Meteorological wind profiler radar emission mapping from orbit
Ground-based wind profiler radars operating between 400 and 1300 MHz cause documented interference into adjacent mobile and satellite bands. Spaceborne RF sensors can map their locations, duty cycles and emission patterns without ground access.
Sensors
- HawkEye 360 cluster satellites: Clusters of three formation-flying microsatellites use time-difference and frequency-difference of arrival (TDOA/FDOA) to geolocate emitters across roughly 100 MHz to 6 GHz. Geolocation accuracy is published as approximately 1 km CEP under good geometry; revisit at any given point is irregular but typically multiple passes per day at mid-latitudes across the growing constellation.
- Sentinel-1 SAR (C-band, 5.405 GHz): Sentinel-1 does not directly detect wind profiler emissions, but its Level-1 products carry RFI characterisation outputs that flag interference from ground emitters in the 400–1300 MHz range coupling into the radar receiver chain. ESA publishes these RFI annotations, providing a georeferenced catalogue of persistent interference sources that can be cross-referenced against known profiler locations.
- Spire Global LEMUR-2 constellation: Spire's RF payload monitors signals across a broad spectrum from LEO. While primarily marketed for AIS, ADS-B and GNSS-RO, the wideband receiver architecture is capable of detecting strong pulsed emitters in the UHF range. Revisit is sub-hourly at high latitudes due to constellation density; temporal sampling supports duty-cycle estimation.
- WMO Global Observing System radar network records: Not a spaceborne sensor, but the WMO OSCAR surface database documents registered wind profiler sites, operating frequencies and nominal power levels. This catalogue serves as essential ground truth for validating orbital detections and for flagging unregistered or non-compliant emitters.
Why a weather instrument becomes a spectrum problem
Wind profiler radars measure atmospheric wind speed and direction by transmitting pulsed RF energy vertically and detecting backscatter from refractive index gradients and precipitation. The WMO allocates operating bands at approximately 449 MHz, 915 MHz and 1290 MHz for this purpose. Under ITU Radio Regulations footnotes 5.291, 5.317A and 5.329, these allocations share spectrum with mobile services and, in some sub-bands, with Earth exploration satellite services. Sharing is supposed to be coordinated. In practice, many profilers predate modern coordination frameworks, operate at powers that were never formally notified, or run duty cycles that exceed what neighbouring services were told to expect.
The interference mechanism is straightforward. A profiler transmitting at 449 MHz with a peak power of 500 W to several kilowatts, even with a narrow vertical beam, produces sidelobes that can couple into nearby base stations and uplink receivers. At 1290 MHz, the overlap with L-band satellite uplinks and GPS L2 is close enough that several national regulators have logged complaints. The aggregate effect across a national network of a dozen or more profilers is not trivial, and it is not always visible from the ground because the interfering geometry is elevation-dependent.
What an orbital pass actually detects
A spaceborne RF receiver in LEO sees the ground from an elevation angle that changes continuously through a pass. For a wind profiler with a nominally vertical beam, the satellite intercepts sidelobe energy rather than the main lobe. At 449 MHz, the free-space path loss to a 500 km orbit is roughly 146 dB. Detecting a 1 kW transmitter through a sidelobe 20 dB down from the main beam therefore requires a receiver with a noise figure and antenna gain combination that makes this marginal but feasible for purpose-built RF intelligence payloads. HawkEye 360's published detection floor covers emitters in this power class at these frequencies.
The pulsed nature of profiler emissions is actually an advantage for detection. Profilers transmit in coded pulse sequences with pulse repetition frequencies typically between 20 and 100 Hz and pulse widths from a few microseconds to around 100 microseconds. This temporal signature distinguishes them from continuous-wave emitters and from other pulsed sources such as maritime radars. A receiver with sufficient time resolution can characterise the pulse repetition interval, pulse width and inter-pulse coding, which together form a waveform fingerprint. That fingerprint can then be matched against published profiler specifications to confirm the emitter type before a geolocation fix is attempted.
Geolocation geometry and its honest limits
TDOA/FDOA geolocation using a cluster of three satellites separated by tens of kilometres gives a position fix whose accuracy depends on the signal-to-noise ratio of the intercept, the baseline between receivers and the number of passes contributing to the solution. For a strong, persistent emitter like a wind profiler operating continuously during a weather event, multiple passes over hours or days can reduce position uncertainty well below 1 km. For a profiler that operates intermittently or at reduced power during fair weather, the intercept probability per pass drops, and the position fix may take days to converge.
A single TDOA pair from two satellites gives a hyperbolic line of position, not a point. A third satellite collapses this to a point, but the geometry still degrades at low elevation angles and when the emitter is near the sub-satellite track. Users should expect position uncertainties of 0.5 to 3 km depending on geometry and signal strength. That is adequate for identifying which site is responsible for interference but insufficient for distinguishing two profilers at the same meteorological station. The Sentinel-1 RFI outputs offer a complementary constraint: because SAR geometry is well-characterised, the RFI stripe in a SAR image can be projected back to a ground range that narrows the candidate emitter location independently of the TDOA solution.
