RHESSI Image Archive Strategy

Last Update 29-Jan-2023, Kim Tolbert (kim.tolbert at nasa.gov)
This version uses the updated flare catalog created in September 2022 based on the updated aspect software and database.

The RHESSI Image Archive provides science-ready images for all RHESSI flares that meet time and flux thresholds. For each flare that meets the threshold, high quality images have been generated for an optimized grid of times and energies customized for that flare, using six of the RHESSI image reconstruction algorithms. FITS files of the image cubes are provided for user download, and a host of prepared image and ancillary plots are available for quick viewing.

The entry into the archive is a collection of monthly listings of links, with dropdown menus to choose year and month to load. The different products available for each flare are described in detail in the Guide to RHESSI Image Archive.

In addition, in the RHESSI Browser, there is a link in the left-hand column called 'RHESSI Image Archive' that opens the image archive page for the flare during the time the red arrow points to. If the red arrow is not in a flare interval, the 'RHESSI Image Archive' link will have a strikethrough.


Caveat

Some of the images in this archive are invalid. Usually this is obvious from the lack of coherent structure in a single image, or by inconsistency between successive times or energies. There are several reasons for this.

If you encounter problems with images in the archive, please let me know (kim.tolbert at nasa.gov).



Organization

In the RHESSI flare catalog, separately listed flares with different numbers are sometimes contiguous in time and have the same source location on the Sun. We have combined these separate events for the Image Archive. The Image Archive data directory organization is based on the 'concatenated' flare times.

The metadata for each flare are contained in directories named
    flare_images_v2/year/month/day/yyyymmdd_hhmm_hhmm
where yyyymmdd_hhmm_hhmm indicates the start and end time of the 'concatenated' flare. The products for each image reconstruction algorithm are in subdirectories named by the algorithm.

The imagecube FITS files archive is in a parallel directory structure named
    imagecube_fits_v2/year/month/day/yyyymmdd_hhmm_hhmm

The eventlist and visibilities used to make the imagecube are stored in FITS files in the Eventlist FITS file archive and the Visibility FITS file archive which are in directory structures named
    eventlist_fits_v2/year/month/day
    visibility_fits_v2/year/month/day

The IDL software used to generate the image archive is all online in the hessi branch of the SolarSoft tree. New routines were written to control the pipeline processing for the archive, but the image object software is exactly the same as that used by all users. Scripts are provided in the metadata for two purposes - to show the parameters used to make the images, and for users to download and perhaps modify to make their own customized images. Two sets of scripts are provided - one showing all parameter settings in the image object, and the other showing only the non-default parameter settings.



Image Settings

Detailed descriptions of the different image reconstruction methods are available in the RHESSI Imaging Software page on the RHESSI web site. We have chosen this representative subset of the available image reconstruction algorithms to make images for the archive:

All images are made with 101x101 1.5x1.5 arcsec square pixels.

Clean and Back Projection use the 'Cart' modulation profile strategy (uses cartesian coordinates intead of annular sector - annsec - coordinates).

Detectors 3-9, front segments are nominally used. No rear-segment data are used for any of the images. The detector status database is consulted to avoid detectors that are unreliable or off for each flare. For the 'CLEAN_59 algorithm', we use the Clean algorithm, with Detectors 5-9 nominally.

The nominal energy bands used are the RHESSI quicklook bands 3-6, 6-12, 12-25, 25-50, 50-100, 100-300 keV. Determination of which of these bands to use for a flare is described below in the time/energy binning description. We only use the 3-6 keV band if the entire flare was in the A0 attenuator state (no shutters in front of detectors).

The xyoffset (offset of the map center from Sun center) used for a flare was determined by making a Back Projection image centered at the flare position from the RHESSI flare catalog, and using a weighted average in x and y of the pixels within 90% of the peak to get a better FOV center. This xyoffset is fixed for the imagecube even if the flare centroid moves during the flare time period.

Each imagecube consists of images in a rectangular grid of time/energy bins. The time intervals are based on the flux in the highest energy band observed in the flare. Any time/energy bin in which an image could not be made is zero-filled. If no time/energy bins meet the criteria below, an imagecube is not made for the flare.

Time / Energy bins are determined as follows:

If the signal-to-noise ratio (SNR) of visibilities in a time/energy bin is < a threshold, we don't even attempt to make an image (the image will be zero-filled). SNR is determined as follows:


Remaining Tasks

Make JPEG2000 images for inclusion in Helioviewer.

Add measures of image quality: