Next: 5. Verification of the Up: olp101sws Previous: 3. Analysis

4. Choice of Models

Although the calibration problem of OLP 10 was a deficiency in the models of the standards used, we believe that this deficiency is quite localized to wavelength regions where the models have the least information (CO fundamental and SiO fundamental). All the models used have been extensively validated against band 1 data (2.3 to 4 m). The models are fully described in the PhD Thesis of L. Decin [4] and references therein. Since the models are well documented and validated at the shortest wavelengths, we prefer to use one consistent set throughout the entire short wavelength section of SWS.

Furthermore, the absolute calibration of the synthetic spectra applied the same photometry as the photometry used by Cohen for the absolute calibration of composites and templates. The scaling of the models to the photometry of the composites guarantees the consistency with the use of the Cohen composites and templates for the long-wavelength section.

For the early-type stars used in band 2A and the short-wavelength end of 2C, the differences between the composites/templates by Cohen and the synthetic spectra by Decin are less than two percent. We are confident that the models themselves are not introducing 2% features in any AOT band of the short wavelength section of SWS.

---

Next: 5. Verification of the Up: olp101sws Previous: 3. Analysis

Improvements in the OLP 10 RSRF and Flux Calibration of ISO-SWS Bands 2A and 2C