Using the WMO catalogue as an anchor
The WMO OSCAR database lists surface-based observing systems by type, location and operating frequency. For wind profilers, entries include nominal transmit frequency, peak power and operating schedule where declared. Cross-referencing an orbital detection against OSCAR serves two purposes. First, it confirms that a detected emitter is a registered profiler rather than a different source at a similar frequency. Second, it flags detections that do not correspond to any registered site, which may indicate an uncoordinated or non-compliant installation.
Not all profilers are in OSCAR. National meteorological services in some regions operate networks that are reported to WMO but whose individual site records are incomplete or outdated. A systematic orbital survey can therefore produce a more current picture of the active network than the administrative database alone. This matters for spectrum managers at mobile network operators planning deployments in the 450 MHz or 900 MHz bands, who need to know not just what is registered but what is actually transmitting.
What the data cannot tell you
Orbital RF monitoring resolves location and waveform parameters. It does not directly measure transmit power with high accuracy, because the received signal level depends on the unknown sidelobe pattern and the satellite's elevation angle at intercept. Power estimates derived from received signal strength carry uncertainties of 10 dB or more unless the antenna pattern of the specific profiler model is known. This limits the ability to determine whether a profiler is operating within its licensed power envelope from orbital data alone.
Cloud cover is irrelevant to RF sensors, which is one of the genuine advantages over optical monitoring. But ionospheric dispersion at 449 MHz introduces group delay that can complicate TDOA measurements at low elevation angles, particularly during periods of elevated ionospheric activity. At 1290 MHz this effect is much smaller. Analysts working at the lower end of the profiler frequency range should apply ionospheric corrections derived from GNSS total electron content maps when computing TDOA-based positions.
Turning detections into a coordination tool
The practical output of an orbital wind profiler survey is a georeferenced emitter catalogue: each entry carrying a confirmed or probable location, operating frequency, observed duty cycle and waveform fingerprint, cross-referenced against OSCAR registration status. Spectrum coordinators at a national regulator or a mobile network operator can use this to prioritise bilateral discussions with the relevant meteorological agency, to define exclusion zones for new base station deployments, or to support ITU Article 15 harmful interference complaints with documented evidence.
Satellize can run this analysis pipeline on HawkEye 360 intercept data combined with Sentinel-1 RFI annotation layers, producing a structured emitter database with confidence scores and revisit-based duty-cycle estimates. The methodology follows published TDOA/FDOA geolocation practice and the ESA Sentinel-1 RFI characterisation framework. Organisations that want to understand the interference environment before a spectrum auction or a network rollout in the 400–1300 MHz range can request a scoped survey covering a defined geographic area and time window.
Typical figures
| Target frequency range | 400–1300 MHz (WMO profiler allocations at ~449 MHz, ~915 MHz, ~1290 MHz) |
| Geolocation accuracy (TDOA/FDOA, 3-satellite cluster) | 0.5–3 km CEP depending on geometry, signal strength and number of passes |
| Minimum detectable emitter (indicative) | Emitters in the hundreds-of-watts to kilowatt class at 449 MHz detectable via sidelobe intercept from 500 km orbit; lower-power sources marginal |
| Revisit (HawkEye 360 constellation) | Multiple passes per day at mid-latitudes; sub-hourly at high latitudes; irregular scheduling |
| Duty-cycle estimation window | Reliable estimates require intercepts across multiple passes over 24–72 hours for intermittently operating profilers |
| Sentinel-1 RFI annotation coverage | Global SAR acquisition footprint; RFI flags georeferenced to SAR swath geometry (~5 m range resolution in IW mode) |
| Ionospheric correction requirement | Required at 449 MHz for TDOA accuracy; negligible at 1290 MHz under normal ionospheric conditions |
| Archive depth (Sentinel-1 RFI outputs) | From Sentinel-1A launch in April 2014; Sentinel-1B data available until November 2021 antenna anomaly |
| Delivery format | GeoJSON emitter catalogue, CSV duty-cycle log, PDF coordination report with ITU Article 15 evidence package |
Analytics Satellize can run
| Georeferenced wind profiler emitter catalogue | TDOA/FDOA geolocation from multi-satellite intercept passes, cross-referenced against WMO OSCAR database | GeoJSON layer with location, frequency, confidence score and OSCAR registration status per detected site |
| Waveform fingerprint library | Pulse parameter extraction (PRI, pulse width, inter-pulse coding) from RF intercept time series matched against published profiler specifications | Structured database of waveform signatures per detected emitter, supporting emitter-type confirmation |
| Duty-cycle and operational schedule estimate | Presence/absence detection aggregated across multiple orbital passes over a defined observation window | Time-series CSV showing active/inactive periods per site, with confidence intervals reflecting intercept geometry |
| Unregistered emitter flag report | Automated comparison of detected emitter locations against WMO OSCAR registered sites within a frequency-and-location tolerance | Alert list of detections with no matching OSCAR entry, formatted for submission to national spectrum authority |
| Sentinel-1 RFI stripe attribution | Geometric back-projection of RFI contamination stripes in Sentinel-1 Level-1 products to ground-range candidate zones, correlated with TDOA fixes | Map overlay showing RFI-affected SAR acquisitions with attributed source locations and dates |
| Spectrum coordination evidence package | Aggregation of geolocation fixes, duty-cycle data and RFI attribution into a structured interference case file following ITU Article 15 documentation conventions | PDF report with supporting data annexes suitable for bilateral coordination or formal ITU complaint |
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.