Using the evolution of dimming regions to probe the global magnetic field topology

A new interpretation of the 12th May 1997 event

G. Attrill1, M. Nakwaki2, L. Harra1, L. van Driel-Gesztelyi1, C. Mandrini2, S. Dasso2, J. Wang3

1 MSSL, UCL

2 Instituto de Astronomia y Fisica del Espacio

3 National Astronomical Observatory, Chinese Academy of Sciences, Beijing

Coronal Dimming Regions

• Dimming seen as a decrease in intensity in both EUV and X-ray images.

• Dimmings appear relatively suddenly (10s minutes)

• Good correlation with CME events

• 1st observation by Skylab mission (1973-74): “Transient Coronal Holes” (Rust, 1983)

POSSIBLE CAUSES:

• Density depletion due to an evacuation of plasma

• Temperature variation

Link with CMEs

• “Double Dimmings” – footpoints of erupted magnetic flux rope?

• Mass studies

• Dimming typically observed 30 minutes before CME seen in LASCO

Motivation: Given the close relationship between coronal dimming

regions and CMEs; drive to understand magnetic nature of CMEs naturally requires the magnetic nature of coronal dimming regions…..

This Study • Analyse intensity of coronal dimming regions using 195 A data from SOHO/EIT.

• Measure magnetic flux in dimming regions using SOHO/MDI data.

• Compare to magnetic flux in the associated magnetic cloud.

• Probe magnetic character of coronal dimming regions

• Asymmetrical temporal evolution of two main dimmings

• “Unidentical Twins”!

• Extend overview using Yohkoh soft x-ray data.

• Present a model to explain the observational measurements

• Indirect calculation of the magnetic reconnection rate

Demonstrate that study of the evolving magnetic nature of coronal dimming regions can be used to probe the large-scale magnetic

structure involved in the eruption of a CME

Event of 12th May 1997

• Simple magnetic structure

• AR 8038

• LDE flare GOES class C 1.3

• Strong coronal wave signature

• Brightening along (& shrinkage of) north polar coronal hole boundary

• Filament eruption

• Full Halo CME

• Associated magnetic cloud reaches Earth 15th May 1997

Pre-eruption EIT on 11th May 1997 Magnetogram

Northern end of sigmoidal structure embedded in negative polarity

Southern end embedded in positive polarity

Base difference images

05:07 - 00:12 UT

Bright coronal wave front

05:41 – 00:12 UT

Maximum extent of the dimming

Defining the boundary of the dimming regions

A transient coronal hole is defined as a region where there is an intensity decrease from a “normal” intensity to a region with an intensity close to that of a coronal hole, so we set our contours to lie halfway between the intensity of an area of quiet Sun and the intensity of an existing coronal hole.

Potential source of error is selected base difference image:

Base difference images showing the main dimming regions. The left panelshows the base difference image (05:41 UT - 00:12 UT). The centre panel shows the base difference image (05:41 UT - 03:59 UT).

Selected Regions

Base difference image (05:41 UT - 00:12 UT) at the maximum extent of the dimming, showing the regions selected for analysis defined by our contour boundary method.

Temporal intensity evolution of the dimming regions

Temporal variation in EUV intensity (counts/pixel) - a tool that allows us to visualise the physical restructuring of the magnetic field topology.

Evolving dimming regions• The boundaries of transient coronal holes are constantly evolving

• Overlaid our contours from images at successive intervals throughout our dataset

Asymmetric temporal evolution of the unidentical twin dimmings

Fast expansion Slow Contraction

Evolution of the area of the two main dimming regions. The dashed lineshows the change in area of region 1 and the solid line shows the change in area ofregion 2.

The graph clearly shows the rapid expansion of the dimming regions andthe gradual recovery. Region 1 (the northern-most dimming region) starts to recover before dimming region 2, the (southern-most dimming region).

Region 1

(northern dimming)

Region 2

(southern dimming)

Time of maximum dimming

Intensity line profilesA change in the average intensity of a dimming region can be causedby two things: • An actual change in the intensity of the region• A change in the size of the region

Change in intensity of each pixel with respect to its location (rather than the averaged intensity)

Change in intensity WITHIN the dimming regions is clearly visible.

Increase in intensity is partly due to the contraction of the dimming region boundary, but there is also a contribution from some internal mechanism. Likely candidate is the

reconnection related process of chromospheric evaporation.

Measuring the magnetic flux within the dimming regions

• Method described in Mandrini et al. (2005) – filtering quiet sun flux for noise

• MDI significantly underestimates the magnetic flux. • Following the work of Green et al. (2003), we used the correction factors:

MDI full-disk image at 06:28 UT on 12 May 1997. The colour white indicates regions of positive polarity and black indicates regions of negative polarity.

The net flux is the sum of the magnitude of the positive and the magnitude of the negative fluxes

The “open” flux is the difference between the flux of the two polarities

Magnetic flux within each dimming region at the maximum extent of the dimming

Since the majority of small-scale mixed polaritiesclose in their direct vicinity, calculating the “open” flux provides an estimate of how much magnetic flux is potentially free for connection with magnetic flux outside of the selected region.

If the interpretation of Webb et al. (2000), is correct, so that the main dimming regions (1 and 2) do indeed mark the footpoints of the magnetic flux rope that erupts as the core of the CME, our calculation of the total net flux should be halved to account for the fact that we are measuring the flux through both footpoints, and so effectively measuring the flux twice. This yields a net flux of 2.02 x1021 Mx and is substantially larger than the “total linked flux” 1.0 ± 0.2 x1021 Mx result of Webb et al. (2000).

