Meteor astrometry: what accuracy do we need?

Talk presented at the European Space Agency's 2010 Meteor Orbit Determination workshop.

Transcript

1. Meteor astrometry
   What accuracy do we need?
   Geert Barentsen - Armagh Observatory - gba@arm.ac.uk
   MOD 2010
   

   View Slide

2. ~ 90x50 deg
   Watec cameras
   Used to study meteors
   (PI: Tolis Christou)
   ~ 3x3 degrees (85mm f1.4)
   Starlight Xpress HXV-16 CCD
   (30s + 3s readout)
   Used to study long-term
   stellar variability
   1 pixel = ~0.001 deg = ~3 meter
   1 pixel = ~0.1 deg = ~300 meter
   The roof of Armagh Observatory ...
   

   View Slide

3. 22 November 2009, 00h57 UT
   It's the same meteor !
   

   View Slide

4. View Slide

5. → What accuracy do we need, to make a similar graph for
   meteoroid streams?
   3:2 resonance
   7:4 resonance
   

   View Slide

6. Perseids 2007 (ESA/RSSD Campaign)
   Large range in semi-major axis ...
   Poynting-Robertson or just errors?!
   109P
   Jupiter
   

   View Slide

7. 15 AU
   41.4 km/s
   10 AU
   41.0 km/s
   5 AU
   39.9 km/s
   Semi-major axis
   Speed at 1 AU Sun
   Semi-major axis is very sensitive to the velocity
   

   View Slide

8. (R
   1
   ,T
   1
   )
   (R
   18
   ,T
   18
   )
   (R
   2
   ,T
   2
   )
   (R
   3
   ,T
   3
   )
   (...)
   ➔ |v| = (R
   18
   -R
   1
   ) / (T
   18
   -T
   1
   )
   = 59.3 km/s
   ➔ |v| = (R
   16
   -R
   3
   ) / (T
   16
   -T
   3
   )
   = 58.7 km/s
   ➔ |v| = (R
   13
   -R
   6
   ) / (T
   13
   -T
   6
   )
   = 61.2 km/s
   ➔
   |v| = avg (R
   a
   -R
   b
   ) / (T
   a
   -T
   b
   )
   = 59.4 km/s
   Velocity is very sensitive to errors !
   R: Position vector
   T: Time
   

   View Slide

9. Fakeor simulations
   ●
   Fakeor = Fake meteor
   ●
   Input
   ●
   Radiant (ra, dec, velocity)
   ●
   Time + entry point
   ●
   Output
   ●
   Position of two observing
   stations with 100km baseline
   – Meteor at same distance from
   each station
   – Optimized convergence angle
   ●
   Synthetic astrometry from
   these stations
   

   View Slide

10. Uncertainty propagation
    Repeat 1000x
    (Monte Carlo)
    x
    x
    x
    x
    x
    x
    SPICE
    Astrometric error 1/a error
    We use the synthetic astrometry to recompute the (known) orbit
    Idea: quantify sensitivity of the semi-major axis by progressively
    adding errors to the synthetic astrometry.
    

    View Slide

11. Empirical result: O(error 1/a) ≥ O(error astrometry) − 5
    100
    10-1
    10-2
    10-3
    10-4
    10-5
    10-6
    10-7
    10-8
    10-9
    10-3 10-2 10-1 100 101 102 103
    Error astrometry (arcseconds)
    Error in 1/a (AU-1 )
    Result: error of 1/a in function of astrometry error
    (condition: optimal convergence angle)
    

    View Slide

12. Semi-major axis (a)
    What accuracy do we really need?
    This is what the Taurids might look like according to David Asher ....
    Taurids 7:2 resonance
    cf. Jeremie's presentation: we need ~0.1-0.01 uncertainty in semi-major axis
    

    View Slide

13. Observed distribution of Taurids semi-major axis in the IAU database ...
    

    View Slide

14. sigma(1/a) = 0.01 => Need astrometry better than ~10 arcminutes
    We add errors to the Taurids model to show the effect of poor orbits:
    IAU?
    

    View Slide

15. O(error radiant) ≥ O(error astrometry)
    What about radiants ?
    Error of radiant position in function of astrometry error
    

    View Slide

16. 4-rev
    9-rev
    Leonids 2001 model (McNaught & Asher)
    What accuracy do we need to separate the radiants of 2 Leonid trails ?
    

    View Slide

17. Need astrometry better than ~1 arcminute to separate trails.
    Increasing the number of meteors may not help in the case of large errors ...
    We add errors to the theoretical radiant distribution:
    

    View Slide

18. ~3 degree
    1 pixel = ~0.001 deg = ~3 meter
    ~90 degree
    1 pixel = ~0.1 deg
    Shower identification Trail identification
    

    View Slide

19. Conclusions
    ●
    At least arcminute-level accuracy is required to uncover
    the substructure of a meteoroid stream
    – In the case of optimized geometry
    – In the case of error-free centroiding
    – Not necessary to identify showers
    (e.g., great results from Sirko and SonotaCo)
    ●
    Further work: build a database of “synthetic meteor
    observations” to validate algorithms.
    – Allows to test trajectory algorithms and data quality parameters
    – May be used to create synthetic videos to test photometry and
    astrometry algorithms
    

    View Slide