Climate Modeling and Analysis: Difference between revisions

No edit summary
(35 intermediate revisions by 5 users not shown)
''This page is about the intersection of climate science and machine learning in the context of climate change adaptation. For an overview of climate science as a whole, please see the [ Wikipedia page] on this topic. For an overview of physics-driven climate models, please see the [ Wikipedia page] or this [ Carbon Brief Q&A].''[[File:Clim prediction.png|thumb|The main avenuesAvenues through which machine learning can support climate science, as described in "Tackling Climate Change with Machine Learning." <ref name=s":0">{{Cite journal|last=Rolnick|first=David|last2=Donti|first2=Priya L.|last3=Kaack|first3=Lynn H.|last4=Kochanski|first4=Kelly|last5=Lacoste|first5=Alexandre|last6=Sankaran|first6=Kris|last7=Ross|first7=Andrew Slavin|last8=Milojevic-Dupont|first8=Nikola|last9=Jaques|first9=Natasha|last10=Waldman-Brown|first10=Anna|last11=Luccioni|first11=Alexandra|date=2019-11-05|title=Tackling Climate Change with Machine Learning|url=|journal=arXiv:1906.05433 [cs, stat]}}</ref>|alt=]]
AsClimate describedmodels inare "Tacklingphysics-driven Climatenumerical Changemodels withthat Machineconsist Learning,"<refof name=":0"different /><blockquote>Thecomponents firstof globalthe warmingclimate predictionsystem was(e.g., madeatmosphere, inland, 1896ocean, whensea-ice, Arrhenius estimatedetc.) that burningare fossilconnected fuelsby coulddifferent eventuallyfeedbacks releaseand enoughexchange CO2of tocarbon, warmwater the Earth byand 5°Cenergy. TheClimate fundamentalmodels physicscan underlyingbe those calculations hasof notdifferent changedcomplexity, butranging ourfrom predictionssimple havezero becomeor farone-dimensional moreenergy detailedbalance andmodels Thefully predominantcoupled predictivecomprehensive toolsEarth are climatesystem models, known asor General Circulation Models<ref>{{Cite (GCMs)web|url=|title=What oris Eartha System ModelsGCM?||access-date=2021-03-16}}</ref> (ESMs or GCMs, respectively). These models informare localable to simulate historical climate changes<ref>{{Cite web|url=|title=Analysis: How well have climate models projected global warming?|date=2017-10-05|website=Carbon Brief|language=en|access-date=2021-01-25}}</ref> and nationalare governmentused decisionsin assessment reports of the Intergovernmental Panel on Climate Change (IPCC)<ref name=":1">{{Cite book|url=|title=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|author=IPCC|coauthors=|date=|first=|publisher=|year=2018|isbn=|location=|pages=}}</ref><ref name=":2">{{Cite book|url=|author=IPCC|title=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|date=2014|first=|publisher=|year=|isbn=|location=|pages=}}</ref><ref name=":4">{{Cite book|url=|title=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|author=IPCC|date=2014|first=|publisher=|year=|isbn=|location=|pages=}}</ref>), helpthat peopleprovide calculatepolicy-relevant theirinformation regarding the current state of the climate system and [[future climate projections]] under different [[emission scenarios]]. Climate models also help with asessing climate risks (see [[Policy, Markets, and Decision Science]] and [[Climate Change Adaptation]]). and allow us to estimate the potential impacts of [[Solar Geoengineering|solar geoengineering]].
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<ref>{{Cite web|url=|title=Earth Observation Gateway|date=|access-date=|website=|last=|first=|archive-url=|archive-date=|url-status=live}}</ref><ref>{{Cite web|url=|title=NASA EarthData|date=|access-date=|website=|last=|first=|archive-url=|archive-date=|url-status=live}}</ref>. Second, massive climate modeling projects are generating petabytes of simulated climate data<ref>{{Cite web|url=|title=Coupled Model Intercomparison Project|date=|access-date=|website=|last=|first=|archive-url=|archive-date=|url-status=live}}</ref>. Third, climate forecasts are computationally expensive<ref>{{Cite journal|title=Position paper on high performance computing needs in Earth system prediction|last=Carman|first=T|coauthors=|url=|year=2017|date=|journal=NOAA Institutional Repository|volume=|pages=|via=}}</ref> (some simulations have taken three weeks to run on NCAR supercomputers<ref>{{Cite journal|title=The Community Earth System Model (CESM) Large Ensemble project|year=2015|url=}}</ref>), 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.</blockquote>
Recent advancements in computational resources create opportunities for use of ML methods to aid climate models development, analysis of output, and make use of the vast amount of observational data<ref name=":0" />.
