Climate Modeling and Analysis

From Climate Change AI Wiki
Revision as of 21:43, 22 January 2021 by Kasia tokarska (talk | contribs) (major text updates)

This page is about the intersection of climate science and machine learning. For an overview of climate science as a whole, please see the Wikipedia page on this topic.
The main avenues through which machine learning can support climate science, as described in "Tackling Climate Change with Machine Learning." [1]
As described in "Tackling Climate Change with Machine Learning,"[1]
"The predominant predictive tools are climate models, known as General Circulation Models (GCMs) or Earth System Models (ESMs). These models inform local and national government decisions[2][3][4]), help people calculate their climate risks (see Policy, Markets, and Decision Science and Climate Change Adaptation) and allow us to estimate the potential impacts of solar geoengineering [and exploring Earth System response to different emission future scenarios, and under differnet assumptions]. Recent trends have created opportunities for ML to advance the state-of-the-art in climate prediction. First, new and cheaper satellites are creating petabytes of climate observation data[5][6]. Second, massive climate modeling projects are generating petabytes of simulated climate data[7]. Third, climate forecasts are computationally expensive[8] (some simulations have taken three weeks to run on NCAR supercomputers[9]), while ML methods are becoming increasingly fast to train and run, especially on next-generation computing hardware. As a result, climate scientists have recently begun to explore ML techniques, and are starting to team up with computer scientists to build new and exciting applications."

Machine Learning Application Areas

Climate modelling and climate data analysis

  • Accelerating climate models: Physical constraints are key ingredients for different components of climate models, including cloud parametrisation, aerosol representation, dynamic vegetation changes, among many other components of GCMs. Traditional solutions to these physics-based models are computationally expensive, but ML components can help with emulating some of the processes, especially on smaller-scales that are currently parametrized in climate models.
  • Physically-constrained projections: Hybrid modelling[10], by incorporating physical-constraints into data-driven ML or deep learning models is a promising field of leveraging the large amounts of data available from observational products, while making robust projections and extrapolating well outside of the training data[11]. The output from climate models can be used for a "perfect model test" of the ML models, before the ML model is applied to make projections based on the observations[12].
  • Downscaling climate models: Climate models often are run on a coarser grid (for computational speed). Downscaling climate projections for smaller grids or specific regions is an important source of information for local impact assessments. ML and deep learning can be useful for interpolation and approximating the fine-scale regional responses based on such coarser climate model output[13],[14].
  • Evaluating and emulating climate models: Climate models can be extremely complex, and climate predictions are often made using the outputs of 20+ climate models. ML can help streamline existing climate models. For instance, ML can help identify and leverage relationships between variables within climate models[15], which could narrow down the spread in the future climate projections[12]. ML can also help to combine the outputs of multiple climate models in order to simplify computation with these ensembles.
  • Collecting observation-based data for climate models: Assimilation of diverse observation-based data sources can improve climate models, and machine learning can transform raw sensor output into more relevant derived data. Relevant applications include sensor calibration and analyzing information in remote sensing data[16]. Well-curated benchmark datasets have the potential to advance several geoscience problems.
  • Filling-in gaps in the observations: Historical record provides valuable information for evaluating performance of climate models given the observations. However, especially early historical observations are available for sparse regions. ML can help with filling-in the gaps in observations to provide a complete record for different climate variables, such as ocean carbon uptake[17][18] or surface air temperature.
  • Detection and attribution of anthropogenic climate change: Separating forced signal (due to anthropogenic climate change) from the natural climate variability has been a challenging task, given only one realisation of observational record[19]. Large ensemble simulations[20], where a given climate model is run multiple times with different initial conditions but identical radiative forcing, are one way of separating the anthropogenic signal from the total response (that is a combination of natural and anthropogenic signal). ML methods provide another avenue for this signal-to-noise problem to detect the anthropogenic signal and attribute it to a given forcing[21][22], now allowing to detect anthropogenic climate change from a single day[23].


