Skip to content

GCRI · Nexus Foundry · Foundations

Risk computing, geospatial systems and decision infrastructure: the foundations

Risk computing and decision infrastructure, with the geospatial systems that link them, is the discipline that turns data about hazards, places and people into tested models, maps and warnings. This foundational guide explains data, Earth observation, compute, models, early warning, lineage, AI and cyber risk, and the instruments that govern them.

As of
16 November 2026
वाचन
19 minutes
पुरावा
26 sources, cited in order
पद्धत
A synthesis of public frameworks, standards, lists and status reports, led by the World Development Report 2021, the UN Integrated Geospatial Information Framework, OGC standards, the TOP500 and Green500 lists, ECMWF's operational AI forecasting release, the WMO and UNDRR global status of early warning systems and the NIST, OECD and EU frameworks for cyber and AI risk.
डेटा
Every value in the article comes from a published source listed in the references. Computing figures come from the November 2026 TOP500 and Green500 lists; connectivity from ITU; early warning coverage from WMO and UNDRR; occurrence counts from the public GBIF API, read on 16 November 2026.
NEXUS TWIN · WF-1 · GLOBESEED fe3528b9G09.PILLARxyzSCHEMATIC · GLOBEST-01 · 0.672 0.623
WF-1 · globe Schematic wireframe twin (globe, climate system). An illustration, not data.

Methods and evidence

प्रश्न
What does the discipline of risk computing, geospatial systems and decision infrastructure cover, how is its quality tested, and which instruments govern it?
Approach
A synthesis of public frameworks, standards, lists and status reports, led by the World Development Report 2021, the UN Integrated Geospatial Information Framework, OGC standards, the TOP500 and Green500 lists, ECMWF's operational AI forecasting release, the WMO and UNDRR global status of early warning systems and the NIST, OECD and EU frameworks for cyber and AI risk.
पुरावा
26 sources, cited in order and listed under References
Data availability
Every value in the article comes from a published source listed in the references. Computing figures come from the November 2026 TOP500 and Green500 lists; connectivity from ITU; early warning coverage from WMO and UNDRR; occurrence counts from the public GBIF API, read on 16 November 2026.
मर्यादा
Benchmarks measure narrow tasks and coverage indicators measure presence, not quality. Evidence on whether decision systems change outcomes is thinner than evidence on their capacity. The article is not a product description, a security assessment or advice for any decision.

01Definition

What risk computing, geospatial systems and decision infrastructure mean

The discipline that builds the data, models, maps and services behind risk decisions, and tests whether they can be trusted.

Risk computing and decision infrastructure, together with the geospatial systems that join them, is the discipline that designs, builds and tests the technical systems that turn observations into decisions about risk. It covers the data pipelines that collect and clean evidence, the geospatial systems that place it on a map, the models and simulations that project what may happen, and the services that deliver results to the people who must decide.

Each part answers a different question. Risk computing asks how to calculate the chance and consequence of harm at the scale and speed a decision needs, from a catastrophe model to an epidemic simulation. Geospatial systems ask where: where hazards occur, where people and assets are, and how the two overlap. Decision infrastructure asks how results reach a decision-maker in time, in a form they can understand, with a clear record of how they were produced.

The discipline draws on computer science, statistics, geography, Earth observation, systems engineering and software safety. Its products include early warning systems, digital maps of exposure, scenario models, dashboards and data registries. Its standards concern quality, interoperability, security and accountability rather than any single tool.

This article is the hub of the Nexus Foundry platform, the technical work environment of The Global Centre for Risk and Innovation (GCRI), and links to a foundational article on each of its eight topics. It explains the discipline and the public evidence on it; it does not describe any client, deployment or contract.

02Stakes

Why decision infrastructure matters now

More data, more computing and more automated decisions raise the value of getting the infrastructure right.

Data are abundant, but their value is often lost. The World Bank's World Development Report 2021 found that much of the value of data remains untapped, and that data collected for one purpose can create value far beyond it 9. The barriers it names are misaligned incentives, incompatible data systems and a lack of trust 9. Each is a problem of infrastructure, not of data volume.

Access is unequal. The International Telecommunication Union (ITU) reports that almost three-quarters of the world's population are online, yet 2.2 billion people remain offline, most of them in low and middle income countries 2. A warning that cannot reach a phone, or a model that cannot run where it is needed, protects no one.

