Vector Magnetic Field Synoptic Charts from the Helioseismic and Magnetic Imager (HMI)
Yang Liu Submitted: 2017-01-17 11:24
Vector magnetic field synoptic charts from the Helioseismic and Magnetic
Imager (HMI) are now available for each Carrington Rotation (CR) starting
from CR 2097 in May 2010. Synoptic charts are produced using 720-second
cadence full-disk vector magnetograms remapped to Carrington coordinates.
The vector field is derived from the Stokes parameters (I, Q, U, V)
using a Milne-Eddington based inversion model. The 180° azimuth
ambiguity is resolved using the Minimum Energy algorithm for pixels in active
regions and for strong-field pixels (the field is greater than about 150 G) in
quiet Sun regions. Three other methods are used for the rest of the pixels: the
potential-field method, the radial-acute angle method, and the random method.
The vector field synoptic charts computed using these three disambiguation
methods are evaluated. The noise in the three components of vector magnetic
field is generally much
higher in the potential-field method charts. The component noise levels
are significantly different in the radial-acute charts. However, the noise
levels in the random-method charts are lower and comparable. The assumptions
used in the potential-field and radial-acute methods to disambiguate
the weak transverse field introduce bias that propagates differently into the
three vector-field components, leading to unreasonable pattern and artifacts,
whereas the random method appears not to introduce any systematic bias.
The computed current sheet on the source surface, computed using the potential-field
source-surface model applied to random-method charts, agrees with the best
solution (the result computed from the synoptic charts with the minimum
energy algorithm applied to each and every pixel in the vector magnetograms)
much better than the other two. Differences in the synoptic charts determined with the
best method and the random method are much smaller than those from the best method
and the other two. This comparison indicates that the random method
is better for vector field synoptic maps computed from near-central meridian data.
Thus, the vector field synoptic charts provided by the Joint Science Operations
Center (JSOC) are produced with the random method.
Authors: Yang Liu, J. T. Hoeksema, Xudong Sun, Keiji Hayashi
Publication Status: Solar Physics, accepted.
Last Modified: 2017-01-18 11:12