Climate Models & Frameworks — Cross-Cutting Comparison

This note compares the climate and Earth-system modeling frameworks a researcher actually chooses between when designing an experiment — full coupled CMIP6 ESMs vs convection-permitting regional models vs intermediate-complexity EMICs vs component-only models (ocean, land, sea-ice, ice-sheet, carbon-cycle) vs integrated assessment models vs the new ML-based neural ESMs — across every ClimateScience library note that touches simulation. Each family has a unified table mapping models against the axes that drive selection (spatial resolution, coupling complexity, computational cost, calibration target, openness). The closing decision tree picks the model class given the research question.

See also

1. The five axes every model choice answers

Climate-model selection looks like five orthogonal questions before any code runs.

AxisCheap endExpensive endWhy it matters
Spatial resolutionEMIC (~500 km)LES / CRM (~100 m)physics resolved vs parameterized
Couplingatmosphere-onlyatmosphere+ocean+land+sea-ice+ice-sheet+chemistry+veg dynamics+CO₂feedback completeness
Time horizonhours (weather)10⁹ yr (deep-time paleoclimate)spin-up + integration cost
Computational costEMIC: hr on desktophigh-res ESM: month on 10⁴ coreswallclock per scenario
Calibration targetpaleoclimate / pre-industrial / historical / futurewhich observations constrain tuningpropagates into projection skill

The textbook ordering says “pick the simplest model that captures your feedbacks”. The practical reality: pick the model whose vetting cost (you understand it, the community trusts it, IPCC AR cites it) matches your timeline. A bespoke high-res run nobody else has done buys little if your reviewer trusts CESM2 outputs.

2. CMIP6 Earth-system models — the canonical coupled set

The Coupled Model Intercomparison Project Phase 6 (CMIP6) is the IPCC AR6 (2021) backbone — ~100 modeling groups submit standardized experiments (historical, SSP1-1.9 through SSP5-8.5, DECK, AMIP, RFMIP, GeoMIP, ScenarioMIP). The lineage continues with CMIP7 (in progress, IPCC AR7 deadline ~2028).