Computing is changing what can be known in time. Machine-learning weather models now run operationally, and the European Centre for Medium-Range Weather Forecasts reports that its AI system cut the energy used per forecast by about 1,000 times 4. Faster, cheaper forecasts make more frequent and more local risk estimates possible.

Automation raises the stakes of error. When models feed directly into alerts, prices or resource allocation, a flaw in data or code can spread quickly. Governments have responded with new rules: the European Union's Artificial Intelligence Act entered into force on 1 August 2024 8, and the UN General Assembly adopted the Global Digital Compact on 22 September 2024, as an annex to the Pact for the Future, with the goal of an inclusive, open, sustainable, fair, safe and secure digital future for all 10.

03डेटा

Data: the raw material of risk decisions

Official statistics, open science data and live feeds, and the principles that make them reusable.

Risk decisions draw on three kinds of data. Official statistics describe population, economy and society. Scientific and environmental data describe hazards, ecosystems and the physical world. Operational data describe what is happening now: river levels, earthquakes, power outages and hospital beds.

Official statistics rest on agreed rules. The Fundamental Principles of Official Statistics, which the UN General Assembly backed by resolution in 2014, begin with relevance, impartiality and equal access 11. These principles matter for risk work because a model built on biased or inaccessible statistics inherits their flaws.

Science data follow the FAIR principles: data should be findable, accessible, interoperable and reusable 12. Open archives show the scale this allows. The Global Biodiversity Information Facility reported a count of 3,905,124,889 species occurrence records when its API was read on 17 November 2026 1. Copernicus, the Earth Observation component of the European Union's space programme 13, makes most of its data available on a free, full and open basis 14.

Live data arrive as feeds. The U.S. Geological Survey publishes real-time earthquake feeds for the past hour, day, week and month, each updated every minute in a documented format 15. Feeds like these let warning systems and dashboards update without manual work, but they also require checks for gaps, duplicates and revisions.

The discipline treats every dataset as having a provenance, a licence and a known quality. The topic on evidence, ontology and lineage explains how these are recorded.

04Geospatial

Earth observation and the geospatial stack

Satellites, coordination bodies, national frameworks, open standards and grids.

Satellites give risk work its global view. The Group on Earth Observations (GEO), established in 2005, brings together more than 100 national governments and over 150 participating organisations 16.

In a disaster, satellite data are shared quickly. Under the International Charter Space and Major Disasters, registered national authorities can request satellite data through a single access point at any hour and at no cost, and may receive analysed maps within hours or days 17.

Countries organise their own geospatial information through national frameworks. The UN Integrated Geospatial Information Framework sets eight goals and nine strategic pathways, from governance and policy to data, innovation and capacity 18. Its premise is that geospatial information is national infrastructure, like roads or power.

Open standards let systems exchange maps and data. The Open Geospatial Consortium (OGC) maintains more than 80 open geospatial specifications 19, and its newer OGC APIs are designed so that anyone can publish and use geospatial data on the web 20.

Analysis also needs a common grid. A hierarchical hexagonal grid divides the whole world into cells at 16 levels, from 122 cells at the coarsest level to 569,707,381,193,162 at the finest 21. Every level has exactly 12 pentagons among its cells 21. The system provides exact logical containment but only approximate geometric containment across levels 22. That limit matters: a value summed over coarse cells is not exactly the sum over the fine cells it represents. The topic on maps and GIS covers how approved public map views are made.

05Compute

Compute: from supercomputers to the edge

How computing capacity, energy efficiency and connectivity shape what risk models can do.

Large risk models need large computers. On the November 2026 TOP500 list, the fastest system reached 2.198 exaflops on the standard benchmark, and five systems have now passed the exascale mark 3. Weather, climate, seismic and catastrophe models are among the main users of such machines.

Energy is a growing constraint. The Green500 list ranks the same systems by performance per watt, and the most efficient reached 73.28 gigaflops per watt in November 2026 23. Efficiency decides how much modelling a public budget can buy and how much power a data centre draws from a grid that may itself be at risk.