== Machine Learning Application Areas ==
=== UnitingClimate data, ML,modelling and climate sciencedata analysis ===
*'''[[Accelerating climate models]]''': Small-scale processes cannot be directly represented in climate models (due to their coarse resolution), and thus often approximated as parametrisations, such as cloud parametrization, convection, aerosols, dynamic vegetation changes, among many other components of GCMs. Traditional solutions to representation of these processes in GCMs are computationally expensive. ML can help with emulating some of these sub-grid processes<ref>{{Cite journal|last=Rasp|first=Stephan|last2=Pritchard|first2=Michael S.|last3=Gentine|first3=Pierre|date=2018-09-25|title=Deep learning to represent subgrid processes in climate models|url=|journal=Proceedings of the National Academy of Sciences|language=en|volume=115|issue=39|pages=9684–9689|doi=10.1073/pnas.1810286115|issn=0027-8424|pmc=PMC6166853|pmid=30190437}}</ref>, such as vegetation changes<ref>{{Cite journal|last=Dagon|first=Katherine|last2=Sanderson|first2=Benjamin M.|last3=Fisher|first3=Rosie A.|last4=Lawrence|first4=David M.|date=2020-12-22|title=A machine learning approach to emulation and biophysical parameter estimation with the Community Land Model, version 5|url=|journal=Advances in Statistical Climatology, Meteorology and Oceanography|language=English|volume=6|issue=2|pages=223–244|doi=10.5194/ascmo-6-223-2020|issn=2364-3579}}</ref>, and clouds parametrization and convection <ref>{{Cite web|url=|title=Towards physically-consistent, data-driven models of convection|last=Beucler|first=T. et al.,|date=2020|website=|url-status=live|archive-url=|archive-date=|access-date=}}</ref><ref>{{Cite journal|last=Seifert|first=Axel|last2=Rasp|first2=Stephan|date=2020|title=Potential and Limitations of Machine Learning for Modeling Warm-Rain Cloud Microphysical Processes|url=|journal=Journal of Advances in Modeling Earth Systems|language=en|volume=12|issue=12|pages=e2020MS002301|doi=10.1029/2020MS002301|issn=1942-2466}}</ref>.
* Data for climate models
*'''[[Physically-constrained ML projections]]''': Hybrid modelling<ref>{{Cite journal|last=Reichstein|first=Markus|last2=Camps-Valls|first2=Gustau|last3=Stevens|first3=Bjorn|last4=Jung|first4=Martin|last5=Denzler|first5=Joachim|last6=Carvalhais|first6=Nuno|last7=Prabhat|date=2019|title=Deep learning and process understanding for data-driven Earth system science|url=|journal=Nature|language=en|volume=566|issue=7743|pages=195–204|doi=10.1038/s41586-019-0912-1|issn=1476-4687|via=}}</ref>, 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 use of physical constraints present in the climate system, to ensure robust projections and extrapolating well outside of the training data<ref>{{Cite journal|last=Zhao|first=Wen Li|last2=Gentine|first2=Pierre|last3=Reichstein|first3=Markus|last4=Zhang|first4=Yao|last5=Zhou|first5=Sha|last6=Wen|first6=Yeqiang|last7=Lin|first7=Changjie|last8=Li|first8=Xi|last9=Qiu|first9=Guo Yu|date=2019|title=Physics-Constrained Machine Learning of Evapotranspiration|url=|journal=Geophysical Research Letters|language=en|volume=46|issue=24|pages=14496–14507|doi=10.1029/2019GL085291|issn=1944-8007}}</ref>. The output from physics-driven GCM 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<ref name=":3">{{Cite journal|last=Schlund|first=Manuel|last2=Eyring|first2=Veronika|last3=Camps‐Valls|first3=Gustau|last4=Friedlingstein|first4=Pierre|last5=Gentine|first5=Pierre|last6=Reichstein|first6=Markus|date=2020|title=Constraining Uncertainty in Projected Gross Primary Production With Machine Learning|url=|journal=Journal of Geophysical Research: Biogeosciences|language=en|volume=125|issue=11|pages=e2019JG005619|doi=10.1029/2019JG005619|issn=2169-8961}}</ref>.
