Cosmos 2612 and ICEYE-X36: what was measured, what was predicted.

Cosmos 2612 was launched from Plesetsk on April 16, 2026 (international designator 2026-083H, NORAD 68762). In May, according to the U.S. Space Force's space threat fact sheet, Russia used fuel-intensive maneuvers to place Cosmos 2610 through 2613 close to a Western commercial radar satellite used by Ukraine. CSIS analysts report that another Russian satellite, Cosmos 2609, passed within about 2.5 km on May 26, and ICEYE confirmed to Reuters on October 9 that the satellite involved in May was its own.

From September 30, Cosmos 2612 made a series of maneuvers. Independent analyst Marco Langbroek dates them to September 30 and October 12, 15 and 16, and writes that they synchronized its altitude, eccentricity and orbital period with ICEYE-X36's. CSIS analysts wrote on October 15 that the maneuvers had restored its alignment with the ICEYE satellite's orbital plane.

On October 16 the commercial tracking company LeoLabs put the two satellites about three miles apart, against more than 200 miles a day earlier. It said the movement "appears to be a deliberate approach" and that the intent remains uncertain (Axios). Langbroek estimated from public element sets a close approach around 01:35 UTC on October 18, within a few hundred meters or possibly much closer, and reported no sign in ICEYE-X36's elements that it had been disturbed.

Element sets published on CelesTrak on the afternoon of October 18 place both satellites in the same plane, at 97.85 degrees inclination and about 540 by 546 km. Propagated to their own epochs, they show Cosmos 2612 trailing by roughly 10 km. The Defense Department and ICEYE declined to comment, per Axios, and as of October 18 no government had published a statement on the October approach.

Three kinds of number have been reported, and they are not interchangeable. LeoLabs' figures come from its own radar tracking. Langbroek's and others' come from public element sets, the compact orbit descriptions the U.S. Space Force publishes for each catalogued object. Orbital Radar's 26 meters is a forecast, generated automatically from the same public element sets for a pass on October 19. And the distance itself kept changing: the satellites closed, passed, and drifted apart.

Public element sets cannot settle meters. ESA's Space Debris Office found their errors can reach about 5 km along the direction of travel in some orbital regions (Flohrer and colleagues, 2009). A 2010 study reported that its improved method cut the growth of their error about tenfold, to roughly 100 meters a day, which puts the standard model at about a kilometer a day (Levit and Marshall). Any separation under a kilometer computed from them is an estimate inside the noise, and a forecast a day ahead is softer still, because the error keeps growing with every hour past the moment the element set describes.

Diplo Space ran the check anyone can run. Propagating the October 18 element sets with SGP4, the standard model, gives a closest approach of about 160 meters near 01:38 UTC on October 18, in line with Langbroek's estimate and with Integrity ISR's Greg Gillinger, who estimated under 150 meters from elements of October 17. That figure is not a measurement. Langbroek reports a small Cosmos 2612 maneuver around then, so the true distance may differ by more than a kilometer either way.

This is a rendezvous, not a crossing. In the 2009 collision between Iridium 33 and Cosmos 2251, the two satellites met at about 11.7 km per second on orbits that crossed. Here the two share one orbit and move at nearly the same speed: astronomer Jonathan McDowell described one pass on October 16 at a few kilometers and about 20 km per hour. That is why the right picture of this event is the relative-motion view analysts use (how far ahead or behind, above or below, one satellite sits from the other), not a flash-by.

Close approaches happen by design in satellite servicing and inspection, and commercial operators that perform them have published principles built on the target owner's consent and on transparency (CONFERS). ICEYE has not said whether it was consulted; it declined to comment. McDowell called this the first rendezvous in low Earth orbit between two active satellites without the owner's approval. That characterization is his.

Russian satellites have approached others before, including U.S. government satellites in 2020, an episode U.S. Space Command called "unusual and disturbing" (CNBC). The record codes such events by the response they draw, not by how alarming the geometry looks.

No treaty sets a minimum distance between satellites. Article IX of the 1967 Outer Space Treaty asks states to conduct their activities "with due regard to the corresponding interests of all other States Parties." A state that believes another state's planned activity could cause potentially harmful interference may request consultation. The treaty does not define due regard.

The UN's guidelines for the long-term sustainability of outer space activities, adopted by consensus in COPUOS in 2019, ask states to share contact information and information on orbital events (Guideline B.1) and to perform conjunction assessment during all orbital phases of controlled flight (Guideline B.4). They are voluntary. The UN satellite-operator contact hub that opened in September 2026 builds on B.1.

The U.S. Defense Department's 2021 Tenets of Responsible Behavior in Space include maintaining safe separation and safe trajectory. They direct U.S. defense components, not other states. As of October 18 the record codes this event at rung 0: logged, with no coordinated response documented. A government statement naming it would move it to rung 1; discussion in a shared forum, such as the UN General Assembly's First Committee, to rung 2.

On October 21 the United States named the satellite at the UN. In a statement to the General Assembly First Committee's thematic discussion on outer space, delivered by Brian Hardesty, Senior Advisor to the Conference on Disarmament, it said Cosmos 2612 "exhibited threatening behavior by intentionally maneuvering dangerously close to a commercial satellite," called the activity "deliberately reckless," and asked Russia "to be transparent about what this satellite was doing in orbit." The statement gave no distance and did not name ICEYE. The UN's record of the meeting notes no Russian reply on the satellite (GA/DIS/3774).

LeoLabs, reported by Via Satellite on October 19, put the closest approach at roughly 500 meters and noted that exact estimates may be uncertain. It said the close proximity activity had ended as Cosmos 2612 moved away on October 18 and 19. The report gives no time for the closest point. ICEYE told the publication it was aware of the situation and tracking it closely.

On October 27, at the General Assembly's Fourth Committee, the European Union said uncoordinated rendezvous and proximity operations pose a risk to space safety and voiced concern about the "pursuit and harassment of other satellites," without naming one (GA/SPD/843). The automated site whose 26-meter figure this brief sorted as a forecast now labels it "not re-confirmed." No government has published a distance for the approach.

The record now codes the event at rung 1: one government named it. The October 18 text above said discussion in a shared forum such as the First Committee would mean rung 2. One delegation's statement there is coded as one body flagging it, the rule the record's rubric sets for a statement that names an event in a First Committee debate.