Most risk work does not need a supercomputer. It runs in the cloud, on local servers and increasingly at the edge, on devices close to sensors and users. Edge computing keeps services running when networks fail, which is often exactly when risk is highest. It also keeps sensitive data close to the people it concerns.

Connectivity sets the last limit. With 2.2 billion people offline 2, decision infrastructure must work over narrow networks, in local languages and on simple devices. The topic on compute and edge covers how cloud, local nodes and edge capacity are combined, and the Data and Compute platform covers compute as public-interest infrastructure.

06Models

Models, simulation and machine learning

Physics, statistics and learning, and the tests each must pass.

Risk models come in three broad families. Physical models simulate the processes that drive a hazard, such as the atmosphere, a river or a fault. Statistical models estimate probabilities from historical records, such as the chance of a flood of a given size. Machine-learning models learn patterns from large datasets and can be very fast once trained.

The families now work together. In February 2025 the European Centre for Medium-Range Weather Forecasts took its Artificial Intelligence Forecasting System into operations beside its physics-based system 4. It reports gains of up to 20 per cent on measures such as tropical cyclone tracks, with a grid spacing of 28 kilometres 4. The physics-based model remains in use, and the two are compared every day.

Simulation lets analysts test what has not happened yet. Agent-based models, network models and Monte Carlo methods explore how failures spread through infrastructure, markets or populations. The value of such models depends on stating their assumptions, testing them against observed events and reporting their uncertainty.

Every model has a domain where it is valid. A model trained on past floods may fail on a flood larger than any in its data; a model calibrated in one country may not transfer to another. The topic on modelling and simulation explains reproducibility and limits, and the Systems Science and Modelling platform covers the underlying methods.

07Early warning

Early warning: decision infrastructure in action

The clearest test of the discipline is whether a warning reaches people in time.

An early warning system links four steps: knowing the risk, monitoring and forecasting the hazard, disseminating warnings, and preparing people to act. Each step depends on data, models, networks and institutions working together. A failure anywhere breaks the chain.

Coverage is growing. The Global Status of Multi-Hazard Early Warning Systems 2025 reports that 119 countries, 60 per cent of all countries, now report a multi-hazard early warning system, a 113 per cent increase over 10 years 5. Comprehensiveness scores rose by an average of 45 per cent across regions 5.

Gaps remain where risk is high. Only 43 per cent of small island developing States report having such systems 5. The UN's Early Warnings for All initiative aims to protect everyone on Earth with early warning systems by the end of 2027 24.

Early warning shows the discipline's core lesson: a better forecast is worth little without the infrastructure that carries it to people and the institutions that act on it. The Climate and Hazards platform covers the hazard science behind such warnings.

08पुरावा

How decision infrastructure is tested and measured

Benchmarks, verification, availability and the evidence that a system works.

A decision system is judged on several kinds of evidence. Accuracy asks whether outputs match what happened, through forecast verification and back-testing. Timeliness asks whether results arrive when a decision is due. Availability asks whether the service runs when needed. Reach asks whether the intended users receive and understand the output.

Benchmarks make computing comparable. The TOP500 and Green500 lists use the same standard benchmark to rank speed and energy efficiency 323. Benchmarks measure a narrow task, so they show capacity rather than the value of any particular model.

Forecast verification is routine in weather services. ECMWF reports the performance of its machine-learning system against its physics-based model on stated measures 4. The same discipline applies to any risk model: name the measure, the period, the comparison and the uncertainty.

Coverage indicators track whether infrastructure exists at all. The early warning survey counts countries reporting multi-hazard systems and scores their comprehensiveness 5, and ITU tracks who is online 2. Such indicators show presence, not quality, and need field evidence on whether warnings led to action.

What it measuresIndicatorLatest valuePublished by
Peak computingFastest system on the TOP500 list2.198 exaflops, November 2026 3TOP500
Energy efficiencyMost efficient system on the Green500 list73.28 gigaflops per watt, November 2026 23TOP500
ConnectivityPeople offline2.2 billion, 2025 2ITU
Early warning coverageCountries reporting a multi-hazard early warning system119 countries, 60 per cent 5WMO and UNDRR
Open biodiversity dataOccurrence records in GBIF3,905,124,889, read 17 November 2026 1GBIF
Earth observation coordinationMember governments of GEOMore than 100 16GEO
Table 1. Selected indicators of computing, data and decision infrastructure, the latest value found and the body that publishes it, as of 16 November 2026.

