Ionospheric scintillation monitoring via GNSS from LEO

Plasma bubbles in the ionosphere cause GNSS signal amplitude and phase fluctuations that can ground aircraft, disrupt precision agriculture and mimic jamming. LEO radio-occultation constellations now map scintillation events globally, giving operators the evidence to tell space weather from hostile interference.

Sensors

Spire GNSS-RO constellation: Over 100 LEO CubeSats carrying TriG and STRATOS receivers; delivers roughly 10,000 to 15,000 radio-occultation profiles per day globally. Scintillation indices S4 (amplitude) and sigma-phi (phase) computed along each occultation arc. Latency to processed profile typically under three hours in near-real-time mode.
COSMIC-2 / FORMOSAT-7: Six-satellite joint NOAA/NSPO constellation in low-inclination orbits (24°), optimised for equatorial and tropical coverage where scintillation is most severe. Provides roughly 5,000 occultation profiles per day in the equatorial band, with S4 index data publicly archived through UCAR COSMIC Data Analysis Center.
PlanetiQ GNSS-RO: Commercial constellation using multi-GNSS receivers (GPS, GLONASS, Galileo, BeiDou), increasing the number of signal paths per orbit pass. Multi-constellation tracking improves temporal sampling of scintillation events and reduces the chance that a single-system outage is misread as a receiver fault.
COSMIC-1 / FORMOSAT-3 archive: Six-satellite predecessor constellation operational from 2006 to approximately 2020; provides a 14-year archive of S4 and sigma-phi data, enabling climatological baselines for scintillation frequency by season, local time and solar cycle phase. Essential context for separating anomalous events from background variability.

What plasma bubbles actually do to a signal

The ionosphere between roughly 60 km and 1,000 km altitude is not uniform. At equatorial latitudes, after local sunset, the F-region develops density irregularities called plasma bubbles, structures that can span hundreds of kilometres horizontally and persist for several hours. At polar latitudes, particle precipitation from the magnetosphere produces a different class of irregularity, smaller in scale but capable of equally severe disruption. When a GNSS signal passes through either type of structure, diffraction and refraction scatter the wavefront. The receiver sees rapid fluctuations in signal amplitude and phase: scintillation.

The S4 index quantifies amplitude scintillation as the normalised standard deviation of received signal power over a 60-second window. Values above 0.3 are considered moderate; above 0.6, severe. Sigma-phi measures phase fluctuation in radians over the same window. At S4 values approaching 1.0, a standard GNSS receiver can lose lock entirely, producing outages indistinguishable at the receiver level from those caused by jamming. That ambiguity is the central operational problem this monitoring addresses.

Why LEO occultation beats ground networks for this problem

Ground-based scintillation monitors, such as those in the CHAIN network in Canada or the LISN network in South America, provide continuous time series at fixed points. They are excellent for deep characterisation of local conditions but sparse over oceans, polar ice and the developing-world equatorial belt where scintillation is worst and ground infrastructure is thinnest.

A GNSS-RO receiver aboard a LEO satellite observes a GNSS satellite rising or setting through the ionosphere as seen from the spacecraft. The signal path sweeps through the ionosphere at a rate of roughly 2 to 3 km per second at the tangent point, sampling a vertical profile of electron density and, crucially, recording amplitude and phase fluctuations along the arc. Spire's constellation produces global coverage without the geographic gaps that make ground networks unreliable for event attribution. The trade-off is that each occultation is a snapshot, not a continuous time series: the same location is revisited only when orbital geometry brings a suitable GNSS-LEO pair overhead, which varies from minutes to a few hours depending on constellation density.

Scintillation versus jamming: the diagnostic the receiver cannot provide

A jammed receiver and a scintillation-degraded receiver can look identical from the inside: elevated noise floor, loss of lock on one or more signals, increased position dilution. Operators who see only receiver telemetry cannot distinguish the two. The consequences of misdiagnosis are significant. A maritime operator who assumes jamming may report an incident to a national authority, triggering an investigation. An aviation operator who assumes space weather may delay a flight unnecessarily, or worse, assume the problem will pass when it will not.

Space-based scintillation data provides the external reference. If an S4 event of appropriate magnitude is recorded by a GNSS-RO satellite over the affected region within the relevant time window, and the affected receiver's signal loss correlates with the scintillation geometry, the ionospheric explanation becomes strongly supported. If no scintillation event is recorded, the probability of a terrestrial interference source rises sharply. This is not a binary proof: scintillation can be spatially patchy, and a single occultation arc may miss a localised bubble. But combined with multiple arcs from a dense constellation and ground-network data where available, the diagnostic confidence is practically useful.

One honest limit: GNSS-RO scintillation data is not real-time. Near-real-time latency from Spire is typically two to three hours. For post-event attribution, that is adequate. For in-flight decision support, it is not, and operators should not treat this data as a live warning system without additional ground-based augmentation.

