Skip to content

GCRI · Systems Science and Modelling

Model intercomparison

No single model captures a complex system fully. Model intercomparison runs several models on the same experiment with the same inputs, so that their agreement and disagreement can be seen and explained. It is how climate, crop, hydrology and epidemic modelling communities learned what their results can and cannot support. This page sets out how intercomparisons are designed, what they reveal, and what evidence serves the institutions that rely on model ensembles.

01Register

Model intercomparison on the record.

02The question

Model intercomparison: how the risk develops.

Comparing models of the same system, and what disagreement tells you. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.

Mechanism

Common protocols isolate differences

When models share inputs, scenarios and output formats, remaining differences point to model structure and parameters rather than set-up.

Mechanism

Spread shows structural uncertainty

The range of results across credible models is a measure of how much the answer depends on modelling choices.

Mechanism

Ensembles are not independent

Models share code, data and ideas, so an ensemble can agree for shared reasons and overstate confidence.

Mechanism

Evaluation against observations weighs models

Comparing models with observations shows which reproduce key processes, which can inform how ensemble results are interpreted.

03Geography

Hazards and exposure in North America.

Flood water, population density and daytime surface heat over North America, from NASA GIBS. Each layer carries its own date.

Ramani shirikishi inahitaji JavaScript. Orodha ya tabaka na vyanzo imetolewa hapa chini.