09वंशावळ

Reproducibility, lineage and provenance

Why every result must be traceable to its data, code and assumptions.

A risk estimate that cannot be reproduced cannot be checked. Reproducibility means that another analyst, given the same data, code and settings, gets the same result. It requires versioned data, versioned code, recorded parameters and a record of the computing environment.

Lineage records the path from source to result: which dataset, which version, which transformation, which model run. When a source is corrected, lineage shows every result that must be updated. When a result is challenged, lineage shows where to look.

Provenance and licence travel with the data. The FAIR principles ask that data carry rich metadata and clear conditions of reuse 12, and the Fundamental Principles of Official Statistics ask that statistical methods and sources be made public 11. Without these records a map or a model output cannot be cited, audited or corrected.

The practical test is simple: could a reviewer, a year later, rebuild the figure that informed a decision? The topic on project workstations covers persistent project context and permissions, and the STEM page on data engineering and stewardship covers the methods.

10Infrastructure risk

When the infrastructure itself is the risk

Cyber attack, model failure, concentration and dependence.

Decision infrastructure can fail, and its failure can cause harm. A ransomware attack can take a warning service offline; a corrupted feed can trigger false alerts; a model error can misdirect relief. Systems built to manage risk must therefore manage their own.

Cyber security is the first line. The NIST Cybersecurity Framework 2.0, published in February 2024, organises cyber risk management into six functions: Govern, Identify, Protect, Detect, Respond and Recover 7. The addition of Govern places responsibility with leadership, not only with technical teams.

AI adds new failure modes: biased training data, drift as conditions change, opaque reasoning and misuse. The NIST AI Risk Management Framework organises AI risk management into four functions, to govern, map, measure and manage AI risks 6. In April 2026 NIST released a concept note for a profile on trustworthy AI in critical infrastructure 25.

Concentration is a quieter risk. When many institutions rely on the same cloud provider, dataset or model, a single failure can affect them all at once. Diversity of suppliers, local fallbacks and tested recovery plans reduce that exposure. The Cybersecurity and Digital Resilience platform covers these threats in depth.

11Automation

Agents and automation: keeping people accountable

Automated tools can speed analysis, but authority must stay with people.

Software agents and AI assistants can now search data, write code, run models and draft reports. They can speed routine work and widen access to technical tools. They also raise the question of who is responsible when an automated step goes wrong.

International principles answer that people remain accountable. The OECD Recommendation on AI, which the OECD describes as the first intergovernmental standard on AI and which sets out the OECD AI Principles, was adopted in 2019 and updated in May 2024 26. The NIST framework treats governance as a function that runs through every stage of an AI system's life 6, and the EU AI Act sets risk-based rules for AI developers and deployers 8.

Good practice sets clear limits. An automated tool should act only within permissions granted to a named person, log every action, and stop for human approval before any step that changes a public output, spends money or affects people's rights. The topic on agents and automation covers authority and risk controls, and the Artificial Intelligence and Autonomous Systems platform covers AI risk more broadly.

12साधने

The instruments that govern risk computing and decision infrastructure

Principles, frameworks, standards and laws, and their status.

No single treaty governs the discipline. It is shaped by a set of instruments that each cover one layer: statistics, geospatial information, data sharing, cyber security, AI and early warning.

For data, the Fundamental Principles of Official Statistics 11 and the FAIR principles 12 set expectations for quality, impartiality and reuse. For geospatial information, the UN Integrated Geospatial Information Framework 18 guides national systems and OGC standards 1920 make them interoperable.

For security and AI, the NIST Cybersecurity Framework 2.0 7 may be adopted voluntarily, and the AI Risk Management Framework 6 is intended for voluntary use, the OECD Recommendation on AI 26 is described by the OECD as the first intergovernmental standard on AI, and the EU Artificial Intelligence Act 8 is a regulation that entered into force on 1 August 2024. The Global Digital Compact 10 is annexed to the Pact for the Future, which the General Assembly adopted.

The Global Risks Forum platform on AI and digital governance follows how these rules are made and enforced, and the Global Risks Alliance platforms on AI and financial model risk and actuarial and quantitative risk cover how models are governed in finance.

