Solar Image Processing A Multiscale View C Alex

Solar Image Processing: A Multiscale View C. Alex Young NASA/GSFC UCLA - IPAM (January 28, 2004)

Outline • My motivations - where I got interested in multiscale methods. • A gallery of data. • A standard analysis - that surprised me. • My approach - an attempt at something better. • Some other problems.

Outline • My motivations - where I got interested in multiscale methods. • A gallery of data. • A standard analysis - that surprised me. • My approach - an attempt at something better. • Some other problems.

The high energy sun. 20 Jan 2000 Solar Flare BATSE and COMPTEL (yellow) 1 -30 Me. V image and 100 ke. V /1 -30 Me. V lightcurves

The EUV/radio sun. Flare in EUV 171 Å pass band Flare in radio

Outline • My motivations - where I got interested in multiscale methods. • A gallery of data. • A standard analysis - that surprised me. • My approach - an attempt at something better. • A suggestion by J. Starck • Some other problems.

From the photosphere to the corona.

The visible sun.

Sunspot evolution.

Swedish Vacuum Telescope.

Solar flare quake.

Solar oscillations.

Oscillations in the MDI high-res. f. o. v.

Helioseismology.

Sub-surface structure.

Backside imaging.

Longitudinal Magnetic field

304 Å Sun

195 Å Sun

195 Å difference image of an EIT wave

Solar flare in UV cont. and 2 EUV passbands seen by TRACE

Filament eruptions in TRACE 171 Å passband

Magnetic loops in TRACE 171 Å passband

A flare in TRACE 1600 Å cont. and 171 Å pass band

Solar tornados

Largest flare ever recorded in 171 Å pass band.

The outer corona seen in white-light by LASCO C 2 2 -15 solar radii

LASCO C 3 7 -30 solar radii

Active sun seen in LASCO C 3

Outline • My motivations - where I got interested in multiscale methods. • A gallery of data. • A standard analysis - that surprised me. • My approach - an attempt at something better. • Some other problems.

TRACE 195 A

Images at the time of the first brightening in the TRACE movie. (Gallagher 2003) Top panels: TRACE 195 Å difference images created by subtracting each image from a frame taken at 00: 42: 30 UT. Bottom panels: LASCO C 2 and C 3 images showing the similar morphology of the eruption as it propagates away from the solar surface. (Gallagher 2003)

A Standard Analysis

Outline • My motivations - where I got interested in multiscale methods. • A gallery of data. • A standard analysis - that surprised me. • My approach - an attempt at something better. • Some other problems.

scale = 2 scale = 4 scale = 8 scale = 16

A set of edges from 00: 46: 34 UT to 01: 58 UT at 2 scales • The top image shows the 30 multiscale edges at scale 8 (green) and 16 (red) over the image at 00: 46: 34 UT. • The bottom images zoom in on the fronts.

Following a front in TRACE.

Outline • My motivations - where I got interested in multiscale methods. • A gallery of data. • A standard analysis - that surprised me. • My approach - an attempt at something better. • Some other problems.

high-energy solar flare imaging

Sunspot Classification Multifractal measure ? Shapelets ?

EIT Calibration calibration lamp for flat field and filter grid. Wavelets, Ridgelets, and Curvelets representation?

EIT 284 Å images Cosmic Rays, Bright Points, Coronal Holes.

Large increase in data volume. • • Data from just EUV imaging 1 passband x 136 images x (1 Mb) = 136 Mb 4 passbands x 1440 images x (16. 6 Mbs) = 95. 6 Gbs Data from the entire spacecraft increases from ~1 Gbyte/day to ~1 Tbyte/day

The sun in STEREO Tomography in space, reconstructing a 3 -D view with two spacecraft