* Accelerating climate models
*'''[[Climate model evaluation]]''' and narrowing down [[Future climate projections|'''future climate projections''']]: Climate models can be extremely complex, and involve interactions and feedbacks among different components of the climate system. The resulting climate predictions are often made using the outputs of 20+ different climate models, which leads to a wide spread of future climate projections. However, since some components are shared among some climate models, the multi-model mean response is not truly independent<ref>{{Cite journal|last=Knutti|first=Reto|date=2010-10-01|title=The end of model democracy?|url=|journal=Climatic Change|language=en|volume=102|issue=3|pages=395–404|doi=10.1007/s10584-010-9800-2|issn=1573-1480}}</ref>. ML can help identify and leverage relationships between variables within climate models<ref>{{Cite journal|last=Nowack|first=Peer|last2=Runge|first2=Jakob|last3=Eyring|first3=Veronika|last4=Haigh|first4=Joanna D.|date=2020-03-16|title=Causal networks for climate model evaluation and constrained projections|url=|journal=Nature Communications|language=en|volume=11|issue=1|pages=1415|doi=10.1038/s41467-020-15195-y|issn=2041-1723}}</ref><ref>{{Cite journal|last=Schlund|first=Manuel|last2=Lauer|first2=Axel|last3=Gentine|first3=Pierre|last4=Sherwood|first4=Steven C.|last5=Eyring|first5=Veronika|date=2020-12-21|title=Emergent constraints on equilibrium climate sensitivity in CMIP5: do they hold for CMIP6?|url=|journal=Earth System Dynamics|language=English|volume=11|issue=4|pages=1233–1258|doi=10.5194/esd-11-1233-2020|issn=2190-4979}}</ref>, which, together with the observed climate changes (i.e., observational constraint) could narrow down the spread in the future climate projections<ref name=":3" />.
* Working with climate models
*'''[[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<ref>{{Cite journal|last=Amato|first=Federico|last2=Guignard|first2=Fabian|last3=Robert|first3=Sylvain|last4=Kanevski|first4=Mikhail|date=2020-12-17|title=A novel framework for spatio-temporal prediction of environmental data using deep learning|url=|journal=Scientific Reports|language=en|volume=10|issue=1|pages=22243|doi=10.1038/s41598-020-79148-7|issn=2045-2322}}</ref>,<ref>{{Cite journal|last=Heinze-Deml|first=Christina|last2=Sippel|first2=Sebastian|last3=Pendergrass|first3=Angeline G.|last4=Lehner|first4=Flavio|last5=Meinshausen|first5=Nicolai|date=2020-10-28|title=Latent Linear Adjustment Autoencoders v1.0: A novel method for estimating and emulating dynamic precipitation at high resolution|url=|journal=Geoscientific Model Development Discussions|language=English|pages=1–39|doi=10.5194/gmd-2020-275|issn=1991-959X}}</ref>.
