ClimateBench v1.0: A Benchmark for Data‐Driven Climate Projections

Watson-Parris, Duncan; Rao, Yuhan; Olivié, Dirk; Seland, Øyvind; Nowack, Peer; Camps‐Valls, Gustau; Stier, Philip; Bouabid, Shahine; Dewey, M. C.; Fons, Emilie; Gonzalez, Jessenia; Harder, Paula; Jeggle, Kai; Lenhardt, Julien; Manshausen, Peter; Novitasari, Maria; Ricard, Lucile; Roesch, Carla

doi:10.1029/2021ms002954

Cited by 20 publications

(14 citation statements)

References 77 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here, the required functionality is an ability to represent diverse, yet realistic, evolutions of emissions of scattering and absorbing aerosols in different regions, beyond the combinations existing in widely studied pathways such as the main SSP realizations, in combination with a prescribed level of GHG-induced warming. The emulator should be able to rapidly produce a statistical sample of regional weather conditions, taking into account both local and remote effects of the resulting combined GHG and aerosol forcing [102][103][104]. The output of such emulators could be valuable in, for example, improving aerosol-specific statistical downscaling methodologies that account for the time-varying and pattern-dependent effects of aerosols and a range of other impact applications.…”

Section: The Way Forwardmentioning

confidence: 99%

Rapidly evolving aerosol emissions are a dangerous omission from near-term climate risk assessments

Persad

Samset

Wilcox

et al. 2023

Environ. Res.: Climate

View full text Add to dashboard Cite

Anthropogenic aerosol emissions are expected to change rapidly over the coming decades, driving strong, spatially complex trends in temperature, hydroclimate, and extreme events both near and far from emission sources. Under-resourced, highly populated regions often bear the brunt of aerosols’ climate and air quality effects, amplifying risk through heightened exposure and vulnerability. However, many policy-facing evaluations of near-term climate risk, including those in the latest Intergovernmental Panel on Climate Change assessment report, underrepresent aerosols’ complex and regionally diverse climate effects, reducing them to a globally averaged offset to greenhouse gas warming. We argue that this constitutes a major missing element in society’s ability to prepare for future climate change. We outline a pathway towards progress and call for greater interaction between the aerosol research, impact modeling, scenario development, and risk assessment communities.

show abstract

Section: The Way Forwardmentioning

confidence: 99%

Rapidly evolving aerosol emissions are a dangerous omission from near-term climate risk assessments

Persad

Samset

Wilcox

et al. 2023

Environ. Res.: Climate

View full text Add to dashboard Cite

show abstract

“…To name a few, Mamalakis et al (2022) provide a framework to create synthetic data sets designed for problems in geosciences. And Watson-Parris et al (2022) introduced ClimateBench, as a benchmark for data-driven climate projections.…”

Section: Introductionmentioning

confidence: 99%

A climate index collection based on model data

Landt-Hayen

Rath

Wahl

et al. 2023

Environ. Data Science

View full text Add to dashboard Cite

Machine learning (ML) and in particular deep learning (DL) methods push state-of-the-art solutions for many hard problems, for example, image classification, speech recognition, or time series forecasting. In the domain of climate science, ML and DL are known to be effective for identifying causally linked modes of climate variability as key to understand the climate system and to improve the predictive skills of forecast systems. To attribute climate events in a data-driven way, we need sufficient training data, which is often limited for real-world measurements. The data science community provides standard data sets for many applications. As a new data set, we introduce a consistent and comprehensive collection of climate indices typically used to describe Earth System dynamics. Therefore, we use 1000-year control simulations from Earth System Models. The data set is provided as an open-source framework that can be extended and customized to individual needs. It allows users to develop new ML methodologies and to compare results to existing methods and models as benchmark. For example, we use the data set to predict rainfall in the African Sahel region and El Niño Southern Oscillation with various ML models. Our aim is to build a bridge between the data science community and researchers and practitioners from the domain of climate science to jointly improve our understanding of the climate system.

show abstract

“…Also, the authors acknowledge that there are numerous other paths being advocated and followed to address the issue of complex climate models. Replacing climate models with emulators or other machine‐learning generated surrogates is an emerging yet contested field (e.g., Beusch et al., 2022; Kasim et al., 2020; Knüsel, 2020; Nonnenmacher & Greenberg, 2021; Rudin, 2019; Watson‐Parris et al., 2022). On a smaller scale, machine‐learning is also used to replace or improve single parameterizations or schemes (e.g., A. Seifert and Rasp (2020); Gettelman et al.…”

Section: Introductionmentioning

confidence: 99%

Addressing Complexity in Global Aerosol Climate Model Cloud Microphysics

Proske

Ferrachat

Klampt

et al. 2023

J Adv Model Earth Syst

View full text Add to dashboard Cite

In climate modeling, our task is to represent an immensely complex system which we wish to understand and learn about. Model development inherently faces tradeoffs between epistemic values like tractability, interpretability, validation potential, and specificity (Beucler et al., 2021;Larsen et al., 2016;Undorf et al., 2022), which manifest themselves in the ever present question of where to draw the line for details. Understanding nature requires a simplification of the mechanisms at play, but wanting to be precise calls for more details that increase complexity (Tapiador et al., 2019). For the term "model complexity" we here follow the "loose definition" mentioned in García-Callejas and Araújo (2016) that a model becomes more complex the more difficult to comprehend its computations are and the more processes/computations there are (Baartman et al. (2020) andRandall et al. (2003)).Climate modeling has followed the mirror view (Parker, 2022), meaning that it wants to mirror the Earth system in the model representation. Different modeling approaches would be following for example, the predictive or

show abstract

ClimateBench v1.0: A Benchmark for Data‐Driven Climate Projections

Cited by 20 publications

References 77 publications

Rapidly evolving aerosol emissions are a dangerous omission from near-term climate risk assessments

Rapidly evolving aerosol emissions are a dangerous omission from near-term climate risk assessments

A climate index collection based on model data

Addressing Complexity in Global Aerosol Climate Model Cloud Microphysics

Contact Info

Product

Resources

About