Obtaining pointing information from EIS data#

Deriving precise pointing coordinates for EIS data can be a little complicated due to a number of issues:

  1. there are spatial offsets between different wavelengths in the Y-direction;
  2. raster images are obtained over a time period during which the instrument pointing is affected by satellite jitter and instrument jitter.

This document provides a guide for users seeking to understand the pointing for their data-set.

EIS Software Note #9 has more detailed information about technical issues relating to EIS pointing.

Improved EIS pointing information Update #

From 2014 a new EIS routine yields the spatial offset between EIS and AIA. It is called as

IDL> xy=eis_aia_offsets(date)

At present the correction needs to be applied manually to the pointing values described in the sections below. So, for example,

IDL> xcen=xcen+xy[0]
IDL> ycen=ycen+xy[1]

The EIS-AIA offsets are determined regularly by co-aligning EIS slot images with AIA images. They should lead to much-improved pointing accuracy, to the level of 5" or better. In particular, the large variations seen during the eclipse season are corrected for.

Note that the EIS-AIA offsets do not correct the short-term (orbit period) jitter of EIS, only the longer-term (days) pointing variation.

First approximation#

After reading your data file into an IDL object, you can obtain the approximate position of the center of your raster by doing:

IDL> xcen=data->getxcen(/raster)
IDL> ycen=data->getycen(wvl,/raster)

where WVL is the wavelength you are interested in. If WVL is not specified then 195.12 (the wavelength of the strong Fe XII line) is assumed. Note that the raster center in solar-Y varies with wavelength by up to 18" due to the tilt of the EIS grating relative to the detector and a spatial offset between the two CCDs.

The values XCEN and YCEN are obtained by taking the pointing of the center of the slit (in both X and Y) of the first exposure of the raster (the exposure at the right-hand side of the raster) and correcting this for satellite jitter. For the X-direction, the value of half of the field-of-view is then subtracted to yield the center of the raster. E.g., if a raster was obtained with the 1" slit with 100 slit positions, then the field-of-view in X is 100" and so 50" is subtracted from the position of the first exposure.

Why is this an approximate pointing? Rasters can take several minutes to several hours to complete and, during this time, the Sun is rotating opposite to the raster direction. This has the consequence that the field-of-view in X is effectively reduced. For observations at disk center the reduction will be 10" per hour of observation. An additional factor is that each individual exposure is affected by satellite and instrument jitter, which may have a systematic trend over the course of the raster although this will affect the raster field of view by a few arcseconds at most.

Field of view#

The field of view (FOV) of an EIS raster can be obtained by doing:

IDL> fovx=data->getfovx()
IDL> fovy=data->getfovy()

or by reading the values from the FITS header via

IDL> fovx=data->getinfo('FOVX')
IDL> fovy=data->getinfo('FOVY')

so if you want to overplot a box on an AIA image (for example) to show the EIS FOV for a raster, you can do:

IDL> x0=xcen-fovx/2
IDL> x1=xcen+fovx/2
IDL> y0=ycen-fovy/2
IDL> y1=ycen+fovy/2
IDL> oplot,[x0,x1,x1,x0,x0],[y0,y0,y1,y1,y0]

More detailed information#

A pointing coordinate can be obtained for each Y-pixel of each exposure by using the following object methods:

IDL> xpos=data->getxpos()
IDL> ypos=data->getypos(wvl)

XPOS has the same number of elements as there are exposures in the raster, while YPOS has the same number of elements as there are pixels along the slit. XPOS includes the satellite jitter, but YPOS does not.

The time-dependent behavior of the Y-jitter is obtained as follows:

IDL> ycen=data->getycen(wvl)
IDL> y_jitter=ycen-ycen[0]

Therefore if you want the coordinates for a particular Y-pixel, j, of a particular exposure, i, to be corrected for satellite jitter, then they are:

IDL> x=xpos[i]
IDL> y=ypos[j] + y_jitter[i] 

Instrument jitter#

In addition to the satellite jitter there also appears to be jitter internal to the EIS instrument. This can be measured by, e.g., looking at a time series of slot images and comparing the jitter obtained by co-aligning the images to the satellite jitter: the residuals represent the instrument jitter. This jitter has not been characterized yet and no software tools exist to yield the jitter values. However the magnitude of the instrument jitter is believed to be smaller than the satellite jitter.

What about XCEN and YCEN in the file header?#

As with most solar data, an EIS data-set has a header that contains XCEN and YCEN values. These can be obtained by doing:

IDL> xcen=data->getinfo('XCEN')
IDL> ycen=data->getinfo('YCEN')

In general the user should not use these values as they are set when the FITS file is created, whereas the values obtained via the object methods are computed when the method is called and so will always contain the most up-to-date pointing information.

Checking what the EIS Chief Observer intended#

It can sometimes be useful to look at what the EIS Chief Observer intended to observe when he/she created the EIS observing plan, in order to compare with what the raster image actually looks like.

eis_image_tool is a widget-based routine used by the EIS Chief Observer (CO) to perform pointing for EIS studies. When the CO performs the pointing he/she is typically using solar images obtained 1-3 days before the observation takes place. For the scientist studying EIS data, eis_image_tool can be used to see where EIS actually pointed using a solar image taken at the same time as the EIS data. The procedure is as follows:

IDL> s=fix_zdbase(/eis)
IDL> eis_mk_plan

These commands set up the EIS databases and start the EIS planning tool. Two windows pop up with names 'EIS Make Plan Tool' and 'EIS Plan Timeline'. For the latter window, go to the 'Display Start' text box and type in the date your observation was taken. The timeline will change to show the studies on this date. Find the study for which you want pointing information by clicking on the timeline entry. If you can't find the study easily, go to the Make Plan Tool widget, click on 'Tools' and then 'Plan summary...'. A text list of the studies for the day will be shown and you can click on your study to select it.

After the study has been selected, within the Make Plan Tool widget the bottom half has a list of rasters for that study (the headings are 'ID', 'Acronym', etc.). Click on your raster (usually there will only be one listed). Below the raster list there are several tabs. Click on 'Pointing' and then 'Send to image_tool'. The eis_image_tool widget pops up.

For this widget you need to load an image. Click on 'File' and 'Import Images'. A new widget called show_synop appears. Type in start and end times that bracket your observation. Next to 'remote sites' there is a drop-down list. Select 'SOHO/EIT (prepped)' and then hit the search button.

From the list of files that appear find the filter/time that is most appropriate for your data-set and click the 'Download' button. After it is downloaded to a local directory, select the file from the list at the bottom of the widget and then click 'Display'. This sends the image to eis_image_tool. You can now close show_synop.

From the eis_image_tool widget, click on 'Plot all rasters' and the location of your raster on the EIT image will be shown.

If EIT images are not available for your date, try looking for a TRACE mosaic at:

http://trace.lmsal.com/mosaic_archive/fits/