साधनIssuerस्थितीहे काय करते
Fundamental Principles of Official Statistics 11UN General AssemblyResolution 68/261, 2014Rules for impartial, accessible official statistics
Integrated Geospatial Information Framework 18UN-GGIMAdopted framework, 2018Eight goals and nine strategic pathways
OGC standards and APIs 1920Open Geospatial Consortiumसध्याचेOpen specifications for geospatial exchange
FAIR principles 12GO FAIRPublished 2016Findable, accessible, interoperable, reusable data
Cybersecurity Framework 2.0 7NISTPublished February 2024Six functions of cyber risk management
AI Risk Management Framework 1.0 6NISTPublished January 2023Govern, map, measure and manage AI risk
OECD AI Principles 26OECDUpdated May 2024Intergovernmental standard on trustworthy AI
Artificial Intelligence Act 8युरोपियन युनियनAdopted June 2024Binding rules for AI on the EU market
Global Digital Compact 10संयुक्त राष्ट्रेAdopted by the General Assembly, 22 September 2024Political commitments for an inclusive, open, safe and secure digital future
Table 2. Instruments that shape risk computing, geospatial systems and decision infrastructure, and their status as of 16 November 2026.

13Frontiers

What is changing in the discipline

Six areas where methods and practice are moving fastest.

Machine learning in operational forecasting. AI weather models now run beside physics-based models, with large energy savings 4.

Exascale and energy limits. Five systems have passed the exascale mark 3, and the Green500 list ranks systems by energy efficiency 23.

AI governance becoming law. The EU AI Act sets risk-based rules for AI developers and deployers 8, the OECD principles were updated in 2024 26, and NIST has released a concept note for a critical-infrastructure profile of its voluntary AI framework 25.

Universal early warning. The 2027 goal of Early Warnings for All 24 tests whether data, models and dissemination can reach the places with the weakest systems 5.

Open, web-native geospatial standards. OGC APIs 20 and global grids of hexagonal cells 22 are two ways to organise geospatial data, and in this guide's view the limits of such grids should be stated.

Closing the connectivity gap. With 2.2 billion people offline 2, designing for low bandwidth and the edge remains a core engineering problem.

14संघ

Three faces of decision infrastructure across the Consortium

Science and engineering, governance and finance for the same systems.

The same forecast model looks different to the engineer who runs it, to the official who must issue a warning from it and to the insurer who prices risk with it. Within the Consortium, GCRI leads on science, technology and evidence through the Nexus Foundry platform, its STEM foundation and pages on geospatial information systems, high-performance, cloud and edge computing and artificial intelligence and machine learning. The canonical method pages for geospatial systems and computing sit in the STEM Commons. Nexus Foundry is operated at national and regional level by Nexus Consortiums through their host and lead organisations, within a multilateral programme led by these three pillar not-for-profit institutions on a public-private-planet partnership.

The Global Risks Forum (GRF) leads on governance, leadership and cooperation. Its AI and Digital Governance platform covers the rules for AI, data and digital systems.

The Global Risks Alliance (GRA) translates technical evidence into the questions finance asks, without lending, underwriting or rating. Its platforms on actuarial science and quantitative risk and AI and financial model risk ask how models are validated and governed where money depends on them.

Work moves between the three in a cycle: a question is stated, evidence is assembled, analysis is built and tested, findings are published, competent authorities decide, and lessons reset the next cycle. The counterparts page lists the matching platforms, and the field gateways for data and compute and methods and modelling show the same subjects across all three faces.

What this article does not claim: it explains a discipline and the public evidence on it. It is not a product description, a performance claim for any system, a security assessment or advice for any procurement or deployment decision. Figures are those of the cited bodies, for the stated date. GCRI does not regulate or certify, GRF does not regulate or finance, and GRA does not lend, underwrite or rate. Naming an organisation, framework, standard or dataset is not a claim of partnership, endorsement or deployment.

15विषय

Explore the platform topics

The eight foundational topics of the platform, each with its own article.

This article is the hub for eight foundational topic articles. Project workstations covers onboarding, persistent context and permissions; the twenty technical systems covers the canonical system inventory; maps and GIS covers approved public map views; and evidence, ontology and lineage covers how data and their history are recorded.