Geographic and temporal patterns worth knowing before you commission monitoring

Equatorial scintillation follows a strong local-time dependence: it peaks between roughly 20:00 and 02:00 local time, driven by the post-sunset plasma bubble generation mechanism. It is worst during solar maximum and during equinoctial months (March and September), when the geometry of the geomagnetic field relative to the sunset terminator is most favourable for bubble seeding. The South Atlantic Anomaly adds a further complication: reduced magnetic field strength in that region increases energetic particle flux, producing a distinct scintillation signature that differs mechanistically from equatorial plasma bubbles.

Polar scintillation is less predictable. It is driven by geomagnetic activity and can appear with little warning during substorm events. The auroral oval, typically between 65° and 75° magnetic latitude, is the primary risk zone, but during major geomagnetic storms the affected region can expand equatorward significantly. COSMIC-2's low-inclination orbit provides limited polar coverage; Spire's higher-inclination satellites fill that gap. Any operator with infrastructure above 60° latitude, including Arctic shipping routes and high-latitude aviation corridors, should weight polar scintillation risk explicitly.

What the analytics produce and where the limits sit

From a processed stream of GNSS-RO profiles, the core analytic outputs are event maps showing S4 and sigma-phi distributions by geographic cell and time window, climatological baselines derived from multi-year archives, and anomaly alerts when observed indices exceed a threshold defined against that baseline. For attribution work, the output is a structured report comparing receiver-reported outage timing and location against the nearest occultation arcs, with a confidence assessment.

Satellize runs this class of analytics on open and commercial GNSS-RO data streams, structuring outputs for clients who need space-weather context alongside their RF monitoring picture. The approach is the same one applied in our Tonga crop-estimation programme: open data, rigorous method, delivered as a decision-ready product rather than a raw archive.

The limits are worth stating plainly. Spatial resolution of scintillation mapping is constrained by occultation geometry: the horizontal footprint of a single arc's scintillation measurement is broad, often 100 km or more, which means fine-scale patchiness is smoothed out. Revisit at a specific point depends on constellation size and orbital geometry; even with Spire's constellation, a given equatorial location may see only a handful of relevant occultation arcs per night. Cloud cover is irrelevant here, which is one genuine advantage over optical monitoring. Archive depth extends to 2006 for COSMIC-1 data, giving nearly two full solar cycles of baseline.

Typical figures

Horizontal resolution of scintillation mapping	Approximately 100–300 km per occultation arc tangent-point footprint; finer effective resolution achieved by aggregating multiple arcs over time
Vertical sampling interval	~100 m along the occultation ray path through the ionosphere (F-region, ~150–800 km altitude)
Global occultation profiles per day	Spire: ~10,000–15,000; COSMIC-2: ~5,000 (equatorial-focused); combined multi-source: up to ~20,000+
Near-real-time data latency	2–3 hours (Spire NRT product); COSMIC-2 research-grade archive typically 24–48 hours
Scintillation indices delivered	S4 (amplitude, dimensionless 0–1+) and sigma-phi (phase, radians), computed over 60-second windows
Frequency bands monitored	GPS L1 (1575.42 MHz), L2 (1227.60 MHz); multi-GNSS receivers also cover GLONASS, Galileo E1/E5, BeiDou B1/B2
Geographic coverage	Global; equatorial band (±30° latitude) best sampled by COSMIC-2; polar regions (>60° latitude) better covered by Spire high-inclination satellites
Archive depth	COSMIC-1 from 2006; COSMIC-2 from 2019; Spire commercial archive from approximately 2019
Minimum detectable scintillation	S4 > ~0.1 reliably distinguished from receiver noise; operationally significant events typically S4 > 0.3
Delivery formats	NetCDF (UCAR archive standard), CSV event tables, GeoJSON anomaly layers, structured PDF attribution reports

Analytics Satellize can run

Scintillation event map	Aggregation of S4 and sigma-phi indices from GNSS-RO profiles into geographic grid cells; thresholded by severity level	GeoJSON or GIS raster layer, updated daily or per-orbit pass, showing moderate and severe scintillation zones
Climatological baseline report	Statistical analysis of multi-year COSMIC-1/2 and Spire archives, stratified by local time, season and solar flux index (F10.7)	PDF report and data tables giving expected scintillation probability by region, month and solar cycle phase
Outage attribution assessment	Temporal and spatial cross-correlation of reported GNSS receiver outage with nearest GNSS-RO scintillation arcs; confidence scoring against climatological background	Structured attribution report with confidence tier (ionospheric, indeterminate, or inconsistent with ionospheric cause)
Anomalous event alert	Real-time comparison of incoming NRT scintillation indices against regional baseline; alert triggered when observed S4 exceeds baseline 90th percentile	Automated alert (email or API push) with event location, severity index and estimated duration based on historical bubble persistence data
Polar scintillation risk bulletin	Integration of GNSS-RO scintillation data with geomagnetic activity indices (Kp, AE) to produce conditional risk assessment for high-latitude corridors	Twice-daily bulletin in PDF and structured JSON, covering auroral oval position and scintillation probability for defined Arctic or Antarctic operational areas
Solar cycle trend analysis	Long-term regression of scintillation event frequency against solar flux and sunspot number using COSMIC-1 archive from 2006 onward	Annual trend report with forward projection of scintillation risk through the current solar cycle, relevant to infrastructure planning

Sources

Who does the work

We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.