 |
 |
| CIRES' Annual, Institute-wide Symposium [ Home | Poster Submission | Agenda | Posters | Archives ] |
|
Science Rendezvous > Posters On the retrieval of effective temperature of the ozone profile from ground-based spectroradiometric measurements Peter Kiedron (1) and Joseph Michalsky (2) (1) CIRES (2) NOAA Ozone column from Dobson and Brewer spectroradiometers is obtained from extinction measurements at a few discrete wavelengths. Ozone cross-sections are a function of temperature. Dobsons and Brewers use Bass and Paur cross-sections at 227°K and 229°K, respectively. If ozone sonde data are not available, these temperatures cannot be verified. Subsequently, the absolute error in ozone column measurements cannot be estimated. Since June 2003 we have operated the rotating shadowband spectroradiometer (RSS) at Table Mountain, Boulder, Colorado. The RSS provides one-per-minute quasi-instantaneous direct flux spectra over 734 pixels covering 297-385 nm with a resolution between 0.3-0.6 nm and stray light less than 5*10-6. Within the 305-340 nm range, where noise is much lower than the measured optical depth, the ozone cross-sections can be determined with sufficient accuracy to estimate the temperature. We also compare the cross-sections from Brion-Malicet and Burrows. The limiting factors of temperature retrieval are noise and aerosol optical depth. Several approaches are used: (1) optical depth obtained with calibration from recursive Langley regressions is analyzed, (2) differential optical depth for two air masses is used and (3) 1st and 2nd derivatives of optical depth are calculated. The last approach minimizes aerosol effects at the expense of noise. The effective temperature can be defined as the profile’s first moment or as the best single temperature cross-section fit to the effective cross-section for a given profile and SZA. The latter is slightly SZA dependent. These two temperatures are correlated, but different. We compare them with data from ozone sondes launched in Boulder. |