Tabaka 3

Hazards and exposure, for this page

Kuhusu tabaka hili
Ushahidi
Imeonekana (mawingu huacha mapengo)
Chanzo
Bidhaa ya Mafuriko ya Kimataifa ya VIIRS (NASA LANCE)
Mtoaji
NASA LANCE
Leseni
Data huria ya NASA
Tarehe ya data
kila siku; Ugunduzi wa setilaiti; mawingu huacha mapengo
Ya hivi karibuni wakati wa ujenzi
2026-10-14
Ubora wa picha
gridi ya takriban mita 250
Kutokuwa na uhakika
Ugunduzi wa setilaiti; mawingu huacha mapengo. Raster ya setilaiti au modeli kama mtoaji anavyoichapisha. Mawingu, mapengo ya mkondo wa setilaiti na makosa ya upimaji huacha mapengo na kelele; thamani ni viashiria, si vipimo mahali fulani.
Imefikiwa kupitia
kipeperushi cha data cha tovuti hii, kutoka gibs.earthdata.nasa.gov
Kuhusu tabaka hili
Ushahidi
Imekadiriwa
Chanzo
Gridded Population of the World v4.11 (NASA SEDAC (CIESIN, Chuo Kikuu cha Columbia))
Mtoaji
NASA SEDAC (CIESIN, Chuo Kikuu cha Columbia)
Leseni
CC BY 4.0
Tarehe ya data
null; Makadirio yanayotokana na sensa, isiyobadilika (2020)
Ubora wa picha
gridi ya takriban kilomita 1
Kutokuwa na uhakika
Makadirio yanayotokana na sensa ya watu kwa kila kilomita ya mraba mwaka 2020, yanayoonyeshwa kama muktadha wa mfiduo; si hesabu ya watu walioathiriwa na janga lolote.
Imefikiwa kupitia
kipeperushi cha data cha tovuti hii, kutoka gibs.earthdata.nasa.gov
Kuhusu tabaka hili
Ushahidi
Imeonekana (mawingu yanayoendelea bado huacha mapengo)
Chanzo
VIIRS NOAA-20 Ngazi ya 3, siku 8 (NASA LANCE)
Mtoaji
NASA LANCE
Leseni
Data huria ya NASA
Tarehe ya data
siku 8; Mchanganyiko wa setilaiti wa siku 8; mawingu yanayoendelea bado huacha mapengo
Ya hivi karibuni wakati wa ujenzi
2026-09-14
Ubora wa picha
gridi ya takriban kilomita 1
Kutokuwa na uhakika
Joto la uso wa ardhi kutoka setilaiti kwa siku 8 zinazoishia tarehe iliyoonyeshwa, kiashiria cha mfiduo wa joto; si joto la hewa, faharasa ya joto au onyo la joto.
Imefikiwa kupitia
kipeperushi cha data cha tovuti hii, kutoka gibs.earthdata.nasa.gov
Kuhusu tabaka hili
Ushahidi
Imeonekana
Chanzo
Mwangaza uliorekebishwa wa VIIRS NOAA-20 (NASA LANCE)
Mtoaji
NASA LANCE
Leseni
Data huria ya NASA
Tarehe ya data
kila siku; Picha ya satelaiti
Ya hivi karibuni wakati wa ujenzi
2026-10-14
Ubora wa picha
gridi ya takriban mita 250
Kutokuwa na uhakika
Picha ya satelaiti. Raster ya satelaiti au ya muundo kama mtoa huduma anavyoichapisha. Mawingu, mapengo ya ukanda na makosa ya upatikanaji huacha mapengo na kelele; thamani ni viashiria, si vipimo mahali fulani.
Imefikiwa kupitia
kipeperushi cha data cha tovuti hii, kutoka gibs.earthdata.nasa.gov
Kuhusu tabaka hili
Ushahidi
Modelled (latest year GIBS publishes)
Chanzo
MODIS Terra and Aqua land cover (MCD12Q1) (NASA LP DAAC)
Mtoaji
NASA LP DAAC
Leseni
Data huria ya NASA
Tarehe ya data
yearly; Satellite classification; latest year GIBS publishes
Ya hivi karibuni wakati wa ujenzi
2024-01-01
Ubora wa picha
gridi ya takriban mita 500
Kutokuwa na uhakika
Satellite land cover classes for the year shown, including croplands; a classification with known errors, not a farm register.
Imefikiwa kupitia
kipeperushi cha data cha tovuti hii, kutoka gibs.earthdata.nasa.gov
Kuhusu tabaka hili
Ushahidi
Imekadiriwa
Chanzo
GPM IMERG (NASA GES DISC)
Mtoaji
NASA GES DISC
Leseni
Data huria ya NASA
Tarehe ya data
kila siku; Makadirio ya satelaiti, karibu na wakati halisi
Ya hivi karibuni wakati wa ujenzi
2026-10-12
Ubora wa picha
gridi ya takriban kilomita 2
Kutokuwa na uhakika
Makadirio ya satelaiti, karibu na wakati halisi. Raster ya satelaiti au ya muundo kama mtoa huduma anavyoichapisha. Mawingu, mapengo ya ukanda na makosa ya upatikanaji huacha mapengo na kelele; thamani ni viashiria, si vipimo mahali fulani.
Imefikiwa kupitia
kipeperushi cha data cha tovuti hii, kutoka gibs.earthdata.nasa.gov

Ya karibuni zaidi inayopatikana

NASA GIBS / EOSDIS · ASTER GDEM (METI, NASA) · Blue Marble (NASA) · © wachangiaji wa OpenStreetMap (ODbL) · Natural Earth

Inapakia picha za hivi karibuni za NASA.

Buruta ili kusogeza. Ctrl au ⌘ na usogeze, bana au tumia vitufe ili kukuza; elea juu ili kuona thamani zilizo chini ya kiashiria; weka upya hurudisha kwenye eneo.

Hazards and exposure, framed on North America.Satellite and model products as NASA publishes them, each with its own date. They are indicators, not assessments of any place.

