create new tag
view all tags

AIA Meeting Minutes

Agenda: eSDO and AIA
22 June 2005
UCL Cruciform Building, B.4.01

10:00 Welcome / coffee EA
10:15 eSDO algorithms all
11:00 Short break all
11:15 Pipeline operation NH, KS
11:45 AIA data products NH, KS
12:15 US / UK data centres all
12:45 Future meetings EA, NH, KS
13:00 Finish  

eSDO Algorithms

Image Processing Helioseismology
Loop recognition Mode frequency analysis
Magnetic field extrapolation Mode asymmetry analysis
Helicity computation Subsurface flow analysis
CME dimming region recognition Perturbation map generation
DEM computation Local helioseismology inversion
Small event detection Solar rotation profiling (time permitting)
  Oscillation source analysis (time permitting)

  • Image processing algorithms discussed:
    • Loop recognition - Inclusion in pipeline if successful, particularly if automatic detection works in the absence of magnetic extrapolation. Develop OCM algorithm (dependent on mag extrapolation) as fallback and offer as service on a UK machine.
    • Magnetic field extrapolation - inclusion in pipeline if non-linear magnetic forcefield extrapolation is successful.
    • Helicity computation - will be included in non-linear force field magnetic extrapolation (if that is successful). Inclusion in Lockheed pipeline in non-linear force field mag. extrapolation is successful.
    • CME dimming regions - inclusion in Lockheed pipeline
    • DEM computation - Multiple solutions could be made available through AIA pipeline. Offer as service on a machine in the UK, possible inclusion in Lockheed pipeline.
    • Small event detection - run continuously over 30 minute intervals on machine in the UK; store results in UK data centre. Might be offered as an AIA service at Lockheed.

Pipeline Processing

  • Processing location of eSDO image processing algorithms: Stanford or Lockheed?

The pipeline at Stanford will automatically produce level 1 products for AIA and HMI and level 2 / 3 products for HMI. The processing software at Lockheed that produces high level AIA products will be mostly automated with some human interaction.

  • Obtaining copies of necessary modules for development pipeline at MSSl

Technical details of setting up a development pipeline at MSSL (for work on eSDO algorithms) will be discussed further in September.

AIA data products

  • Data products available as inputs to eSDO image processing algorithms

Level 1 products will be available as inputs. Suggested use of TRACE, SOHO-MDI and artificial data as test input; creation of artificial test data will be discussed at joing AIA / HMI meeting in Feb. 06. Level 1 inputs:

    • Data catalogue (possibly event catalogue too)
    • full flatfield images in all wavelengths
    • full resolution 500 by 500 active region movies
    • quicklooks: thumbnails and low resolution versions of flatfield images

  • Cadence and size of AIA high level data products

Total size of AIA data products per year: 20 TB. The following high level products and approximate cadence will be produced:

    • thermal maps - 4 to 10 images / day
    • DEM measure - 4 to 10 images / day
    • irradiance estimates - string of numbers several times / day
    • potential field magnetic extrapolations - image 4 times / day
    • non-linear force field magnetic extrapolations - depending on complexity, image 1 / day

US / UK data centres

  • AIA high level data products archived at Stanford, at Lockheed, or at all?

AIA high level products listed above will be archived permanently online at Lockheed.

  • Grid users’ method of generating high level data products at US data centre

Not discussed - access to online data products will be through the VSO portal (possibly AstroGrid web service as well; alternatively, AstroGrid access may be through VSO)

  • UK data centre considering archive of some high level products & movies of filtergrams

No user traffic model for SDO yet, but experience with TRACE shows an increase in data interest over lifetime of mission. Disk cache of SDO data centre in UK will take pressure off US data centre.

Future meetings

  • Elizabeth Auden in Palo Alto September 6-9 - visits to Stanford and Lockheed will be scheduled.
  • AIA team meeting – rescheduled for winter 2006? - eSDO developers will attend meeting if attendance at relevant breakout sessions is possible for specific algorithms (so far, possible breakout sessions for DEM, loop recognition and magnetic field extrapolation)
  • LWS 2007 – SDO presence - eSDO will have a presentation slot for algorithms / data centre demonstrations

-- ElizabethAuden - 29 Jun 2005

Edit | Attach | Watch | Print version | History: r2 < r1 | Backlinks | Raw View | More topic actions
Topic revision: r2 - 2005-06-29 - 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