From: Matson, Robert (
Date: Tue Jun 06 2000 - 11:18:16 PDT

  • Next message: Ron Lee: "Re: Accurate Lat/Long Info"

    > hello,I have just completed my predictions for UARS and the following
    > UARS which by the way I predicted perfectly tonight at -1, any of the NOAA
    > satellites,DMPS satellites,Telstar, and Orbcomm satellites.  Even A TRMM
    > pass will be a predicted flare.
    I don't want to discourage you, Kevin, but two necessary assumptions
    for glint prediction are an accurate 3-dimensional model of the
    candidate satellite's specular surfaces, and accurate knowledge
    of that satellite's orientation at any time.  While the first
    is usually possible, at least to some degree of fidelity, the
    second almost never is.
    Take UARS for instance.  This satellite has an extremely complicated
    3-D structure, so your work is already cut out for you just doing
    the geometrical modeling.  But don't bother.  UARS' orientation is
    always changing depending on mission operations.  It would be a
    pointless exercise (no pun intended!) to attempt prediction of
    UARS glints since you won't have any information about which way
    the satellite bus is pointing.
    Even if you could do both of these things, magnitude predictions
    would still be very inaccurate.  You don't know the reflectivities
    of all the satellite surfaces, nor their bidirectional reflectance
    distribution functions (BRDFs), so you won't be able to predict
    the peak magnitude or the brightness function with angle.
    I know you'd like to predict "flares" from satellites other than
    Iridium, but you have to appreciate that the Iridium constellation
    is a VERY SPECIAL case.  The satellite pointing is accurately
    maintained and predictable apriori; this allows the BRDF to be
    crudely estimated after many dozens of observations.  But accurate
    glint prediction for these satellites was no small undertaking.
    I used hundreds of observations by over a dozen observers in the
    "early days" to create the Iridium MMA BRDF.  This BRDF has to be
    convolved with the radiance distribution profile for the sun
    (since the center of the solar disk is quite a bit brighter
    than the limb).  There are other factors like refraction and the
    earth's non-spherical shape that affect predictions, but these
    are minor problems to solve.
    Of course, the accuracy of the results made the whole exercise
    very rewarding for me, as tens of thousands of people worldwide
    have now enjoyed predicting and seeing their own flares, either
    through Chris's Heavens-Above site, my program, or Randy John's.
    Like I said, I don't mean to discourage you, but you're going to
    first want to stick to satellites that maintain specific
    orientations.  UARS and Hubble do not.  Orbcomm does, but not
    to the degree of accuracy that Iridium does.  From the Orbcomm
    Attitude Control: 
      - Nadir Pointed +/-5 degrees using active magnetic
        controls, reaction wheel and gravity gradient 
      - New generation lightweight Earth sensors and magnetometer
    +/- 5 degrees makes a huge difference in both the timing of a
    glint, and its maximum brightness.  This fact alone would be
    a pretty big deterrent for me.  But when I look at the satellite
    geometry relative to its deployed orientation, I see that no
    glints are even possible except during the daytime.  There are
    two circular solar panels, apparently deployed "clam-shell"
    fashion from either side of the disk-shaped satellite, and
    these are the only "large" specular surfaces you have to work
    with.  Since they're supposed to be pointing at the sun at all
    times, you can only receive glints on the ground during the
    I haven't checked TRMM, NOAA, Telstar or DMSP, by I suspect
    if I did I would find problems for glint predictions with all
    of them.  (TRMM is a bad choice anyway, since it is constantly
    being maneuvered).
    If you find a satellite that "behaves" properly, let me know,
    and I'll be happy to assist you in your efforts.
    Best wishes,
    Unsubscribe from SeeSat-L by sending a message with 'unsubscribe'
    in the SUBJECT to

    This archive was generated by hypermail 2b29 : Tue Jun 06 2000 - 11:15:45 PDT