1. High resolution solar reference spectrum Robert Voors KNMI

2. Introduction  Solar Reference spectra used for calibration  Wavelength calibration  Radiometric calibration  Current spectra do not meet requirements  Spectral resolution and sampling  Radiometric accuracy

3. Available solar reference spectra  Chance and Spurr:  composite high resolution  Kurucz:  model spectra  Thuillier et al:  composite medium resolution  SUSIM on ATLAS and UARS:  low resolution <410nm; long time period  ASTM solar standard spectrum  Composite medium resolution  GOME, SCIAMACHY…  Many more…

4. What’s the problem? Bladibla

5. Combine best of both worlds High resolution spectrum: assume relative line strengths are OK Low resolution spectrum: assume absolute radiances are OK.

6. Recipe for a new reference spectrum 1. Convolve a high resolution spectrum with poor radiometric calibration with the best slit function that is available for the lower- resolution solar reference spectrum 2. Interpolate the thus obtained high sampling, low-resolution spectrum on the wavelength grid of the low-resolution reference spectrum 3. Divide that spectrum by the low-res spectrum, to obtain the fraction by which to multiply the original high-resolution spectrum, used in 1 4. Interpolate the fraction from 3 to the high resolution wavelength grid

7. Recipe for a new reference spectrum The original high resolution spectrum How to find the best low(er)-resolution reference spectrum How to determine the instrument transfer function (slit function) of the low-resolution instrument. Wavelength calibration of the low- resolution spectra.

8. The original high resolution spectra (1)Balloon spectrum, corrected for atmospheric effects. (2)Ground-based measurements (Kurucz 1984), corrected for atm. effects

9. The original low resolution spectrum (1)Average of ~1 solar cycle of SUSIM data (2)LPMA balloon measurement from SCIAMACHY validation

10. The instrument transfer function SUSIM: scanning triangular LPMA: non-scanning flat/gaussian

11. Wavelength scale of low resolution spectra

12. Final result CaII H&K Comparison at 1 nm resolution

13. Solar variability

14. Discussion Solar variability not a serious issue, except near 270 nm and at strong Fraunhofer lines CaII H&K lines pose a problem New spectrum compares well with other solar spectra, e.g. from SCIAMACHY and Thuillier. Differences mostly < 4 percent Differences possibly due to Solar reference spectrum Absolute radiometric scale of Thuillier, OMI, SCIAMACHY etc. Assumed low-res slit function Wavelength calibration