6.10.5 — Disaster Digital Twins — maturity: live

Real-Time Hazard Simulation

Fusing live satellite observations into physics-based simulation engines to run continuous, real-time hazard models across flood, fire, seismic and volcanic scenarios.

Auto-drafted reference page — content reviewed for shape, individual references not yet link-checked.

Emergency managers do not fail because they lack data — they fail because the data arrives too late and feeds no running model. A wildfire doubles in area every twenty minutes; a river basin can move from bank-full to catastrophic inundation in under two hours. Static risk maps and post-event assessments are operationally useless. What is needed is a continuously updated simulation engine that ingests live satellite observations — radar-derived soil moisture, thermal anomalies, wind-field retrievals, precipitation rates — and propagates them forward in time, producing probabilistic hazard envelopes that guide evacuation orders before the event peaks.

A sovereign constellation built for this purpose combines SAR microsatellites for all-weather surface change detection, multispectral nanosatellites for thermal and vegetation state, and GNSS-RO instruments for atmospheric profiling. Revisit intervals of 30–60 minutes across the national territory feed assimilation layers in a ground-based simulation stack. The models — hydraulic routing, cellular automaton fire spread, ground-motion ShakeMap, volcanic ash dispersion — run on sovereign GPU infrastructure, not commercial cloud APIs that can be rate-limited or suspended during a major event when global demand spikes.

The operational payoff is a decision-ready hazard picture pushed to emergency operations centres minutes after each satellite pass. Incident commanders see a 6-hour probabilistic forecast of fire perimeter growth or flood inundation extent, updated every overpass. Evacuation zone boundaries are drawn from model output, not intuition. Post-event, the same pipeline generates rapid damage proxies that trigger insurance pay-outs and reconstruction logistics before ground teams can reach affected areas.

What matters

A 30-minute satellite revisit interval can shorten evacuation lead time by 2–4 hours in fast-moving flood or wildfire events — the difference between orderly evacuation and mass-casualty outcomes.
Commercial simulation-as-a-service providers have suspended or throttled access during concurrent multi-country disasters, precisely when sovereign operational continuity is most critical.
Physics-based hazard models require national terrain, soil, land-cover and infrastructure datasets that host nations are legally prohibited from exporting to foreign cloud environments.
ShakeMap, flood-routing and ash-dispersion models must be tuned to national geology, hydrology and prevailing meteorology; generic global products produce unacceptable false-alarm and miss rates.

Sovereignty score: 9/10 — Real-time hazard simulation is a core sovereign life-safety function: the models, the data and the compute must remain under national control when lives and critical infrastructure hang on the output.

Foreign commercial simulation platforms have denied or delayed access during high-demand multi-country disaster events, creating exactly the blackout that evacuation decision-makers cannot afford.
National terrain, hydrological, geological and infrastructure datasets fed into hazard models are subject to data-localisation and national-security law in most jurisdictions, precluding export to foreign cloud environments.
Adversarial actors who know a nation's simulation dependencies can degrade or spoof its hazard picture during a compound crisis — sovereign compute and sovereign satellite feeds remove that attack surface.
Model calibration against national geology, land cover and climate regimes is a years-long scientific investment; outsourcing it to a vendor creates a strategic dependency that cannot be unwound quickly in a crisis.

Reference architecture

Payload: Primary: X-band SAR, 3m stripmap resolution, 80km swath, for all-weather surface change and soil-moisture retrieval; secondary: multispectral-thermal imager, 30m resolution, 8 bands including SWIR and TIR at 11µm, for fire radiative power and flood extent; tertiary: GNSS-RO receiver for tropospheric and ionospheric profiling supporting precipitation and wind assimilation.
Bus class: Mixed fleet — SAR microsatellites at 120kg, 600W payload power, 3-axis stabilised; optical-thermal nanosatellites at 16U cubesat, 24kg, 40W payload power; GNSS-RO nanosatellites at 6U cubesat, 10kg.
Orbit: Sun-synchronous LEO at 520–570km; 18-satellite SAR walker constellation for 40-minute median revisit; 12-satellite optical-thermal constellation interleaved for thermal passes at 55-minute median revisit; 6 GNSS-RO satellites in 550km circular orbit at 35° inclination for atmospheric limb profiling.
Ground segment: 4-station national TT&C network (X-band downlink at 150 Mbps for SAR; S-band uplink for command); X-band direct readout terminals co-located at regional emergency management centres for in-pass latency under 8 minutes; SatNOGS UHF/VHF beacon monitoring for constellation health.
Software & data
Latency
Cost band
Lead time

References

UNDRR Sendai Framework Monitor — Real-Time Risk Information — The Sendai Framework targets require nations to develop and maintain real-time multi-hazard early warning systems. UNDRR explicitly links satellite Earth observation to national risk simulation capacity.
ESA — Digital Twin Earth and Destination Earth Initiative — ESA's Digital Twin Earth programme demonstrates continuous data assimilation from satellite observations into Earth-system simulation models. The programme highlights the latency and data-sovereignty constraints that arise when national hazard agencies depend on third-party cloud platforms.
USGS — ShakeMap Science and Applications — USGS ShakeMap generates near-real-time ground-shaking maps by assimilating seismic waveform and satellite geodetic data into ground-motion prediction equations. The system architecture requires locally hosted terrain and geology databases to achieve sub-10-minute map generation.
NOAA — National Water Model and Flood Inundation Mapping — NOAA's National Water Model runs hydraulic routing over 2.7 million river reaches, assimilating satellite soil-moisture and precipitation retrievals updated every hour. Operational experience shows that model skill degrades sharply when satellite input latency exceeds 45 minutes.
Copernicus Emergency Management Service — Rapid Mapping — The Copernicus EMS Rapid Mapping service demonstrates end-to-end satellite-to-situational-awareness pipelines for floods, fires and earthquakes. Service-level agreements cap activation-to-product delivery at 12 hours, a ceiling that sovereign real-time simulation architectures are designed to break.