Data: NASA GIBS · ASTER GDEM · Blue Marble · © OpenStreetMap contributors · Natural Earth · NASA EONET · U.S. Geological Survey · NASA JPL OPERA DIST-ALERTSources and licences

  • We acknowledge the use of imagery provided by services from NASA’s Global Imagery Browse Services (GIBS), part of NASA’s Earth Science Data and Information System (ESDIS). Colours redrawn from the published colour map; values averaged into hexagonal cells where shown (modified). NASA open data
  • We acknowledge the use of imagery provided by services from NASA’s Global Imagery Browse Services (GIBS), part of NASA’s Earth Science Data and Information System (ESDIS). Colours redrawn from the published colour map; values averaged into hexagonal cells where shown (modified). NASA open data
  • VIIRS near-real-time products (flood, land surface temperature, snow cover, corrected reflectance), NASA LANCE; imagery via NASA GIBS. Colours redrawn from the published colour map; values averaged into hexagonal cells where shown (modified). NASA open data
  • Center for International Earth Science Information Network (CIESIN), Columbia University. 2018. Gridded Population of the World, Version 4 (GPWv4): Population Density, Revision 11. NASA Socioeconomic Data and Applications Center (SEDAC). Licensed under CC BY 4.0. Colours redrawn from the published colour map; values averaged into hexagonal cells where shown (modified). doi:10.7927/H49C6VHW. CC BY 4.0
  • MODIS land products (vegetation indices, gross primary production, land cover), NASA LP DAAC and LANCE; imagery via NASA GIBS. Colours redrawn from the published colour map; values averaged into hexagonal cells where shown (modified). NASA open data
  • GPM IMERG precipitation, NASA Precipitation Processing System; distributed by the NASA Goddard Earth Sciences Data and Information Services Center (GES DISC); imagery via NASA GIBS. Colours redrawn from the published colour map; values averaged into hexagonal cells where shown (modified). NASA open data
  • ASTER GDEM is a product of METI and NASA (ASTER Global Digital Elevation Model V003, NASA LP DAAC); imagery via NASA GIBS. Colours redrawn from the published colour map; values averaged into hexagonal cells where shown (modified). Cite as: NASA/METI/AIST/Japan Spacesystems and U.S./Japan ASTER Science Team (2019). ASTER Global Digital Elevation Model V003. doi:10.5067/ASTER/ASTGTM.003. NASA open data
  • Imetengenezwa kwa Natural Earth. Data huria za ramani za vekta na rasta @ naturalearthdata.com. Kikoa cha umma
  • © OpenStreetMap contributors (ODbL), via NASA GIBS (coastlines, boundaries, roads and place names). Colours redrawn from the published colour map; values averaged into hexagonal cells where shown (modified). ODbL
  • Blue Marble, NASA Earth Observatory; imagery via NASA GIBS. Colours redrawn from the published colour map; values averaged into hexagonal cells where shown (modified). NASA open data
  • Muktadha kwenye tufe na ramani bapa: mwambao, mipaka, barabara na majina ya mahali kutoka tabaka za marejeo za NASA GIBS (data ya OpenStreetMap, ODbL, © wachangiaji wa OpenStreetMap); picha za Blue Marble kama msingi wa hiari; ardhi ya 3D na usomaji wa mwinuko kutoka faharasa ya rangi ya mwinuko ya ASTER GDEM (madaraja ya 5 m hadi 200 m, yaliyokuzwa ili yasomeke). Mipaka imechorwa kama ilivyochapishwa na haimaanishi msimamo wowote kuhusu eneo lolote.
  • Matukio ya moja kwa moja: NASA Earth Observatory Natural Event Tracker (EONET); matukio ni ya kuashiria tu na yanabaki na viungo vya vyanzo vyao asili.
  • Moja kwa moja: NASA EONET. Matukio ni ya kuashiria tu, si tahadhari rasmi za mamlaka.
  • Moja kwa moja: U.S. Geological Survey. Matukio ni ya kuashiria tu, si tahadhari rasmi za mamlaka.
  • Observed (EO): NASA JPL OPERA DIST-ALERT (contains modified Copernicus Sentinel data 2026). We acknowledge the use of imagery provided by services from NASA’s Global Imagery Browse Services (GIBS), part of NASA’s Earth Science Data and Information System (ESDIS).
  • NASA GIBS: Tunatambua matumizi ya picha zinazotolewa na huduma za Global Imagery Browse Services (GIBS) za NASA, sehemu ya Earth Science Data and Information System (ESDIS) ya NASA.
  • Utabiri wa mtiririko wa mito, siku 30 pamoja na masafa (GloFAS kupitia Open-Meteo): Una taarifa za Copernicus Emergency Management Service zilizorekebishwa za 2026. Si onyo la mafuriko.
  • Data ya hali ya hewa kutoka Open-Meteo.com (https://open-meteo.com/), chini ya leseni ya CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).
  • Optional layers (11 sources in the layer list: live, forecast, observed and context) print their provider credit on the map and here when switched on; forecasts are model output and derived indices are estimates.

