3.9  Spatial resolution and binning along the slit

Even after all the possible tilt and wavelength corrections, the line profiles are not strictly gaussian, and residuals to gaussian fits show peculiar spatial patterns still not explained.

It has to be noted that the resolution of the instrument (FWHM of the Point Spread Function or PSF) is at least 5-6^¢¢. Therefore, the pixel resolution (1.68^¢¢) along the slit is much less than the spatial resolution, and the spectra can be binned without losing any spatial information.

Following suggestion from the author, many CDS studies have been designed and used to perform Y-binning on-board, to reduce telemetry time.

S. V. H. Haugan discusses in Anomalous Line Shifts from Local Intensity Gradients on the SOHO/CDS NIS Detector (Solar Physics 185, 275, 1999) the NIS PSF. He suggests that the NIS spectrograph PSF must be slightly elliptical and rotated.

Pauluhn, A. et al., 1999, Applied Optics 38, 7035, also discuss the issue, deriving the pre-loss Point Spread Function (PSF) for CDS/NIS from a comparison with Sumer data. They suggest a slightly elliptical form with FWHM(x) = 6" and FWHM(y) = 8" (when the 4"x240" slit is used).

The post-recovery PSF has apparently degraded. See details in the CDS user guide.

Figure 15: [Figure from the CDS user guide] CDS wide-slit observation of the Mercury coronal transit on Nov 15 1999, and simulations with a telescope PSF Gaussian with (x,y) FWHM of (7.0,13.5) arcsec and a rotation of 39^o from the vertical.

Figure 16: [Figure from the CDS user guide] CDS wide-slit observation of the Mercury coronal transit on Nov 15 1999, and simulations with a telescope PSF Gaussian with (x,y) FWHM of (7.0, 21.5) arcsec and a rotation of 104^o from the vertical.