*'''[[Data Assimilation]]''': 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|remote sensing]] data<ref>{{Cite journal|last=Yuan|first=Qiangqiang|last2=Shen|first2=Huanfeng|last3=Li|first3=Tongwen|last4=Li|first4=Zhiwei|last5=Li|first5=Shuwen|last6=Jiang|first6=Yun|last7=Xu|first7=Hongzhang|last8=Tan|first8=Weiwei|last9=Yang|first9=Qianqian|last10=Wang|first10=Jiwen|last11=Gao|first11=Jianhao|date=2020-05-01|title=Deep learning in environmental remote sensing: Achievements and challenges|url=|journal=Remote Sensing of Environment|language=en|volume=241|pages=111716|doi=10.1016/j.rse.2020.111716|issn=0034-4257}}</ref> or assimilating climate model output with the observations<ref>{{Cite web|url=|title=Machine learning at ECMWF: A roadmap for the next 10 years|last=|first=|date=|website=|url-status=live|archive-url=|archive-date=|access-date=}}</ref>. Well-curated benchmark datasets have the potential to advance several geoscience problems<ref>{{Cite web|url=|title=WeatherBench: A benchmark dataset for data-driven weather forecasting|last=Rasp|first=S., et al.,|date=2020|website=|url-status=live|archive-url=|archive-date=|access-date=}}</ref>.
=== Forecasting extreme events ===
*'''[[Filling in gaps in the observations]]''': Historical record provides valuable information for evaluating the performance of climate models with respect to the observed changes. However, especially early historical observations are available only 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<ref>{{Cite journal|last=Landschützer|first=P.|last2=Gruber|first2=N.|last3=Bakker|first3=D. C. E.|last4=Schuster|first4=U.|last5=Nakaoka|first5=S.|last6=Payne|first6=M. R.|last7=Sasse|first7=T. P.|last8=Zeng|first8=J.|date=2013-11-29|title=A neural network-based estimate of the seasonal to inter-annual variability of the Atlantic Ocean carbon sink|url=|journal=Biogeosciences|language=English|volume=10|issue=11|pages=7793–7815|doi=10.5194/bg-10-7793-2013|issn=1726-4170}}</ref><ref>{{Cite journal|last=Gregor|first=Luke|last2=Lebehot|first2=Alice D.|last3=Kok|first3=Schalk|last4=Scheel Monteiro|first4=Pedro M.|date=2019-12-10|title=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?|url=|journal=Geoscientific Model Development|language=English|volume=12|issue=12|pages=5113–5136|doi=10.5194/gmd-12-5113-2019|issn=1991-959X}}</ref> or surface air temperature using neural networks<ref>{{Cite journal|last=Kadow|first=Christopher|last2=Hall|first2=David Matthew|last3=Ulbrich|first3=Uwe|date=2020|title=Artificial intelligence reconstructs missing climate information|url=|journal=Nature Geoscience|language=en|volume=13|issue=6|pages=408–413|doi=10.1038/s41561-020-0582-5|issn=1752-0908|via=}}</ref>, Kriging<ref>{{Cite journal|last=Cowtan|first=Kevin|last2=Way|first2=Robert G.|date=2014|title=Coverage bias in the HadCRUT4 temperature series and its impact on recent temperature trends|url=|journal=Quarterly Journal of the Royal Meteorological Society|language=en|volume=140|issue=683|pages=1935–1944|doi=10.1002/qj.2297|issn=1477-870X}}</ref><ref>{{Cite journal|last=Morice|first=C. P.|last2=Kennedy|first2=J. J.|last3=Rayner|first3=N. A.|last4=Winn|first4=J. P.|last5=Hogan|first5=E.|last6=Killick|first6=R. E.|last7=Dunn|first7=R. J. H.|last8=Osborn|first8=T. J.|last9=Jones|first9=P. D.|last10=Simpson|first10=I. R.|title=An updated assessment of near-surface temperature change from 1850: the HadCRUT5 dataset|url=|journal=Journal of Geophysical Research: Atmospheres|language=en|volume=n/a|issue=n/a|pages=e2019JD032361|doi=10.1029/2019JD032361|issn=2169-8996}}</ref>, or Empirical Orthogonal Functions<ref>{{Cite journal|last=Benestad|first=R. E.|last2=Erlandsen|first2=H. B.|last3=Mezghani|first3=A.|last4=Parding|first4=K. M.|date=2019|title=Geographical Distribution of Thermometers Gives the Appearance of Lower Historical Global Warming|url=|journal=Geophysical Research Letters|language=en|volume=46|issue=13|pages=7654–7662|doi=10.1029/2019GL083474|issn=1944-8007}}</ref>.
