Browse the RHESSI data!
From RHESSI Wiki
|1st Author:||Albert Shih|
|Published:||16 January 2006|
|Next Nugget:||A Solar Hard X-Ray Halo: Exploring the Quiet Sun 2|
|Previous Nugget:||RHESSI and the Wilson effect|
RHESSI data (and software) are of course freely available to all, but to view the data directly often requires some training or study. Accordingly one needs a simple overview, and Browser - a Web interface - fills this need. This science nugget will not be providing knowledge on some scientific topic, but rather on how to use Browser to its maximum potential in the pursuit of science. As many of you already know, Browser is a very useful tool to browse through many different types of RHESSI-related data plots simultaneously. However, I suspect that many of you may not realize some of its not-so-obvious capabilities. If so, read on!
First, if you are rather unfamiliar with Browser, just play with it! Perhaps pick a time when you know something interesting happened, and start clicking around on various icons and links. Feel free to experiment!
The most useful icons or links may be the question-mark icon at the top and the "Help" link at the bottom. Both will open the help page, which may answer any basic questions you may have. The next most useful icons are the magnifying-glass icons that are littered across the page. Clicking on such an icon will reveal advanced options that may turn out to be indispensable for your method of using Browser. Finally, one should not forget the red-arrow icons, which indicate a particular time on all plots that have a time axis. Combined with the ability to click on plots to select a specific time, one can click on a feature of interest on one plot, and then the red arrow will identify the same time on all the other plots.
I will now present two case studies for how one might use Browser.
Case Study 1: X4.8 Solar Flare on 2002 July 23
I will first discuss the typical use of Browser to look at solar flares, and I have chosen the much-studied X4.8 solar flare on 2002 July 23. Click on the image on the left to open the actual Browser page in a new window. Now to explain its elements... Note that there are quite a few plots open, although these are only a fraction of the possible number. Unsurprisingly, the more plots that are open, the longer the page takes to load completely, so I would suggest a smaller set for casual browsing, and then opening other plots when necessary. However, if you ordinarily view only one plot at a time in Browser, I would strongly encourage you to try browsing with multiple plots open.
The first plot  shows the GOES lightcurves with the RHESSI times for eclipse and radiation-belt passages superposed (no data and bad data, respectively). For many flares, one can use these plots to determine whether a given flare has observed. This plot is also useful to jump between multiple flares on a single day (see the red-arrow icon discussion above).
Next, the quicklook lightcurve plots  show the corrected and uncorrected rates for the orbit. "Quicklook" is rich collection of reduced secondary databases - lightcurves, images, and spectra. A brief digression: when you see interesting behavior on the corrected-rates plot, be sure to check the uncorrected-rates plots to make sure that the behavior was not simply a correction artifact. For example, the attenuation by the shutters was corrected for in this flare, but note that the 7000-20000 keV band has picked up a small spike at 0041 UT that is clearly a correction artifact (there is nothing to correct for at such high energies).
The next quicklook plot  shows a sampling of the data from the particle counters on RHESSI (on board mainly to track Earth's radiation belts). The top and bottom halves show lightcurves for the detectors that usually respond to electrons and protons, respectively. In this particular example, there is a short burst of activity after 0100 UT. By using the red-arrow icons to identify the same time on the quicklook lightcurve plots, we see the particle activity did in fact produce an effect visible in the RHESSI lightcurves, although fortunately well after virtually all the interesting flare emission.
The next three plots  are various types of quicklook images for this flare, and the following plot  (gaudy and complicated) combines GOES, RHESSI, and WIND particle data, and also includes the RHESSI data in spectrogram form.
The final three plots  above are examples of monitor-rates plots. The first plot shows the "slow valid" (i.e. good) counts in each of the nine front segments. Of course, the quicklook lightcurve plots are usually much more useful, but this plot has its uses. For example, the artifact in detector 8 due to interference from the aft antenna can be clearly seen (e.g. 1650 UT). The following two plots show the livetimes in the front segments and the rear segments. Note that since flare photons tend to stop in the front segments before reaching the rear segments, the rear segments have substantially more livetime than the front segments. However, this relationship does not hold for non-solar photons such as those from nearby particle precipitation events. There are many more monitor-rate plots, but they are probably too technical for most users.
Case Study 2: Magnetar Burst on 2004 December 27
Now let us take a look at an event that was not a solar flare, specifically the magnetar burst on 2004 December 27. (Even though RHESSI flagged this event a solar flare, it is most definitely not one.) Looking at the quicklook plots , one can see an egregious correction artifact where many of the lightcurves shift to a new baseline after the magnetar burst. Again, one should always remember to check the uncorrected-rates plots when seeing strange behavior.
The next two quicklook plots  show the count lightcurves for the nine front segments and nine rear segments, respectively. One can use these plots in much the same way as the corresponding monitor-rates plots, and, in addition, flares are marked by lines on these plots.
The next plot  is an attempt to display a quicklook image plot for this event. Since this non-solar event cannot be imaged, an error message appears instead.
Finally, there is the GOES/RHESSI/WIND plot . Note that the GOES data shows that the burst went virtually unrecorded in the low-energy channel. It would have been difficult to pick out this event on the usual GOES-with-RHESSI-times plots. Also note the appearance of the burst in the spectrogram, and compare it against the totally unrelated particle precipitation event that occurred an hour later at 2340 UT.
The first advanced feature that you may not have been aware of is that Browser automatically synchronizes links to corresponding data on other websites at the same time as currently selected in Browser. For example, if you have the first case study open, you can click on the Active Region Monitor link on that page to bring up SOHO data for the 2002 July 23 flare. Another advanced feature is the ability to create links to a specific time with particular plots open (like the two case studies above). Once you have arranged Browser in the way you want, you can click the "Direct URL" button (or right-click and save the link's location) to access the custom shortcut. One can then bookmark the page, or perhaps send it to a colleague for consultation. Incidentally, if one clicks the "Reset Date" button immediately before using the "Direct URL" button, then the time information is erased, and the direct URL stores only the current view layout (e.g. my preferred view).
Finally, there is a major timesaver in the form of text fields for the flare number or time (click the magnifying-glass icon next to the year-selection box). Although these fields are somewhat useful as output fields, their real power comes from using them as input fields. One can enter a specific flare number or a specific time (in most Solarsoft-compatible formats), and Browser will jump to the corresponding time. This feature works great when analyzing a list of events. Note that there is an eraser button provided to aid copy-and-pasting on systems where selecting automatically overwrites the contents of the clipboard.
I hope this discussion introduced you to Browser and some new facets that you had not been aware of. Provide me with comments and suggestions on Browser, and if I happen across any free time, I will see what I can do. But, as is, Browser has helped my own research immensely, and I hope that you will be able to say the same.
Biographical note: Albert Shih is a graduate student at UC Berkeley, working on a PhD thesis analyzing solar gamma-ray flares.