Nadge MEUNIER, the BASS2000 team

1
MSDP DATA PROCESSING AT BASS2000

• Nadège MEUNIER, the BASS2000 team Pierre MEIN

MSDP Workshop, Tarbes, 18-20th January 2006
2
Data processing

• Objective to attract users !!!
• Data levels  clean spectra  Icont, V//, B//
  maps Stokes profiles vect B
• Steps
• Definitions of the data levels
• Collection of informations about the data
• Robustness tests of the code automatisation
• Implementation of the processing
• Development of a BASS2000 interface run of
  large quantities of data, additional formatting
• Processing and checks of the results

3
History

• Visit P. Mein october 2004 first definitions
• Purpose processing by BASS2000
• with almost no manual intervention
• Homogeneity
• Introduction in the on-line catalogue
• Choice of the processing options
• Only one standard mode (but variation of the
  code / parameter files depending on the year)
• Stokes profiles (I, Q, U, V) on 17 points in the
  line
• B//, V//, I maps bissector position at 2??,
  4?? from line center, sum and difference
  blue-red at 3?? (depending on the wavelength)
• 2 levels are saved in fits files before and
  after ? centering
• Complementary informations (co-spatiality)
• Choice of on-line jpeg images variable
  depending on ? and year (related to data quality)

4
The code

• Organization of the code (P. Mein)
• A sequence.par file one line per sequence and ?
• Parameters fixed (default)
• IDL routine / call of the fortran code
• Output files not fits files
• Intensive test of the code at BASS2000 on 2004
  and 2003
• Objectives
• Identification of the main problems
• Identification of the limits
• Modification of the fortran code by P. Mein to
  solve them automatically when possible
• Modification of the IDL code to avoid
  interruption and alloowing to go to the next
  sequence (example if bad naxis3)
• Remark Visual verification of the direction of
  the scan and B// with MDI (because no indication
  in the data headers) too difficult in some case
  (especially quiet regions) and long to be done
  by the data users!

5
BASS2000 interfaces

• Realization of a routine ot automatize the
  processing (file organization)
• Automatic creation of the sequence.par files (1
  per day) from the file listing and header
  informations (exposure time, time ) for the
  choice of calibration files for each sequence 2
  versions, one for 2004 (2 camera / wavelength)
  and other years (one camera per wavelength)
• Possible to start the code on the whole DLT
• Formatting of the output files
• New keywords in the headers (for the catalogue or
  for information)
• Standard units (m/s, gauss)
• Creation of the jpeg, fits files
• Creation of the html files (on-line quicklook,
  containing the jpeg and other links) and history
  of the processing information (parameter files,
  important outputs)

6
Make sequence.par
msdpauto
Rep auto
7
Results

• Visual verifications, iterations avec Pierre
• 3 years processed
• 2004 6 runs, 296 (280) sequences 130 ok
  (without counting 1 run, not received)44
• 2003 4 runs, 170 (155) sequences 17 ok for
  the 2 lines 33 ok 1 line only (includes bad
  seeing and clouds not kept in the catalogue) 22
• 2002 5 runs, 169 (153) sequences 48 ok 31
• Main problems SEE PRACTICAL SESSION !
• Exotic scans (prominences, bad scan keywords due
  to  manual  observation) not processed
• Bad file names (exchange between calibration file
  names)
• Bad data (low intensity level / dark current,
  problem in the polarization keyword)
• Calibration, geometry
• Interruptions du code
• Future
• 2000, 2001, 2005 to be processed

8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)