ModelGroupAtm resOcn resCouplingEquilibrium climate sensitivity (ECS, °C)Linked note
CESM2NCAR (US)~1° (CAM6)1° (POP2)atm+ocn+land(CLM5)+ice(CICE6) + carbon + atm chem (CAM-chem) + WACCM(stratosphere)5.2physical-climate-system / climate-sensitivity-and-feedbacks
CESM2.2 / CESM3NCAR~0.25° HR variant0.1° HR variantas CESM2 + improved aerosolbeing released 2025-2026physical-climate-system
E3SM v2 / v3US DOE (LLNL+ANL+ORNL+PNNL+SNL)regionally refined unstruct (MPAS-A)unstruct (MPAS-O)atm+ocn+ice+land+river+ice-sheet (MALI)4.0physical-climate-system
GFDL ESM4 / SPEARNOAA (US)1° (AM4)0.5° (MOM6)atm+ocn(MOM6)+ice(SIS2)+land(LM4)+atm chem + biogeochem2.7atmospheric-chemistry-and-aerosols
HadGEM3-GC31Met Office (UK)60 km (N216)0.25°atm(UM)+ocn(NEMO)+ice(CICE)+land(JULES); UKESM1 adds chem + ocn BGC5.3 (UKESM1)physical-climate-system
UKESM1UK consortium (Met Office + NCAS + NERC)60 km0.25°full Earth system (atm chem UKCA + MEDUSA ocn BGC + TRIFFID veg)5.3carbon-cycle-and-greenhouse-gases
MPI-ESM 1.2Max-Planck (Germany)1.9° (ECHAM6)0.4° (MPIOM)atm+ocn+land(JSBACH)+ice + carbon3.0carbon-cycle-and-greenhouse-gases
MPI-ESM HRMax-Planck0.5°0.4°as MPI-ESM 1.2 high-res3.0physical-climate-system
IPSL-CM6A-LRIPSL (France)2.5° (LMDZ6)1° (NEMO)atm+ocn+land(ORCHIDEE)+ice+atm chem(REPROBUS/INCA)4.6atmospheric-chemistry-and-aerosols
ICON-A / ICON-ESMDWD + MPI-M (Germany)unstruct icosahedral 80-160 km; 5 km in ICON-Sapphire CRM mode0.4°atm+ocn(ICON-O)+land(JSBACH)~3atmospheric-dynamics-deep
MIROC6JAMSTEC + NIES + U-Tokyo (Japan)1.4°1° (COCO)atm+ocn+land(MATSIRO)+ice2.6physical-climate-system
MIROC-ES2Las above2.8°+ carbon cycle + aerosol + chem2.7carbon-cycle-and-greenhouse-gases
NorESM2-LM / MMNorway consortium2° / 1°1° (BLOM)atm(CAM-OSLO aerosol)+ocn(BLOM)+ice(CICE)+land(CLM5)2.5atmospheric-chemistry-and-aerosols
AWI-CM-1-1-MRAWI (Germany)1° (ECHAM6)unstruct FESOM2 (varies)atm+ocn(unstruct)+ice3.2physical-climate-system
CMCC-ESM2CMCC (Italy)1° (CAM5.3)0.25° (NEMO)atm+ocn+land(CLM4.5)+ice3.6physical-climate-system
BCC-ESM1Beijing Climate Center (China)T42 (~2.8°)1° (MOM4)atm+ocn+land+ice+aerosol3.3physical-climate-system
CanESM5Environment CanadaT63 (~2.8°)~1° (NEMO)atm+ocn+land(CLASS-CTEM)+atm chem (CMAM-chem)5.6 (high-ECS outlier)climate-sensitivity-and-feedbacks
EC-Earth3European consortium~1° (IFS)0.25° (NEMO)atm(IFS)+ocn(NEMO)+ice(LIM3)+land(HTESSEL)4.3physical-climate-system
ACCESS-ESM1-5Australia (CSIRO + BoM)~1° (UM)1° (MOM5)atm+ocn+land(CABLE)+ice + carbon3.9carbon-cycle-and-greenhouse-gases
CNRM-ESM2-1Météo-France + CERFACST127 (~1.4°)1° (NEMO)atm(ARPEGE-Climat)+ocn(NEMO)+ice(GELATO)+land(SURFEX-ISBA)4.8physical-climate-system
INM-CM5-0INM RAS (Russia)0.5°atm+ocn+land+ice1.9 (low-ECS outlier)climate-sensitivity-and-feedbacks
KIOST-ESMKIOST (Korea)T62~1°atm+ocn+land+ice3.4physical-climate-system
FIO-ESMFirst Inst of Oceanography (China)T42adds explicit wave model3.5physical-climate-system
MRI-ESM2-0Met Research Institute (Japan)T127~1°atm+ocn+land+ice + atm chem + aerosol3.2atmospheric-chemistry-and-aerosols
SAM0-UNICONSeoul National (Korea)as CESM2 + UNICON convection4.0physical-climate-system

A typical CMIP6 historical+SSP1-1.9 through SSP5-8.5 ensemble for one of these models is ~5-10 Mcore-hours on a Tier-1 HPC system. CMIP6 archive is on ESGF (Earth System Grid Federation, esgf-node.llnl.gov and mirrors) at >20 PB. Access via cdo, xarray, intake-esm, pangeo. The python-Pangeo + Zarr stack on AWS / GCP is the de-facto modern access path.

Notable ECS spread: CanESM5 (5.6 °C) and HadGEM3/UKESM1 (5.3 °C) sit high; INM-CM5 (1.9 °C) sits low. The IPCC AR6 “very likely range” of ECS = 2.0–5.0 °C (best 3.0 °C) deliberately narrower than the model spread because emergent-constraint analyses tightened the range — see climate-sensitivity-and-feedbacks.

3. Convection-permitting and regional models

Below ~10 km grid spacing, deep convection is partially resolved rather than parameterized. This regime is where most regional climate-change adaptation work happens.

