Asteroid Zoo Talk

item from NASA NEWS...

  • stonepenny by stonepenny

    where does this leave Asteroid Zoo?
    Asteroid that flew past Earth today has moon! Asteroid 2004 BL86 passed Earth today at 16:19 UTC (11:19 a.m. EST). This "movie" of the asteroid was generated from our radar data. Twenty individual images were used. Radar observations allow scientists to better measure an asteroid's size, rotation and location. As a bonus, radar observations of 2004 BL86 detected a tiny moon orbiting the asteroid. Details:


  • hightower73 by hightower73

    well, i think it leaves us in the " wasting our time " department.

    Its taking so long to sort out what are good canidates, what are asteroids and what are flying shopping trollies that at this rate, well discover a N.E.O long before the data is processed and sorted out that well all be dead.

    sorry to be a defeatist, but its infomation like this that makes our little community here seem worthless and something to giggle at in quiet corners.

    we should be taken seriously and it seems as though were here to do this task and save some computer power in some universaity somewhere and save the sciencetists some money.

    when are we going to be taken seriously and publish some results of all our work??? instead of treating us like mushrooms???

    are we simply time and money savers or are we sciencetists in our own way looking through this data???


  • Dr.Asteroid by Dr.Asteroid scientist, admin

    The entire goal of AsteroidZoo is to understand where can we use human perception to find asteroids that the computers miss. We have specific examples known that cross behind stars or do odd things near frame edges.

    Once we have the data where humans catch things the computers miss - that might help us improve the observatories' programs to catch everything. Plus we'll find some new asteroids.

    I can absolutely confirm I'm waiting on feedback from the Minor Planet Center. When I have the official word, I will post the results - and we'll be moving forward form there.

    One of the things we're figuring out - and have to figure out run time is how people are different than computers in terms of abilities to sort through noise. We're getting a much better idea of what you (the AsteroidZoo community) are better at. And this will help inform what the most logical next step will be. We didn't know ahead of time where the most scientifically important differences would be.

    My goal is to understand the best way for the community to improve the state of the art - should we look through data looking for asteroids that were missed? Should we send in confirmation data (the computer isn't sure there's an asteroid here - can you help? - that would be the question)

    Thank you for your time.


  • Barbalbero by Barbalbero in response to Dr.Asteroid's comment.

    Thank very much for the reply.
    Let us hope to have good news from the Minor Planet Center soon.

    As the people involved in this project know, in the images we are studying it happens we find known asteroids, we observe moving objects which are potentially unknown asteroid or we do not observe nothing except the stars and galaxies or the images.

    For the set of images with known asteroids, I think there is no need to make a selection of them: we already know there are known asteroids there. These images can be used to obtain more information about the asteroids observed, so I think they can be sent to the Minor Planet Center without problem so we can have hopefully good news and some nice results.

    For the images with unknown asteroids, the situation is a bit different. I studied more than 50000 images up to now, sometimes the presence of a moving object is clear, sometimes it is difficult to understand if there is a movement of an object or it is just noise. I suppose it is possible to make a selection of the images, and after the first reply of the Minor Planet Center we can send another set of images with the evident presence of something interesting.

    My advices are these:

    1. The images with known asteroids can be sent, so the Minor Planet Center can also see the good job we are doing and we can start to have some results about these asteroids (we know sometimes we need more than one set of images of an asteroid to have exact information about its physical properties and its real existence)

    2. It is possible to sent images where it is almost sure there is something interesting and moving. I marked almost 1000 images where there are interesting objects (known and unknown), for some of them I know it is difficult to understand if it is noise or a moving object, other times it is clear we are observing real objects. For example, in this image:
      it is clear there is what I suppose is a comet. 10 different people already marked its presence, so I think this image can be sent for analysis to the Minor Planet Center without problem, and the same for similar images.

    I will wait for your comments


  • grums by grums

    Yes, as you say Barbalbero, some asteroids are obvious and unlikelely to be anything else where others may be very doubtful. I tried to address this by systematising the observations in "Classifying Asteroid Observations" in the "Objects" section to enable clear asteroids to be treated with greater priority than ones that may be a selection of background noise fluctuations or that may, indeed, be a dim asteroid. I got the impression that Dr A. preferred the wide spread (no pre-processing by the crowd) though I am unclear what, if any, preselection is done after our submissions. I would guess it maybe having multiple persons finding the same asteroid but I don't know. If all of our selections are treated equally (including the flaky ones which may or may not be noise) then the hit rate will be quite low, which seems a shame and a waste of someone's time. I am not saying don't look at these, but just prioritise the obvious ones.

    Dr A., I think it is reasonably well understood how humans (and even other animals) can discern patterns (and moving patterns) based on many Psychology experiments. And actually also why humans see patterns where there are none (canals on Mars is a famous example). And, having played with these image sequences myself, I can also appreciate why it is hard for a computer program to pick these things out but, in some cases, how this could be improved. I note there are also quite a few academic papers on this. As a vast amount of the problem is due to the atmosphere, satellite imagry would be a massive, massive improvement; but this, I appreciate, is expensive and, to do it well and quickly, is probably beyond your present budget. Even a single, good CCD detector array without too many defects would cost over $250k let alone a reasonable (preferably fairly wide aperture) telescope, steering arrangement, temperature control system, communication and control systems, power systems etc. etc., plus the minor problem of getting the whole payload into orbit. Demonstrating some quicker success at this exercise may be a route to getting more funding; I don't know, but the board meetings have to be interesting 😃.