Talk:Asteroid impact prediction

Blind side
We're suggesting that the clustering of scanners only a few thousand km apart is a bad thing, because dangerous rocks can sneak in when those places are in daylight. However, I also notice that no site is scanning its entire sky in a night; most are taking a week or more. This tells me the evildoers can sneak in anyway, on a night when the watchers are checking another part of their available sky. So, for the blind side of the globe to be a major weakness, the vigilantes would have to be working faster, inspecting their entire sky at least once during their night. Even better if they could look at some bits of sky twice, several hours apart. Is my amateurish analysis going badly wrong? Jim.henderson (talk) 23:55, 20 November 2018 (UTC)


 * Hi, you raise a good point. There are certainly subtleties here, which the article doesn't really bring out. Yes, it's true that no single survey scans the entire sky in one night, not even the part of the sky it can see. ATLAS comes closest, scanning once every two clear nights. For the few asteroids we detect before they hit us, we typically only have hours notice, so yes ideally want to be scanning the entire sky multiple times per night, and by the same token it would be ideal if we were scanning at all longitudes - just to spot as early as possible. However, the more significant point is the current grouping of latitudes. Grouping by longitude means you need to wait till the Earth has rotated before you get to scan again. The fact that there is currently no survey in the southern hemisphere however means that there are whole sections of the night sky that are never surveyed at all (as the Earth only rotates East-West not North-South). The reason this is so significant is that the two strategies we have (cataloging large asteroids to give years notice and final approach detection of small asteroids) are both affected by this. The East-West grouping still allows larger (brighter) asteroids to be detected on a previous orbit as we can see them from a greater distance and therefore for longer.


 * Another question (and I don't have any sources that address this, but there may be some out there): is there any coordination between surveys to maximise coverage? What I mean is, given the different speeds they operate at, it's easy to imagine them scanning the sky on independent schedules, so that on some nights the same patch of sky gets scanned by multiple surveys at the same time. Meanwhile an asteroid approaches in an adjacent patch of sky that none of them are looking at... but could have been if they'd coordinated their efforts. Telescopes within the same survey do co-operate with each other, but I'm not sure about between surveys. Part of the problem is that most of the surveys get some of their funding for carrying out other tasks and so are not 100% dedicated to detecting near Earth asteroids. If they were all fully coordinated, I think there is just enough capacity to scan the entire Sky once per night ATLAS 50% + ZTF 33% + CSS 14% + Pan-STARRS 3% = 100%. Of course they all have different strengths and weaknesses so it possibly doesn't make sense to organise them that way, but some coordination would presumably help. ZTF has made a massive difference to the number of asteroids spotted since it came on line earlier this year. ZTF scanning every three nights and ATLAS every two would make a lot of sense to coordinate together.


 * I guess the question is what update to make to the article? Any thoughts? Rafflesgluft (talk) 16:39, 24 November 2018 (UTC)

We have no more information than before,, just more doubt, so I only see that we ought to be less insistently distressed about the longitude problem. Part of the lookout problem is, these are pretty much scientific surveys, not defensive military surveillance. There is no quickness, no urgency, no planned response. And not a lot of reason for those things, since there isn't as much to do about a sighting as there was in the days of watching out for bombers and ICBMs. Spotting an incoming rock can mean hours of warning, and probably not days. Small rocks, not easily detectable, make bangs comparable to atom bombs, but they aren't targeted, so they'll usually hit ocean and wilderness. Search resources thus are rationally not generously provided, so the same survey instruments that can do quick, frequent searches also devote time for longer, deeper exposures which are better for finding plutonoids, young supernovas, exoplanet transits, and other cool scientific transient objects.

Thank goodness we don't have the job, but my answer if there were more urgency would be, get things going in Australia and South Africa, and put up a spotter satellite. For the fastest among the existing watchers, the method is, if I'm the northern team then I get the Circumpolar; you get the sky as far south as practicable, and we split the middle so the load is about equal. My two telescopes further subdivide the sky and scan from the west at twilight, hopping to catch something at eastern elongation, finishing in the east at midnight. Then start again at the west and scan eastward to morning twilight, now hoping for western elongations since sunward is the real, inherent blind spot. So, much of the sky near opposition gets scanned twice, but those are the relatively easy finds anyway.

