A particularly obvious example of daily changing background noise level

Constructing the BEST High-Resolution Synoptic Maps from MDI

J.T. Hoeksema, Y. Liu, X.P. Zhao, A. Amezcua Stanford University

Synoptic maps provide a global view of the solar magnetic field.  However, what constitutes the 'best' possible synoptic map

depends on the application. Traditional charts are compiled from data observed close to central meridian from magnetograms

observed over a 27-day solar rotation. But with higher resolution data, a whole new set of details must be addressed when

assembling such maps.  Small-scale features move and evolve on the time scale over which the maps are constructed.  The

fluctuating background noise level is comparable to the smallest features.  Projection effects and sensitivity variations of

different sorts complicate the effort.  And what about the polar field? (See accompanying poster by Liu et al.)  Parts of the Sun

cannot be seen.  We describe methods for assembling the best maps possible by accounting for image sensitivity, image scale,

corrupt pixels, differential rotation, geometric field projection, zero offset, varying noise characteristics of MDI

magnetograms, and polar field interpolation.

Process for Constructing Synoptic ChartProcess for Constructing Synoptic Chart

Announcing New Level 1.8 MDI MagnetogramsAnnouncing New Level 1.8 MDI Magnetograms

For each 96-minute MDI magnetogram, we correct forFor each 96-minute MDI magnetogram, we correct for

Uniform Zero Offset due to Shutter Jitter (Existing)Uniform Zero Offset due to Shutter Jitter (Existing)Bad Cosmic Ray Pixels (New, rare)Bad Cosmic Ray Pixels (New, rare)MDI saturation (Not yet implemented)MDI saturation (Not yet implemented)Calibration using UCLA rescaling mask (New)Calibration using UCLA rescaling mask (New)

In heliographic remapping processingIn heliographic remapping processing

Convert from Line-of-sight to Convert from Line-of-sight to radial radial (assumption).(assumption).Adjust Carrington mapping for differential rotationAdjust Carrington mapping for differential rotation Ulrich & Boyden (Solar Phys. 235, 17, 2006); Zhao et al., in prep, 2006/7)Ulrich & Boyden (Solar Phys. 235, 17, 2006); Zhao et al., in prep, 2006/7)

Radial Field Synoptic Chart ConstructionRadial Field Synoptic Chart Construction

Combine remapped radial magnetogramsCombine remapped radial magnetogramsSelect for uniform effective integration timeSelect for uniform effective integration time

(equivalent to 20 one-minute magnetograms)(equivalent to 20 one-minute magnetograms)Use data closest to central meridianUse data closest to central meridianStatistically test and eliminate cosmic ray corruptionStatistically test and eliminate cosmic ray corruption

Options:Options:

Polar field interpolation (See adjacent poster)Polar field interpolation (See adjacent poster)

Variance at Each Longitude Variance at Each Latitude

Top: Standard Synoptic Chart; Middle: Differential Chart; Bottom: Constant Integration Time Differential Chart

Variance depends on the noise level. In CR 2041, more 5-minute integrations contributed near 360 degrees, lowering the noise level at the left edge. Requiring 20-minutes of observations at each longitude lowered and made the variance more uniform.

Rows: Different areas of radial field synoptic chart CR 1988 shown in each row at high latitude and equator. Columns:Left: standard synoptic chart; some polar features blur.Differentially corrected mapping improves poles, center.Right column shows best equal integration time chart.

Saturation of strong field regions in MDI occurs because of the on-board algorithm for computing the field strength. Identification and approximate correction for saturation is possible using associated intensity images, but this correction not yet implemented (Liu, et al.)

Scale correction map for MDI magnetograms derived by Tran, Bertello, Ulrich, &Evans, (Ap J Supp, 145:259, 2005) from Mt Wilson observations. A smoothed version is used to calibrate individual magnetograms

Data from 96-minute magnetograms combine to make synoptic chart.Some 1-minute and some 5-minute observations are collected. Sensitivity and instrumental noise vary across the disk.

Calculation of open field foot-point locations with a potential-field – source surface model provides a sensitive test of large-scale data quality. The calculation is highly sensitive to way in which the polar field is interpolated (see accompanying poster by Liu et al.) WSO & MDI computations agreed well in CR 2029 in May, 2005.

Carrington Rotation 2055 – April 2007 RADIAL FIELD