Performance of the Michelson Doppler Imager Instrument on SOHO

P Scherrer, R Bogart, R Bush, T Duvall, J T Hoeksema, A Kosovichev, J Schou
 W.W. Hansen Experimental Physics Laboratory, Stanford University
M Morrison, T Tarbell, A Title 
 Lockheed Martin Palo Alto Research Laboratory

Launched on SOHO in December 1995, the MDI instrument took its 10
millionth filtergram in early April, 1997.  The instrument and
spacecraft have performed admirably since commissioning, providing over
a year of virtually uninterrupted time series of velocity and intensity
measurements at moderate resolution, a continuous 60-day time series of
full disk 4" velocity and line depth maps, monthly 72+ hour time series
in various observables, a host of daily 8-hour campaigns, and full-disk
magnetograms every 96 minutes throughout.  Another uninterrupted 90-day 
interval of nearly full data recovery is scheduled to be completed in 
mid July.  Various scientific results using MDI data are being presented 
at this meeting.

About a dozen terabytes of data sets have been created and archived and
normal pipeline processing is now completed soon after retrieving the
data, typically less than a month after the observations are made.
Most of the data products are generally available on the WWW, see
http://soi.stanford.edu.  Selected data are available in near real
time.  The SOI team welcomes collaborations.

Routine and extraordinary calibrations along with analysis of
scientific data sets allow us to make good estimates of the noise and
understand many of the sources of systematic errors in the instrument.
In almost every respect the instrument performs as well or better than
expected before launch, the primary limitations being photon noise on
the short term and fixed or slowly varying offsets on the long term.
We have found that the Michelsons are somewhat more sensitive to
operational temperature variations than was expected, adding some
additional constraints on our observing sequences.