Hmm, I wonder if the antique Baker-Nunn satellite cameras could be refurbished to help. Highly agile machinery, lovely optics, though with lesser aperture than modern survey gear, and something must be done about the deeply curved focal plane. Jim.henderson (talk) 02:45, 25 November 2018 (UTC)


 * Hi, OK I think you're right. Reading the text back again, if the reader doesn't understand astronomy, it almost reads as though there are parts of the sky which are not viewable from the current sites, but would be at a different longitude, which is obviously misleading! Moreover it really doesn't bring out the North/South hemisphere issue. I think I'll rewrite and refocus it, would you be happy to re-review when I have? I really appreciate your input by the way, thanks. I want this article to be the best it can be! If we focus on the exclusively on the North/South issue then we can even bring out subtleties like seasonal variation of hours of daylight, an important advantage of having telescopes in both hemispheres. Rafflesgluft (talk) 09:59, 26 November 2018 (UTC)

A bit busy with edit-athons and even a bit of real life, but here's a stab at it:


 * Asteroids are detected by repeated sky surveys by astrographs, camera telescopes with a wide field of view so they can see a large part of the sky. Image differencing software compares a recent photograph with earlier ones, detecting things that have moved, brightened, or appeared. Other software distinguishes between asteroids and other things that suddenly appear, such as artificial satellites. Cameras with a wide enough field, high enough sensitivity, and the ability to work quickly are few, and some do other work besides watch for approaching asteroids. All are in the Northern hemisphere, which means a large part of the southern sky is hidden from them. All are between 110 and 160 degrees latitude in the Western hemisphere, which means that for several hours per day all of them are in daylight and cannot see asteroids. The majority are near each other in Hawaii, which means weather can blind more than one of them at the same time. Like much astronomical work, the watch is handicapped during full moon. All this makes large gaps in coverage, and asteroids can approach in parts of the sky that won't be checked that night or in the next few nights.

There; it's more a list of several shortcomings than a detailed discussion of one. After this paragraph, we go on about how future watchers might tighten the watch. I haven't swapped it into the article yet, hoping for corrections and constructive criticism. Jim.henderson (talk) 04:35, 30 November 2018 (UTC)


 * So, I did a bit more reading. A small part of what I wrote before was simply wrong; I have corrected it. Good thing I was too cautious to put it directly in the article. Didn't realize that ATLAS is all about approaching asteroids; still don't know whether its telescopes do this full-time or have other work. The ATLAS article devotes much space to describing how an unexpected blast from an asteroid can be a bad thing; maybe that material belongs in a more general article. Must find time for more study. Jim.henderson (talk) 04:40, 1 December 2018 (UTC)

Hi, sorry been doing other things the last week or so. I need to re-read the article as I can see there have been a lot of edits by others in the interim, including some really useful edits, which I'm very grateful for. Once I've been through them I'm going to see if I can do the update - thanks for your words! Rafflesgluft (talk) 07:30, 3 December 2018 (UTC)

Hi, sorry, had an interruption for a few days but back again now. I've been mulling in the meantime and feel unsure about what exactly to say. One thing I have done is nabbed some of your above text and put it in the overview section as it's excellent info on the overall process, but not about the clustering. I'll need to think about that aspect a bit more! --Rafflesgluft (talk) 22:21, 5 December 2018 (UTC)


 * With a few editors who each know the topic better than I do, things seem to be rolling along nicely without my active participation. I shall eventually check for style, links and other matters of form. Naturally this will improve my understanding of the topic, which might inspire me to say something about the division of information between this article and others. Jim.henderson (talk) 18:06, 9 December 2018 (UTC)


 * Thanks . I made some minor changes to it a couple of nights ago, as a start, but they were done in a bit of a rush. Someone else has kindly tidied it a bit. I think like you say, others are looking at this now and the ball is rolling, which is good. I think we need to get your above point about the full moon in there somehow though, it's not explicitly mentioned in the article text at the moment, only in footnotes. A couple of the references mention that the surveys they are describing don't operate at all for 4-5 days per month when the moon is full, which is not mentioned at present. I'll see if I can include it. Rafflesgluft (talk) 08:38, 10 December 2018 (UTC)

Dubious? Overview becoming inaccurate
Recently the overview section was modified to make it clearer. Part of the text now says:


 * Current mechanisms for detecting asteroids on final approach rely on ground based telescopes with wide fields of view which currently can monitor the sky at most every second night. They therefore still miss most of the smaller asteroids that more commonly impact Earth, which are bright enough to detect for less than two days.