Tabaka na tarehe

  • Maji ya mafuriko, mchanganyiko wa siku 3: VIIRS Global Flood Product, NASA LANCE · 14 Okt 2026
  • Population density, 2020: Population: CIESIN, Columbia University, Gridded Population of the World v4.11 (2018), NASA SEDAC, CC BY 4.0 · static
  • Land surface temperature, day, 8-day composite: VIIRS NOAA-20 Level 3, 8-day, NASA LANCE · 14 Sep 2026
  • Rangi halisi: VIIRS NOAA-20 mng'ao uliosahihishwa, NASA LANCE · 13 Okt 2026
  • Land cover type (IGBP), yearly: MODIS Terra and Aqua land cover (MCD12Q1), NASA LP DAAC · 2024
  • Kiwango cha mvua, wastani wa kila siku: GPM IMERG, NASA GES DISC · 12 Okt 2026

04Across the domains

Intercomparison is how each domain learns where its models disagree.

05Monitor

Coupled model runner.

The screen a team would use to follow this subject: its measures, states and decisions. Names are examples and values are illustrative.

Coupled model runner

Model intercomparison · Coupled model A · Scenario set 3

Ensemble runningGCRI · Systems Science and Modelling

Model coupling graph

nodes: models · edges: exchanged variables

Coupled components

Hidrolojia1 dayrunoffYes
Power system1 hdemand, outagesYes
Uchumi1 quarteroutput, pricesPartial
Idadi ya watuMwaka 1exposureYes

Ensemble output

Coupled model runner: the screen’s structure, terms and states for Model intercomparison.Example names and illustrative values; a design, not a record or a live system.
06Viashiria

What is measured, and how well.

Each indicator carries its unit, its cadence, the kind of source behind it and the evidence label a value would carry. No values are shown: a programme supplies them under its own authority.

IndicatorUnitCadenceSource typeLabel
Models participating in an intercomparisoncountper roundIntercomparison recordsIMERIPOTIWA
Ensemble spreadrange or standard deviation across modelsper experimentIntercomparison outputsIMEUNDWA KIMUUNDO
Agreement on sign of changepercent of modelsper experimentIntercomparison outputsIMEUNDWA KIMUUNDO
Skill against observationserror metric by modelper evaluationModel evaluationIMEKADIRIWA
Shared code lineage between modelscount of related model familiesper assessmentModel genealogy studiesIMEKADIRIWA
Outputs published in standard formatspercent of outputsper roundArchive metadataIMEONEKANA
07Twin and cascade simulator

Treat model intercomparison as a system. Run the cascade.

Assembled from the Nexus Standards area profile Compound and cascading hazard standards: the hazards it names, the systems and sectors that carry the loss, the controls that dampen it and the data that observe it. Choose the components, pick an initiating hazard and step through the cascade. The Map view places the run on an example geography with open data. Every coupling weight is a stated assumption you can change. A teaching model, not a forecast, an assessment or a certification. To model a place of your own choosing, open the full System Builder.

