GNSS interference
Aircraft whose ADS-B position integrity has collapsed, per 0.25° × 0.5° cell, grouped into interference events with the jammer's likely location where the aircraft's radio horizons meet. Spoofing is flagged when a position, speed or GNSS height is impossible, and ships in the Gulf of Finland are watched the same way. One aircraft with an old transponder is noise; many losing it in the same place is jamming.
How this is measured
Every ADS-B position carries the aircraft's own integrity figures: NIC (the containment radius the navigation system guarantees) and NACp (its position accuracy). A healthy airliner reports NIC 7-8 and NACp 8-11. An aircraft counts as degraded here when NIC or NACp is 4 or less (worse than about 0.6 NM / 1 NM), or when the receiver network itself marked its GPS as lost. TIS-B positions and aircraft without ADS-B are excluded: they are not GNSS.
The share is aircraft-hours: each aircraft is counted once per hour in each cell it was seen in. Cells with fewer than three are not drawn. A cell where more than 10 % of aircraft were degraded is outlined and most likely jammed. The data covers wherever this site polled adsb.lol (always the Gulf of Finland and the southern Baltic, elsewhere only while someone was watching), so the absence of a colour is the absence of data, not the absence of interference.
Interference events follow the method SeRo Systems published with EANS and TTJA (ION GNSS+ 2023). Each aircraft's loss of GNSS and its recovery is logged with where it happened, and a point every 5 minutes while it stays degraded. Three filters drop what is not jamming: aircraft whose losses happen where every other aircraft around them is fine (their own avionics), losses in a bank steeper than 35° (the antenna looks away from the sky), and losses no other aircraft shares: losses of different aircraft within 50 km and 30 minutes join one event, and an event needs two aircraft. Its radius runs from the centre to the farthest loss.
Likely source. A strong jammer denies GNSS out to its radio horizon. An aircraft losing or regaining GNSS is crossing that horizon, and horizons work both ways: at that moment the jammer sits on the aircraft's own horizon circle (4.12 × (√ aircraft height + √ jammer height) km, heights in metres, a 50 m mast assumed). Where the circles of three or more aircraft meet, and the jammer can still see every aircraft it jammed, is the likely source area; the ring shows how far the equally good answers spread. Two jammers at once, or one too weak to reach its horizon, leave the source blank rather than guessed. It is an area to look into, not a position.
Spoofing. Four checks, each needing two polls in a row: a position jump further than the aircraft can fly (15 km and twice its speed); a ground speed (from GNSS) more than 220 kt away from the true air speed (from the air data computer), or an airliner above FL150 almost standing still; a GNSS height that steps more than 1 200 ft away from every neighbour's at the same time and place while the aircraft still reports healthy integrity; two aircraft parked on one 1 km spot at altitude. The line runs back to the last position the aircraft reported with healthy integrity.
Ships. Fintraffic's open AIS feed covers about a thousand ships in Finnish and Estonian waters. AIS says almost nothing about GNSS health, so what is counted is ships jumping further than they could sail; a ship jumping back and forth all hour is two transmitters sharing one identity and is left out.
Context. The planetary Kp index (NOAA SWPC): at Kp 5 or more a geomagnetic storm can degrade GNSS at these latitudes with no jammer at all. GNSS NOTAMs from EANS are listed when they apply. MLAT positions count as degraded when the aircraft has ADS-B: adsb.lol shows it by MLAT only because its own position is gone.
Airport performance
Movements and delay indicators in the style of Eurocontrol's Performance Review Unit, from ADS-B. Times are good to about ±10 s; ground movements exist only where receivers see the ground.
How this is measured
Landing and take-off times are when the aircraft crosses the threshold and lifts off, from the last fix on final (projected along the runway at its ground speed) or the first fix after lift-off (projected back at its climb rate). Where the archive has no ground fixes, the time is projected from the last airborne fix and marked est.
Additional ASMA time (Arrival Sequencing and Metering Area): the time from entering a 40 NM circle to landing, minus an unimpeded reference, which is the 20th percentile of the last 30 days for the same entry sector and runway (Eurocontrol uses a comparable reference). It is the queueing in the air. It needs at least five comparable flights, so it fills in over the first days.
