eSDO 1121: CME Dimming Region Recognition

This document can be viewed as a PDF.
Deliverable eSDO-1121: CME Dimming Region Recognition
V. Graffagnino, A. Fludra
07 September 2005


A coronal mass ejection results in a ‘dimming’ of the hot (EUV and X-ray) corona as a result of the opening of magnetic field lines, the expansion of coronal material and the resulting density reduction that occur during a CME. In attempting to identify the source region of CME events, a number of studies have characterised the phenomenon of coronal dimming (see for example the reviews of Hudson and Webb 1997, Harrison and Lyons 2000). These studies have been based on two types of difference images in which dimmed regions appear as dark features with a reduced intensity. The two types are:

  1. Running difference images, which are obtained by subtracting the preceding image from the current one. These emphasise changes in the brightness, localisation and structure of sources that have occurred during the interval between successive images. However, artefacts such as spurious dimmings can occur if the intensity of a bright feature decreases.
  2. Fixed difference images, which are produced by subtracting a single pre-event image from all subsequent images. Any changes that occur during an event are particularly clear. However unless solar rotation is taken into account potentially spurious dimmings and brightenings can be produced since the pre-event reference image may have been taken hours before subsequent frames.

Previous Coronal Dimming investigations have made use of wide-band soft X-ray imaging, narrowband EUV filters and EUV spectra. AIA narrow-band filters are therefore expected to allow the detection of such events. The identification of a CME signature such as Coronal

Dimming is critical to fully investigate the underlying physics of the CME onset. The goal of this algorithm is therefore to identify regions of coronal dimming in order to carry out follow up investigations of subsequent CME events.


  • Multi-wavelength AIA images - 10 channels of full disk / low resolution images, full disk / high resolution images, and region-of-interest / high resolution images.


  • Processed images indicating CME Dimming Region locations in each of the filter wavelengths;
  • The output will be entered into the CME Dimming Region Catalogue where statistical data extracted from the images will be presented. This statistical data includes:
    • Size of the Dimming Region
    • Duration of the dimming,
    • Intensity variations during the Dimming
    • Wavelength in which event has been detected
  • FITS file produced with table extensions.
  • Catalogue access possibly via webservice/ASTROGRID

Test Data

  • TRACE and SOHO EIT images in multiple wavelengths

Tool Interface

  • commandline: input of AIA images, output of FITS files containing images and statistical data.
    1. AstroGrid CEA web service: this algorithm will be deployed as a CEA service hosted in the UK. The algorithm will run continously to generate the CME dimming region recognition catalogue, and users can call the web service to process datasets on the grid.
    2. SolarSoft routine: the C module will be wrapped in IDL and distributed through the MSSL SolarSoft gateway. Users will access to a SolarSoft installation can call the routine from the commandline or GUI to process locally held data.
    3. JSOC module: the C module will be installed in the JSOC pipeline. Users can access the routine through pipeline execution to operate on data local to the JSOC data centre.

Science Use Case

The user would like to obtain physical and statistical information on the properties of possible CME dimming regions observed in the AIA images at a number of wavelengths. The goal of the algorithm is to identify these dimming regions, both on disk and above the limb. Follow up investigations of these regions resulting in the actual identification of a CME event is left to the user.

  1. The user identifies a series of images from one or more of the AIA channels taken during the specified time period.
  2. The user inputs the flatfield AIA images to one or more of the automated CME Dimming Region recognition algorithms.
  3. The user specifies the constraints used to identify the level of dimming
  4. The algorithm runs and returns a FITS file to the user.
  5. The user can view an image within the FITS file displaying the location of the CME Dimming region.
  6. A fits table will also be available to allow standard statistical analysis to be carried out and also allows the user to carry out further analysis on each event(s) as desired.

Technical Use Case

The technical aspect of the CME dimming region recognition procedure has much in common with the Small Event Detection algorithms being investigated, so much code will be reusable. Both algorithms will use the methods of Berghmans et al. (1998) and Aschwanden et al. (2000); the primary difference is that while the small event detection algorithm will look for short-lived increases in localised areas, the CME dimming region recognition procedure will identify drops in intensity over large areas that last for tens of minutes up to several hours.

In comparison to the small event detection algorithm, the CME dimming region recognition code will also need to group together a much larger number of pixels that undergo simultaneous dimming to find the extent of the dimming region. The CME dimming region recognition processing can also use fixed difference AIA images in order to clearly identify regions of dimming. For on-disk areas, corrections to take into account solar rotation will also have to be applied to these difference images in order to reduce the number of spurious dimming regions that might be identified.

Modified Method of Berghmans et al. (1998)

  1. The automated CME Dimming Region recognition algorithm receives a series of AIA flatfield images and fixed difference AIA images as input.
  2. An average light curve for each pixel over a set time period is derived to define a background reference emission.
  3. CME dimming events are defined as pixels where intensity drops near-simultaneously and significantly below this background value.
  4. The spatial and temporal extent of the event is then defined.
  5. Relevant data are extracted for each event and tabulated.
  6. A FITS file is produced and returned.
  7. In automated runs, the CME dimming region recognition catalogue is updated.

Modified Method of Aschwanden et al. (2000)

In this method, a spatio-temporal pattern recognition code is used which extracts events with significant variability. This is achieved as follows:

  1. The automated CME Dimming Region recognition algorithm receives a series of AIA flatfield images and fixed difference AIA images as input.
  2. Each full image is rebinned into a number of macropixels.
  3. An event is then spatially defined to include a number of neighbouring pixels that undergo coherent time variability, within a certain tolerance limit.
  4. For each macropixel the time series is examined and the maximum, minimum fluxes and corresponding times are extracted and a flux variability defined as the difference is derived.
  5. These difference values are then ordered.
  6. The macropixel with the largest flux decrease is then chosen and neighbouring pixels examined for variability and temporal coincidence of the minimum of the lightcurve within the tolerance limit. This continues until no further neighbouring pixel is found which corresponds to the appropriate tolerance limits.
  7. This collection of pixels is then defined as an event and these pixels are marked so as not to be included in subsequent events searches.
  8. The remaining pixels are resorted in order of flux variability and the process is repeated. In this way a number of dimming events are defined, although it is expected that simultaneous, spatially independent dimming events will occur rarely.
  9. The relevent statistical information is derived.
  10. An output FITS file is produced and returned.
  11. In automated runs, the CME dimming region recognition catalogue is updated.

Quicklook Products

  • Entry in CME Dimming Region Recognition catalogue:
    • low resolution full disk image of the Sun indicating positions of dimming regions
    • table entry with relevant statistical data

Support Information

  1. K.P. Dere, G.E. Brueckner, R.A. Howard et al. Solar Physics. 175, 601-612 (1997).
  2. I.M. Chertok and V.V. Grechnev, Astronomy Reports. 47, 139-150 (2003).
  3. I.M. Chertok and V.V. Grechnev, Astronomy Reports, 47, 934-945 (2003).
  4. D. Berghmans, F. Clette, and D. Moses. Astron.Astrophys. 336,1039-1055 (1998)
  5. M.J. Aschwanden, R.W. Nightingale, T.D. Tarbell, and C.J. Wolfson, Ap.J. 535,1027-1046 (2000)
  6. Hudson, H.S., Webb, D.F., in: Coronal Mass Ejections, Geophys. Monograph Ser., AGU. 1, (1997).
  7. Harrison, R.A., Lyons, M., A&A, 358, 1097, (2000)

-- ElizabethAuden - 08 Sep 2005

This topic: SDO > PhaseACMEDimmingRegionRecognition
Topic revision: r6 - 2005-09-30 - ElizabethAuden
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2020 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback