This page (revision-57) was last changed on 07-Dec-2016 14:14 by David R Williams

This page was created on 09-Jul-2008 07:16 by David R Williams

Only authorized users are allowed to rename pages.

Only authorized users are allowed to delete pages.

Page revision history

Version Date Modified Size Author Changes ... Change note
57 07-Dec-2016 14:14 17 KB David R Williams to previous
56 15-Jul-2008 13:54 17 KB David R Williams to previous | to last
55 15-Jul-2008 13:27 16 KB David R Williams to previous | to last
54 15-Jul-2008 13:19 16 KB David R Williams to previous | to last
53 15-Jul-2008 13:19 16 KB David R Williams to previous | to last
52 15-Jul-2008 13:18 16 KB David R Williams to previous | to last
51 15-Jul-2008 13:18 16 KB David R Williams to previous | to last
50 15-Jul-2008 13:14 16 KB David R Williams to previous | to last
49 15-Jul-2008 11:43 14 KB David R Williams to previous | to last
48 15-Jul-2008 10:05 14 KB David R Williams to previous | to last
47 15-Jul-2008 10:02 14 KB David R Williams to previous | to last
46 15-Jul-2008 09:53 14 KB David R Williams to previous | to last
45 15-Jul-2008 09:39 14 KB David R Williams to previous | to last
44 15-Jul-2008 09:36 13 KB David R Williams to previous | to last
43 15-Jul-2008 09:23 13 KB David R Williams to previous | to last
42 15-Jul-2008 09:06 13 KB David R Williams to previous | to last
41 15-Jul-2008 08:19 12 KB David R Williams to previous | to last

Page References

Incoming links Outgoing links
EisXstudy
EisXstudy

Version management

Difference between version and

At line 1 added 2 lines
[{ALLOW edit EISMainUsers}]
[{ALLOW view Anonymous}]
At line 38 changed 2 lines
%%information
One of the nice things about EIS is that you can recycle a study already designed by someone else for a similar — or even different — purpose, without having to go through this design step.%%
%%information One of the nice things about EIS is that you can recycle a study already designed by someone else for a similar — or even different — purpose, without having to go through this design step.%%
At line 51 changed one line
If you have some idea about the study you're looking for, ''e.g.'' it has ''CH'' in the name, or it was written by George Doschek, you can use the search facility to save you time.
If you have some idea about the study you're looking for, ''e.g.'' it has ''EL'' in the name, or it was written by George Doschek, you can use the search facility to save you time.
At line 53 changed one line
Below the Study Description panel, you can search on one or more fields, namely ''Acronym'' (the name of the study), ''Title'' (confusingly, a description of the study), or ''Author'' (self-explanatory). Since I've mentioned George Doschek, let's try a search for studies he's designed:
Below the Study Description panel, you can search on one or more fields, namely ''Acronym'' (the name of the study), ''Title'' (confusingly, a description of the study), or ''Author'' (self-explanatory). Since I've mentioned George Doschek, let's try a search for studies he's designed. I don't know any other Doscheks on the EIS team, so I'll just use his surname:
[{Image src='Search_Doschek.png' }]
At line 57 added 188 lines
You can see that this results in four studies showing up, each of which was authored by someone called Doschek.
You can clear the search fields by clicking ''Clear'' on the right-hand side.
Changing the ''Study Type'' drop-down list on the right from ''Science & Engineering'' to ''Science'' excludes all engineering-type studies from our search (although this tends to be more useful when looking for specifice engineerings studies hidden amongst the forest of science studies).
!Raster panel
Below the search area is the ''Raster'' panel, where you can start to look at the details of the study's execution. Let's take another example to look at this.
For the purposes of this example, my memory's a bit shaky, so I just remember that the study I'm interested in has the letters ''CH'' in the acronym. Searching for these letters in the ''Acronym'' field produces these results:
[{Image src='Search_CH_sample.png'}]
and I recognise the study I'm interested in: PRY_CH_density. I then click on that study in the list, to highlight it.
This also makes the raster panel go from being empty to showing the rasters in the selected study (if no study is highlighted, {{{eis_xstudy}}} can't know which rasters to display).
A study is essentially a package: it's an observing programme that contains a pre-defined order of executable rasters. {{{eis_xstudy}}} allows you to look inside this package, at the rasters that a study contains, as well as the details of those rasters. You'll see that Study #268 {{{PRY_CH_density}}} contains raster ID #257 {{{CH_density}}}. In most practical cases, studies contain only one raster. But the study is what is actually makes it onto the ''timeline'' (schedule) for EIS.
%%information It's important not to confuse raster and study IDs when requesting observations. Please always use the __Study ID or full study acronym__.%%
!Viewing raster details
If we want to see what happens when {{{PRY_CH_density}}} runs, the best thing to do from here is to click on each raster in the raster panel (in this case, there's only one), and then click the ''View Selected Raster Details...'' button just above.
This opens up a separate window which details how the raster will execute on-board the space-craft (only omitting pointing details, which are flexible). The window should look something like this:
[{Image src='Raster_details_CH_density2.png'}]
This is also a two-panel window. The reason for that is that, similar to the way that a study can't exist without a raster, a raster can't exist without a line list. More on that later...
The upper panel here shows the key information that we need to know about to describe the raster, namely:
||Parameter Name||value
|Title| ''to be discussed later''
|Author|Peter Young
|Raster Type|Scanning
|Slit/Slot|2"
|Compression|DPCM
|Window Height|200 pixels
|Number of Raster Pointings|34
|FM Step Size|2.0"
|Number of Exposures/step|1
|Exposure time(s) in seconds|100.0
|Delay times (ms)|0.0
So, what to do with all this information? Let's put it in context.
__Author__\\
The author of this ''raster'' is Peter Young. Fairly easy
__Raster Type__\\
The term [raster|http://en.wikipedia.org/wiki/Raster_graphics] usually conjures up the idea of a rectangular image. And that's what EIS does when it makes its observations. It's just that in some cases the ''X'' dimension is time, rather than space.
EIS has two modes of raster. The first is ''scanning'' raster, which is what we see in this example. EIS takes a restricted-field image of the Sun and then disperses it, perpendicular to the length of the slit. Then, a fine mirror (FMIR) adjusts the X position on the Sun where that image comes from (by the amount indicated in ''FM Step size'') and makes another image. This is how EIS builds up [spectroheliograms|http://en.wiktionary.org/wiki/spectroheliogram] (spectral image of the Sun).
The second mode is the so-called ''fixed slit'' or ''sit'n'stare'' raster type, whereby the same slit or slot-size portion of the Sun is repeatedly made into an image and dispersed. This is qualitatively the same as the scanning raster, except for the fact that we are usually more interested in the time-dependent changes at a particular location.
__Slit/Slot__\\
''a.k.a. SLA — SLit/slot Assembly''\\
EIS makes images of different widths every time it takes an exposure (and disperses it). Slit spectra contain direct spectroscopic information, but take time to build up information over areas (see ''Raster Type'' above). To make slit spectra, EIS uses two slits: 1" and 2" in width.
The 1" slit has less spatial coverage (because it is narrower) so there is sharper spectral resolution; meanwhile, the advantage of the 2" slit is that it is twice as wide, and thus it can accumulate photons twice as quickly as the 1" slit. (Note that this argument does not extend linearly to the width of the slots).
For more image-based information, at the expense of line profile information, we can make slot images (scanning or sit'n'stare). In both cases the images are dispersed on the detector. The slot images are a sort of convolution of a spectrum with an image. However, the narrow slot — 40" wide — has been chosen so that essentially monochromatic images of the Sun can be made in many lines, without significant blending with images made in nearby lines.
The width of the wider slot — 266" — was chosen for ground calibration reasons, rather than scientific considerations. However, the Fe XV line at 284 Å is sufficiently removed from other active region lines that we can essentially make images of only Fe XV in bright coronal loops. In other lines, the images resemble the ''overlappograms'' produced by the ''Skylab'' S020 spectrograph (see Figure 2 of [http://solar.physics.montana.edu/reiser/] for a full-Sun idea of what this means), though with smaller spatial extent.
