A researcher has an unidentified smudge on screen. David Schlegel, a Lawrence Berkeley National Laboratory researcher and co-lead of the Legacy Surveys, describes what happens next: astronomers "often start by pulling up the Legacy Imaging Viewer to see what you're looking at," and he calls the surveys "part of the fabric of astronomy research now" [11]. That description is worth more than any traffic figure, because it places the tool at the start of somebody else's work, well before any press cycle plays out.
The pitch and the practice are different products. Arjun Day, co-director of the Legacy Surveys and an astronomer with the National Science Foundation, says the skies belong to everyone and that the survey gives everyone the chance to marvel at them [12], and Wired's write-up invites readers to poke around the universe [21]. Marvelling is free and harmless. The load-bearing evidence is narrower: upwards of 1,800 papers already reference the data [10], which works out near 360 a year if you count only the five observing years [20]. No figure in the release says how long a first-time user needs to reach an answer they would put in a document, or how many people using the viewer sit outside the collaboration.
The scale numbers get more useful when divided. 5.6 trillion pixels [4] spread across roughly 4 billion catalogued objects [5] averages about 1,400 pixels an object [18]. That is the budget a user has for deciding whether a faint thing is a galaxy or an asteroid trail, and since most of any survey image is empty sky, the pixels that carry the decision are fewer still. Getting there took 18 years [19]: the five-year survey [1] sat on 13 years of earlier groundwork [8], with more than 160 scientists on data collection [9] and DESI's "5,000 mechanical eyes" aimed out through the top or bottom of the Milky Way to dodge our own dust [7].
Two tests apply to anyone shipping a dataset as a product. First, the gap between a stranger arriving and that stranger holding an answer good enough to cite. Second, whose Monday breaks if you switch the thing off. DESI's imaging clears the second on Schlegel's own account, and clearing it is expensive, because a default first step cannot be revised quietly. The full five-year cosmology analysis due in 2027 [15] and the comparisons other instruments will draw against DESI [16] both land on top of imaging decisions already baked into those 1,800 papers. A survey that people open before they do anything else now owns its calibration on behalf of everyone downstream.
