Toward Eliminating the Decades‐Old “Too Zonal and Too Equatorward” Storm‐Track Bias in Climate Models

Abstract

This page was generated automatically upon download from the ETH Zurich Research Collection. For more information, please consult the Terms of use. tions of regional land surface climate" (WG1 Chapter 10). For the projection of future regional land precipitation changes downstream of the North Atlantic storm tracks, this implies increased uncertainty, since regional precipitation changes are mainly determined by large-scale circulation changes. As a result, the sign of, for example, changes in fall but also annual mean precipitation under the SSP5-8.5 scenario over western and central Europe remains unclear which can be conveniently verified using the new IPCC WGI Interactive Atlas accessible via https://interactive-atlas.ipcc.ch/regional-information (last accessed March 2022). In addition, climate models tend to underestimate in particular the cases of rapid cyclone intensification (Priestley & Catto, 2022). Overall, the biases are thus best summarized by "too zonal, too equatorward and too weak for extreme cases."

The Northern Hemisphere Wintertime Storm Track Simulated in the High‐Resolution Community Earth System Model

“…The poleward shift of storm tracks with higher resolution is also found in other models (Schemm, 2023; Sein et al., 2018; Willison et al., 2013). Their findings suggested that better resolved diabatic processes may enhance the poleward propagation of extratropical cyclones (Coronel et al., 2015; Tamarin & Kaspi, 2017) and emphasized the impact of increasing atmospheric resolution (e.g., see Figure 8 in Sein et al., 2018).…”

“…Over the North Atlantic, the storm track in LR is too weak over the Nordic Seas region, Europe, and the regions adjacent to the Grand Banks (by 14-18 gpm) compared to ERA5. As mentioned before, previous studies reported a too-zonal bias of the North Atlantic storm track with an overestimation over Europe and an underestimation over the Nordic Seas region (e.g., Harvey et al, 2020;Schemm, 2023). HR reduces the negative bias over Europe and the Nordic Seas region (of up to 9-13 gpm).…”

“…This is indicative of a combination of weakened instability (the deficit in sensible heat flux) and reduced moisture content (due to the insufficient latent heat flux) (Athanasiadis et al., 2022). These changes could be related to the resolution‐induced bias of the surface storm track, suggesting the importance of the diabatic processes for improving the North Atlantic storm track simulation (Schemm, 2023; Willison et al., 2013).…”

“…Many studies based on different models reported that increasing atmospheric resolution, and that of the coupled ocean, will improve the simulations of the midlatitude storm activity and thus provide more realistic storm tracks in the NH (e.g., Foussard et al., 2019; Hewitt et al., 2017; Lee et al., 2018; Small et al., 2019; Willison et al., 2013). A recent study (Schemm, 2023) showed that increasing model resolution improves local diabatic processes and thus reduces the “too‐zonal and too‐equatorial” bias in the North Atlantic storm track. Although high‐resolution models have been proven to improve the simulated storm track to some extent, considerable differences remain across model families (Chassignet et al., 2020; Priestley et al., 2023).…”

Regional climate change: consensus, discrepancies, and ways forward

“…Storm resolving model data is now becoming regionally available, for example over Europe through CORDEX FPS convection (Coppola et al, 2020) or for the Lake Victoria area through CORDEX FPS Lake Victoria (van Lipzig et al, 2023), and globally in the next years through the digital twin developed within the Destination Earth project (Hoffmann et al, 2023). A few examples for processes where storm resolving simulations bring demonstrated benefit are the dynamics of the North Atlantic jet (Schemm, 2023), precipitation in complex orography (Ban et al, 2021), convection and circulation over Lake Victoria (Van de Walle et al, 2020), local features around islands in the Caribbean (Martinez et al, 2024), the subtropical Atlantic (Gao et al, 2023) and modeling lake-effect snow around the Tibetan Plateau (Lin et al, 2023). Higher resolution models, particularly for the ocean, have also been shown to have more fidelity in capturing SST trends than lower resolution models (Yeager et al, 2023).…”