This package: (i) Applies new time-normalization processing to time series of images (Morgan & Hutton, ApJSupp, under review 11/2017). This processing effectively reveals extremely faint motions, oscillations and disturbances in chromospheric and coronal data. It has not yet been tested on photospheric images. (ii) Applies a modified Lucas-Kanade optical flow method to the time-normalized images to gain a vector field, in the image plane, of the dominant directions of motions. The method currently works well as applied to AIA/SDO data, and seems to give decent results when applied to chromospheric data (IRIS slitjaw images) – although this has not been tested sufficiently. There are three main challenges to adapting this package to DKIST data: (1) Whilst the structure of the resulting vector fields (i.e. direction of the field) is reliable, the absolute magnitude of the field (i.e. velocity of motions at the Sun in physical units of km/s) is scaled according to the choice of certain parameters in the method (e.g. width of smoothing kernels). So this requires independent calibration, possibly using spectroscopic Doppler shifts or other estimates of flow velocities. (2) The method has not been tested sufficiently on photospheric or chromospheric data. (3) The LK part of the package needs to be adapted to work on a GPU. It is rather slow on a CPU.

• Tool name: Time-normalization & LK flow diagnostics
• Developers: Morgan & Hutton (Aberystwyth)
• Main Contact: Huw Morgan (hmorgan@aber.ac.uk)
• Basic description: See above – full method paper under review
• Language: IDL
• Resource needed to use: Desktop (but slow, see point (3) above)
• Host location:
• Current status: Works but needs improvement & testing
• 6-month plan to availability: (1) Test on photospheric & chromospheric data & (2) Improve efficiency (including use of GPU).
• Status of documentation: Method paper under review. More technical/practical documentation currently sparse (does not exist).
• Test status: OK on AIA/SDO.
• How to reference tool in publication: Morgan & Hutton, ApJSupp, under review 11/2017