The crater resulting from this impact on the far side of the moon has been found by the Lunar Reconnaissance Orbiter, about three months after impact. Here's their commentary on what they found.
Comparison of prediction to actual impact location: Before the actual impact on 2022 March 4, I'd computed a prediction for the impact location, as had the Jet Propulsion Laboratory. Our predictions differed by about 8 km, which didn't really surprise either of us. Their position was close enough to mine, and mine to theirs, to be consistent with the data we had.
The actual impact location was uncertain by about a dozen kilometers, largely because our last observations were made about four weeks before impact. (After that, the object was too close to the sun in the sky to be able to point a telescope at it.) The telescopic observations were of good quality and gave us an excellent idea of what the trajectory was at that time.
The problem was that spacecraft and space junk are gently pushed by sunlight, in a way that depends on how the objects are oriented as they tumble end over end. It's a small push, but over the four weeks, we knew it could push the object a dozen or so kilometers one way or the other, in a poorly-determined direction. It's a bit like predicting where an empty trash bag will go in a windstorm. You know it'll go downwind, but not exactly where it'll go.
So we were fairly confident that the LRO crew would have to search a bit around our predicted impact sites. In fact, the impact crater was found eight km from where JPL put it and about 16 km from where I'd put it. (It is very difficult to be more accurate than JPL.) It's a little further off than either of us had expected, but not surprisingly so.
In any case, I'm very pleased that the LRO folks were able to locate this. I must confess that I'd naively thought it would be easier to find and would have been located shortly after impact. However, this story about finding the impact site of the Apollo 16 booster (which had a similar level of uncertainty about where it was) makes it clear that finding one small crater among hordes of craters isn't all that easy. That crater was found about six years after the other Apollo booster impacts. Compared to that, having to wait about three months looks pretty good.
LRO project scientist Noah Petro has kindly pointed out to me that there's yet another complication : LRO gets a daytime pass over a given spot roughly every four weeks, but the illumination will be somewhere between nearly vertical and nearly horizontal. (I'm not sure you can get a sun-synchronous orbit around the moon -- it lacks the oblateness that lets you do so around the earth -- and in any case, a variety of illumination angles would actually be useful for figuring out what you're looking at.) For comparison to an archived image, you'd want the "before" and "after" illuminations to be fairly close. He lists the following post-impact passes made by LRO :
2022 Mar 28: Sun 83.3 degrees above horizon 2022 Apr 24: Sun 61.5 degrees above horizon 2022 May 21: Sun 31.9 degrees above horizon 2022 Jun 17: Sun 1.4 degrees above horizon
Why I (and JPL) were confident in the impact location: I write this bit because I've been asked several times if I was pleased or surprised about the crater being found fairly close to where it was predicted to be.
This isn't our first impact calculation, by a good margin. I won't say it's entirely routine, but I've run one previous prediction for a lunar impactor (the completely intentional LCROSS mission), and I've computed impact locations for the comet impacts on Jupiter in 1994. I made the first public prediction of a (very small) asteroid to hit the earth back in October 2008 (it hit the earth 17 hours later, in Sudan. Big enough to result in scientifically useful chunks to study, and probably quite impressive if you were close enough to see the meteor, but not at all dangerous.) We've had three similar instances since then, and I've run impact calculations for all of them.
Back in 2015, I ran impact calculations for an unidentified bit of junk that re-entered near Sri Lanka, and there have been several other instances where impacts of junk were predicted using telescopic data.
So I've had a fair number of chances to check such computations before. JPL, of course, has had even more opportunities to do so, having tackled the even more challenging problem of hurling objects at the moon, Mars, Venus, asteroids, etc. and either hitting them at the intended spot or doing flybys at the desired location. Basically, these things hit where Isaac Newton says they will.
A couple of minor corrections to the LRO post : The uncertainty was about a dozen kilometers, not 100 km. I think the LRO folks didn't realize that the impact predictions really were as good as was claimed. Apparently, they had another impact prediction about 100 km away. That was simply wrong (given the data, it just wasn't going to be that far off). But the LRO folks didn't know that, and took the scatter among the predictions as an indicator of uncertainty.
Also, the object is quite conclusively identified as the Chang'e 5-T1 booster. (I can completely understand the LRO folks not wanting to say much on that point, though. It's fine for me, as an unaffiliated researcher in the field, to lay out the evidence and say there's no real question about it. It could be messier for them.)
There was some confusion about this for a while, much of it my own fault. I'd initially mis-identified this as a booster for NASA's DSCOVR mission, one that actually escaped into orbit around the sun. Then the Chinese Foreign Ministry thought we were talking about the Chang'e 5 booster rather than the 5-T1 booster (should add that from their wording, I think they just got confused about which object they were being asked about). They are two different objects. The CE-5 booster former was tracked coming back from the moon and hit the middle of the Pacific less than a week after launch. The CE-5-T1 booster spent over seven years wandering around the earth-moon system, before hitting the far side of the moon on 2022 March 4.
Comments on odd double-crater appearance: I'm a little puzzled by this. But I am in no way an expert on what happens in high-speed impacts like this, except to know that they can have some very strange, non-intuitive results.
The theory has been going around that if this were an impact at a shallow angle, it might create an elongated or double crater (as does sometimes happen, and accounts for some elongated/double lunar craters). But the Chang'e 5-T1 booster came in at about 15 degrees from vertical. So that's not the explanation for this one.