Group B: Solar System

Participants

Alan Fitzsimmons, Stephen Lowry, Sam Duddy

Science Project

Asteroids and comets in AG

Much solar system science can be done by identifying images of known asteroids and comets in archived surveys. Science projects can include measurement of lightcurves, searches for cometary activity, and refinement of orbits for dynamical investigations.

How many modern surveys can be probed in this manner?

How do we generate positions?

How are those positions searched via AstroGrid?

An example project is below:

Pan-STARRS Precoveries

Project background

Current Discovery rate of NEOs is ~60 per month. This will increase by a factor 2-3 with PS1, and a factor >10 with PS4.

Initial orbits may only cover an arc of ~few days or 1 month. Task is to obtain "precoveries" i.e. images of the new objects years before they were identified by Pan-STARRS. Some will appear on DSS/POS plates, but current telescope archives are an unused resource at present.

Science goals

• Accurate orbit refinement to allow future predictions of position and close approaches.
• Search for cometary activity as found for 107P/Wilson Harrington (http://cometography.com/pcomets/107p.html)

To perform such work either orbit propagation must be inbuilt into astrogrid, or a suitable interface must be arrived at between astrogrid and an external orbit integrator (i.e. OrbFit, NEODyS, Horizons). Note that a problem with SkyBoT is that it does not seem to be able to produce a true ephemeris i.e. positions for a list of dates.

To do this in astrogrid is multi-stage process:

• Extrapolation of orbit and sky error bars back in time (do we need a cut-off date/sigma here?)
• Retrieval of optical/NIR images along the sky track

Meeting Notes

Notes below added by AnitaRichards:

SkyBot documentation (mostly in French but fairly obvious).

The main problem I can see is that most astronomical images only have the date hidden in the FITS header, if anywhere. We attempt to persuade data providers to give this information but it is not under the control of the VO!

Notes during meeting:

SkyBot positions appears accurate when propgated back a couple of decades.

Need to identify large scale surveys whose images contain MJDs or some form of date/time information.

Possible procedure:

• Know the time and area over which surveys were performed.
• Calculate position of asteroid over that span.
• If asteroid was within survey area during timespan of survey then:
  • Obtain date information for all survey images covering those positions (once per day over survey span?).
  • Then go through dates one at a time and see when corresponding image was taken.
  • Was object within 0.5 degrees of survey field? (note - need size of survey fields)
  • If a match occurs then: * extract precise time data was taken
    • re-calculate position and brightness of body for that time.
    • If body is definitely in field then report.
• IMPORTANT: If a match occurs then all images at that position should be returned to aid identification.

Apparently this could be done through a workflow?

Another possibility would be to generate an observing log for each survey at the start i.e. a catalogue of which positions were imaged at which time. This would be a valuable resource for other time-domain science.

2MASS sources in region defined by position/JD range below

J-band 2Mass image retrieved at this position using Montage service overlaid with SkyBot information for the epoch

Unfortunately the Montage images lose date information and the only other available 2MASS images are quick-look at degreded sensitivity - although that might be good enough. These do contain UT data and time information in the header.

It is probably not useful at the start to include much space-based instrumentation due to their small fov. Should only concentrate on large area sky surveys.

Also need limiting magnitude information and exposure times of survey frames. Magnitudes in different passbands can be calculated to first order but taking into account possible +/- 0.5 mags from rotation. Effective limiting magnitude of the survey frames will be brighter by a factor 2.5*log10[T. dtheta/dt], where T(sec) is the effective exposure time for the survey image and dtheta/dt is the rate of motion of the body in angular resolution elements/sec.

Overall the simplest method may be to simply define a conservative limiting magnitude in each pass band for each survey, and only search positions when the body has a predicted magnitude brighter than this limit.

* Large-area surveys with current date information*

A number of major surveys do have either observing logs or lists of extracted sources in the forms of tables containing RA, Dec and data (although not always in the same table nor in the same data format):

• 2MASS twomass_psc (column jdate)
• SDSS PlateX (column mjd)
• UKIDSS CurrentSpatialMap (column startString)
• INT-WFS dqc catalogue (column obstime)
• DSS2 (column epoch)

Test: 52 Europa is in the 2MASS survey on Feb 25th 2000, so here's the predicted positions. Can we find it?

Date RA Dec Vmag

2000-Feb-23 00:00 12 21 16.79 +05 54 20.0 10.79

2000-Feb-24 00:00 12 20 49.11 +06 00 41.9 10.77

2000-Feb-25 00:00 12 20 20.31 +06 07 08.2 10.75

2000-Feb-26 00:00 12 19 50.43 +06 13 38.6 10.73

2000-Feb-27 00:00 12 19 19.49 +06 20 12.8 10.71

In JD and decimal degrees:

2451597.500000000 185.31995 5.90556 10.79

2451598.500000000 185.20461 6.01164 10.77

2451599.500000000 185.08462 6.11895 10.75

2451600.500000000 184.96011 6.22740 10.73

2451601.500000000 184.83119 6.33690 10.71

Notes on OrbFit

Fortran90 Source code obtained from Orbfit

Along with this you will need to latest list of orbital elements from Lowell observatory called astorb.dat (This is updated with new objects and revised orbits once per 24 hours)

OrbFit is a multipurpose code that can do lots of things, to do something simple like generate the ephemerides for a specific object, check out example 1996RO13_ephem3 in Orbfit3.3.2/tests/orbfit/README.examples.