*'''[[Detection and attribution of anthropogenic climate change]]:''' Separating forced signal (due to anthropogenic climate change) from the "noise" due to natural climate variability has been a challenging task, given only one realization of observational record<ref>{{Cite web|url=|title=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.|last=Bindoff,|first=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|date=2013|website=|url-status=live|archive-url=|archive-date=|access-date=}}</ref><ref>{{Cite journal|last=Gillett|first=Nathan P.|last2=Kirchmeier-Young|first2=Megan|last3=Ribes|first3=Aurélien|last4=Shiogama|first4=Hideo|last5=Hegerl|first5=Gabriele C.|last6=Knutti|first6=Reto|last7=Gastineau|first7=Guillaume|last8=John|first8=Jasmin G.|last9=Li|first9=Lijuan|last10=Nazarenko|first10=Larissa|last11=Rosenbloom|first11=Nan|date=2021-01-18|title=Constraining human contributions to observed warming since the pre-industrial period|url=|journal=Nature Climate Change|language=en|pages=1–6|doi=10.1038/s41558-020-00965-9|issn=1758-6798}}</ref>. Large ensemble simulations<ref>{{Cite web|url=|title=Multi-Model Large Ensemble Archive|last=|first=|date=|website=|url-status=live|archive-url=|archive-date=|access-date=}}</ref>, 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 the natural and anthropogenic signals). ML methods provide another avenue for addressing this signal-to-noise problem, to aid with detecting the anthropogenic signal and attributing it to a given forcing<ref>{{Cite web|url=|title=A direct approach to detection and attribution of climate change|last=Szekely|first=Eniko et al.,|date=2019|website=9th International Workshop on Climate Informatics|url-status=live|archive-url=|archive-date=|access-date=}}</ref><ref>{{Cite journal|last=Barnes|first=Elizabeth A.|last2=Hurrell|first2=James W.|last3=Ebert‐Uphoff|first3=Imme|last4=Anderson|first4=Chuck|last5=Anderson|first5=David|date=2019|title=Viewing Forced Climate Patterns Through an AI Lens|url=|journal=Geophysical Research Letters|language=en|volume=46|issue=22|pages=13389–13398|doi=10.1029/2019GL084944|issn=1944-8007}}</ref>. Statistical learning also allows detecting anthropogenic climate change from a single day<ref>{{Cite journal|last=Sippel|first=Sebastian|last2=Meinshausen|first2=Nicolai|last3=Fischer|first3=Erich M.|last4=Székely|first4=Enikő|last5=Knutti|first5=Reto|date=2020|title=Climate change now detectable from any single day of weather at global scale|url=|journal=Nature Climate Change|language=en|volume=10|issue=1|pages=35–41|doi=10.1038/s41558-019-0666-7|issn=1758-6798|via=}}</ref>.
=== Forecasting of seasonal variations and extreme events ===
* Storm tracking
* Local forecasts
*'''[[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<ref>{{Cite journal|last=Ham|first=Yoo-Geun|last2=Kim|first2=Jeong-Hwan|last3=Luo|first3=Jing-Jia|date=2019|title=Deep learning for multi-year ENSO forecasts|url=|journal=Nature|language=en|volume=573|issue=7775|pages=568–572|doi=10.1038/s41586-019-1559-7|issn=1476-4687|via=}}</ref><ref>{{Cite journal|last=Toms|first=Benjamin A.|last2=Barnes|first2=Elizabeth A.|last3=Ebert‐Uphoff|first3=Imme|date=2020|title=Physically Interpretable Neural Networks for the Geosciences: Applications to Earth System Variability|url=|journal=Journal of Advances in Modeling Earth Systems|language=en|volume=12|issue=9|pages=e2019MS002002|doi=10.1029/2019MS002002|issn=1942-2466}}</ref><ref>{{Cite web|url=|title=Forecasting El Niño with Convolutional andRecurrent Neural Networks|last=Mahesh|first=A., et al.,|date=2019|website=|url-status=live|archive-url=|archive-date=|access-date=}}</ref><ref>{{Cite web|url=|title=Graph Neural Networks for Improved El NiñoForecasting|last=Cachay|first=S. R. et al.,|date=|website=|url-status=live|archive-url=|archive-date=|access-date=}}</ref><ref>{{Cite journal|last=Guo|first=Yanan|last2=Cao|first2=Xiaoqun|last3=Liu|first3=Bainian|last4=Peng|first4=Kecheng|date=2020|title=El Niño Index Prediction Using Deep Learning with Ensemble Empirical Mode Decomposition|url=|journal=Symmetry|language=en|volume=12|issue=6|pages=893|doi=10.3390/sym12060893|via=}}</ref>.
