Moving Object Update
Rob Jedick
Wri May 30th - Moon Room - 2:00pm


During this meeting I would like to have a discussion on progress made and
directions to be taken on the detection, linkage and orbit determination of
moving objects in PanSTARRS.  Please read my 'Moving Object Roadmap'
distributed after this email and be prepared with comments and suggestions.
The roadmap is intended merely as a preliminary stab at major issues
associated with moving objects.  I still need to develop details, time
lines, personnel requirements, etc.

In particular I'd like to discuss the relationship between transient
stationary and transient moving objects.  Ken Chambers and I had a
discussion yesterday which I think raised a couple interesting issues:

1) If PanSTARRS is to live up to its moniker of a Rapid Response system,
presumably mostly in relation to the near real-time detection and subsequent
world wide notification of an optical transient, we need to develop a means
of differentiating quickly between transient stationary and moving objects.
Differencing the current image with the static sky image detects transients
but does not distinguish between stationary and moving objects unless the
exposures are long enough to significantly trail the asteroid images.  Near
asteroid stationary points the confusion between stationary and moving
transients could be high in even longish exposures.  Distant asteroids don't 
trail nearly as much as the NEOs and Main Belt asteroids.  Since roughly 1/3 of
the PanSTARRS surveying will be in 30s solar system mode it would be
difficult to distinguish between stationary and moving transients unless
  another image is obtained in relatively quick succession.  The desire for
Rapid Response pushes PS towards successive exposures obtained as near in
time as possible of O(minutes).  The desire for a good velocity vector for
moving objects indicates a longer time baseline of O(tens of minutes).  The
actual amount of time required between exposures will depend on the location
of the survey region with respect to opposition which determines the mean
rates of motion of asteroids in the field.

2) People interested in transient phenomenon are interested in using data
obtained during non-solar system survey modes of operation.  e.g.  Asteroid 
science
will benefit from obtaining asteroid colors and also more astrometric
measurements of asteroids during other science objectives.  e.g.  Optical
transient searches will benefit from the increased area available through
use of the other search programs.  Asteroids could be detected during long
exposures through their non-PSF trailed shapes but I am concerned about the
efficiency of detecting severely non-PSF trails.  Transient stationary
phenomenon could be identified through their PSF-like profiles in static-sky
differenced images.  But this ability will be sky-plane position dependent
due to nearby asteroid motions appearing to be almost stationary and distant
asteroids being nearly stationary in even relatively long exposures.  Thus,
in order to utilize other science objective's searches for transient
detections I/we propose that all science objectives and the pipeline adopt a
process whereby images are divided in time to allow the detection of
transient phenomenon.  We suggest that every filter be associated with a
'Base Time Exposure' where sky-noise dominates read noise and that no single
exposure be longer than the BTE so that addition of successive images,
perhaps even taken with a longer time baseline between them, allows
transient object detection.

If there is time ;) I would like to highlight some recent progress in:

i) researching available orbit/ephemeris generation software (Heasley &
Jedicke)

ii) proposals to obtain CFHT Megacam G-filter 30s images for testing
transient object detection (Heasley & Jedicke)

iii) preliminary tests of a general technique to link detections of objects
on multiple nights. (Jedicke)

iv) determining orbital element accuracy for objects detected in a single
lunation. (Jedicke)