Asteroid Zoo Talk

what happens after tagging?

  • LifeSVR by LifeSVR

    As soon as I tag an asteroid which is 'already known' It prompts so on the screen (but how do you know that if it says it's the first time someone sees these pics?)

    But what happens when someone does find an object that is new?

    Posted

  • nadirt by nadirt

    Yes, it would be nice to get an answer to this question. As well as to the question about what will happen to the newly discovered asteroid ? Any benefit for those who has discovered it ?

    Posted

  • Staceva by Staceva

    I hope the tagger would at least be informed if we managed to classify an unknown? I'd like to know if I'd found one.

    Posted

  • tashipoo by tashipoo

    I've been wondering too. I have several that were not labeled as "known". I'm sure it takes time--just wondering about the timeline.

    Posted

  • scibuff by scibuff

    I assume, once enough people tag a new asteroid, someone will have to manually go to the original data and take astrometric measurements manually (as the assumption is that the software missed it). The users who tagged the object then may appear on the official report to the Minor Planet Center (among the measurers) - but that's just a theory.

    Looking at the known objects and seeing that the last time they were observable (elongation > 60) was in Q4 of last year, unless the same object is then found by users in images from a nearby night (within ~7 days), the found object will simply be lost. At least two observation during a single opposition are required to attain the discovery credit but the hopes of a direct recovery during future oppositions is virtually zero without at least ~4 weeks of observations during the discovery opposition.

    It is far more likely, that (at least at this stage) new objects originally missed by the software will be used for machine learning, i.e. to find better parametrization / a metric for the software to automatically find moving objects in the data and to distinguish them from artifacts

    Posted

  • nadirt by nadirt in response to scibuff's comment.

    It is far more likely, that (at least at this stage) new objects originally missed by the software will be used for machine learning, i.e. to find better parametrization / a metric for the software to automatically find moving objects in the data and to distinguish them from artifacts

    Exactly...

    Posted

  • LifeSVR by LifeSVR

    Well then, as much as I understand your information, what's the point?

    I mean, the users find asteroids, which are already known, so I suppose they have been detected by the machine, too. Or at least confirmed on the pics by data and known positions.

    But if no-one confirms a tag by calculating the trajectory and making a confirmation observation, how can you teach the machine that it has missed an object? As far as this goes, you are merely teaching the machine to detect something that humans THINK it might be an asteroid...

    We could be discovering our asses off, and nothing is done with that? At least feed us data which will make it possible to make a confirmation observation and keep it interesting for everyone...

    As cool as it is to put on my resume 'tought a computer to detect asteroids in pictures' ... It is WAY cooler to have something out there with your name on it ;o)

    Posted

  • Emili_Sancha by Emili_Sancha

    one point to pay attention is that we only are searching in a older database of images, not in recent images ( at the moment ). If we find a known asteroid, i think this data is not a relevant because day by day there are amateurs around the world working with his telescopes. May happen that an amateur or professional astronomer finds a new asteroid, we (asteroid zoo) can steal the paterny of this discovery ( I think ) if we find it in older images. 😉

    In the other hand if we find an unkown asteroid the process would be...(I think)

    • Two users or more finds an unkown asteroid.
    • CSS confirms visually the find and determine preliminary orbit of asteroid.
    • CSS search in older and new images the presence of the asteroid. Refines orbit. If not, a new session with the telescope to search the asteroid. If positive...
    • CSS sends data to MPC for confirmation
    • If positive, MPC publish a circular with the new find and his orbitals parameters. Credits to CSS staff and the users who find it. The name of the object remains to CSS Telescopes.

    (sorry for my bad english)

    Posted

  • scibuff by scibuff in response to Emili Sancha's comment.

    1 - As I said in my comment, the data is now at least six month old and there's virtually no chance of direct recovery unless the same object is found in data taken within ~7 nights (and then at least once again, but ideally two, during the next lunation

    2 - Discovery credit is based on when the object is reported, if we then found it in data taken before the discovery night, that would be considered a "pre-covery", but does in no way affect the discovery credit

    3 - Overall, I think there is a much higher threshold for discoveries than 2 users

    4 - The Zoo stuff will no doubt be looking at objects found by the users here, but I doubt the work will go any further than machine learning - definitely, CSS has no interest in yet another regular main belt objects, of which is has discovered tens of thousands

    Posted

  • Dr.Asteroid by Dr.Asteroid scientist, admin in response to scibuff's comment.

    1. In this set of data the chance of a discovery are modest. We are looking to find out ways to reward folks. Keep in mind, this is territory that has never been used before - we're making precedent with the IAU - an organization that took a long time to define what a planet is. (or is not) Stay tuned. Everyone involved would like to be able to reward the best asteroid finders.

    2. Correct.

    3. We're working out how many users it will take. Right now we don't have a good idea of how you wonderful folks are at this project - since it's never been done with enough people to generate statistics. It's always been done more or less by one person (if that - mostly automated). So, probably more than 2 people, but we'll be submitting to MPC - and making sure that we get it right.

    4. Oh - not at all - there will be evolutions to the CSS algorithms based on the results here. While I agree, the discovery of yet another 1 km wide main belt isn't the science driver, being good at finding near earth objects will be really good at finding big stuff out in the main belt.

    Posted

  • Dr.Asteroid by Dr.Asteroid scientist, admin in response to LifeSVR's comment.

    I left a comment further down, we completely agree - the diligent folks should get an asteroid name. We're working out how to do that - the IAU is pretty slow to change and we're working to establish a precedent.

    Posted

  • Emili_Sancha by Emili_Sancha

    There are a similar project with citizens involved: http://www.laeff.cab.inta-csic.es/projects/near/main/index.php

    They report to MPC...

    Posted

  • Dr.Asteroid by Dr.Asteroid scientist, admin

    They do report to the MPC - it's the naming convention that is new. Typically the person who took the image gets the credit - but... this is a different situation.

    Posted

  • mboschmd by mboschmd

    Important Question: If only one or two people out of 10 see an asteroid on a set, does it get reviewed manually before being tossed out? Question is relevant because of the highly variable quality of the image sets.

    This question has been up for over three months without a response.

    We need the scientists to have presence on this project! We have good, relevant questions that need answering, and while we understand the difficulties created by the failed Antares launch and the lost satellite, you have a team of dedicated people working very hard for no compensation and are getting no support from the Scientists whose project this is.

    Posted