A cautionary note on the use of EESC-based regression analysis for ozone trend studies

Jayanarayanan Kuttippurath (1) , G. E. Bodeker (2) , H. K. Roscoe (3) , Prijitha J. Nair (4)

1 STRATO - LATMOS
2 Bodeker Scientific
3 BAS - British Antarctic Survey
4 NCESS - National Centre for Earth Science Studies [India]

Equivalent effective stratospheric chlorine (EESC) construct of ozone regression models attributes ozone changes to EESC changes using a single value of the sensitivity of ozone to EESC over the whole period. Using space-based total column ozone (TCO) measurements, and a synthetic TCO time series constructed such that EESC does not fall below its late 1990s maximum, we demonstrate that the EESC-based estimates of ozone changes in the polar regions (70–90°) after 2000 may, falsely, suggest an EESC-driven increase in ozone over this period. An EESC-based regression of our synthetic “failed Montreal Protocol with constant EESC” time series suggests a positive TCO trend that is statistically significantly different from zero over 2001–2012 when, in fact, no recovery has taken place. Our analysis demonstrates that caution needs to be exercised when using explanatory variables, with a single fit coefficient, fitted to the entire data record, to interpret changes in only part of the record.

Mots clés: PWLT regression; regression analysis; EESC-based regression; ozone recovery; ozone trends

Domaines: Physique Atmosphérique et Océanique [physics.ao-ph]

Jayanarayanan Kuttippurath, G. E. Bodeker, H. K. Roscoe, Prijitha J. Nair. A cautionary note on the use of EESC-based regression analysis for ozone trend studies. Geophysical Research Letters, 2015, 42 (1), pp.162-168. ⟨10.1002/2014GL062142⟩. ⟨hal-01130187⟩