This is probably the clearest this part of the text has ever been, which is a good thing. However it also seems to be to be inaccurate. There are two aspects to this, one of which is new and the other which has been in the text for a while now, but the latest change exacerbates.

The first aspect is the two day claim. It seems to assume that each survey operates in isolation, therefore the best chance of detecting the fleeting passage of a dim asteroid is limited by the fastest single survey (which at the moment is ATLAS at 2 days). However, as the article itself explains, this is not how asteroid impact prediction works. All surveys submit their observations to the MPC which does the orbit calculations, and NEODyS and CNEOS use this data to do the impact / close approach calculations. Therefore it is not the survey with the highest frequency of sky coverage that matters, but the frequency of coverage between all the surveys. I.e. adding more surveys helps, even if they are slower than the current fastest. ZTF coming online this year has made this very clear!

The second aspect is the implication that if the sky coverage frequency alone could be increased then the detection of small asteroids would always happen. There are a huge number of reasons why we fail to detect near earth asteroids, some of which are mentioned in the article (the brightness of the moon, weather, daylight, the opposition effect and airglow). There are other effects not mentioned such as atmospheric distortion, the fact that asteroid albedos tend to be very low on the whole etc.

For these two reasons I think this text should be updated. Conditions need to be pretty much perfect for us to spot a small asteroid, which is why so few impacts are predicted. I'm unsure whether in the Overview we want to list all of the factors that need to line up for success or whether we should just say "unless conditions are just right", but the text quoted above seems inaccurate to me.

I'm not sure who made the edit so I can't tag you here to garner your thoughts, instead I'll await a reply. Looking at the history log the edit appears to have been carried out on 13 Dec by somebody at IP address 134.171.86.111, which google tells me is ESO, so certainly a distinguished source! If no reply appears, I'll add a dubious tag to the main text in an attempt to draw attention to this discussion. Thanks in advance for your thoughts. --Rafflesgluft (talk) 14:54, 15 December 2018 (UTC)

Some more food for thought: looking at confirmed close approaches within 1LD, the stats for 2018 so far are: Mt Lemmon : 34 Other CSS : 21 ATLAS    :  9 ZTF      :  8 Pan-STARRS: 1 JAXA SSS :  1 Amateur  :  1

Out of all these surveys, Mt Lemmon is the slowest capturing just 1 square degree per 30 second exposure vs 30 square degrees for ATLAS. The JAXA and Amateur telescopes are both below 0.5m. Not sure why Pan-STARRs has done so poorly this year, and more importantly, what IS the secret ingredient in the CSS/Lemmon telescopes that makes them so effective. I do kind of feel like the article is missing something fairly major here in not documenting why these two telescopes are so much more productive than the rest. Does anybody know why? --Rafflesgluft (talk) 23:12, 15 December 2018 (UTC)

Wiki style
Back from an out of town trip and other distractions of real life, I noticed a few things. First the article has grown pretty large, large enough to make me wonder whether some of the details ought to be spun off to existing detail articles. Second, the lead is a single small paragraph, not fitting very well the ideas in Manual of Style/Lead section. The "Overview" is perhaps a bit large to be an intro as is, but some material could be moved into detail sections and the "Overview" header removed to make a proper lead. Third, some material is poorly organized. For example, I got confused with two separate discussions of the delayed Western Australia station.

More vaguely, I am not sure the distinction between the asteroid catalog method and the incoming alarm method sufficiently informs the various aspects discussed. Also without an idea for a precise solution, the name of the article is about "prediction" but the substance is almost all about simply finding these darn elusive rocks. Jim.henderson (talk) 15:56, 19 December 2018 (UTC)


 * Thanks Jim. Yes the intro para got rather short after a huge edit that was done recently that improved the article in a lot ways but also had some slightly dubious aspects to it. For example the NEOCAM section heading appeared, and while the text below it does mention NEOCAM that's not actually what the paragraph is about. I've been meaning to fix that for a while too! Great that you've had a chance to look at it with a fresh pair of eyes. I definitely have the familiarity problem. I'll see if I can fix it up, although currently very busy. Thanks for the comments --Rafflesgluft (talk) 19:31, 19 December 2018 (UTC)

Tokyo Kiso Observatory
Looking at the pie chart in List of asteroid close approaches to Earth in 2021 (see also 2020), the Tokyo Kiso Observatory is making a significant contribution to discoveries of new asteroids with its 1m class Schmidt. Suggest adding a section on this observatory. This seems to be due to a new instrument added towards the end of 2019, Tomo-e Gozen, which is specifically design to detect "short time-scale phenomena including faint meteors, near-earth asteroids..." using a high-speed, high-cadence, wide-field observations (20 deg2).