Modelling and simulation covers workloads, reproducibility and limits; agents and automation covers AI tools, authority and risk controls; compute and edge covers cloud, local and edge capacity; and usage and billing covers service units, quotas and charges. The STEM pages on systems engineering and safety-critical software explain the engineering methods the topics use.

प्रश्न

प्रश्न आणि उत्तरे

What is decision infrastructure?

It is the combination of data, models, computing, networks and institutions that turns evidence into a decision in time. An early warning system, which links monitoring, forecasting, dissemination and preparedness, is a clear example.

What is a geospatial system?

A system that stores, analyses and displays information tied to locations on Earth. It combines maps, satellite imagery, sensor data and statistics, and uses open standards such as those of the Open Geospatial Consortium to exchange data 19.

Are AI weather forecasts reliable?

They are now used operationally beside physics-based models. ECMWF reports gains of up to 20 per cent on measures such as tropical cyclone tracks for its AI system, while it continues to run and compare its physics-based model 4.

Why do risk models need so much computing power?

Because they simulate large physical systems at fine resolution, many times over, to estimate uncertainty. The fastest systems on the TOP500 list now exceed an exaflop, a million trillion calculations per second 3.

What does FAIR data mean?

FAIR data are findable, accessible, interoperable and reusable, following guiding principles published in 2016 12. In practice this means persistent identifiers, rich metadata, open formats and a clear licence, so that people and machines can find and reuse the data.

How is AI risk managed?

Through frameworks such as the NIST AI Risk Management Framework, which organises the work into governing, mapping, measuring and managing risk 6, and through law such as the EU Artificial Intelligence Act 8.

Does this article describe a product?

No. It explains a discipline and the public evidence on it. The Nexus Foundry platform pages describe the platform's own scope and limits.

अटी

Glossary

Decision infrastructure
The data, models, computing, networks and institutions that turn evidence into timely decisions.
Geospatial information
Information tied to a location on Earth, such as maps, imagery and located statistics.
Exaflop
A computing speed of a million trillion floating-point operations per second.
वंशावळ
The record of the sources, versions and steps that produced a result.
FAIR data
Data that are findable, accessible, interoperable and reusable.
Edge computing
Computing done close to sensors and users rather than in a distant data centre.
Forecast verification
The systematic comparison of forecasts with what actually happened.

Reuse and cite

Text CC BY 4.0. Cite the page with its date; data tables are linked under each figure.

