HETGS Line Profiles  Validation: Wavelength  

WavelengthAs mentioned earlier, we utilize four generic functions (2 Gaussians and 2 Lorentzians) to parameterize a line profile. The position of each function generally follows the trend described by the grating equation, For each of the four functions, the centroid position is derived in the dispersion angle (in radian). At a very small angle, the dispersion angle is proportional to the wavelength value. So we fit a linear function to the anglevs.wavelength trend and then derive the scale of deviation of data points from the linear trend. The deviation points are fitted with piecewise quadratic functions in order to ensure continuity on position and slope at each junction defined by the measured data points (see Figures 7 and 8 for the results of the fitting). Then later this nonlinear term is added to the linear function to complete the analysis. I need to add the same kind of plots for minus 1st order (both HEG & MEG) here. (TBD)
This step in the processing also minimizes any significant nonlinearity in the wavelength/dispersion scale resulting from the instrumental design. Both HRCS and ACISS, for example, form their detector surfaces by ajoining 3 to 6 planersurface detectors (MCPs or CCDs), i.e., not precisely a Rowland torus (but mimicking closely by design). This type of nonlinearity can be implicitly taken out with our new procedure. Line Profile: WavelengthOnce again we run a number of monoenergetic marx simulations and analyze the line spectra by fitting a delta function with isis. The measured line centroid positions are then compared with the expected values to see the difference. Figures 9 and 10 illustrate the difference (Expected  Measured).I need to add the same kind of plots for minus 1st order (both HEG & MEG) here. (TBD)
Note that the measured offsets are below the defined instrumental accuracy of HEG or MEG. Addetum: J. Drake and D. P. Huenemoerder et al. inform me that the growing trend in offset seen in Figures 9 and 10 may be real and primarily due to some coding bug introduced within CIAO. The same effect is more noticeable in LETG/HRCS (which would seriously affect users' scientific analysis with LETG/HRCS).

This page was last updated Apr 11, 2003 by Bish K. Ishibashi. Accessibility To comment on it or the material presented here, send email to bish@space.mit.edu.