ModelGroupRangeNative resWhere usedLinked note
WRF (Weather Research & Forecasting)NCAR (US)regional, ~50 km–100 m1 km common; 100 m LESdominant US regional (NCAR, EPA, DOE, universities)regional-climate-and-downscaling
WRF-ChemNCAR / NOAAregional + atm chem1 kmair quality + climate feedbackatmospheric-chemistry-and-aerosols
RegCM4 / RegCM5ICTP (Italy)regional10–25 kmCORDEX (Coordinated Regional Climate Downscaling) workhorseregional-climate-and-downscaling
COSMO-CLMClimate Limited-area Modelling Communityregional2.8 km commonEU + Germany regionalregional-climate-and-downscaling
REMOHelmholtz-Zentrum Geesthacht / HZGregional12–25 kmEU CORDEXregional-climate-and-downscaling
ALADIN-ClimateMétéo-France + ALADIN consortiumregional12 km common; HARMONIE 2.5 kmEU operational + climateregional-climate-and-downscaling
HadRM3 / HadREM3Met Office (UK)regional12 km / 2.2 kmUK climate projections (UKCP18, UKCP-Local)regional-climate-and-downscaling
RACMO2KNMI (Netherlands)regional + ice sheet5.5–11 kmAntarctica + Greenland surface mass balanceglaciology-and-cryosphere
MAR (Modèle Atmosphérique Régional)U-Liège (Belgium)polar regional7.5–25 kmGreenland + Antarctic SMB; ice-sheet projectionsglaciology-and-cryosphere
NICAM (Nonhydrostatic Icosahedral)RIKEN + JAMSTEC + U-Tokyoglobal cloud-resolving870 m / 3.5 kmglobal CRM Aqua-planet + DYAMOND-classatmospheric-dynamics-deep
ICON-SapphireMPI-M (Germany)global CRM5 km / 2.5 kmDYAMOND + nextGEMS Earth Virtualization Enginesatmospheric-dynamics-deep
X-SHiELDNOAA-GFDLglobal CRM3.25 kmDYAMOND classatmospheric-dynamics-deep
SCREAM (Simple Cloud-Resolving E3SM Atm Model)DOE (US)global CRM3 kmE3SM CRM-mode, GPU-native (Frontier)atmospheric-dynamics-deep

CORDEX (Coordinated Regional Climate Downscaling Experiment) is the regional analog of CMIP — standardized regional simulations for 14 CORDEX domains (Africa, Europe, North America, etc.). EURO-CORDEX is the most active subset.

4. EMICs — Earth-system Models of Intermediate Complexity

Below CMIP6 ESMs lies a class of models that trade resolution for speed, enabling deep-time paleoclimate (10⁵–10⁸ yr), large ensembles for uncertainty quantification, and IAM coupling.

ModelGroupWhat’s simplifiedRangeWhere usedLinked note
UVic ESCMU-Victoria (Canada)1.8°×3.6° energy-moisture-balance atm; full ocnup to 10⁵ yrclimate-carbon-cycle feedback, deep-time, MIP6 (CMIP6 contribution)carbon-cycle-and-greenhouse-gases
CLIMBER-2 / CLIMBER-XPIK (Potsdam, Germany)sectoral statistical-dynamical atm; full ocn; ice sheetup to 10⁶ yrtipping points, glacial cycles, AMOC collapsepaleoclimate-and-deep-time
Bern3DU-Bern (Switzerland)2D zonally-averaged atm; 3D ocnup to 10⁵ yrcarbon cycle, ocean BGC, paleoclimatecarbon-cycle-and-greenhouse-gases
GENIE / cGENIEU-Bristol + UEA (UK)reduced-cmplx atm (EMBM); GOLDSTEIN ocnup to 10⁸ yrdeep-time paleoclimate, PETM, snowball Earthpaleoclimate-and-deep-time
LOVECLIMUCLouvain (Belgium)quasi-geostrophic atm (ECBilt); CLIO ocnup to 10⁵ yrHolocene, last deglaciationpaleoclimate
CESM (low-res FullPhys-CESM)NCAR4° + 3° ocnup to 10⁴ yrbridges EMIC to full ESMpaleoclimate-and-deep-time
MAGICCClimate Resource (Australia)global-mean energy balance + reduced ocn + carbon cycleseconds-to-minutes per scenarioreduced-complexity for IAM use; emulates CMIP6 spreadipcc-scenarios-and-integrated-assessment
FaIRUK consortium (Smith et al.)impulse responseseconds per scenarioreduced-complexity for AR6 + IAM couplingipcc-scenarios-and-integrated-assessment
OSCARLSCE / IIASAlinearized boxseconds per scenarioAR6 Probabilistic Earth-System Emulatoripcc-scenarios-and-integrated-assessment
HectorPNNL1D boxseconds per scenarioIAM-coupled (GCAM partner)ipcc-scenarios-and-integrated-assessment

