\documentclass[11pt]{article} \usepackage[dvips]{graphicx} %% FOR INCLUDING GRAPHICS \usepackage{natbib} \usepackage{epsf} %% ---------------------------------------------------------------------- %% Add specials here %% ---------------------------------------------------------------------- \def\ion#1#2{#1\,{\sc #2}\relax} %% %% ---------------------------------------------------------------------- %% Modify: %% Note number, version number and date %% Title %% Author and affiliation %% ---------------------------------------------------------------------- \setlength{\oddsidemargin}{0in} % 1 inch (0 + 1) from left. \setlength{\topmargin}{-0.5in} % 1 inch (1.5 - 0.5) from top. \setlength{\textwidth}{6.5in} % 1 inch (8.5 - 1 - 6.5) from right. \setlength{\textheight}{9in} % 1 inch (11 - 1 - 9) from bottom. \addtolength{\parskip}{0.5\baselineskip} % The environment centerpage centers text vertically on the page. This is % useful within a titlepage environment. \newenvironment{centerpage}{\mbox{} \protect\vspace*{\fill}}{ \protect\vspace*{\fill} \mbox{} \protect\\ \mbox{}} \newcommand{\lam}{$\lambda$} \newcommand{\arcsec}{$^{\prime\prime}$} \newcommand{\kms}{km~s$^{-1}$} \begin{document} %\bibliographystyle{plain} % Define bibliography style. \begin{titlepage} \hrule \noindent\textbf{EUV IMAGING SPECTROMETER}\\ \vspace{-0.7\baselineskip}\\ {\noindent\Huge\bf Hinode} \vspace{2mm} \hrule \vspace{3mm} \centerline{\bf EIS SOFTWARE NOTE No. 5} \vspace{3mm} \hrule \noindent Version 1.2 \hfill 18 August 2021 \vspace{2mm} \hrule \begin{centerpage} \begin{center} {\Large\bf Orbital drift of the EIS wavelength scale}\\ \mbox{}\\ \mbox{}\\ Suguru Kamio\\ Max-Planck-Institut f\"ur Sonnensystemforschung\\ Max-Planck-Str. 2\\ 37191 Katlenburg-Lindau\\ Germany\\ \mbox{}\\ \mbox{}\\ Terje Fredvik\\ Institute of Theoretical Astrophysics\\ University of Oslo\\ P.O. Box 1029 Blindern\\ N-0315 Oslo\\ Norway\\ \mbox{}\\ terje.fredvik@astro.uio.no\\ \mbox{}\\ \mbox{}\\ \mbox{}\\ Peter Young\\ NASA Goddard Space Flight Center\\ U.S.A.\\ \mbox{}\\ \end{center} \end{centerpage} \end{titlepage} %% ---------------------------------------------------------------------- %% The document! %% ---------------------------------------------------------------------- \textbf{UPDATE (18-Aug-2021)}: The method described by \citet{kamio10} no longer works correctly and should not be used for data obtained from 2018 onwards. Generally you can recognise that the method is not working if you see vertical bands in the Doppler maps with amplitudes of around 10~km~s$^{-1}$. The best solution is to fit the emission line of interest and then identify a band of quiet Sun. By averaging the line centroid over this band you will see the orbital variation, which can then be subtracted . The text below is retained from the previous version of the document (1.1). \section{Overview} If a Gaussian is fit to an emission line profile over an entire raster, then plotting the resulting image of the centroid variation shows a strong oscillation in the X-direction. An example is shown in Fig.~\ref{fig.orbital}, where the alternating bands of red and blue correspond to red and blue shifts of the emission line. The period of the oscillation is 98~mins and corresponds to the orbital period of the spacecraft, thus the oscillation is clearly due to thermal effects related to the satellite's orbit around the Earth. In this document we refer to this effect as ``spectrum drift'' and we describe methods of correcting it. \begin{figure}[h] \centerline{\epsfxsize=12cm\epsfbox{orbital.eps}} \caption{A velocity map derived from an EIS raster. EIS rasters from right-to-left, and the vertical bands of blue and red correspond to the spectrum drift during the raster.} \label{fig.orbital} \end{figure} \section{What causes the oscillation?} The Hinode spacecraft is oriented such that it always points at the Sun, and it does not rotate relative to the Sun. This means that each side of the spacecraft receives a variable illumination from the Earth as the spacecraft orbits the Earth. For EIS this means that in some parts of the orbit the instrument is directly illuminated by the Earth, and in other parts it is shadow behind SOT. The movement of the emission lines on the CCD is thus understood as being due to an oscillation in the angle of the reflecting surface of the EIS grating. The changing temperature of the instrument in the vicinity of the grating leads to a small rotation of the grating position. The effect corresponds to a $\pm$1.5 pixel amplitude oscillation in the positions of the lines on the detector. The EIS pixels have a size of 13.5~$\mu$m and the grating--detector separation is 1.44~m thus the tilt in the grating position is only 0.0008$^\circ$. However, Fig.