Introduction
Next: Correcting Data - The Up: VLBA TEST MEMO 69 Previous: VLBA TEST MEMO 69
Introduction
A significant issue has been found with the Earth Orientation Parameters (EOP: UT1-UTC and pole position) used in VLBA correlator job scripts since May 2003. The correlator is supposed to use rapid service values of EOP, that are based on measurements, when they are available at the time jobs are generated. Instead, because of a software bug, it was using predictions from about 2 weeks before the observe date. The Earth's rotation rate and pole position are sufficiently unpredictable that this can make a difference of tens of milliarcseconds in the observation geometry.
The problem adversely affects projects that depend on an accurate correlator model. Such projects include those that use phase referencing, especially at high frequencies or with large separations between calibrator and target or with very stringent position determination requirements. Projects that use sources scattered around the sky to solve for the atmosphere (AIPS task DELZN (Mioduszewski & Kogan 2004)) can have the atmospheric solutions degraded. Also spectral line observations that use a line feature for phase calibration but depend on a calibrator to take out phase slopes (delay) will be impacted at a low, but perhaps significant, level. Projects that are not affected include those that depend primarily on self-calibration, such as imaging of strong sources. Also unaffected are geodetic and astrometric projects that use the total delays from the correlator.
While investigating this bug, it was realized that some projects get processed with predicted rather than measured EOP even when the system is working normally. This usually happens when correlator jobs are prepared before the rapid service EOP values are available in Bulletin A. For this reason, users with projects outside the time period in which the bug was in effect might want to check the EOP used for their project and, in affected cases, make corrections.
This memo starts with a quick recipe on how to correct data. Users who are not interested in the details, but who want to make corrections can read that section only, or even just the CLCOR help file in AIPS. Later sections describe the problem in detail and present test results that show the possible effects and show that the AIPS correction works.
Next: Correcting Data - The Up: VLBA TEST MEMO 69 Previous: VLBA TEST MEMO 69 Craig Walker 2005-10-06
Connect with NRAO