Taxi-out runs from the aircraft first moving after standing (off-block) to take-off; taxi-in from landing to the first stop off the runway. Only flights whose stand was seen count. Runway occupancy is landing to the first fix clear of the runway. Go-around: below 1 000 ft on final, then a climb of 600 ft or more without touching down. Holding: 360° or more of turning one way within 12 minutes inside 40 NM.
Spacing: where the follower was when the leader crossed the threshold. Pairs behind a heavy (wake category H) are marked.
Airspace · Tallinn FIR (EETT)
Traffic through the Tallinn flight information region: where it comes from and goes to, how busy it is, and how Estonia's danger, restricted and temporary areas were used while active.
How this is measured
FIR boundaries are VATSpy's (close to, not identical with, the AIP). An entry is counted when consecutive fixes cross the boundary within five minutes; where the aircraft came from is the FIR of the fix before. A departure or arrival at an Estonian aerodrome shows as DEP or ARR instead. Occupancy is the number of distinct aircraft with at least one fix inside the FIR in each 10-minute slot.
Area activations are the ones this site read from the EANS NOTAM bulletin and the airspace use plan (AUP) during the day. "Inside while active" means an ADS-B fix within the area's shape and vertical limits while it was active. That is not an infringement verdict: a controller may have cleared the flight through, and the vertical limits of NOTAM areas are sometimes given in a form this reads approximately. Military aircraft often fly without ADS-B, so an active area with nobody inside may still have been in use.
Flight efficiency, fuel and CO₂
Continuous descents and climbs, the fuel burnt queueing and taxiing, and the fuel and CO₂ of everything flying through the recorded regions.
Fuel and CO₂ in the recorded regions
How this is measured
Level flight: consecutive fixes at least 20 s apart with less than 300 ft/min of altitude change. For arrivals it is counted from the top of descent within 200 NM to 10 NM before the airport; for departures from take-off to the top of climb within 200 NM. A continuous descent (CDO) or climb (CCO) has no level flight below FL100. Only flights the archive saw from at least 100 NM out count toward the percentages.
Fuel: jets with OpenAP (TU Delft, open source), from ground speed (standing in for true airspeed), altitude and vertical rate at 80% of maximum take-off mass. The A220 is modelled as an Embraer E-Jet of similar size, which OpenAP has; turboprops (ATR, Dash 8, Saab) use typical published fuel flows for climb, cruise, descent and taxi. Taxiing is the engines' certified idle fuel flow (ICAO emissions databank). CO₂ is 3.16 kg per kg of fuel. These are estimates for comparison, not an inventory: wind, actual mass and engine variants are not known.
Holding fuel is the fuel burnt inside holding stretches; additional ASMA fuel is the share of the arrival's fuel inside 40 NM that matches its additional ASMA time.
Noise screening
How many flights each day were loud enough to interrupt a conversation (65 dB(A) outdoors, N65) and how many at night could wake people (60 dB(A), N60, 23:00-07:00 local). A screening estimate from flight tracks, not a certified noise contour.
How this is measured
For every arrival and departure, the track below about 8 000 ft is filled in every 2 s (with the runway point at the landing or take-off time, where receivers often miss the last few hundred feet). The loudest level at each 500 m cell is estimated from the slant distance to the aircraft: a reference level at 1 000 ft for the aircraft class and phase (heavy jet 94/89 dB(A), medium jet 88/84, turboprop 82/79, light 75/70, departure/arrival), spherical spreading and 5 dB/km of air absorption. Ground effects, lateral attenuation, thrust settings and buildings are ignored.
That is the shape of an N-above count, the metric Australian airports publish for communities, but not its precision: real levels differ by several decibels. Use it to compare days, runways and places, not to settle a dispute. Places are approximate centres of districts.
Replay this day on the map · Data: ADS-B © adsb.lol contributors (ODbL), archived by Ampron Sky from 16 Oct 2026. Ships: Source: Fintraffic / digitraffic.fi, license CC 4.0 BY. Kp: NOAA SWPC. Airspace © openAIP (CC BY-NC), NOTAM and AUP © EANS, FIRs © VATSpy, map © OpenFreeMap / OpenStreetMap. Estimates from public data, not for operational use and not for navigation.