An Analysis of the Power Spectra of Total Ozone Data

View More View Less
  • 1 System Sciences Division, Computer Sciences Corporation, Silver Spring MD 20910
© Get Permissions
Full access

Abstract

An analysis is made of the power spectrum of total ozone data using the window closing procedure of Jenkins and Watts (1968). It is concluded that it is presently difficult to distinguish statistically between true periodicities in the data and a “red noise” spectrum for periods greater than one year. Nevertheless, statistical procedures are outlined to distinguish between red noise and a significant quasi-biennial oscillation (QBO) in ozone. Some evidence for a QBO in ozone was found in the Arosa total ozone data, but the period or periods of the QBO could not be firmly established due to the short time span of data available.

Abstract

An analysis is made of the power spectrum of total ozone data using the window closing procedure of Jenkins and Watts (1968). It is concluded that it is presently difficult to distinguish statistically between true periodicities in the data and a “red noise” spectrum for periods greater than one year. Nevertheless, statistical procedures are outlined to distinguish between red noise and a significant quasi-biennial oscillation (QBO) in ozone. Some evidence for a QBO in ozone was found in the Arosa total ozone data, but the period or periods of the QBO could not be firmly established due to the short time span of data available.

Save