Skip to content

GCRI · Systems Science and Modelling

Optimisation under uncertainty

Plans for infrastructure, reserves and portfolios must work across futures that cannot be predicted. Optimisation under uncertainty chooses decisions that perform well on average, hold up in bad cases or stay acceptable across many scenarios. This page sets out the main approaches, from stochastic to robust optimisation and decision-making under deep uncertainty, how their results are tested, and what evidence serves the planners who rely on them.

01Register

Optimisation under uncertainty on the record.

02The question

Optimisation under uncertainty: how the risk develops.

Robust and adaptive decisions when futures are deeply uncertain. 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

Optimal for one future is fragile

A plan optimised for a single forecast can perform badly if the future differs. Representing uncertainty explicitly avoids this trap.

Mechanism

Stochastic methods use probabilities

Stochastic programming optimises expected performance across scenarios with assigned probabilities, which needs those probabilities to be credible.

Mechanism

Robust methods guard the worst case

Robust optimisation chooses decisions that perform acceptably for all scenarios within a set, trading average performance for protection.

Mechanism

Adaptive plans keep options open

Plans with decision points and triggers adapt as uncertainty resolves, avoiding commitments that lock in poor outcomes.

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
Imeonekana (mawingu na muda wa kupita kwa setilaiti huacha mapengo)
Chanzo
VIIRS NOAA-20 mita 375 (FIRMS) (NASA LANCE FIRMS)
Mtoaji
NASA LANCE FIRMS
Leseni
Data huria ya NASA
Tarehe ya data
kila siku; Ugunduzi wa setilaiti; mawingu na muda wa kupita kwa setilaiti huacha mapengo
Ya hivi karibuni wakati wa ujenzi
2026-10-13
Ubora wa picha
gridi ya takriban mita 500
Kutokuwa na uhakika
Ugunduzi wa setilaiti wa maeneo ya joto kwa siku iliyoonyeshwa, ikiwemo miale ya gesi na joto la viwandani; si moto uliothibitishwa, ripoti ya tukio au onyo.
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
  • NASA Fire Information for Resource Management System (FIRMS), VIIRS 375 m active fire product, NASA LANCE; vector tiles 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
  • Moto unaowaka na hitilafu za joto: VIIRS NOAA-20 375 m (FIRMS), NASA LANCE FIRMS · 13 Okt 2026

04Across the domains

Optimisation under uncertainty supports decisions in every domain.

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

Optimisation under uncertainty · 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 Optimisation under uncertainty.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
Scenarios evaluatedcountper analysisAnalysis documentationDECLARED
Regret of chosen plandifference from best plan in each scenarioper analysisRobustness analysisIMEUNDWA KIMUUNDO
Value of the stochastic solutiondifference between stochastic and deterministic plansper analysisOptimisation outputsIMEUNDWA KIMUUNDO
Satisficing sharepercent of scenarios meeting objectivesper analysisRobustness analysisIMEUNDWA KIMUUNDO
Adaptation triggers definedcountper planPlan documentationDECLARED
Out-of-sample performanceobjective value on held-out scenariosper analysisTestingIMEUNDWA KIMUUNDO
07Twin and cascade simulator

Model optimisation under uncertainty 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
Society for Decision Making under Deep Uncertainty resourcesMethods and case studies for robust and adaptive planningPublic, terms to check
Open-source optimisation tools such as Pyomo and JuMPModelling languages for stochastic and robust optimisationOpen source, BSD and MPL licences
EMA WorkbenchExploratory modelling and analysis toolkitOpen source, BSD licence
Operations research literature on stochastic programmingMethods and applicationsLicence per publication
Published adaptive pathway plans, such as delta programmesApplied adaptive planningPublic records, licence per government
09Methods and evidence

How optimisation under uncertainty 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 optimisation under uncertainty, 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
Infrastructure and resource plannersWhich plans hold up across plausible futuresRobustness analyses; pathway designs
Analysts across Nexus platformsWhich optimisation approach suits their questionMethod selection guides
Finance and audit bodiesWhether planning analyses considered uncertainty adequatelyAnalysis documentation

Outputs this topic produces

  • Method selection guides for planning under uncertainty
  • Robustness and regret analyses
  • Adaptive pathway designs with triggers
  • Teaching material on optimisation under uncertainty
11Core questions

What experts ask about decisions under uncertainty.

  • Is the aim to perform best on average, to avoid the worst outcomes, or to perform acceptably across many futures?
  • Which uncertainties can be described with probabilities, and which cannot?
  • Which decisions can be delayed, adapted or reversed as information arrives?
  • How do trade-offs between objectives change across scenarios?
  • What signals would indicate that a strategy needs to change?
12Mechanisms and drivers

Why optimal can be fragile.

Classical optimisation assumes a known model and known inputs. A strategy optimised for a single forecast can be finely tuned to that forecast and fragile if the future differs. In infrastructure planning, for example, a design optimised for one climate projection may fail under another.

Stochastic optimisation handles uncertainty described by probability distributions, seeking the best expected performance or managing risk measures. Robust optimisation seeks solutions that perform well across a set of possible conditions without probabilities.

Under deep uncertainty, when experts cannot agree on models or probabilities, methods such as robust decision making and adaptive pathways explore many futures, identify where strategies fail, and design plans that can change course when signposts are observed.

13Methods and data

Methods and their limits.

  1. 01

    Stochastic programming

    Optimising expected outcomes over scenarios, using optimisation and operations research. Requires credible probabilities.

  2. 02

    Robust optimisation

    Optimising worst-case or regret across uncertainty sets. Can be conservative.

  3. 03

    Exploratory modelling

    Running models across many futures to find vulnerabilities, using decision analysis. Computationally demanding.

  4. 04

    Adaptive pathways

    Sequencing actions with triggers for change. Requires monitoring and institutional capacity.

These methods support judgement, they do not replace it. The choice of objectives and acceptable risks is a value choice for decision-makers.

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 robust optimisation?
An approach that finds solutions that remain feasible and perform acceptably for all values of uncertain parameters within a defined set.
What is deep uncertainty?
A situation where parties cannot agree on models, probabilities or how to value outcomes, so planning seeks robust and adaptive strategies.
What is an adaptive pathway?
A sequence of actions with decision points that lets a plan change course as conditions evolve, keeping future options open.
Is a robust strategy always more expensive?
Often it costs more upfront but reduces the risk of costly failure. The trade-off is made explicit.
Does the platform choose strategies?
No. It develops methods. Choices remain with decision-makers.
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.