Our measurement of the open flux through region 1 is in agreement with that measured by Webb et al. (2000) for this event, but our measurement through region 2 is double their result. (They found the flux for region 1 to be -9 x1020 Mx and region 2 to be 1.2 x1021 Mx). From these measurements, a “total linked flux” of 1.0 ± 0.2 x1021 Mx is estimated by Webb et al. (2000).

Evolution of the quiet Sun magnetic flux in the 2 main regions from 04:50 UT to 14:59 UT defined by the evolving contours

• Only the magnetic flux measurements for the quiet Sun component of the regions are shown.• The active region component has been removed due to its uncertainty.• Region 1 shows an increase in flux, peaking at 05:41 UT and then a gradual decrease. • Region 2 shows an increase, peaking at the same time, before a slight decrease is indicated. • Region 2 does not immediately decrease sharply after its maximum (as does region 1), but rather remains at a plateau.

Building a detailed picture of the global nature ofthe event

2D intensity profile made along the thick white line shown in the left panel of this Figure. The centre panel shows the intensity change along the selected line profile with time. The right panel is a contour plot of the

intensity profile. The shinking of the north polar coronal hole and the brightening along the shrinkingboundary is clearly visible.

Calculation of the magnetic flux contained in the 15th May 1997 magnetic cloud

Total axial magnetic flux = 4.8 ± 0.8 x1020 Mx

Total azimuthal magnetic flux = 1.3 ± 0.6 x1021 Mx (assuming a 1 AU length)

The total magnetic flux contained within the cloud is calculated at 1.8 ± 0.7 x1021Mx (assuming a 1 AU length).

c.f. Estimate by Webb et al. (2000):

Axial magnetic flux = 7.35 x1020 Mx

Magnetic topology scenario

Using our analysis of the magnetic character and evolution of the dimmings, we propose a scenario which links the formation of the CME with the formation of the coronal dimming regions (TCHs).

• Webb et al. (2000) propose that the two main dimming regions of 12 May 1997 mark the footpoints of the flux rope that erupts to form the CME.

• Kahler & Hudson (2001) state that the formation and expansion of TCHs occurs by the opening of closed magnetic field and that the contraction must be at least partially due to magnetic reconnection. Interestingly, they suggest that larger-scale newly opened magnetic field does not re-close in the arcade or even in the vicinity of the neutral line, but with an independent source of open magnetic field of opposite polarity. They propose that such a magnetic field could be found inpreviously existing coronal holes.

Applying these ideas to our event suggests the following scenario with respect to the evolution of the global magnetic field topology….

Magnetic topology scenario

North Polar

Coronal HoleRegion 2

Region 1

Dashed lines represent the pre-event magnetic structure and the solid lines the post-event magnetic

structure. The hashed regions represent the main dimming regions.

Implications of our scenario

• The total open flux from dimming region 2 is measured as 2.12 x1021 Mx. • The total magnetic flux contained within the cloud is calculated as 1.8 ± 0.7 x1021Mx (assuming a 1 AU length).

Therefore the magnetic flux contained in the magnetic cloud measured at Earth on 15 May 1997 comes mainly from dimming region 2, the southern-most, long-lived, positive polarity dimming region.

The absence of bi-directional electron streams in this magnetic cloud is noted by Webb et al. (2000). In view of our new interpretation of this event, we suggest that this absence is most likely due to the source of the magnetic cloud coming only from the southern-most dimming region (one footpoint of the erupted flux rope), rather than jurisdiction for a questionable correspondence between bi-directional electron flows and such solar ejecta.

Interaction with the north polar coronal hole magnetic field effectively closes dimming region 1 (the northern most dimming region).

Thus dimming region 2 (the southenmost dimming region) that remains “open”, becomes the main source region of the developing magnetic cloud at the (solar) Western leg of the expanding CME.

Rotation of the Filament

Partial Filament Eruption

N-S alignment

E-W alignment

Webb et al. (2000) COUNTER CLOCK-WISE ROTATION

Rotation of Coronal “Wave” signature

Podladchikova & Berghmans (2005)

Entire structure of CME from core to “skirt” exhibits writhe as it erupts.

COUNTER CLOCK-WISE ROTATION

N-S alignment changes to E-W alignment

Webb et al. (2000)

S Flux rope

0 degree pitch angle electrons.

Electrons travel parallel to B from the (solar) West

Given the CCW rotation, solar WEST maps back to the Sun as SOUTH. Ie. South leg is connected.

Electrons travel parallel to B from the (solar) West. Ie. West leg is connected to the Sun, East leg is disconnected.

Indirect calculation of the magnetic reconnection rate

Assuming the validity of our proposed model, then the contraction of the negative polarity outer boundary of region 1 (O1) occurs directly as a result of reconnection with the positive open magnetic field of the north polar coronal hole (OCH). The rate of contraction for the north polar coronal hole boundary is measured as 2 km s−1.

Conservation of magnetic flux:

This scenario is derived from study of the evolution of the coronal dimming regions.

The total open flux from dimming region 2 is measured as 2.12 x1021 Mx.

The total magnetic flux contained within the cloud is calculated as

1.8 ± 0.7 x1021Mx (assuming a 1 AU length).