Majanga

Controls

  • NS-0 Katiba na kanuni
  • NS-3 UNI za rekodi za hatari (sarufi, ulinganishi wa urithi, mamlaka ya ugawaji)
  • NS-9 UNI za kimuktadha (utekelezaji, matukio, matukio ya programu)
  • NS-11 Uthibitishaji na ufuatiliaji (ushahidi, uthibitisho, uchunguzi wa ulinganifu usio na imani)
  • NRS-H61 Familia ya majanga: Kushindwa kwa mfululizo (NHS-1 sys.cascade)
  • NRS-H62 Familia ya majanga: Matukio mchanganyiko na ya majanga mengi (NHS-1 sys.compound)
  • NRS-H60 Familia ya majanga: Kukatizwa kwa mnyororo wa ugavi na mitiririko (NHS-1 sys.supply-chain)
  • NRS-H94 Familia ya majanga: Matukio ya kimtandao ya kimfumo (NHS-1 cyber.systemic-cyber)
  • NRS-H107 Familia ya majanga: Mkusanyiko na ufanano wa miundombinu ya kidijitali (NHS-1 tech.digital-concentration)
  • NRS-H7 Familia ya majanga: Hitilafu za mzunguko mkubwa wa angahewa (NHS-1 atmo.climate-mode)

Observation

  • Copernicus EMS rapid mapping activations (listed for reference)
  • OpenStreetMap planet and replication (listed for reference)
  • ERA5 reanalysis (Copernicus Climate Data Store) (listed for reference)
  • Direct economic loss attributed to disasters in relation to global GDP
  • Number of disruptions to basic services attributed to disasters

Example geography

  • Bangladesh: 134 hexagonal cells
  • Night lights tonight, outages (VIIRS Day/Night Band), NASA GIBS
  • Carbon monoxide, smoke transport (AIRS), NASA GIBS
  • Vegetation disturbance alerts, 30 m (OPERA DIST-ALERT), NASA GIBS
08Vyanzo

Where the evidence comes from.

Named public datasets and institutional records, each with its licence class. Sources by agreement are used only under their terms and never republished.

ChanzoWhat it givesLicence class
CMIP6 archive via the Earth System Grid FederationClimate model outputs from common experimentsOpen, CC BY 4.0 for most models (to check per model)
ISIMIP repositoryImpact model outputs for water, agriculture, health and other sectorsOpen, licence per dataset (to check)
AgMIP crop model intercomparisonsCrop model ensemble protocols and resultsPublic, terms per project
Scenario Modeling Hub publications for epidemicsMulti-model epidemic scenario projectionsPublic, terms per hub
Literature on model genealogy and ensemble independenceAnalysis of shared model componentsLicence per publication
09Methods and evidence

How model intercomparison is modelled, measured and tested.

The model, its assumptions, its data and its uncertainty, in the form a scientist or engineer would check. Structures follow published methods. The values drawn are illustrative, not findings.

Model, equations and assumptions

The model structure used for model intercomparison, the assumptions it makes and how each is tested.

Risk of element iRi = f(Hi, Ei, Vi, Ci)hazard, exposure, vulnerability and coping capacity
Expected annual lossEAL = ∑k pk · Lkover events k with annual probability pk
Coupled systemsxt+1 = A xt + B ut + εtA carries the couplings, u the shock, ε the error
AssumptionWhy it is madeHow it is tested
Hazard and exposure are independent within an eventKeeps the loss integral tractableCompare with a copula fit on joint records
Couplings are linear near the operating pointLets shocks propagate through APerturb beyond the linear range and compare
Vulnerability curves transfer between placesLocal curves are rarely availableValidate against observed losses where they exist
Inputs
versioned datasets with provenance
Matokeo
distributions, never single numbers
Code
released with the model card and tests
Reproduction
same data, code, environment and seed
Standard risk and state-space formulations; the assumptions listed are typical of the method, stated for review, not results.
FAIR guiding principles · GO FAIR and FORCE11W3C PROV · World Wide Web ConsortiumGuidance on consistent treatment of uncertainties · IPCCCEOS Analysis Ready Data · Committee on Earth Observation Satellites
10Decisions

Who decides, and what the evidence serves.

Systems Science and Modelling prepares evidence. The decisions below are taken by the authorities named, under their own rules, and never by the platform.