  • Seasonal forecasting: Seasonal variations, such as those due to El Niño/Southern Oscillation (ENSO) are difficult to predict using traditional methods. ML and deep learning can be useful for multi-year ENSO forecasting[24][25][26][27][28].
  • Local forecasts of extreme events: Storms, droughts, fires, floods, and other extreme events are expected to become stronger and more frequent as climate change progresses. Machine learning can be used to refine what are otherwise coarse-grained forecasts (e.g., generated from climate or weather prediction models) of these extreme weather events. These high-resolution forecasts can guide improvements in system robustness and resilience.
  • Storm tracking: While climate models can forecast long-term changes in the climate system, separate systems are required to detect specific extreme weather phenomena, like cyclones, atmospheric rivers, and tornadoes. Identifying extreme events in climate model outputs can inform scientific understanding of where and when these events may occur. ML can help classify, detect, and track climate-related extreme events such as hurricanes in climate model outputs.

Background Readings


  • Introduction to climate dynamics and climate modeling (2010)[29]: A technical treatment of the climate system, energy balance, climate modeling, and climate perturbations. Available here.
  • Principles of Planetary Climate (2010)[30]: An introduction to the physics of climate, with examples in python.


Online Courses and Course Materials

  • An Introduction to Climate Modeling (2014)[32]: A video lesson from Climate Literacy's Youtube channel. Available here.

Conferences, Journals, and Professional Organizations

Major conferences

  • American Geosciences Union (AGU) Fall Meeting: A yearly conference organised by the AGU, usually takes place in December, location varies across different states of the USA. Website here.
  • European Geosciences Union (EGU) General Assembly: A yearly conference organised by the EGU, usually takes place in April or Early May in Vienna, Austria. Website here.
  • International Meeting on Statistical Climatology (IMSC): Meetings occurring approximately every three years, usually around June or July, in different locations worldwide. IMSC is organised by statisticians, climatologists and atmospheric scientists aiming to transfer knowledge among different communities (e.g., see the previous meeting)
  • Climate Informatics (CI): annual workshop series, usually occurring in September in different locations worldwide (e.g., see the previous meetings).

Major journals

Applications of Machine Learning in different domains of climate science appear in various climate-related journals, such as:

  • Bulletin of the American Meteorological Society (BAMS): A journal published by the Americal Meteorological Society (AMS). Available here.
  • Earth System Dynamics (ESD): An open-access journal of the European Geophysical Union. Available here.
  • Environmental Research Letters (ERL): An open-access journal from IOPscience publishing group. Available here.
  • Environmental Data Science: A new open-access interdisciplinary journal. Available here.
  • Geophysical Research Letters (GRL): A journal of the American Geophysical Union. Available here.
  • Journal of Climate: A journal published by the Americal Meteorological Society (AMS). Available here.
  • Proceedings of the National Academy of Sciences (PNAS): A wide-reaching journal often featuring climate science. Available here.
  • Springer Nature journals: Often feature climate science topics, recently also with applications of machine learning -e.g., Nature Climate Change, Nature Geoscience, Nature Communications (open access), Communications Earth and Environment (open access), npj Climate and Atmospheric Science (open access).

Major societies and organizations

  • American Geophysical Union (AGU): An organization supporting work across the geophysical sciences. Website here.
  • Climate Informatics (CI): An organization dedicated to computing in climate science. Website here.
  • European Geosciences Union (EGU): An organization supporting research in Earth, planetary, and space science in Europe. Website here.
  • Intergovernmental Panel on Climate Change (IPCC): A United Nations body that assesses climate change and provides policy-relevant information. IPCC provides Assessment Reports and Special Reports that provide a comprehensive summary of the state-of-the-art developments and findings. Website here.

Libraries and Tools


The largest climate prediction datasets are ensembles of many climate simulations. These include simulations with varied physics, architectures, or initial conditions, and they are used to explore the range of possible climate futures. The largest ensembles include:

  • The Coupled Model Intercomparison Project (CMIP): A gateway to climate models in use and development, available here. CMIP is associated with the Earth System Grid Federation, which also provides data analysis tools and tutorials:
  • The CESM Large Ensemble: Read about it in The Community Earth System Model (CESM) Large Ensemble Project. Available here.
  • Google Cloud Weather and Climate Datasets: Petabyte-scale weather and climate datasets from sources like NOAA’s NEXRAD and NASA/USGS’s Landsat, made available for free as part of Google Cloud’s Public Datasets Program. Available here.
  • EARTHDATA: NASA's gateway to earth science data. Data are available at multiple levels of processing. Available here.

N.B. Climate model data is typically presented in netcdf4 format. These may be smoothly converted to csv files or pandas dataframes, but be aware that the data lies on irregular 3D spherical grids.

