SHINE 2008

Vector Magnetic Fields from

the Helioseismic and

Magnetic ImagerSteven Tomczyk (HAO/NCAR)

Juan Borrero (HAO/NCAR and MPS)

SHINE 2008

From Juan Borrero

SHINE 2008

Basic problem is to infer properties of

magnetic field from observations of the

Stokes profiles.

Historically there have been two classes

of instruments.

Magnetographs and

Spectropolarimeters…

SHINE 2008

Magnetograph

Instrument with typically poor spectral

resolution and good spatial resolution. Few

samples across line profile.

Usually only Stokes V - LOS Flux Density

Examples: BBSO, GONG, KPNO, MDI

Pro: Fast Temporal cadence

Con: LOS Flux density, saturation (systematic

errors)

SHINE 2008

Spectropolarimeter

Instrument with typically good spectral resolution

and moderate spatial resolution. Many samples

across line profile. Usually spectrograph, I, Q, U,

V

Allow inference of vector field (B and geometry)

Examples: ASP, Hinode SP, IVM, SOLIS

Pro: Very precise

Con: Slow to acquire map (Hinode SP - 15

minutes, partial Sun), complicated and slow data

inversion

SHINE 2008

HMI - Imaging Spectropolarimeter

Full Sun image - 1 arcsecond resolution (4096x4096)

Full Stokes I, Q, U, V with 6 samples across FeI 617.3

nm line

Allows inference of vector magnetic field

Full magnetic information every 90 seconds - 10

minute cadence for standard data product

Precision is not as good as Hinode SP but very fast

SHINE 2008

SHINE 2008

HMI will have:

– Full Disk Coverage

– High Temporal Cadence

– Continuous Observations

● Evolution of magnetic structure of sunspots and active regions

● Magnetic shear accumulations and impulsive release

● Evolution of helicity in active regions

SHINE 2008

● Full Stokes I,Q,U,V allows inference of Vector Magnetic Fields

● More accurate velocities in the presence of magnetic fields through removal of Q, U & V to I crosstalk

– Flows around sunspots, Sunspot seismology

● More accurate flux determinations

– Total flux, as opposed to longitudinal

● True full disk magnetometry

– Limb fields, polar fields

Vector polarimetry with HMI

SHINE 2008

Observations

Initial guess model Spectral synthesisSynthetic Stokes

spectrum

Comparison

Final model

Model parameter modification

Yes

No

Stokes

Inversion

SHINE 2008

HMI inversion code

Very Fast Inversion of the Stokes Vector (VFISV) by

Juan Borrero

Provides inference of:Magnetic Field StrengthTotal Magnetic FluxInclinationAzimuth and ErrorsFilling FactorLine-Of-Sight VelocityThermodynamic Properties

Milne-Eddington radiative transfer

Initial guess with Neural Network and Weak Field

Approximation

Quick look and full inverted data product

Macroturbulence and damping not fit

Look up table for Voigt/Faraday functions

Works with high or low spectral sampling data

Very fast

SHINE 2008

How precisely will HMI and VFISV

infer the properties of the vector

magnetic field?Simulations by Juan Borrero

SHINE 2008

σB

(Gauss)

σInclination

(deg)

σAzimuth

(deg)

σVlos

(m/s)

Sunspot 4-5 0.2 0.2 20

Plage 10 1 1 30

Network 40-50 2-3 5-7 60-90

HMI Expected

Errors

We expect to see saturation starting at about 4000

Gauss

SHINE 2008

SHINE 2008

180º Ambiguity

Resolution

Is a fundamental problem

Growing confidence in

our ability to resolve it

HMI will provide resolved

vector field