excluded months. We demonstrate that the common assumption of bias cancellation (invariance) in climate change projections can have significant limitations when temperatures in the warmest months exceed 4–6 °C above present day conditions.
The authors state in the Summary and Conclusions:
We have found that model biases have the potential to grow when used for climate change simulations under global warming conditions. We have demonstrated that in warmer and wetter climatic conditions, it is quite conceivable that a significant systematic model bias will dominate increase, generally resulting in an enhanced warm bias with increasing temperature and hence lead to an overestimation of the projected warming. Similarly, systematic precipitation biases depending on the precipitation amounts themselves may lead to either an over- or underestimation of the projected precipitation changes. These tendencies will be model specific, and here we have made a simplistic approach to capture the intrinsic model behaviours.