Nani anaamuaThe questionEvidence this page serves
Modelling groupsWhich experiments and protocols to adoptProtocol designs; evaluation reports
Institutions using model resultsHow much confidence to place in ensemble findingsEnsemble summaries with spread
Research fundersWhether to fund intercomparison infrastructureEvidence on value of intercomparison

Outputs this topic produces

  • Intercomparison protocol designs for cross-sector models
  • Ensemble summaries with spread and agreement stated
  • Model evaluation reports against observations
  • Method notes in Reports Methods and Technical
11Core questions

What experts ask about intercomparison.

  • Do independently built models agree, and on which outputs?
  • Where they disagree, is it because of structure, parameters, inputs or errors?
  • Is the spread of an ensemble a meaningful measure of uncertainty?
  • Are the models truly independent, or do they share code, data and assumptions?
  • How should results be weighted when some models perform better against observations?
12Mechanisms and drivers

What disagreement tells you.

Every model embodies choices: which processes to include, how to represent them and how to calibrate them. Two well-built models can therefore produce different answers. When models agree, confidence increases, though not to certainty, since they may share the same flaw. When they disagree, the disagreement is information: it reveals structural uncertainty that no single model can show.

Interpreting an ensemble is not straightforward. Models are rarely independent; many share components, data or lineage. An ensemble’s spread may underestimate true uncertainty if the models share errors, or overstate it if poor models are included. Methods for weighting models by performance and independence are an active research area.

Intercomparison also improves models. Diagnosing why one model differs often reveals errors or missing processes, and shared protocols raise standards across a community.

13Methods and data

Methods and their limits.

  1. 01

    Common protocols

    Shared inputs, scenarios, output variables and formats. Designing them takes time and agreement.

  2. 02

    Ensemble statistics

    Means, spreads and distributions across models, using probability and statistics. Spread is not a probability unless justified.

  3. 03

    Performance and independence weighting

    Weighting models by skill and distinctness, using uncertainty quantification. Weights depend on chosen metrics.

  4. 04

    Process diagnostics

    Comparing internal variables to find sources of difference. Requires access to model internals.

Agreement among models is not proof of correctness. Intercomparison results state which models took part and how they relate.

14Mbinu

How it is studied, and what the methods cannot say.

15Where this page’s references

Where this page’s references lead.

The bodies and documents this page is read against. Status and editions change; each link goes to the issuing body.

16Maswali

Questions and answers.

What is a model intercomparison project?
A coordinated effort in which several modelling groups run the same experiments with shared inputs and output formats, so that their results can be compared systematically.
Does agreement between models mean they are right?
Not necessarily. Models can share errors through common components and assumptions. Agreement increases confidence only when models are evaluated against observations.
What is structural uncertainty?
Uncertainty that comes from the way a model is built, such as which processes it includes and how, which shows up as differences between models.
Does intercomparison find the best model?
It can show which models perform better on chosen tests, but different models may be better for different purposes.
Why not just average the models?
Because models are not independent and not equally good. Simple averages can mislead without considering both.
Can private models take part?
Intercomparisons can be designed to include them, with rules on what is disclosed. Each protocol states its rules.
Is the Universe exercise an intercomparison?
The designed cross-system integration exercise uses intercomparison methods, described on the Universe page.
RipotiChangia

Changia ushahidi kutoka ukurasa huu.

Reports ni chapisho la kila robo mwaka la Muungano. Unachojua kuhusu somo hili kinaweza kuwa ushahidi, kisa au sahihisho katika toleo lijalo, kikitajwa kwa jina lako na kuchapishwa chini ya CC BY 4.0 pale ambapo ni kazi yako mwenyewe.

GCRISayansi ya Mifumo na Uundaji ModeliMbinu na KiufundiHali Tarajiwa na Jaribio la Shinikizo
Sayansi ya Mifumo na Uundaji Modeli

Bring a method, a dataset or a question.

Information and enquiries are open today. An enquiry receives a reference and names the next step; applications, payments, compute allocations and registrations open only when their team, terms and review capacity exist.