स्रोत

References

26 sources, numbered in the order they are first cited
  1. Global Biodiversity Information Facility (2026-11-17). GBIF occurrence count (GBIF API). api.gbif.org. Accessed 2026-11-17.
  2. International Telecommunication Union (2025 edition). Measuring digital development: Facts and Figures 2025. www.itu.int. Accessed 2026-11-16.
  3. TOP500 (2026-11-16). TOP500, November 2026 (68th edition). top500.org. Accessed 2026-11-16.
  4. European Centre for Medium-Range Weather Forecasts (2025-02-25). ECMWF's AI forecasts become operational. www.ecmwf.int. Accessed 2026-11-16.
  5. World Meteorological Organization and UNDRR (2025 edition). Global Status of Multi-Hazard Early Warning Systems 2025. wmo.int. Accessed 2026-11-16.
  6. National Institute of Standards and Technology (Version 1.0, January 2023). Artificial Intelligence Risk Management Framework (AI RMF 1.0), NIST AI 100-1. nvlpubs.nist.gov. Accessed 2026-11-16.
  7. National Institute of Standards and Technology (2024-02-26). The NIST Cybersecurity Framework (CSF) 2.0, NIST CSWP 29. nvlpubs.nist.gov. Accessed 2026-11-16.
  8. European Union, EUR-Lex (2024-06-13). Regulation (EU) 2024/1689 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act). eur-lex.europa.eu. Accessed 2026-11-16.
  9. World Bank (2021 edition). World Development Report 2021: Data for Better Lives. www.worldbank.org. Accessed 2026-11-16.
  10. United Nations General Assembly (2024-09-22). Pact for the Future, Annex I: Global Digital Compact (A/RES/79/1). documents.un.org. Accessed 2026-11-16.
  11. United Nations Statistics Division (2014-01-29). Fundamental Principles of Official Statistics (General Assembly resolution 68/261). unstats.un.org. Accessed 2026-11-16.
  12. GO FAIR (GO FAIR page, current as of 16 November 2026). The FAIR Guiding Principles. www.go-fair.org. Accessed 2026-11-16.
  13. European Commission, Copernicus programme (Programme web page). About Copernicus. www.copernicus.eu. Accessed 2026-11-17.
  14. European Commission, Copernicus programme (Programme web page). Access to Copernicus data. www.copernicus.eu. Accessed 2026-11-17.
  15. U.S. Geological Survey (USGS feed documentation, current as of 16 November 2026). Real-time earthquake feeds (GeoJSON summary format). earthquake.usgs.gov. Accessed 2026-11-16.
  16. Group on Earth Observations (GEO page, current as of 16 November 2026). About GEO. earthobservations.org. Accessed 2026-11-16.
  17. International Charter Space and Major Disasters (Charter page, current as of 16 November 2026). About the Charter. disasterscharter.org. Accessed 2026-11-16.
  18. United Nations Committee of Experts on Global Geospatial Information Management (2018-07-24). Integrated Geospatial Information Framework, Part 1: Overarching Strategic Framework. ggim.un.org. Accessed 2026-11-16.
  19. Open Geospatial Consortium (OGC standards page, current as of 16 November 2026). OGC Standards. www.ogc.org. Accessed 2026-11-16.
  20. Open Geospatial Consortium (OGC API site, current as of 16 November 2026). OGC API: building blocks for location. ogcapi.ogc.org. Accessed 2026-11-16.
  21. H3 project (open-source geospatial index) (Documentation, current as of 17 November 2026). H3 documentation: tables of cell statistics across resolutions. h3geo.org. Accessed 2026-11-17.
  22. H3 project (open-source geospatial index) (Documentation, current as of 16 November 2026). H3 documentation: introduction. h3geo.org. Accessed 2026-11-16.
  23. TOP500 (2026-11-16). Green500, November 2026. top500.org. Accessed 2026-11-16.
  24. World Meteorological Organization (WMO page, current as of 16 November 2026). Early Warnings for All. wmo.int. Accessed 2026-11-16.
  25. National Institute of Standards and Technology (NIST page, current as of 16 November 2026). NIST AI Risk Management Framework: programme page. www.nist.gov. Accessed 2026-11-16.
  26. OECD (OECD.AI page, current as of 16 November 2026). OECD AI Principles. oecd.ai. Accessed 2026-11-16.

नोंद

Editorial record

Facts and policies are stated as of 16 November 2026. This article is reviewed at least every 180 days and whenever a cited list, framework or law changes, including the next TOP500 and Green500 lists, the next status report on multi-hazard early warning systems and NIST profiles under the AI Risk Management Framework.

Published by the GCRI editorial desk for the Nexus Foundry platform under the editorial integrity policy. To report an error or request a correction, use the corrections and concerns route; corrections are dated and listed with the article.

Text is published under the Creative Commons Attribution 4.0 licence (CC BY 4.0): reuse it with credit to GCRI and a link to this page. Third-party data keep their own terms.

Cite this article

GCRI editorial desk (2026, November 16). Risk computing and decision infrastructure: a guide. Foundry. https://therisk.global/nexus-foundry/risk-computing-geospatial-systems-and-decision-infrastructure/

Republish

Republish or translate this article free of charge, in print or online, under CC BY 4.0: keep the byline and Foundry, link to the original and note any change. Figures, maps and twins made by the institutions carry the same licence; third-party data keeps its own credit line.

Attribution to paste
<p>Risk computing, geospatial systems and decision infrastructure: the foundations — GCRI editorial desk, Foundry. First published 16 November 2026 at <a href="https://therisk.global/nexus-foundry/risk-computing-geospatial-systems-and-decision-infrastructure/">https://therisk.global/nexus-foundry/risk-computing-geospatial-systems-and-decision-infrastructure/</a>. Republished under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).</p>

Related across the Consortium

Share

EmailLinkedInBlueskyMastodon

The Global Centre for Risk and Innovation (GCRI Canada) is a Canadian not-for-profit organisation, not a registered charity, with its registered office in Ontario.