← The Constellation
The Observatory · space & science

The galaxy's only helium nova, and the spy satellite that erased the missile gap

Sunday · July 26, 2026 · The Milky Way's only helium nova is caught firing gas at 8,900 kilometres a second, a possible seed of a Type Ia supernova; and CORONA, the film-return satellite that turned a feared Soviet missile lead into a countable few dozen.
I · Now observing

The galaxy's only helium nova, unmasked at last, is firing oxygen-rich gas at 8,900 kilometres a second

Astronomers finally saw through the dust that had hidden the only star system of its kind in the galaxy, and found it firing gas at almost nine thousand kilometres a second.[1][3] At the Royal Astronomical Society’s national meeting in the week of July 22, a Warwick doctoral student, John Mills, presented the deepest look yet at V445 Puppis, the Milky Way’s single confirmed helium nova.[1][2] The team combined infrared images from a large ground telescope in Chile, optical images from a space telescope, spectra from a southern observatory, and precise brightness measurements from an orbiting photometer to reconstruct a binary that an expanding dust disk had veiled for about twenty-five years.[1][3] What emerged were narrow, oxygen-rich clumps of gas, seen in no other nova.[1]

white dwarfcompanion (helium star)gas “bullets”, up to ~8,900 km/sTwo opposed jets, more than a trillion miles tip to tip. Not to scale.
The system, sketched. Helium pulled onto the white dwarf detonates and drives collimated jets in opposite directions.[1]
~8,900 km/s
top speed of the gas “bullets”
3.7 days
orbital period, about twice the old estimate
2000
year of the eruption now cleared of dust

01 What happened

A nova is a thermonuclear flash on the surface of a white dwarf, the dense cinder left when a Sun-like star dies, as it pulls gas off a companion until the bottom of that layer ignites.[3] In almost every case the stolen gas is hydrogen. V445 Puppis is the lone confirmed exception, where the accreted layer is helium drawn from a stripped helium star, so the runaway burns helium directly.[1][4] The eruption itself went off in late 2000, but the debris threw up a thick disk of dust that blocked a clear view of the two stars for a generation.[1] Only now, with the dust thinned, could the team clock the pair at a 3.7-day orbit, roughly double the earlier figure, and measure the jets racing outward at up to about 8,900 kilometres a second.[1][3]

02 Why it is hard, and why it matters

The physics that makes a helium nova rare is the same physics that makes it consequential. A hydrogen nova tends to blow off slightly more mass than it gains, so the white dwarf slowly loses ground.[3] A helium layer can burn in a way that lets the white dwarf keep and even add mass, creeping toward the Chandrasekhar limit near 1.4 times the mass of the Sun, the ceiling above which the star cannot support itself.[3][4] Cross that line and the white dwarf detonates as a Type Ia supernova, the class of explosion used as a standard brightness to measure cosmic distances and, in the late 1990s, to reveal that the universe’s expansion is speeding up.[3] Which ordinary stars actually produce those explosions is still unsettled, so a nearby system caught in the act of possibly building toward one is a rare piece of evidence.[3]

03 What to watch

The result was presented at a conference, and the peer-reviewed paper is still to come; the figures that matter most for the supernova question, the white dwarf’s exact mass and distance, were not restated in the announcement and rest for now on earlier work that placed the star at roughly twenty-six thousand light-years with a mass already near the limit.[5] The claim that the clumps are oxygen-rich is described as likely rather than settled, and the 8,900-kilometre figure is an upper bound, not a single speed.[1] The next concrete step is that paper, with a firmer mass and a clearer verdict on whether V445 Puppis is truly on course to become a Type Ia supernova.[1]

II · From the record

CORONA: the film-return spy satellite that photographed the missile gap out of existence

For eighteen months a secret American program threw camera after camera into orbit and lost them, until one August day in 1960 a cargo plane snatched a falling capsule of film out of the sky and gave the United States its first look at the Soviet Union from space.[4] The program was CORONA, hidden inside a civilian cover called Discoverer, and the pictures it brought back did something rare for an intelligence system: they settled a national argument with numbers.[3][2] A feared Soviet missile lead, sized in the hundreds, turned out to be a few dozen.[2][5]

01 The story

In May 1960 a Soviet missile brought down an American U-2 spy plane over the Urals, and overflying the Soviet Union with piloted aircraft became politically impossible.[5] That left a dangerous blind spot, because estimates of Soviet intercontinental missiles were largely guesswork, and the fear of the day held that Moscow might field between 140 and 200 of them by 1961.[5] CORONA’s answer was a satellite that carried a panoramic camera, exposed film in orbit, and dropped that film back to Earth in a small reentry capsule for a mid-air catch by aircraft.[3] The engineering fought back: more than a dozen launches over roughly a year and a half failed, on lost boosters, balky upper stages, and film that tore in the cold of space.[4] Discoverer 13 finally returned a capsule from orbit in August 1960, the first object ever recovered from space, though it carried no film; days later Discoverer 14 flew clean, and a C-119 plucked its film bucket from the air.[4][3]

140–200feared, pre-CORONA10–25measured, Oct. 196163 US missiles deployedEstimated Soviet ICBMs ready in 1961
The missile gap, before and after the cameras. A feared Soviet lead of well over a hundred missiles resolved, once photographed, into a couple of dozen, against sixty-three the United States had already fielded.[2][5]
1.65M
sq nautical miles imaged on one 1960 flight
~800,000
images returned over the program’s life
~$850M
program cost across twelve years

02 The hard part

Two problems defined the effort, and both were new to engineering.[4] The first was getting anything back from orbit at all, which meant a heat shield, a precisely timed deorbit burn, a parachute, and finally a physical catch by a passing aircraft, a chain in which any weak link lost the mission.[4] The second was taking a sharp photograph from a moving platform hundreds of miles up with the film of 1960, which cracked and fogged in vacuum and cold; the static sparks that marred early frames, called corona discharge, gave the program its name.[4][5] There was no way to send a picture down a wire, so the only path from orbit to an analyst’s light table ran through the atmosphere on a strip of film.[3]

03 Why it mattered, measurably

The single flight of Discoverer 14 photographed about 1.65 million square nautical miles of Soviet territory, more than every U-2 mission of the preceding years combined.[3][4] Once analysts mapped the actual missile sites, the estimate of Soviet intercontinental missiles ready in 1961 fell from that feared 140-to-200 range to just 10 to 25, against the 63 the United States had already deployed; the gap not only closed but ran the other way.[2][5] Over twelve years the program flew roughly 130 launches, returned about 800,000 images, and improved ground resolution from around 35 to 40 feet to 6 to 10 feet, for a total cost near 850 million dollars.[4][1] Announcing declassification in 1995, Vice President Al Gore said the satellite coverage had given the country the confidence to pursue arms control, and CORONA’s method of watching from orbit became the national technical means of verification later written into the 1972 SALT accord.[1]

04 Echoes today

Every imaging reconnaissance satellite and every commercial Earth-observation constellation since descends from CORONA’s template, even as the film bucket gave way to real-time digital downlinks.[1] Its declassified frames, now public, also serve science as the earliest global high-resolution record for tracking how glaciers, forests, and coastlines have shifted over sixty years.[1]