The MAGICC + FaIR + OSCAR + Hector class is what IAMs actually couple to — full CMIP6 ESMs are too slow for IAM-loop iteration. AR6 ChapAtlas chapter constructs probabilistic projections from MAGICC + FaIR calibrated to CMIP6 spread.

5. Specialized component models

A modular view — these are the components that combine into CMIP6 ESMs above, but are also run standalone for component-specific science.

5.1 Ocean-only

ModelGroupNative resWhere usedLinked note
MOM6NOAA-GFDL + Princeton + NCAR0.1° (HR), 0.25° (MR), 1° (LR)de-facto US ocean model; CESM, GFDL, NCARocean-biogeochemistry
NEMONEMO Consortium (CNRS + Met Office + INGV + Mercator Ocean)0.25° (ORCA025) → 1/12°dominant EU ocean (UKESM1, IPSL, EC-Earth)ocean-biogeochemistry
MITgcmMIT1/3° → eddy-resolvingECCO ocean state estimate; Lagrangian analysisocean-biogeochemistry
POP2NCAR (legacy)pre-MOM6 CESM oceanocean-biogeochemistry
MIROC OGCM (COCO)JAMSTECMIROC ocean componentocean-biogeochemistry
FESOM2AWI (Germany)unstruct (varies)AWI-CM coupled, eddy-permitting in key regionsocean-biogeochemistry
MPAS-OLANL + DOEunstructE3SM ocean componentocean-biogeochemistry

5.2 Atmosphere-only

ModelGroupNative resWhere usedLinked note
CAM6 / CAM7NCAR~1°CESM atmosphere; CAM-chem variant + WACCM stratosphereatmospheric-chemistry-and-radiative-transfer
IFS (Integrated Forecasting System)ECMWF9 km HRES (TCo1279) → 0.1° AIFSECMWF medium-range + EC-Earth coupledatmospheric-dynamics-deep
MPAS-ANCAR / LANLunstruct (varies)E3SM + research; mesh-refinementatmospheric-dynamics-deep
FV3NOAA-GFDL13 km (UFS)NOAA UFS operational; SHIELD HRatmospheric-dynamics-deep
COSMODWD + COSMO Consortium2.2 km operationalDWD operational + COSMO-CLMatmospheric-dynamics-deep
UM (Unified Model)Met Office10 km operationalHadGEM atmosphere + GloSea seasonalatmospheric-dynamics-deep
ARPEGE-ClimatMétéo-FranceT127–T255CNRM-CM/ESM atmosphereatmospheric-dynamics-deep
ECHAM6MPI-MT63/T127MPI-ESM atmosphereatmospheric-dynamics-deep
LMDZLMD (France)2.5°IPSL atmosphereatmospheric-dynamics-deep

5.3 Land-surface

ModelGroupWhere usedLinked note
CLM5NCARCESM + NorESM + CMCC; gold-standard for plant functional types + hydrology + biogeochemhydrology-and-water-cycle
JULESMet Office + CEHHadGEM + UKESMhydrology-and-water-cycle
ISBA (SURFEX-ISBA)Météo-FranceCNRMhydrology-and-water-cycle
ORCHIDEEIPSL + LSCEIPSL ESMcarbon-cycle-and-greenhouse-gases
JSBACHMPI-MMPI-ESMcarbon-cycle-and-greenhouse-gases
Noah-MPNCAR + NOAAWRF + NOAA UFShydrology-and-water-cycle
CABLECSIRO (Australia)ACCESScarbon-cycle-and-greenhouse-gases
CLASS-CTEMEnvironment CanadaCanESMcarbon-cycle-and-greenhouse-gases

