ASPCS
 
Back to Volume
Paper: Determining Wavelength Scales for the EUVE Spectrometers
Volume: 61, Astronomical Data Analysis Software and Systems III
Page: 323
Authors: Boyd, W.; Jelinsky, P.; Olson, E.; Abbott, M.; Christian, C.
Abstract: The Extreme Ulraviolet Explorer satellite, launched by NASA on June 7, 1992, contains three EUV spectrometers with grazing-incidence optics built at UC Berkeley. Measurement of the resolution and determination of wavelength scales for these instruments has been complicated by several factors. Because the imaged spectral lines are distinctly non-Gaussian, simple Gaussian fits to the lines yield inaccurate results of the line widths, so the actual line width must be measured directly for each line; a custom IDL routine was written to make this process semi-automatic. The optical system has been ray-traced on a computer to produce a theoretical wavelength scale for each of the three spectrometers. The true scales differ from theory because of distortion due to the solid-state detector anodes, among other causes, and so real data must be used to characterize these deviations. Pre-launch spectra of EUV sources were taken but did not cover all three spectrometers completely, hence in-orbit spectra of real astrophysical plasmas have been necessary to fill in the gaps. Measurement of many line centroids for emission spectra from many angles of incidence have been analyzed with a simple chi-squared minimization technique and the theoretical scale's free parameters have been optimized to produce reasonably correct wavelength maps. These maps typically predict wavelengths to within one resolution element (RMS) which corresponds to about 0.5 A for the Short-Wavelength Spectrometer (70 - 180 A), 1.0 A for the Medium-Wavelength Spectrometer (150 - 370 A) and 2.0 A for the Long-Wavelength Spectrometer (290 - 750 A); the scales are much worse when they are applied near the detector edges. Recent efforts have focused on measuring and correcting departures from the theoretical maps, such as a curvature of the spectrum in the imaging direction and a compression of the spectrum along the dispersion axis near the detector edges; both of these effects are attributed to the detector design. To the extent that these distortions are repeatable, large-scale effects, we are able to correct for them, with a noticable improvement in the wavelengths of lines near the detector edges. The wavelength solution is then incorporated into the EUVE Guest Observer calibration data set. Calibrated spectra are produced by applying this solution on a photon-by-photon basis. This research has been supported by NASA contract NAS5-29298.
Back to Volume