__Compression Type__\\
In order to make the process of transferring data to the ground more efficient, data from all Hinode instruments can be compressed on-board the spacecraft by the mission data processor or ''MDP''.
The MDP allows several choices of compression:
%%sortable
||Scheme||Compression Type
|Uncompressed|None
|[DPCM|~DPCMScheme]|Lossless
|[JPEG98|~JPEG98Scheme]|lossy (Q factor 98)
|[JPEG95|~JPEG95Scheme]|lossy (Q factor 95)
|[JPEG92|~JPEG92Scheme]|lossy (Q factor 92)
|[JPEG90|~JPEG90Scheme]|lossy (Q factor 90)
|[JPEG85|~JPEG85Scheme]|lossy (Q factor 85)
|[JPEG75|~JPEG75Scheme]|lossy (Q factor 75)
|[JPEG65|~JPEG65Scheme]|lossy (Q factor 65)
|[JPEG50|~JPEG50Scheme]|lossy (Q factor 50)
%%
%%information The links in the above table each take you to a page where the estimates of the Data Compression (''DC'') factor are listed. These values vary depending on the target and SLA choice. They will be updated from time-to-time until the team are satisfied that they are adequate enough for planning purposes.%%
In addition to these webpages showing the more involved analysis results, the current values used for DC factors in making observations are stored in {{{$SSW/hinode/eis/idl/spacecraft/decompression/eis_get_compression_factor.pro}}}
__Window Height__\\
Whereas the width of each spectral window can vary (see Line Lists below), the height of the image ''y%%sub size%%'' read out from the CCDs is necessarily fixed for each window. So each raster has the same height in all spectral windows. This height is expressed in arcseconds. In EIS's design, it was decided that a 512" image height would be sufficient (1 pixel = 1"), so this is the maximum image height that can be made at any one time.
__Number of Raster Pointings__\\
''Scanning raster-specific''\\
This is a little confusing, but it tells you the number of times ''N'' that the mirror must move between observations to build up a scanning raster, i.e. 1 less than the number of positions at which EIS makes an exposure.
__Working out the size of a raster__\\
If ''N+1'' is the number of positions at which the slit makes an image, Σ ''t%%sub exp%%'' is the sum of the exposure times (usually there is only one) at each position (in seconds), and ''x%%sub step%%'' is the size (in arcseconds) of the step, then it takes ''n'' × Σ ''t%%sub exp%%'' × ''x%%sub step%%'' seconds to build up a spatial image of an area that is ''x%%sub size%%'' = ( (''n'' × ''x%%sub step%%'') + SlitSize ) arcseconds across.
Putting this together with the ''x%%sub size%%'', worked out above, tells you the spatial extent of the observation that this raster produces.
__Exposure Times__\\
EIS allows its user to make more than one exposure at each position on the Sun, and these exposures can have independent shutter-open times. In fact, you can have up to eight such exposures at each fine mirror position.
__Delay Times__\\
Delay times can be up to 655.35 seconds long (2^16 -1 × 10-ms units), and come after each exposure. The number of delay times has to be the same as the number of exposure times, but typically these delays are set to zero.
For example, with two exposure times of 10 and 120 seconds, and delays of 100 seconds and 200 seconds, a typical ordering of events could go like this:
||Event
|Move fine mirror to starting position
|Take an exposure for 10 seconds
|Read out exposure
|Delay for 100 seconds
|Take an exposure for 120 seconds
|Read out exposure
|Delay for 200 seconds
|
|Take an exposure for 10 seconds
|Read out exposure
|Delay for 100 seconds
|Take an exposure for 120 seconds
|Read out exposure
|Delay for 200 seconds
|
|Move fine mirror to next position
|''etc.''
__Raster Title__\\
Knowing what all the above information means, you can also get a quick idea of most of the characteristics of a raster from the single-string raster ''Title'' field in {{{eis_xstudy}}}.
This "summary", if you like, is composed of a series of square-bracketed terms in the following order:
|RA:RasterID
|Scan:NumberOfPositons-1(SizeOfStep)steps,
|ss:SLAchoice
|wH:WindowHeight,nWins:NumberOfSpectralWindows,LL:LineListID
|ExpT(Delay):ExpTimeNumberOneInSeconds(DelayNumberOneInHundredthsOfASecond), ...
and can be handy when you understand how it works, because it means you don't need to delve into the raster description to find out how large it is, which slit or slot it uses, how long the exposure times(s) is/are, ''etc.''
%%information Note that since engineering studies are created by hand, rather than through the planning tool, it's not normally possible to view their contents (such as number of exposures, or exposure time) in the Raster panel.%%
\\
\\
----
!Line List
The most critical part of the raster description that __isn't__ summarised in the ''Raster Title'' is the Line List. This is because the line list is too variable to easily summarised. So, it gets its own panel in the raster description.
Note that, like studies and rasters, line lists have and ID and acronym.
\\
__Rationale of a Line List__\\
When EIS makes a spectral image of a slit- or slot-sized portion of the Sun, it exposes the whole of both detectors, which amounts to some 2 × 2048 × 1024 pixels. At 16 bits per pixel (2 reserved for marking which half of which detector a pixel belongs to), this would create a massive 8 MB for all data from the detector, __at each position on the Sun!__
So, in order to limit telemetry need — even before the transition to the S-band antenna — EIS was designed so that only the spectral portions of interest on the detector would be telemetered to the ground for a given observation.
Line lists are described in terms of absolute wavelength, according to the standard wavelength calibration, so you don't need to work things out in terms of pixel co-ordinates.
__Title__\\
You'll see that, like the raster, a Line List has a one-line so-called ''Title'' string associated with it. Since this is truncated, it doesn't serve much purpose, and we'll skip it here.
__Author__\\
Again, fairly self-explanatory\\
:-)
One point to perhaps reiterate, though, is that you can use a line list created by anyone else in your study. All line lists, rasters, and studies in the OSDB are there to be re-used and cannibalised by anyone wanting to propose an observation.
__Number of lines__\\
There can be up to 25 spectral windows in a line list, but the minimum is normally three: the core lines. The core lines are
||Ion||Wavelength (Å)
|Fe XII |195.12
|Ca XVII|192.82
|He II |256.32
__Line List Detail__\\
Each spectral window can have a width that is an integer multiple of 8, from 8 to 1024 (half a CCD width), and there can be up to 25 windows, each with independent widths. In the example image above, you can see that most windows have a width of 24 pixels (a little over 0.5 Å), which is usually adequate. But a few windows are wider (32, 40 or 48 pixels). Wider windows often (but don't necessarily) indicate that the line list author was trying to capture other lines by using a wider spectral window.
This also helps to explain why the core lines of {{Fe XII}} (195.12 Å) and {{Ca XVII}} (192.82 Å) aren't explicitly included in this line list: they are already covered by the {{Fe XII}} window at 194.91 Å and the {{O V}} window at 192.84 Å, respectively.
This page (revision-57) was last changed on 07-Dec-2016 14:14 by David R WilliamsTop