5.4 Sea-ice

ModelGroupWhere usedLinked note
CICE6LANL + NCAR + Met OfficeCESM + UKESM + NorESM + CanESM + EC-Earthglaciology-and-cryosphere
LIM3UCLouvainNEMO-based (IPSL-CM, EC-Earth)glaciology-and-cryosphere
SIS2NOAA-GFDLGFDL ESM4glaciology-and-cryosphere
GELATOCNRMCNRM-ESM2glaciology-and-cryosphere

5.5 Ice-sheet

ModelGroupWhere usedLinked note
PISM (Parallel Ice Sheet Model)U-Alaska + PIKresearch workhorse; Antarctic + Greenland projectionsglaciology-and-cryosphere
ISSM (Ice-Sheet System Model)JPL / NASAadjoint capability for inverse problems, satellite assimilationglaciology-and-cryosphere
BISICLESLBNL + BristolAMR (adaptive mesh refinement) for shelves + grounding linesglaciology-and-cryosphere
Elmer/IceCSC Finland + IGE Grenoblefinite-element, full Stokes; thermomechanicalglaciology-and-cryosphere
CISMLANL + NCARCESM-coupled ice-sheetglaciology-and-cryosphere
MALI (MPAS-Albany Land Ice)DOE (LANL+SNL+ORNL)E3SM-coupled; unstruct gridsglaciology-and-cryosphere
ÚaMITgcmvariousgrounded-ice + ice-shelf MITgcm-coupledglaciology-and-cryosphere

ISMIP6 (Ice Sheet MIP for CMIP6) is the standardized intercomparison; produced AR6 ice-sheet sea-level contributions.

5.6 Carbon cycle

ModelGroupWhat it doesLinked note
CASANASA + Stanfordterrestrial primary productivity from satellitecarbon-cycle-and-greenhouse-gases
CTE (CarbonTracker Europe)Wageningenatmospheric inverse for CO₂ + CH₄carbon-cycle-and-greenhouse-gases
CarbonTrackerNOAA + CIRESatmospheric inverse for CO₂ flux (operational since 2007)carbon-cycle-and-greenhouse-gases
TM5 (Transport Model 5)KNMI + Wageningenatmospheric transport for inversionscarbon-cycle-and-greenhouse-gases
TM5-MPKNMI + JRCmassively-parallel TM5carbon-cycle-and-greenhouse-gases
GEOS-ChemHarvard + communityglobal atm chem + carbonatmospheric-chemistry-and-aerosols

6. Integrated Assessment Models (IAMs)

IAMs couple climate (a reduced-complexity emulator like MAGICC or FaIR) with economy + energy + land-use + technology. The IPCC AR6 WG3 used 7 IAM frameworks to build the SSP-RCP scenario space.

ModelGroupTypeWhere usedLinked note
GCAM (Global Change Analysis Model)PNNL + Joint Global Change Research Institutepartial-equilibrium energy-economic + land + waterAR6, US DOE policyipcc-scenarios-and-integrated-assessment
IMAGEPBL Netherlands Environmental Assessmentpartial-equilibrium energy-land-economyAR6, OECD workipcc-scenarios-and-integrated-assessment
MESSAGE-GLOBIOMIIASAlinear-programming energy + GLOBIOM landAR6, SDP scenariosipcc-scenarios-and-integrated-assessment
REMIND-MAgPIEPIKgeneral-equilibrium energy + MAgPIE landAR6, EU-Green-Deal-classipcc-scenarios-and-integrated-assessment
WITCHRFF-CMCC (Italy)game-theoretic optimal-growth + energy + climateclimate policy + game theoryclimate-finance-and-investment
AIM/CGENIES + Mizuho (Japan)computable-general-equilibriumAR6, Japan / Asia scenariosipcc-scenarios-and-integrated-assessment
POLESJRC (EU) + Enerdatapartial-equilibrium energyEU climate policy + JRCipcc-scenarios-and-integrated-assessment
DICE / RICE (Nordhaus)Yalecost-benefit growth modelbenchmark for SCC (Nordhaus, Nobel 2018)ipcc-scenarios-and-integrated-assessment
FUNDAnthoff + Tolenumerated impact-cost regionsSCC sensitivityipcc-scenarios-and-integrated-assessment
PAGEHope (Cambridge)probabilistic damage + SCCUK Stern review + sensitivityipcc-scenarios-and-integrated-assessment
GEM-E3EU consortiumEU-policy CGEEU 2050 long-term strategyipcc-scenarios-and-integrated-assessment