Some references which may help provide material for the article:
 * - http://www.ioa.s.u-tokyo.ac.jp/kisohp/OVERVIEW/overview_e.html
 * - https://global.canon/en/news/2019/20190930.html
 * - http://www.mtk.ioa.s.u-tokyo.ac.jp/kisohp/NEWS/pr20190930/pr20190930e.html
 * - https://tomoe.mtk.ioa.s.u-tokyo.ac.jp/index.html

Not sure if the limiting magnitude is mentioned in any of these - all though this is metre class, so collects four times as much light as the ATLAS telescopes, the short exposure time suggests it will not be very dim. May be worth looking at the apparent magnitudes quoted in the minor planet centre observations where the asteroid was discovered by the Kiso observatory e.g. http://www.minorplanetcenter.net/db_search/show_object?object_id=2021+CC7 shows that it can detect down to at least apparent magnitude 17.4. There are also the following:
 * https://minorplanetcenter.net/mpec/K20/K20P97.html 17.6,
 * https://minorplanetcenter.net/mpec/K21/K21GC0.html 17.7,
 * https://minorplanetcenter.net/mpec/K21/K21KA0.html 17.9,
 * https://www.minorplanetcenter.net/mpec/K20/K20F60.html 17.9,
 * https://www.minorplanetcenter.net/mpec/K20/K20Q81.html 18.0,

The limiting magnitude seems to be around 18.0 in practice. EDIT this agrees with their slides see: http://www.mtk.ioa.s.u-tokyo.ac.jp/kisohp/NEWS/pr20190930/fig19en.PNG

There is an existing wikipedia page for the Kiso Observatory, however it doesn't seem to contain any information on the Tomoe Gozen instrument, so probably worth updating that page too whilst doing this one, including linking to the Tomoe Gozen page about a famous (fictional?) Samurai after whom the telescope was named. There is a long list of references to her in the "In fiction and culture" section of the article, but the telescope name is not cited. Suggest adding it to the list.

In a similar vein the Kiso observatory should be added to the world map showing planned / active systems (which is now a little out of sync with the article so could do with updating anyway).Rafflesgluft (talk) 12:53, 2 July 2021 (UTC)

Map update - thoughts?
Recently, the world map showing major surveys was updated to replace Kiso observatory with "Tomo-e Gozen" and "BATTeRS". This creates a problem as the article the diagram was created for (Asteroid impact prediction) refers multiple times to Kiso Observatory, rather than Tomo-e Gozen. The page for BATTeRS, seems to be referring to Bisei Spaceguard Center. Whilst this is a notable historic observatory, unlike Kiso it doesn't seem to be making any significant contributions to surveying the skies at present. My feeling is therefore that it might be better to revert the changes to the diagram. Thoughts? Rafflesgluft (talk) 12:50, 19 December 2022 (UTC)


 * Some more info: 1. the minor planet centre credits BATTERS with 445 minor planet discoveries, the last of which was in 2010 (there is also one discovery credited to the observatory "Bisei SG Center" by the minor planet centre, which was back 2007). As mention above, these are separate from the the Spaceguard discoveries and may make it worth mentioning in the main article for historic reasons, but seems not to belong in the diagram. 2. The name Tomo-e Gozen is the name of the instrument, not the observatory. As well as not matching with the article text, the observatories listed by the minor planet centre name it as Kiso, not Tomo-e Gozen. E.g. 2022 ET discovered earlier this year  Rafflesgluft (talk) 14:16, 19 December 2022 (UTC)

Sar2736
Apparently Entered over germany about half an hour ago. Will needed to be added once some RS notice.©Geni (talk) 00:56, 21 January 2024 (UTC)