RHESSI Optical Images
From RHESSI Wiki
|1st Author:||H. Jabran Zahid|
|2nd Author:||Hugh Hudson|
|Published:||29 October 2008|
|Next Nugget:||The Good Guys and the Rascals|
|Previous Nugget:||Flare Plasma Abundances - New X-ray Observations|
RHESSI's sole scientific instrument is its array of X-ray and gamma-ray imagers, as described from many points of view in these Nuggets. To make this high-resolution imaging possible, though, even higher-resolution optical images are required. Only in this way can we follow the natural Poinsot motions and thereby determine, photon by photon, the X-ray image structure.
This requirement dictated the invention of a solar aspect sensor (SAS), which consists of three simple lenses forming solar images on three linear CCDs, with 2048 x 1.73 pixels. The set of three line images (plus independently obtained stellar aspect data) uniquely determines the instantaneous angular position of the RHESSI imager, to a resolution considerably finer than that of the X-ray images themselves. In this Nugget we show that these 1D images can be assembled into rather clean 2D images of the Sun in the 670-nm visible continuum. This unusual way of making images is of some scientific value, as will be related in future Nuggets. The merits of doing this are (a) that RHESSI is in space (no astronomical seeing whatsoever), and (b) RHESSI is rapidly rotating (radically different and more tractable flat-field problem).
RHESSI normally provides about 1,000 line images per day, and they are spread around over all position angles (image azimuths), in the manner of sticks piling up on a map (somewhat resembling Buffon's Needle Problem). For each line image we determine the limb intercepts, thus defining a chord length, and map the pixel contents onto the 2D solar image according to the roll angle of the spacecraft motion. The spacecraft nutates as well as rotates, and so the actual sampling pattern is fairly complicated. The sampling is roughly sufficient to generate one clean image per day, but the longer the integration, the smoother the result.
Figure 1 shows this in the form of a one-month average and a high-pass version of the same thing. Here "high pass" means high spatial frequencies; the Sobel filter we have used strongly reduces the smoothly varying parts of the disk and thus enhances its fine structures - spots, faculae, and the limb itself.
Note the noise patterns in the image on the right (click to enlarge). Other than the active-region traces themselves, there is also a non-uniform granularity of the images, plus excess noise in a small region near disk center. These properties result from the manner of RHESSI's rotation and nutation, but it is a somewhat subtle matter that we will not explain here.
Of what use are images of this sort? We are just working on applications that will be described in future RHESSI Science Nuggets and, we hope, research papers. These will include the characterization of the facular limb-darkening function and the global temperature structure of the photosphere, both problems of outstanding importance. There is at least one better optical telescope viewing the Sun from a satellite, the 50-cm SOT on the Hinode spacecraft. The advantages of our RHESSI images include its whole-Sun imaging and superior photometric performance, thanks to the telescope rotation. We have not yet quantified the photometric precision sufficiently but believe that it will be extremely good, because each pixel redundantly observes many positions on the solar disk. In the meanwhile we are just pleased that this imaging techniques works so well and are presenting it because it is novel in this application.
Biographical note: Jabran Zahid is now a graduate student at the University of Hawai'i, but was a researcher at Berkeley when much of this work was done. Hugh Hudson is a senior RHESSI team member in Berkeley.