7. ML-based and neural Earth-system models (2022+)

The post-2022 wave of ML weather + climate models trained on ERA5 reanalysis + CMIP. Speed gain is 10⁴-10⁶× vs traditional NWP/ESMs; skill is now competitive with conventional ECMWF IFS at medium range for many variables. Climate-range projection skill (multi-decadal, response to forcing) is still an open question — most demonstrations stop short of multi-decadal forcing-response.

ModelGroupRangeResolutionArchitectureWhere it winsLinked note
Pangu-WeatherHuawei Cloud (2022)7 d forecast0.25°3D Earth-Specific Transformerfirst ML to beat IFS on RMSE for several variables (Bi et al. 2023 Nature)ai-and-machine-learning-for-climate
GraphCastGoogle DeepMind (2023)10 d forecast0.25°graph neural network on icosahedral meshbeats IFS on 90% of 1380 variable-lead-time pairs (Lam et al. 2023 Science)ai-and-machine-learning-for-climate
AIFS (AI Forecasting System)ECMWF (2024)10 d operational0.25°graph neuralfirst operational ML forecast at ECMWF, deployed Feb 2025ai-and-machine-learning-for-climate
AuroraMicrosoft (2024)5 d weather + air quality + waves + tropical cyclone0.1° (waves)foundation-model transformerfirst multi-domain ESM emulator (Bodnar et al. 2024)ai-and-machine-learning-for-climate
NeuralGCMGoogle (2024)hours-to-decade hybrid1.4° / 0.7° / 2.8°diff-physics + MLbeats CMIP6 ensemble on key climate diagnostics with physics constraint (Kochkov et al. 2024 Nature)ai-and-machine-learning-for-climate
ClimaXMicrosoft Research (2023)climate + weatherflexiblefoundation transformerdownstream-task fine-tuning baselineai-and-machine-learning-for-climate
FourCastNetNVIDIA + Caltech + LBNL (2022)7 d forecast0.25°adaptive Fourier neural operatorfirst ML to operate at IFS scaleai-and-machine-learning-for-climate
StormerNVIDIA + others (2024)medium range0.25°transformerweather forecast at IFS-classai-and-machine-learning-for-climate
NVIDIA Earth-2NVIDIAfull platformflexibleplatform for digital-twin climateNVIDIA Omniverse + Modulus + FourCastNet + CorrDiff downscalingai-and-machine-learning-for-climate
Ai2 Climate Emulator (ACE)Allen Institute for AIclimate-range emulatorC96 (~100 km)spherical harmonic neural operatortrained to emulate CMIP6 forced response (Watt-Meyer 2023)ai-and-machine-learning-for-climate
GenCastGoogle DeepMind (2024)probabilistic ensemble0.25°generative diffusionbeats IFS ensemble on probabilistic skill (Price et al. 2024 Nature)ai-and-machine-learning-for-climate
CorrDiff (NVIDIA)NVIDIAdownscalingkm-scaleconditional diffusionsuper-resolution from km-scale modelsregional-climate-and-downscaling
Prithvi-WxCNASA + IBM (2024)foundation model0.25°transformerNASA’s open foundation weather-climate modelai-and-machine-learning-for-climate

The key open question (still being resolved through 2025-2026) is whether these neural models capture climate-change response rather than just short-range weather. Hybrid (NeuralGCM, Ai2 ACE) and physics-constrained approaches look most promising for decade-to-century projection. Pure-data emulators trained only on ERA5 (a 1940–present record) cannot extrapolate to future forcings they have never seen.

