Increasingly Sophisticated Climate Models Need the Out‐Of‐Sample Tests Paleoclimates Provide

Climate models are becoming increasingly sophisticated as climate scientists continually work to improve the realism with which the processes influencing Earth's climate are represented. One example is the treatment of cloud microphysics: as complexity is added to cloud microphysical schemes, Earth's energy budget can respond to changes in climate forcings, such as carbon dioxide or aerosols, in new ways. This increase in degrees of freedom has illuminated larger spread in climate sensitivity across the latest generation of climate models participating Coupled Model Intercomparison Project, Phase 6, with more high climate sensitivity models (Zelinka et al., 2020, https://doi.org/10.1029/2019gl085782). Whilst the historical record gives us just over a century of data to apply toward climate sensitivity constraints (e.g., Nijsse et al., 2020, https://doi.org/10.5194/esd-11-737-2020), the ocean is still taking up much of the heat trapped by anthropogenic greenhouse gas emissions and the climate system is far from equilibrium which limits our understanding how climate sensitivity might change in response to long-term forced climate change. Here we discuss the valuable tests that paleoclimate reconstructions can provide the latest generation of climate models, as demonstrated by the recent study of Zhu et al., 2022, https://doi.org/10.1029/2021ms002776. Their study provides an example of the benefits for climate model development when climate models are confronted with simulating climates very different from today. Ideally the climate model development stage under future iterations of CMIP will involve such tests as an effort to constrain global climate sensitivity and the regional patterns of climate, such as polar amplification and subtropical aridification.