The Global Centre for Risk and Innovation (GCRI)
Scenarios and methods reports set out the analytical machinery behind the series: scenario frameworks, models, statistical methods and the assumptions that drive them. A scenario is a coherent account of a possible future, not a prediction, and its value depends on whether readers can see and test what went into it. This page sets out how analytical studies are documented, reviewed and reproduced, and how their uncertainty is stated.
Scenario analysis is the study of coherent, plausible accounts of how the future might unfold under stated assumptions, used to explore risk rather than predict it. Its value depends on whether readers can see and test what went into each scenario, through assumptions registers, archived code and data, cited model versions and sensitivity analysis.
The working screen behind this subject: its structure, terms and states. Names are placeholders and values illustrative.
Scenarios and methods · Annual synthesis · Edition 9
| Annual synthesis | Global | Data check | 75 | W13 |
| Country profile 7 | Country | Peer review | 99 | W32 |
| Water outlook | Sectoral | Drafting | 50 | W15 |
| Methods note 4 | Methods | Released | 100 | done |
Transparent analytical studies. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.
Discount rates, baselines, elasticities and exposure data can drive results more than the model does. When assumptions sit in code rather than in the text, reviewers cannot test what matters most.
Illustrative pathways are quoted as expected outcomes, and a middle scenario is read as most likely. Without stated purpose and plausibility, scenario results become false predictions in decision documents.
Models are reused and extended across studies, so an error in a shared component or dataset propagates into every study built on it. Version identifiers make that lineage traceable.
Code and data may be available yet never rerun by anyone else. Without independent reproduction, errors in implementation can persist through review and into the reports that cite the results.
Each term as this page uses it. Where a standard-setting body defines the term, the wording follows that body.
| Term | Definition |
|---|---|
| Scenario | A coherent and plausible description of a possible future under stated assumptions, not a forecast. |
| Shared Socioeconomic Pathways | Five narratives of global socioeconomic development used in climate research, held in the IIASA SSP database. |
| Assumptions register | A list of every assumption a study makes, with its source and the sensitivity of results to it. |
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.
| Indicator | Unit | Cadence | Source type | Label |
|---|---|---|---|---|
| Studies with archived code and data | percent of studies | per version | Repository deposit records | OBSERVED |
| Assumptions documented in an assumptions register | percent of key parameters | per study | Editorial audit | REPORTED |
| Independent reproduction status | category: not attempted, partial, full | per study | Reproduction record | REPORTED |
| Sensitivity analyses reported for key parameters | percent of key parameters | per study | Editorial audit | REPORTED |
| Model and dataset versions cited with identifiers | percent of model and data citations | per version | Reference metadata | OBSERVED |
| Statistical or methods review completed | yes or no, with reviewer role | per study | Review record | DECLARED |
The topic as a computational graph: named sources feed the methods, an evidence review gates the outputs, and the decisions stay with the authorities that hold them.
Fig. 02Nodes carry their type, two metric rows and a state; connectors show what feeds what; dashed frames name the owners; the timeline is the cadence. Design of the pipeline, not a record of a run.
Named public datasets and institutional records, each with its licence class. Sources by agreement are used only under their terms and never republished.
| Source | What it gives | Licence class |
|---|---|---|
| AR6 Scenarios Database (IAMC and IIASA) | Integrated assessment scenarios assessed by the IPCC with metadata | Free with registration, terms per IIASA |
| SSP Database (IIASA) | Shared Socioeconomic Pathway projections of population, economy and emissions | Free with registration, terms per IIASA |
| NGFS climate scenarios | Climate-related financial scenarios with documentation | Free to use, terms per NGFS |
| Zenodo | Archived code, data and model versions with DOIs | Metadata CC0, licence per record |
| Software Heritage | Archived source code with intrinsic identifiers | Code under its own licence |
| OSF Registries | Pre-registered analysis plans and study protocols | Metadata CC0, licence per registration |
Nexus Reports prepares evidence. The decisions below are taken by the authorities named, under their own rules, and never by the platform.
| Who decides | The question | Evidence this page serves |
|---|---|---|
| Central banks and financial supervisors | Which scenario sets are fit for climate and systemic stress testing | Scenario documentation, assumptions and sensitivity of results |
| Planning and environment ministries | How to use scenario results in national plans without reading them as forecasts | Stated scenario purpose and the range across scenarios |
| Statistical and methods review panels | Whether a method is adequately documented and reproducible | Assumptions registers, archived code and reproduction records |
| Research funders | Whether to require code, data and reproduction for funded analytical work | Archive and reproduction indicators across studies |
Findings are only as reliable as the methods behind them. Scenario studies and stress tests are powerful tools for exploring plausible futures and testing resilience, but they are easily misread as predictions or as verdicts on institutions. Publishing methods in full, with their assumptions and validation, lets readers judge what a result means and lets other researchers reproduce or challenge it.
Two series publish analysis and the methods behind it.
Narrative and quantitative scenarios that explore how systems could evolve under stated assumptions.
Tests of how systems respond to defined shocks; they describe systems, not the soundness of any firm.
Model documentation, data pipelines and validation reports for reuse and review.
Good scenario practice distinguishes between exploratory scenarios, which map a range of plausible futures, and stress scenarios, which test a system against a severe but plausible shock. Both are useful, and both are misused when a single path is presented as the expected outcome. Editions therefore present sets of scenarios with their driving assumptions and show where results diverge.
Purpose, structure, inputs, assumptions and known limits, published alongside results.
Back-testing, comparison with independent data and sensitivity analysis, using uncertainty quantification and model validation.
Data and code released where rights permit, with pipelines documented under data engineering, interoperability and stewardship.
Results labelled Scenario, Modelled or Forecast, with ranges rather than single numbers where uncertainty is material.
The current service level is R0: information and enquiries. 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.