MDI Level 1.8 Magnetograms

1
MDI Level 1.8 Magnetograms

• Yang Liu
• Stanford University

2
MDI magnetograms
sa F1 - F3, sc F2 - F4, sum sa sc, a
sum / sa, or a sum / sc, if sum 0. a
velocity B_losc(veloci
ty_lcp - velocity_rcp)
15-bit lookup table
F1
F3
F2
F4
3
MDI on-board computation for Doppler velocity

• saF1-F3, scF2-F4,
• sasa24, scsc24,
• sumsasc.
• sa (sc) reciprocal
• pseudo_asumreciprocal
• pseudo_a pseudo_a22.
• pseudo_a a
• a Doppler velocity

Produce this 15-bit recip-lookup
table tablebig_number/sa big_number3224 if
table gt 32767 then table 32767 Thus, when sa
(or sc) is lower than 96, the table saturates.
The reciprocal for this sa is simply set to be
32767.
15-bit recip-lookup table
extract high 15-bit value
15-bit lookup table
4
Calibration and corrections

• Offset using a Gaussian function to fit the
  distribution of low density pixels of a MDI
  magnetogram. The shift of the gaussian center is
  deemed to be the offset of this magnetogram. The
  offset is subtracted.

5
Calibration and corrections

• Offset

6
Calibration and corrections

• Saturation measurement is saturated in some
  sunspots umbrae.

7
Calibration and corrections

• Saturation correction.

8
Calibration and corrections

• Saturation correction using I-B relationship
  log(B)ablog(I/Ic)

Period a b
1996-1999 2.9 -0.426
2000-2000 2.9 -0.326
2001-2001 2.9 -0.402
2002-2002 2.9 -0.307
2003-2003 2.9 -0.291
2004-2004 2.9 -0.275
2005-2007 2.9 -0.430
9
Calibration and corrections

• Saturation correction one example in AR10882
  near the disk center at 1736 of 2005 November
  18.

10
Calibration and corrections

• Under-estimate of magnetic flux corrected by
  multiplied by a rescaling factor derived by Tran
  et al. (2005).

11
Remaining quirks

• Non-uniform noise over solar disk.

Distortion difference for LCP and RCP images. It
allows granulation and oscillation to leak into
the magnetic signal as noise.
12
Remaining quirks

• Magnetic field reversal during some major flares.

Field reversal during an x-class flare
13
Remaining quirks

• Cosmic rays.

Cosmic ray
14
Comparison with other observations

• Comparison with magnetograms taken by SOLIS, MWO,
  and WSO.

15
Comparison with other observations

• Comparison with WSO magnetograms.

16
MDI magnetograms in cycle 23

• Interplanetary magnetic field versus mean field
  they are correlated (cc0.79 after removing solar
  cycle trend).

17
MDI magnetograms in cycle23

• Sunspot areas, total flux, and unsigned mean
  field after removing solar cycle trend magnetic
  field in sunspots doesnt make significant
  comtribution to IMF.

Sunspot area vs mean field (cc0.03)
Sunspot area vs total flux (cc0.60)
Mean field vs total flux (cc0.23)
18
Conclusions

• MDI level 1.8 magnetograms have made following
  corrections
• Zero offset
• Under-estimation of magnetic flux density
• An empirical method is proposed to correct
  saturation.
• MDI level 1.8 magnetograms still have following
  issues needed to be addressed
• Non-uniform noise level over the disk
• Magnetic field reversal during some major flares
• Cosmic rays.
• From the new MDI magnetograms from 1996 to 2008,
  we found
• a strong correlation between solar mean field and
  interplanetary magnetic field
• a correlation between sunspot area and total
  flux, indicating magnetic field from sunspots
  make a significant contribution to the total
  flux
• no significant correlation between solar mean
  field and sunspot area, indicating that most IMF
  comes from areas outside of sunspots.