~\ref{fig.orbital} shows that this tiny movement dominates the Doppler shifts due to real motions on the Sun. \section{Properties of the spectrum drift} The paper of \citet{kamio10} gives a thorough discussion of the spectrum drift. Some key points are given below. \begin{itemize} \item The amplitude of the oscillation is about 1.5 pixels and is the same for all wavelengths. This means, however, that in velocity space the amplitude varies with wavelength, being around 50~\kms\ for \ion{Fe}{xii} \lam195.12 and 35~\kms\ for \ion{Fe}{xv} \lam284.16. \item During the eclipse season (April--August) the oscillation significantly deviates from a sinusoidal pattern. \item Over a day the line centroids oscillate about a center which itself drifts with time. This drift is seasonal in nature and due to the varying separation of the Sun and the Earth. \end{itemize} \section{Correcting for the spectrum drift} Initial efforts at correcting the spectrum drift involved using the science data themselves. Typically the strongest line in the spectrum, \ion{Fe}{xii} \lam195.12, was fit at each pixel in a raster, and the centroid averaged in the Y-direction. This average centroid yielded the spectrum drift during the course of the raster which could then be used to correct other lines. The method assumes there are no large scale flows within the raster that would distort the spectrum drift. A common variation for active region data is to identify a quiet part of the raster and then only average over this region so that any flows within the active region do not affect the determination of the spectrum drift. Since the EIS instrument contains 31 temperature sensors that yield values with a 10~s cadence, then an alternative method of determining the spectrum drift is to correlate the temperature sensor readings with the spectrum line positions through a model. This work was performed by \citet{kamio10} and, since around 2010 September, this new method is now the default way of correcting for the spectrum drift. Since the temperature sensor data form part of the EIS housekeeping data stream, then the method is referred to as the ``HK method''. Full details on how the model works are provided by \citet{kamio10}, and we give a brief summary below: \begin{itemize} \item The model is based on measurements of the centroid of \ion{Fe}{xii} \lam195.12 line over the entire course of the Hinode mission. \item As new data are obtained, the model correction is continually updated. \item The model sets the average \lam195.12 velocity over the whole mission to be 0~\kms, thus the spectrum drift correction also yields a form of absolute wavelength calibration. \item Following the spectrum drift correction, centroids should be accurate to 4.4~\kms. \item The data needed for performing the spectrum drift correction are stored in the EIS Solarsoft distribution. \end{itemize} \section{How is the spectrum drift correction implemented?} An empirical model has been developed to reproduce the drift of the EIS spectrum using instrumental house keeping temperatures and the relative motion of the spacecraft, see \citet{kamio10}. The wavelength correction can be found using eis\_wave\_corr\_hk.pro, the gethkwavecorr() method of the eis\_data object or eis\_prep.pro. All procedures use specially prepared house keeping data files that are distributed with SolarSoft in the directory \$EIS\_DATA/wave\_corr. If the house keeping files that are needed for correcting a certain EIS raster are missing in your SolarSoft installation, the new routines will search for the files in \$EIS\_WAVE\_CORR\_HK\_DATA (if this environment variable is set), and in the current working directory. If the files are not found, they will automatically be downloaded from the Oslo archive and saved in \$EIS\_WAVE\_CORR\_HK\_DATA (if set) or to the current working directory (if not set). The following three examples demonstrate different methods to obtain the new wavelength correction. The example data-set is specified through % \begin{verbatim} IDL> paths = '~/hinodedata/eis/mission/2007/12/11/eis_l0_20071211_162413.fits.gz' \end{verbatim} % (please modify the path to the location where your Hinode data are stored). \subsection{eis\_getwindata.pro} This routine reads a specified