Impact of increased resolution on long-standing biases in HighResMIP-PRIMAVERA climate models

Abstract

Abstract. We examine the influence of increased resolution on four long-standing biases using five different climate models developed within the PRIMAVERA project. The biases are the warm eastern tropical oceans, the double Intertropical Convergence Zone (ITCZ), the warm Southern Ocean, and the cold North Atlantic. Atmosphere resolution increases from ∼100–200 to ∼25–50 km, and ocean resolution increases from ∼1∘ (eddy-parametrized) to ∼0.25∘ (eddy-present). For one model, ocean resolution also reaches 1/12∘ (eddy-rich). The ensemble mean and individual fully coupled general circulation models and their atmosphere-only versions are compared with satellite observations and the ERA5 reanalysis over the period 1980–2014. The four studied biases appear in all the low-resolution coupled models to some extent, although the Southern Ocean warm bias is the least persistent across individual models. In the ensemble mean, increased resolution reduces the surface warm bias and the associated cloud cover and precipitation biases over the eastern tropical oceans, particularly over the tropical South Atlantic. Linked to this and to the improvement in the precipitation distribution over the western tropical Pacific, the double-ITCZ bias is also reduced with increased resolution. The Southern Ocean warm bias increases or remains unchanged at higher resolution, with small reductions in the regional cloud cover and net cloud radiative effect biases. The North Atlantic cold bias is also reduced at higher resolution, albeit at the expense of a new warm bias that emerges in the Labrador Sea related to excessive ocean deep mixing in the region, especially in the ORCA025 ocean model. Overall, the impact of increased resolution on the surface temperature biases is model-dependent in the coupled models. In the atmosphere-only models, increased resolution leads to very modest or no reduction in the studied biases. Thus, both the coupled and atmosphere-only models still show large biases in tropical precipitation and cloud cover, and in midlatitude zonal winds at higher resolutions, with little change in their global biases for temperature, precipitation, cloud cover, and net cloud radiative effect. Our analysis finds no clear reductions in the studied biases due to the increase in atmosphere resolution up to 25–50 km, in ocean resolution up to 0.25∘, or in both. Our study thus adds to evidence that further improved model physics, tuning, and even finer resolutions might be necessary.

Impact of climate change on site characteristics of eight major astronomical observatories using high-resolution global climate projections until 2050

“…The PRIMAVERA atmosphere-only simulations are forced by HadISST2.2.0.0 sea surface boundary conditions. 19 The PRIMAVERA atmosphere-only simulations provide an opportunity to decide whether performance issues arise from ocean-atmosphere coupling (Moreno-Chamarro et al 2022). PRIMAVERA coupled and atmosphere-only simulations use the same external radiative forcings, as described in the High-ResMIP protocol (Haarsma et al 2016).…”

“…As suggested by Graham et al (2004) and already shown by Cantalloube et al (2020), climate models that project scenarios of anthropogenic radiative forcing into the future may prove useful. To address particular topographic textures as they often appear at astronomical sites, GCMs with the highest possible horizontal resolution are advantageous (Moreno-Chamarro et al 2022). Even if the use of regional climate models (RCMs) would allow for even higher resolution, RCMs do not provide global coverage or global consistency.…”

Increased wintertime European atmospheric blocking frequencies in General Circulation Models with a coupled eddy-permitting ocean

“…Another major outlook from this research is the potential benefit of employing GCMs with an eddy-rich ocean resolution, which can be obtained for nominal resolutions as low as 0.1°x0.1°2 7,28 . Unlike EP ocean grids, which only include mesoscale eddies with broader horizontal extents (100 km), eddy-rich ocean grids incorporate smaller-scale eddies (20 to 100 km), resulting in a significantly higher number of these mesoscale eddies and a better represented mean climate 17,27,28,29,37 . The use of an eddy-rich grid was notably shown to further improve most benefits and bias reductions seen in EP GCMs, notably in terms of SST mean biases 27 .…”

“…At the surface, the horizontal distribution of heat driven by mesoscale eddies importantly contribute to sharp SST gradients, thus acting on the atmospheric boundary layer and higher atmosphere through variations in surface stability on momentum transfer, cloud cover and pressure gradients 26 . They have notably been shown to contribute significantly to the realism of air-sea interactions in GCMs, as well as large-scale oceanic features 27,28,29 .…”

Regional climate over Northern Europe: from observations to high-resolution modeling

“…The spatial resolution of GCMs has commonly been 25-300 km (e.g. Demory et al 2020;Moreno-Chamarro et al 2022) while it has been around 12 km in the state-of-the-art RCM simulations for Europe, such as those of the EURO-CORDEX project (Jacob et al 2020). Recently, climate models have reached even finer grid spacings (Stevens et al 2019;Coppola et al 2020).…”