*'''[[Localized Extreme Event Forecasting|Extreme events predictions]]''': 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.
For use of ML in weather forecasting, see [[weather prediction]].
== Background Readings ==
*'''Introduction to climate dynamics and climate modeling (2010)'''<ref>{{Cite book|title=Climate system dynamics and modeling|last=Goosse|first=Hugues|date=2015|publisher=Cambridge University Press|isbn=978-1-107-08389-9|location=New York, NY}}</ref>: A technical treatment of the climate system, energy balance, climate modeling, and climate perturbations. Available [ here].
*'''Principles of Planetary Climate (2010)'''<ref>{{Cite book|url=|title=Principles of planetary climate|last=Pierrehumbert|first=Raymond T.|date=2010|publisher=Cambridge University Press|isbn=978-0-521-86556-2|location=Cambridge ; New York|oclc=601113992}}</ref>: An introduction to the physics of climate, with examples in python.
=== Other ===
*'''Intergovernmental Panel on Climate Change (IPCC)''' Assessment Reports (e.g., AR5)<ref name=":2" /> and the IPCC Special Report Global Warming of 1.5 ºC<ref name=":1" />
*'''An Introduction to Climate Modeling (2014)'''<ref>{{Cite web|url=|title=5.1 Introduction to Climate Modeling - YouTube||access-date=2020-09-24}}</ref>: A video lesson from Climate Literacy's Youtube channel. Available [ here].
*'''Carbon Brief''' explainers on [ How do climate models work], [ How well have climate models projected global warming], and [ The next generation of climate models (CMIP6).]
*'''Oxford Research Encyclopedia of Climate Science'''<ref>{{Cite web|url=|title=Oxford Research Encyclopedia of Climate Science|website=Oxford Research Encyclopedia of Climate Science|language=en|access-date=2020-11-19}}</ref>: A collection of articles on the climate systems, impacts of climate change, and the methods used in climate science.
*'''Ted Talks''' about how climate models work and what they can be used for, by [ Dr Gavin Schmidt] and [ Dr Kate Marvel].
== Online Courses and Course Materials ==
* '''[ An Introduction to Climate Modeling] (2014)'''<ref>{{Cite web|url=|title=5.1 Introduction to Climate Modeling - YouTube||access-date=2020-09-24}}</ref>: A video lesson from Climate Literacy's Youtube channel. Available [ here].
== Community ==
* '''[ A Climate Modelling course]''': A hands-on lectures notes and Python code by Prof. Biran E. J. Rose.
* '''[ Advanced courses in climate date science]''': ''Research Computing in Earth Science, Introduction to Physical Oceanography, Geophysical Fluid Dynamics'' (with python code) by Prof. Ryan Abernathey.
== 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.
*'''AGU Fall Meeting''': A yearly conference organized by the American Geophysical Union. 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.
*'''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:
Climate science is a journal field. Noteworthy research appears in journals such as
*'''[ Bulletin of the American Meteorological Society] (BAMS)''': A journal published by the AMS.American AvailableMeteorological [ here](AMS).
*'''Geophysical[ ResearchEarth LettersSystem Dynamics] (ESD)''': TheAn open-access journal of the AmericanEuropean Geophysical Union. Available [ here].
