The Global Centre for Risk and Innovation (GCRI)
Many systemic risks travel along networks: supply chains, power grids, financial exposures, transport and contact networks. Network models represent who is connected to whom and simulate how a failure or infection spreads. This page sets out how networks are built and analysed, how cascade models work and fail, and what evidence serves the institutions that use network analysis to manage interconnected risk.
Networks and cascades · Coupled model A · Scenario set 3
nodes: models · edges: exchanged variables
| Hydrology | 1 day | runoff | Yes |
| Power system | 1 h | demand, outages | Yes |
| Economy | 1 quarter | output, prices | Partial |
| Population | 1 year | exposure | Yes |
Fig.Coupled model runner. Design wireframe with placeholder names and illustrative values, not a record or a live system.
Network cascade modelling is the simulation of how failures, infections or shocks spread along the links of a network. Structure shapes spread, nodes fail beyond thresholds and shift load to neighbours, missing links distort results, and in layered systems failures jump between networks.
For networks and cascades, in GCRI's spatial-intelligence reading, the map opens on night lights, true-colour imagery and population density with relief in 3D, framed on North America; every other NASA layer, live feed and forecast waits in the layer list.
The interactive map needs JavaScript; the layer list and sources are below.
Layers 3
Hazards and exposure, for this page
Keyboard, with the map focused
Hold Ctrl or ⌘ and scroll to zoom
Latest available
NASA GIBS / EOSDIS · ASTER GDEM (METI, NASA) · Blue Marble (NASA) · © OpenStreetMap contributors (ODbL) · Natural Earth
Loading the latest NASA imagery.
Drag to move. Ctrl or ⌘ and scroll, pinch or the buttons to zoom; hover for the values under the pointer; reset returns to the region.
For networks and cascades, in GCRI's spatial-intelligence reading, the map opens on night lights, true-colour imagery and population density with relief in 3D, framed on North America; every other NASA layer, live feed and forecast waits in the layer list.Satellite and model products as NASA publishes them, each with its own date; indicators, not assessments of any place; claim ceiling INTEGRATED.
Layers and dates
Network science for dependency, contagion and cascading failure. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.
Highly connected nodes, short paths and clusters determine how fast and far a cascade travels. The same shock spreads differently on different structures.
Nodes fail when stress exceeds a threshold, and failed nodes shift load to neighbours, which can trigger further failures.
Real networks are rarely fully observed. Missing links can make a system look more or less fragile than it is.
Real systems form multilayer networks, such as power and communications, where failures jump between the layers.
Each term as this page uses it. Where a standard-setting body defines the term, the wording follows that body.
| Term | Definition |
|---|---|
| Threshold model | A cascade model in which a node fails or adopts once enough of its neighbours have. |
| Load redistribution | The transfer of a failed node's load to its neighbours, which may cause them to fail. |
| Percolation | The study of how connectivity of a network changes as nodes or links are removed. |
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 |
|---|---|---|---|---|
| Network size | nodes and links | per dataset | Network data | OBSERVED |
| Degree distribution | distribution shape | per dataset | Network analysis | ESTIMATED |
| Centrality of key nodes | centrality score | per analysis | Network analysis | ESTIMATED |
| Cascade size in simulation | share of nodes failed | per scenario | Cascade models | MODELLED |
| Data completeness | estimated share of links observed | per dataset | Data assessment | ESTIMATED |
| Interlayer dependencies | count | per dataset | Multilayer network data | 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 |
|---|---|---|
| Published network datasets such as SNAP and Netzschleuder | Benchmark networks for methods testing | Open, licence per dataset |
| Input-output and trade networks (OECD ICIO, BACI) | Economic dependency networks | Public, OECD terms; open, Etalab licence |
| Power grid network models for research | Grid topology datasets | Licence per dataset (to check) |
| NetworkX, igraph and graph-tool | Open-source network analysis software | Open source, BSD and GPL licences |
| Network science literature on cascades | Models and validation studies | Licence per publication |
Systems Science and Modelling 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 |
|---|---|---|
| Analysts across Nexus platforms | Which network methods suit their dependency data | Construction guides; model comparisons |
| Regulators and operators | How to read network-based risk findings | Sensitivity analyses |
| Data owners | Which links most need to be observed | Missing link sensitivity |
Network science has shown that many real networks have heterogeneous structures, with a few highly connected hubs. Such networks are robust to random failures but vulnerable to targeted loss of hubs. The same structure that makes them efficient makes them fragile in specific ways.
Cascades occur when a failure overloads neighbours, which then fail in turn. In power grids, a line failure shifts flow to other lines; in finance, a default imposes losses on creditors; in supply chains, a missing part halts downstream production. Whether a shock stays local or spreads depends on load, capacity and connectivity.
Interdependent networks add risk. Power systems depend on communications for control, which depend on power. Failures can bounce between networks, producing larger collapses than either network alone would suffer.
Centrality, connectivity and clustering measures, using network science. They describe structure but not dynamics.
Simulating failure propagation with load-redistribution or threshold rules. Results depend on assumed rules.
Coupling networks to study cross-system failure. Data on real interdependencies are scarce.
Designing redundancy, using optimization. Trade-offs with cost and efficiency.
Network data are often incomplete or confidential, and stylised models may not capture operational detail.
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.