Freeman Dyson’s 1960 paper in Science is two pages long and is routinely misremembered. He did not propose that aliens build a solid shell around a star. He proposed something far more useful: that any civilisation which grows its energy use for long enough must eventually intercept a large fraction of its star’s output, and that thermodynamics gives it nowhere to put the leftover heat except back out as infrared. He then suggested we look for that. Sixty-five years later, a survey called Kestrel is doing exactly that with better catalogues than he could have imagined.
Kestrel is not a telescope. It is a cross-match, run by a group at the Nordic Institute for Observational Cosmology, that takes precise distances and stellar parameters from astrometric catalogues, pairs them with mid-infrared photometry, and asks a single question of each of 340,000 stars within about 1,500 light years: is this object emitting more heat than a star of its type and distance has any business emitting? An excess at ten to twenty microns, with no corresponding excess in the near infrared, is what a partial Dyson swarm would look like from here.
Everything looks like an alien until it looks like dust
The problem, as every previous search has discovered, is that the universe is full of warm dust and warm dust looks like everything. A debris disc left over from planet formation produces a mid-infrared excess. So does a background galaxy that happens to sit within the same few arcseconds. So does a stripe of interstellar cirrus, a poorly deblended neighbour, or a detector artefact near a bright source. The first Kestrel data release cut 340,000 candidates down to 62 by machine, then to nine by human inspection, and eight of those nine have since been resolved as ordinary astrophysics.
The honest description of this field is that we are building an excellent catalogue of dust and calling it a search for engineers.
That is Dr Havard Lindqvist, who runs the group and who has spent a decade arguing that the null results are the product, not the disappointment. Each clean non-detection puts a hard number on how much of a star’s energy budget can be captured by anything within a given distance without our noticing. Kestrel’s first pass rules out swarms intercepting more than roughly two per cent of stellar output for the great majority of its sample. That is not a headline, but it is a real constraint on a real hypothesis, which is more than most speculative astronomy manages.
The remaining ninth object is being handled with visible care. It is a K-type star at around 480 light years with a mid-infrared excess that has resisted three attempts at a dust explanation, largely because its near-infrared colours are stubbornly normal. Lindqvist’s team have requested spectroscopic follow-up and have gone out of their way to say, in public and repeatedly, that the overwhelmingly likely answer is a background source that better resolution will separate out. They are almost certainly right. They are also not deleting it.
What is genuinely new is not the idea but the scale. Dyson could only gesture at a method. The current generation of all-sky infrared and astrometric archives makes it possible to run his test on hundreds of thousands of stars as a background task, cheaply, and to repeat it every time a new data release lands. The search has become plumbing rather than an event, and plumbing is how astronomy actually finds things.
Nobody involved expects a result. What they expect is a shrinking box: a set of statements about what cannot be out there in our immediate neighbourhood, tightened by a few per cent every year, until the absence itself becomes an argument worth taking seriously. That is the least dramatic possible version of the search for other minds, and it is the only version that has ever produced data.
Leave a Reply