*'''[ Environmental Research Letters] (ERL):''' An open-access journal from IOPscience publishing group.
*'''Proceedings of the National Academy of Sciences''': A wide-reaching journal often featuring climate science. Available [ here].
*'''[ Environmental Data Science]''': A new open-access interdisciplinary journal.
*'''[ Geophysical Research Letters] (GRL):''' A journal of the American Geophysical Union.
*'''[ Journal of Climate]:''' A journal published by the Americal Meteorological Society (AMS).
*'''[ 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).
*'''[ Artificial Intelligence for the Earth Systems] (AIES)''': A new journal focusing on AI published by the American Meteorological Society (AMS).
=== 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.
*'''[ 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.
== Libraries and Tools ==
'''Pangeo''': An open source python package for geoscience applications, available [ here].
* '''[ Pangeo]''': An open source python package for geoscience applications, available [ here].
** Pangeo also maintains a list of packages useful for [ atmospheric, ocean, and climate science].
''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. [ Xarray]is a commonly used Python package for post-processing of climate data.
* '''Climate Data Operators''' [ CDO] and [ NCO] are command-line tools that can be used in manipulating netcdf files and calculating climatologies.
* '''AI2 Climate Modeling''': The [ ai2cm] climate modelling toolbox provides different Python wrappers to work with climate and weather models provided by the [ Allen Institute for AI].
== Data ==
The [ NCAR Climate Data Guide] is a useful resource for learning about different datasets and where to find source data for different components of the climate system, including the atmosphere, ocean, and land-based climate indices, as well as observation-based product from satellite and reanalysis data sources.
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 output from climate models includes various simulations, driven under historical and different future emission scenarios. Some simulations are idealised (e.g., in response to CO2-only forcing), to aid with inter-comparison of models, or focus on comparing specific components of the climate system (e.g. land or ocean carbon uptake). The Coupled Model Intercomparison Project (CMIP) is an international exercise among different modelling centres to compare the output of climate models for a given set of scenarios and simulations<ref>{{Cite journal|last=Eyring|first=Veronika|last2=Bony|first2=Sandrine|last3=Meehl|first3=Gerald A.|last4=Senior|first4=Catherine A.|last5=Stevens|first5=Bjorn|last6=Stouffer|first6=Ronald J.|last7=Taylor|first7=Karl E.|date=2016-05-26|title=Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization|url=|journal=Geoscientific Model Development|language=English|volume=9|issue=5|pages=1937–1958|doi=10.5194/gmd-9-1937-2016|issn=1991-959X}}</ref>. Other climate model simulations may include with varied physics (e.g., perturbed ensemble simulations), or different initial conditions (often referred to as large ensemble simulations), where the differences in output for each model arise due to natural climate variability. These different types of simulations are used to explore a range of different future climate projections and quantify different sources of uncertainties<ref>{{Cite journal|last=Lehner|first=Flavio|last2=Deser|first2=Clara|last3=Maher|first3=Nicola|last4=Marotzke|first4=Jochem|last5=Fischer|first5=Erich M.|last6=Brunner|first6=Lukas|last7=Knutti|first7=Reto|last8=Hawkins|first8=Ed|date=2020-05-29|title=Partitioning climate projection uncertainty with multiple large ensembles and CMIP5/6|url=|journal=Earth System Dynamics|language=English|volume=11|issue=2|pages=491–508|doi=10.5194/esd-11-491-2020|issn=2190-4979}}</ref>.
*'''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].
Key resources for accessing climate and weather data:
''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 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 Earth and climate science community is also working to create benchmark datasets:
*'''[ Large ensemle simulations]''' for different climate models, also referred to as Single Model Initial Condition Ensemble [ SMILEs], including [ The Community Earth System Model (CESM) Large Ensemble Project]
*'''Climate and weather datasets for ML research''' are listed [ here].
*The Earth and climate science community is also working to create benchmark datasets: [ benchmark datasets]
*'''[ 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 earthEarth scienceScience data. Data are available at multiple levels of processing. Available [ here].
== References ==
<references />