Notes on hmi.ME_720s_e15w1332 ([http://jsoc.stanford.edu/ajax/lookdata.html?ds=hmi.ME_720s_e15w1332 click here]) and hmi.ME_720s_e15w1332_HARP ([http://jsoc.stanford.edu/ajax/lookdata.html?ds=hmi.ME_720s_e15w1332_HARP click here])

These two series contain all ME inversion data processed by means of VFISV of the version ([:VectorPaper:details]) that we think is best as of February 23, 2012.

Our efforts on optimizing the VFISV in the HMI data pipeline are on going, and we will create the series to fill the data processed by the newer codes when ready. We will announce the release of the new data through the Solar News.

• temporal coverage chart of these two series : Charts in [http://jsoc.stanford.edu/doc/data/hmi/coverage_maps/inversion/ this directory] show the 12-minute time slots for which the ME inversion data processed with the VFISV module are available or not at these series. The coverage chart also contain the status flags of the HMI observation thus the quality and availability of input Stokes data. See [:ME720se15w1332CoverChart: details of the coverage charts].

• hmi.ME720s_e15w1332 : For full-disk. Synoptic data, on 2 per day basis at 7 and 19UT daily, are stored in this series. The latest version (Oct. 14, 2011) of VFISV is used to process the full-disk input Stokes data mainly since October 1st, 2011. The latest version of VFISV is being used by the current near-real-time run that is filling this series using the latest definitive Stokes data, on the 2 per day basis, usually a week after the observation. An older version (May 26, 2011) was used to process the Stokes data from May 2010 to September 2011, however, the latest version was used to process data for some of this 17-month period. Checking the version of VFISV codes is strongly recommended.

• hmi.ME720s_e15w1332_HARP : For active region(s). The latest version (Oct. 14, 2011) of VFISV is used to generate the ME data stored in this series. Many of the active regions processed are recognized automatically by [:HARPDataSeries:HARP module], while some are determined in different ways. This series covers many of the ARs that created the X-class flares in 2011, including the one occurring on 2011 February 15 (at NOAA AR11158), with the 12 minutes (720 seconds) cadence. The ME data stored at this series were used for preparing [:ReleaseNotes:the test data release] of the 3-component disambiguated vector data for the AR 11158.

• version of VFISV in the FITS files exported at JSOC database.

  • Stored as a string value of a keyword INVCODEV, or a header card of the FITS file.

  • On IDL, for example, the version will be printed with sxpar,

    IDL> img=fitsio_read_image('img.fits',header)
    IDL> print, sxpar(header,'INVCODEV')

  • The ME data at the two series were processed with one of three versions, labeled with a date, "May 26, 2011", "Oct 14, 2011" or "Nov 27, 2011".

• differences between 2011-May-26 version and 2011-October-14 version

  • VFISV core part
    1. Some of initial guess values for the iterative residual minimization were changed.
    2. Derivatives of residual square-chi is calculated with the best-fitting parameters instead of those at the last iteration stage.
  • Other change in the JSOC context
    1. The 32-bit quality flag for each pixel (segment name : qual_map) had been redefined. This change is adjustable without rerunning VFISV, thus, the published data at the two series have the quality flag map given with the latest definition.

• differences between 2011-October-14 version and 2011-November-27 version

  • The changes were limited to the part of JSOC data interface and MPI parallelism thus did never affect the output data values.