The Global Centre for Risk and Innovation (GCRI)
Droughts and floods are the two ends of hydrological variability, and they increasingly occur together: a dry season that empties reservoirs and dries soils, then intense rain on hard ground that overwhelms drainage. This page sets out how the two extremes develop, how they are measured and forecast, and what the forecasts can and cannot tell the people who must act on them.
Water NexusWireframe of the working environment, not a live screen. Names and values are placeholders.
Compound water extremes, forecasts and limits. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.
A rainfall deficit becomes a soil-moisture deficit, then low river flow and falling reservoirs and groundwater. Heat deepens it: higher evaporation and higher demand for irrigation and cooling at the same time.
Flash floods from intense convective rain, river floods from sustained basin-wide rain or snowmelt, and coastal floods from surge and tide. The same storm can produce all three.
A warmer atmosphere holds more moisture and releases it in heavier bursts. Dry, compacted or burnt ground sheds water instead of absorbing it, so a drought raises the flood risk of the rain that ends it.
Reservoir storage sets hydropower output; turbine curtailment strains the grid; pumping and treatment lose power; hospitals fall back on fuel. Each step carries its own evidence label and its own uncertainty.
The systems this subject touches, the method at its core and the decisions it informs.
Select a scenario. Real geography, placeholder values: the cells show a design of how stress spreads, not a measurement.
Fig. 01Drought and flood on a reference basin with real elevation, coastlines, lakes and rivers. Cell levels are placeholders, not results; a national programme replaces them with its own labelled evidence.
The reference basin in a normal season: storage within rule curves, flows near the median, no indicator in deficit.
Deficit indicators agree inland first; hydropower and treatment feel it weeks later. Each step keeps its own label and confidence.
Intense rain on dry ground overwhelms drainage and inundates treatment works and pumping stations. The same model shows where the two shocks compound.
Levels in the figure are placeholders in a design, not findings. A programme replaces them with labelled evidence it owns.
Where this page’s subject sits in the world: boundaries and context now, the field layers when the interactive map loads.
MapWorld map in the Equal Earth projection, land and boundaries from Natural Earth (public domain). The interactive map loads through the Nexus Foundry plugin (globe or flat view).
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 |
|---|---|---|---|---|
| Standardised precipitation index | dimensionless (3, 6, 12 month) | monthly | Gauge and satellite rainfall | ESTIMATED |
| Soil moisture anomaly | percent of climatology | weekly | Satellite radiometry, land-surface model | MODELLED |
| Streamflow percentile | percent of period of record | daily | Stream gauges | OBSERVED |
| Reservoir storage | percent of live capacity | daily | Operator records | OFFICIAL |
| Groundwater level anomaly | metres below mean | monthly | Observation wells, satellite gravimetry | OBSERVED |
| Flood return period of peak flow | years | per event | Flood frequency analysis | ESTIMATED |
| Inundated area | square kilometres | per event | Synthetic-aperture radar | OBSERVED |
| Population in the flood extent | persons | per event | Gridded population on the flood layer | ESTIMATED |
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 |
|---|---|---|
| CHIRPS rainfall | Daily and monthly gridded rainfall since 1981, gauge-blended satellite estimates | Public domain |
| ERA5 reanalysis (Copernicus) | Hourly atmospheric and land variables, global, 1940 to present | Free and open, attribution |
| Global Runoff Data Centre | Daily river discharge from national gauge networks | Free for research and public tasks under the data policy |
| GRACE and GRACE-FO | Monthly terrestrial water storage anomalies from satellite gravimetry | Public domain |
| JRC Global Surface Water | Surface-water occurrence, change and seasonality from Landsat, 1984 onwards | Free and open, attribution |
| Sentinel-1 radar | All-weather flood extent mapping at ten-metre resolution | Free and open, attribution |
| National hydrological services | Gauge records, reservoir rule curves, warnings | By agreement with each service |
Water Nexus 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 |
|---|---|---|
| Water allocation authority | When to move from voluntary to mandatory restrictions, and in which sectors | Indicator agreement across SPI, storage and groundwater; outlook skill for the next season |
| Grid and hydropower operators | How much hydropower to schedule and when to secure replacement capacity | Reservoir inflow forecasts with confidence bands; the cascade to generation |
| Disaster management agencies | When to warn, pre-position and evacuate, and for which areas | Flood forecasts by reach, the mapped extent of comparable past events, exposure in the extent |
| Finance ministries and insurers | What a one-in-fifty and a one-in-two-hundred year loss looks like, and who carries it | Return-period losses with retained, transferred and uninsured layers kept apart |
Drought develops slowly: a lack of rainfall becomes a soil moisture deficit, then low river flows and falling reservoirs and groundwater. Higher temperatures intensify it by increasing evaporation. Floods can be fast, as in flash floods from intense convective rain, or slow, as in large river floods driven by prolonged rain or snowmelt.
A warmer atmosphere holds more moisture, which tends to increase heavy rainfall intensity. Compound events matter because their joint probability is higher than independent hazards would suggest, and because they defeat defences designed for one hazard at a time. A drought can harden soils and kill vegetation, increasing runoff when rain returns.
extreme value analysis estimates return periods from records. Short or non-stationary records make rare-event estimates uncertain.
Ensemble weather and hydrological forecasts, with forecasting and time-series analysis, give probabilities rather than certainties. Skill falls with lead time and is lower for convective rainfall.
Flood and drought maps built with geospatial information systems and Earth observation. Maps reflect their input data and defences as assumed, not necessarily as built.
| Method | Applied here | Canonical page |
|---|---|---|
| Hydrology and water science | Hydrology and water science in this field | STEM Commons |
| Forecasting and time-series analysis | Forecasting and time-series analysis in this field | STEM Commons |
| Extreme value theory and tail risk | Extreme value theory and tail risk in this field | STEM Commons |
| Earth observation and remote sensing | Earth observation and remote sensing in this field | STEM Commons |
| Uncertainty quantification and model validation | Uncertainty quantification and model validation in this field | STEM Commons |
Forecasts give probabilities, not events. Short records make rare return periods uncertain by a wide margin. Gauge networks are thinning in many basins, so satellite and model products carry more weight and their labels matter more. Compound events are harder to estimate than either extreme alone because their dependence changes with the climate. Water Nexus issues no warnings; competent authorities do, under their own rules.
| Link | Relationship | Why |
|---|---|---|
| Heat and drought | Related topic | Hydrological impact context links to hazard science without equating forecast and observed evidence. |
| Catastrophe models | Related topic | Drought/flood evidence links to independent reinsurance modelling and its uncertainty. |
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.