8. Unified five-axis cross-class summary

A compact view of where each class lives. “Time horizon” is the longest a researcher would run it in production; “Cost” is rough wallclock for one historical+SSP scenario at standard resolution.

ClassResolutionCouplingTime horizonCost (compute)Calibration targetOpenness
CMIP6 ESM (CESM2, UKESM1, …)~1° / ~0.25°full ESM1850–2100 + perturbed5–10 Mcore-hrhistorical SST + atmospheremostly open (LANL, NCAR, MPI, GFDL); some national (HadGEM3)
Convection-permitting global (NICAM, SCREAM, ICON-Sapphire, X-SHiELD)2–5 kmatm only (mostly)months100 Mcore-hrDYAMOND benchmarkopen
Regional CRM (WRF)100 m–4 kmregional1 month–10 yr10–100k core-hrreanalysis lateral boundaryopen
Regional CORDEX (RegCM, REMO, COSMO-CLM)10–25 kmregional1990–2100100k–1M core-hrreanalysis lateral boundaryopen
EMIC (CLIMBER, Bern3D, GENIE, UVic)5° / boxatm + ocn + carbonup to 10⁸ yrhours-dayspaleoclimate recordsopen
Reduced-complexity (MAGICC, FaIR, OSCAR, Hector)global meanenergy-balance + carbonseconds-minutesdesktopCMIP6 ensemble + obsopen
Component (MOM6, NEMO, CLM5, CICE6, PISM, ISSM)variesone componentvaries0.1–5 Mcore-hrreanalysis or proxyopen
IAM (GCAM, IMAGE, MESSAGE, REMIND, WITCH)regional (32 regions)energy+land+economy + emulator2020–2100minutes-hours per scenariohistorical economic + emissionmostly open; some IIASA/PNNL specific
Neural ESM (Pangu, GraphCast, AIFS, GenCast, NeuralGCM, Aurora)0.25° / 0.1°atm (mostly)days for weather; emerging climate-range (Ai2 ACE)inference: secondsERA5 reanalysis + CMIP6open (most); proprietary weights some

9. Recent shifts (2022–2026)

  • Storm-resolving global modeling is now possible — DYAMOND-class runs (3 km global) on Frontier, Aurora, LUMI HPCs. SCREAM, ICON-Sapphire, X-SHiELD all crossed the threshold.
  • AI weather models beat IFS — Pangu-Weather (2022), GraphCast (2023), AIFS (2024 operational at ECMWF). Climate-range skill is the next frontier.
  • Neural ESM emulators — Ai2 ACE, NeuralGCM. Trade fidelity for orders-of-magnitude speed.
  • Probabilistic ML ensembles — GenCast (Google 2024) replaces 51-member IFS ensemble at lower cost.
  • CMIP7 in preparation — IPCC AR7 reports due 2027-2029; CMIP7 design finalized 2024-25, runs 2025-2027.
  • Digital twin Earth — Destination Earth (EU ECMWF + ESA, launched 2022, operational 2024), NVIDIA Earth-2 platform. Each combines storm-resolving simulation + ML downscaling + observational nudging.
  • AR6 emergent constraints tightened ECS — model-spread 1.5–6 °C; AR6 assessed range 2.0–5.0 °C, best 3.0 °C — see climate-sensitivity-and-feedbacks.
  • CDR / geoengineering — GeoMIP under CMIP6/7 increases SRM + CDR scenario coverage — solar-geoengineering-and-cdr.