wavelength window into an IDL structure. By default it also calls the routine eis\_wave\_corr.pro and stores the wavelength correction related to the spectrum drift within the structure, in a tag called wave\_corr\_t. For example: % \begin{verbatim} IDL> wd=eis_getwindata(paths,195.12) IDL> help,wd.wave_corr_t DOUBLE = Array[256] \end{verbatim} % This array gives the wavelength correction for each exposure in the raster required to put the emission line on the absolute wavelength calibration scale of the \citet{kamio10} method. The windata structure has an additional tag called wave\_corr that combines the spectrum drift with the EIS slit tilt (see EIS Software Note \#4) to give a 2D array. The wave\_corr array stored in the windata structure is used by the Gaussian fitting routine eis\_auto\_fit.pro to perform the spectrum drift and slit tilt corrections, so that the velocity map that results from the fit is free from these instrumental effects. \subsection{eis\_wave\_corr.pro} The eis\_wave\_corr routine was modified in 2010 September so that it uses the HK method by default. Previously it has used moments of the \ion{Fe}{xii} \lam195.12 line to determine the spectrum drift. The calling sequence is: % \begin{verbatim} IDL> eis_wave_corr, paths, wvl_corr, dw_tilt, dw_t \end{verbatim} % (the last two parameters are optional). To use the previous method to determine wvl\_corr, set the /OLD keyword. There is a routine called eis\_wave\_corr\_hk which returns wvl\_corr, dw\_tilt and dw\_t in the identical manner to eis\_wave\_corr (in fact it is called by eis\_wave\_corr) but which has the additional two keywords: % \begin{description} \item[wvl\_cube:] If set to a named variable, wvl\_cube will return a 3 dimensional wavelength cube in the format used by mk\_analysis.pro of the Component Fitting System for IDL (cfit) package. \item[lam:] the underlying methods of eis\_data calculate the wavelength correction in pixels, and we assume that the shift in pixels is the same for all lines formed at any location on the two detectors. However, when measured in Angstroms the line centre variation and the slit tilt will vary slightly with wavelength. As default, eis\_wave\_corr\_hk assumes that the corrections are to be calculated for the Fe XII 195~\AA\ line, but the lam keyword may be set to the wavelength of the line you want to investigate. For example, the following code compares the spectrum drift for the Fe XII \lam195.12 and \ion{Fe}{xv} \lam284.16 lines. \end{description} % \begin{verbatim} IDL> eis_wave_corr_hk, paths, wvl_corr_hk, dw_tilt_hk, dw_t_hk IDL> plot, dw_t_hk, /xst, /yst IDL> eis_wave_corr_hk, paths, wvl_corr_hk2, dw_tilt_hk2, dw_t_hk2, wvl_cube=wvl_cube, lam=284.16 IDL> help, wvl_cube IDL> loadct,12 & oplot, dw_t_hk2, color=170 \end{verbatim} % Finally, eis\_wave\_corr\_hk can take a level 0 or level 1 eis\_data object reference as input instead of a file name: % \begin{verbatim} IDL> o = obj_new('eis_data',paths) IDL> eis_wave_corr_hk, o, wvl_corr \end{verbatim} \subsection{gethkwavecorr(), a new eis\_data method} When creating an eis\_data object, the aux\_data will contain two new null pointers, hkpixcorrtilt and hkpixcorrtime: \begin{verbatim} IDL> o = obj_new('eis_data',paths) IDL> help, o->getaux_data(),/str \end{verbatim} The two pointers will become valid when the method gethkwavecorr() is called (i.e. when you try to access the contents of the pointers): \begin{verbatim} IDL> corr = o->gethkwavecorr() IDL> help, corr,/str IDL> help, o->getaux_data(),/str \end{verbatim} The gethkwavecorr() may be called with a wavelength as input parameter, or with the keyword wvl\_cube set, to get the same functionality as for the eis\_wave\_corr\_hk wrapper as explained above. \begin{verbatim} IDL> corr = o->gethkwavecorr(/wvl_cube) IDL> help,corr,/str IDL> corr = o->gethkwavecorr(284.16) IDL> oplot,corr.time,color=0 \end{verbatim} \subsection{eis\_prep.pro, with hkwavecorr set} eis\_prep.pro now accepts the keyword hkwavecorr. If set, the resulting level 1 file will contain the wavelength correction information. \begin{verbatim} IDL> eis_prep,paths,outdir='~/',/hkwavecorr,/default,/quiet,/save IDL> o = obj_new('eis_data','~/eis_l1_20071211_162413.fits') IDL> help,o->getaux_data(),/str IDL> corr = o->gethkwavecorr() IDL> oplot,corr.time \end{verbatim} \begin{thebibliography}{} \bibitem[Kamio et al.(2010)]{kamio10} Kamio, S., Hara, H., Watanabe, T., Fredvik, T. \& Hansteen, V. H., 2010, Sol. Phys., 266, 209 \end{thebibliography} \end{document}