Dynamical Downscaling at 50-12 km

Global climate models (GCMs) are primarily used to assess how the climate is responding to changes in external forcings, and a typical GCM horizontal resolutions is currently of around 100 km. To gain higher-resolution regional climate information, regional climate models (RCMs) are used as a dynamical downscaling tool. It is a common belief that the errors in GCM-RCM model chains behave approximately additive. If this hypothesis was true, the application of model chains would not lead to any intrinsic improvement except for higher-resolution details. A recent publication from our group (Sørland et al., 2018) show that the biases of the RCMs and GCMs are not additive and not independent. Bias patterns and climate change signal are investigated in two GCM-RCMs model chains. Both RCMs are systematically reducing the biases (see Figure) and modifying the climate change signals of the driving GCMs, even on scales that are considered well resolved by the driving GCMs.

Enlarged view: RCM Bias Reduction — Systematic reduction of the GCM bias by the RCMs. The panels display the RCM increments (RCM–GCM) versus the GCM biases (GCM–OBS) for near-surface temperatures. Data points along the dashed diagonal signal an increased performance of the GCM-RCM simulations in comparison to the GCMs, for the Alps (left), and eastern Europe (east). The colors indicate the driving GCM while the symbol is the regional model. The correlation between the RCM increments and GCM biases is given in each sub-figure, calculated separately for the GCM-CCLM and GCM-RCA simulations.

Sørland S., Schär C., Lüthi D. and E. Kjellström, Bias patterns and climate change signals in GCM-RCM model chains, Environ. Res. Lett. 13 074017, 2018, external pagehttps://doi.org/10.1088/1748-9326/aacc77.