19th International CODATA Conference
Category: Multi-disciplinary Use

Studies of the Earth's Ozone from Historic Stellar Spectra

Dr. R. Elizabeth Griffin (Elizabeth.Griffin@nrc.gc.ca)
York University, Toronto, Canada, Herzberg Institute for Astrophysics, Canada

The re-working of certain astronomical observations dating back 50-80 years is yielding fascinating results, not for Astronomy but for Atmospheric Science.

The stratospheric ozone layer protects every aspect of life on Earth, at least upon land.  The growing concern at recent downward trends in ozone concentrations stems from the assumption that all the losses arise from anthropogenic causes.  Natural variations, if any, cannot be determined separately in present-day data, so we cannot assess how much of the decrease is due to anthropogenic causes, nor the extent to which it may be controlled (and possibly reversed).  Evidence of natural variations has to be sought in historic data.

Historic data are sparse; world-wide ozone monitoring did not really start until 1957, by which time the destructive chemicals (principally Cl and NO compounds) were already being released into the atmosphere.  Prior to 1957 only one station (at Arosa, in Switzerland) was making daily ozone measurements, and for most of that time only a single instrument was being used.  Long-term variations can therefore have been confused with systematic errors (indeed, a long-term drift in the instrumental calibration was suspected during the 1950s), and evidence of natural long-term variations in ozone levels could thus have been lost.  "Redundant" datasets need to be located for checking the Arosa measurements.

Ground-based astronomical observations unavoidably carry signatures of the Earth's atmosphere, and at certain wavelengths ozone bands are superimposed on the spectra of stars.  A new database of historic stellar spectra, constituting a unique "redundant" dataset, is now being analysed for ozone concentrations. The results to date from the astronomical data agree closely with the Arosa ones in the 1960s and 70s, but indicate a significant fall during the 1950s which is not apparent in the Arosa data.

The paper describes this original interdisciplinary project, which so perfectly exemplifies the adage, "one man's Noise is another man's Signal".