2024 YR4 is currently Torino level 3 - the first to reach this rating in the history of the Sentry risk table (although Apophis did reach Torino 4). All of the virtual impactor sources generally agree - CNEOS gives 1.2%, NEODyS gives 1.4%, and ESA gives 1.1%. Further, the virtual impact isn't very far from nominal, within 0.4 sigma for all for sources.
The Line of Variations of the impact range covers a range just south of Costa Rica (earth-grazing), passing over Colombia and Venezuela into the Atlantic, through the Gulf of Guinea (an almost vertical impact at up to 83 degrees), Nigeria, Cameroon, Central African Republic, South Sudan, Ethiopia, Yemen, the Arabian Sea, and finally again a grazing impact over central India.
At present, the ephemeris uncertainty is low enough that it's very unlikely that we will be able to confirm or rule out an impact before it fades out of detectability. YR4 will be magnitude 25 by the time the uncertainty reaches 0.1", and magnitude 27 by the time it reaches 0.5". If an impact is not ruled out through extraordinary observation efforts by then, it will take until 2028 for it to be brought enough to observe again, at which point it will become brighter than magnitude 21 and be relatively trivial to recover. Although using current published observations it's possible that an impact may be ruled out with 0.5" of uncertainty, the following section should explain why I personally doubt that will happen.
--
I searched a couple days ago for negative observations of 2024 YR4 during its 2016 approach. Although the virtual impactor has no images deep enough to rule it out (A DECam image is frustratingly only half a magnitude too shallow to potentially detect it) I can rule out a great deal of its non-impacting orbits using a number of Subaru images taken in early August 2016. Among these, I can rule out a great deal of the 'early' side of the line of variations. The asteroid will most likely not make an early perigee anywhere between 120,000 and 1,000,000 kilometers, or between 350,000 and 740,000 on the 'late' side. Working with Daniel Bamberger on calculations, this rules out something between 60% and 80% of possible non-impacting trajectories, and makes the actual impact chance in reality at something between 3% and 6% (closer to 6% if you include the first observation, a singlet from ATLAS on the 25th on which the orbit depends greatly, and closer to 3% if you ignore it). I personally consider the measurement and its associated uncertainty trustworthy, but I can understand others disagreeing so that lower chance is included for completeness.
So: in reality, the impact chance of 2024 YR4 is much higher than currently reported due to negative observations, possibly as high as 6%. I invite people to double check and verify my claims here, because if they are correct this may have some pretty significant consequences for the approach to this potential impactor and may need reflecting in public-facing announcements.
Currently there is no official pipeline by which to submit negative observations and have them be processed, which becomes an issue when their usage becomes a matter of public concern as in this case.