Adjacent

When to pick what

What's the question?
├─ Multi-decade global climate change projection
│   ├─ AR6 reference / official scenario → CMIP6 ESM ensemble (already on ESGF; analyze in Pangeo)
│   ├─ Fast scenario sweep / emulation → MAGICC + FaIR (probabilistic)
│   └─ Need ECS uncertainty → CMIP6 ensemble + emergent constraints
├─ Deep-time paleoclimate (>10 kyr, e.g. glacial cycles, PETM, snowball Earth)
│   ├─ Glacial cycles → CLIMBER-X or LOVECLIM
│   ├─ PETM / Cretaceous / Permian → cGENIE
│   └─ Last deglaciation → LOVECLIM, CESM low-res, or transient TraCE-21k
├─ Regional climate adaptation
│   ├─ Standard CORDEX downscaling → RegCM, REMO, COSMO-CLM, WRF
│   ├─ Polar / ice sheet SMB → MAR or RACMO2
│   └─ Hyper-local hazard → WRF at 1 km or CorrDiff ML downscaling
├─ Storm-permitting global climate
│   └─ ICON-Sapphire / SCREAM / NICAM / X-SHiELD (DYAMOND class) — needs Frontier/Aurora/LUMI access
├─ Component-only science
│   ├─ Ocean (eddies, AMOC) → MOM6, NEMO, MITgcm, FESOM2
│   ├─ Atmosphere (radiation, chemistry) → CAM6, IFS, CAM-chem, GEOS-Chem, WACCM
│   ├─ Land (carbon, hydrology) → CLM5, JULES, ORCHIDEE, Noah-MP
│   ├─ Sea-ice → CICE6
│   └─ Ice-sheet → PISM (research), ISSM (inverse), BISICLES (shelves), Elmer/Ice (full Stokes), MALI/CISM (coupled)
├─ Carbon-cycle / flux attribution
│   └─ CarbonTracker / CTE / TM5 for inversions; CASA for satellite-driven NPP; bottom-up via JULES/ORCHIDEE
├─ Climate-economy / policy / SCC
│   ├─ Long-term IPCC scenario → MESSAGE, GCAM, REMIND, IMAGE, AIM
│   ├─ SCC sensitivity → DICE, FUND, PAGE
│   └─ EU policy CGE → GEM-E3, POLES
├─ Operational weather forecast (1-10 d)
│   ├─ Traditional → IFS, GFS, ICON, UKM, GEM
│   ├─ ML deterministic → AIFS, Pangu-Weather, GraphCast, FourCastNet
│   └─ ML probabilistic → GenCast, AIFS-ENS
├─ Sub-seasonal / seasonal (S2S)
│   ├─ Traditional → SEAS5 (ECMWF), CFSv2 (NOAA), GloSea (Met Office)
│   └─ ML emerging → CorrDiff, ClimaX-S2S
├─ Climate-range ML emulator
│   ├─ Hybrid (physics + ML, multi-decadal) → NeuralGCM
│   ├─ Pure data-driven climate response → Ai2 ACE
│   └─ Multi-domain foundation model → Aurora, Prithvi-WxC
├─ Extreme event attribution
│   ├─ Pseudo-global-warming → WRF-PGW
│   ├─ Counterfactual ensemble → CMIP6 LE + storyline approach (NCAR, KNMI, WWA)
│   └─ Operational → World Weather Attribution + ClimaMeter
├─ Solar geoengineering / CDR
│   └─ GeoMIP / G6solar within CMIP6/7 framework — see [[Sciences/ClimateScience/solar-geoengineering-and-cdr]]
└─ Earth observation + data assimilation
     → ECCO (ocean), ICESat-2 + CryoSat-2 + Operation IceBridge (cryo), Sentinel + Copernicus, GRACE-FO (mass), OCO-2/3 + MERLIN + GOSAT-3 (carbon)

The biggest practical lesson 2010–2026 is that model class matters more than model identity. The within-class CMIP6 spread (e.g. CESM2 vs MPI-ESM vs UKESM1) is smaller than the cross-class difference (CMIP6 ESM vs EMIC vs ML emulator). Pick the class that matches your question’s resolution, time horizon, and feedback needs — then within that class pick the model whose community + documentation + reanalysis-comparison best supports your study. Default to multi-model ensembles for any policy-relevant projection; default to single-model multi-realization ensembles (Large Ensembles like CESM2-LE, MPI-GE, CanESM5-LE) for internal-variability questions.