The Earth and climate science community is also working to create benchmark datasets:


  1. 1.0 1.1 Rolnick, David; Donti, Priya L.; Kaack, Lynn H.; Kochanski, Kelly; Lacoste, Alexandre; Sankaran, Kris; Ross, Andrew Slavin; Milojevic-Dupont, Nikola; Jaques, Natasha; Waldman-Brown, Anna; Luccioni, Alexandra (2019-11-05). "Tackling Climate Change with Machine Learning". arXiv:1906.05433 [cs, stat].
  2. 2.0 2.1 IPCC (2018). Global warming of 1.5C. An IPCC special report on the impacts of global warming of 1.5C above pre-industrial levels and related global greenhouse gas emission pathways, in the context of strengthening the global response to the threat of climate change, sustainable development, and efforts to eradicate poverty.
  3. 3.0 3.1 IPCC (2014). IPCC. Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change.
  4. IPCC (2014). Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change.
  5. "Earth Observation Gateway".
  6. "NASA EarthData".
  7. "Coupled Model Intercomparison Project (CMIP)".
  8. Carman, T (2017). "Position paper on high performance computing needs in Earth system prediction". NOAA Institutional Repository.
  9. "The Community Earth System Model (CESM) Large Ensemble project". 2015. Cite journal requires |journal= (help)
  10. Reichstein, Markus; Camps-Valls, Gustau; Stevens, Bjorn; Jung, Martin; Denzler, Joachim; Carvalhais, Nuno; Prabhat (2019-02). "Deep learning and process understanding for data-driven Earth system science". Nature. 566 (7743): 195–204. doi:10.1038/s41586-019-0912-1. ISSN 1476-4687. Check date values in: |date= (help)
  11. Zhao, Wen Li; Gentine, Pierre; Reichstein, Markus; Zhang, Yao; Zhou, Sha; Wen, Yeqiang; Lin, Changjie; Li, Xi; Qiu, Guo Yu (2019). "Physics-Constrained Machine Learning of Evapotranspiration". Geophysical Research Letters. 46 (24): 14496–14507. doi:10.1029/2019GL085291. ISSN 1944-8007.
  12. 12.0 12.1 Schlund, Manuel; Eyring, Veronika; Camps‐Valls, Gustau; Friedlingstein, Pierre; Gentine, Pierre; Reichstein, Markus (2020). "Constraining Uncertainty in Projected Gross Primary Production With Machine Learning". Journal of Geophysical Research: Biogeosciences. 125 (11): e2019JG005619. doi:10.1029/2019JG005619. ISSN 2169-8961.
  13. Amato, Federico; Guignard, Fabian; Robert, Sylvain; Kanevski, Mikhail (2020-12-17). "A novel framework for spatio-temporal prediction of environmental data using deep learning". Scientific Reports. 10 (1): 22243. doi:10.1038/s41598-020-79148-7. ISSN 2045-2322.
  14. Heinze-Deml, Christina; Sippel, Sebastian; Pendergrass, Angeline G.; Lehner, Flavio; Meinshausen, Nicolai (2020-10-28). "Latent Linear Adjustment Autoencoders v1.0: A novel method for estimating and emulating dynamic precipitation at high resolution". Geoscientific Model Development Discussions: 1–39. doi:10.5194/gmd-2020-275. ISSN 1991-959X.
  15. Schlund, Manuel; Lauer, Axel; Gentine, Pierre; Sherwood, Steven C.; Eyring, Veronika (2020-12-21). "Emergent constraints on equilibrium climate sensitivity in CMIP5: do they hold for CMIP6?". Earth System Dynamics. 11 (4): 1233–1258. doi:10.5194/esd-11-1233-2020. ISSN 2190-4979.
  16. Yuan, Qiangqiang; Shen, Huanfeng; Li, Tongwen; Li, Zhiwei; Li, Shuwen; Jiang, Yun; Xu, Hongzhang; Tan, Weiwei; Yang, Qianqian; Wang, Jiwen; Gao, Jianhao (2020-05-01). "Deep learning in environmental remote sensing: Achievements and challenges". Remote Sensing of Environment. 241: 111716. doi:10.1016/j.rse.2020.111716. ISSN 0034-4257.
  17. Landschützer, P.; Gruber, N.; Bakker, D. C. E.; Schuster, U.; Nakaoka, S.; Payne, M. R.; Sasse, T. P.; Zeng, J. (2013-11-29). "A neural network-based estimate of the seasonal to inter-annual variability of the Atlantic Ocean carbon sink". Biogeosciences. 10 (11): 7793–7815. doi:10.5194/bg-10-7793-2013. ISSN 1726-4170.
  18. Gregor, Luke; Lebehot, Alice D.; Kok, Schalk; Scheel Monteiro, Pedro M. (2019-12-10). "A comparative assessment of the uncertainties of global surface ocean CO2 estimates using a machine-learning ensemble (CSIR-ML6 version 2019a) – have we hit the wall?". Geoscientific Model Development. 12 (12): 5113–5136. doi:10.5194/gmd-12-5113-2019. ISSN 1991-959X.
  19. Bindoff,, N.L., Stott, K.M. AchutaRao, M.R. Allen, N. Gillett, D. Gutzler, K. Hansingo, G. Hegerl, Y. Hu, S. Jain, I.I. Mokhov, J. Overland, J. Perlwitz, R. Sebbari, and X. Zhang (2013). "2013: Detection and attribution of climate change: From global to regional. In Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. T.F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Doschung, A. Nauels, Y. Xia, V. Bex, and P.M. Midgley, Eds. Cambridge University Press, pp. 867-952, doi:10.1017/CBO9781107415324.022".CS1 maint: extra punctuation (link)
  20. "Multi-Model Large Ensemble Archive".
  21. Szekely, Eniko; et al. (2019). "A direct approach to detection and attribution of climate change" (PDF). 9th International Workshop on Climate Informatics. Explicit use of et al. in: |first= (help)
  22. Barnes, Elizabeth A.; Hurrell, James W.; Ebert‐Uphoff, Imme; Anderson, Chuck; Anderson, David (2019). "Viewing Forced Climate Patterns Through an AI Lens". Geophysical Research Letters. 46 (22): 13389–13398. doi:10.1029/2019GL084944. ISSN 1944-8007.
  23. Sippel, Sebastian; Meinshausen, Nicolai; Fischer, Erich M.; Székely, Enikő; Knutti, Reto (2020-01). "Climate change now detectable from any single day of weather at global scale". Nature Climate Change. 10 (1): 35–41. doi:10.1038/s41558-019-0666-7. ISSN 1758-6798. Check date values in: |date= (help)
  24. Ham, Yoo-Geun; Kim, Jeong-Hwan; Luo, Jing-Jia (2019-09). "Deep learning for multi-year ENSO forecasts". Nature. 573 (7775): 568–572. doi:10.1038/s41586-019-1559-7. ISSN 1476-4687. Check date values in: |date= (help)
  25. Toms, Benjamin A.; Barnes, Elizabeth A.; Ebert‐Uphoff, Imme (2020). "Physically Interpretable Neural Networks for the Geosciences: Applications to Earth System Variability". Journal of Advances in Modeling Earth Systems. 12 (9): e2019MS002002. doi:10.1029/2019MS002002. ISSN 1942-2466.
  26. Mahesh, A., et al., (2019). "Forecasting El Niño with Convolutional andRecurrent Neural Networks" (PDF).CS1 maint: extra punctuation (link)
  27. Cachay, S. R. et al.,. "Graph Neural Networks for Improved El NiñoForecasting" (PDF).CS1 maint: extra punctuation (link)
  28. Guo, Yanan; Cao, Xiaoqun; Liu, Bainian; Peng, Kecheng (2020/6). "El Niño Index Prediction Using Deep Learning with Ensemble Empirical Mode Decomposition". Symmetry. 12 (6): 893. doi:10.3390/sym12060893. Check date values in: |date= (help)
  29. Goosse, Hugues (2015). Climate system dynamics and modeling. New York, NY: Cambridge University Press. ISBN 978-1-107-08389-9.
  30. Pierrehumbert, Raymond T. (2010). Principles of planetary climate. Cambridge ; New York: Cambridge University Press. ISBN 978-0-521-86556-2. OCLC 601113992.
  31. "Oxford Research Encyclopedia of Climate Science". Oxford Research Encyclopedia of Climate Science. Retrieved 2020-11-19.
  32. "5.1 Introduction to Climate Modeling - YouTube". Retrieved 2020-09-24.