Abstract. Dobson and Brewer spectrophotometers are the primary, standard
instruments for ground-based ozone measurements under the World
Meteorological Organization's (WMO) Global Atmosphere Watch program. The
accuracy of the data retrieval for both instruments depends on a knowledge of
the ozone absorption coefficients and some assumptions underlying the data
analysis. Instrumental stray light causes nonlinearity in the response of
both the Brewer and Dobson to ozone at large ozone slant paths. In addition,
it affects the effective ozone absorption coefficients and extraterrestrial
constants that are both instrument-dependent. This effect has not been taken
into account in the calculation of ozone absorption coefficients that are
currently recommended by WMO for the Dobson network. The ozone absorption
coefficients are calculated for each Brewer instrument individually, but in
the current procedure the effect of stray light is not considered. This
study documents the error caused by the effect of stray light in the Brewer
and Dobson total ozone measurements using a physical model for each
instrument. For the first time, new ozone absorption coefficients are
calculated for the Brewer and Dobson instruments, taking into account the
stray light effect. The analyses show that the differences detected between
the total ozone amounts deduced from Dobson AD and CD pair wavelengths are
related to the level of stray light within the instrument. The discrepancy
introduced by the assumption of a fixed height for the ozone layer for ozone
measurements at high latitude sites is also evaluated. The ozone data
collected by two Dobson instruments during the period of December 2008 to
December 2014 are compared with ozone data from a collocated double
monochromator Brewer spectrophotometer (Mark III). The results illustrate the
dependence of Dobson AD and CD pair measurements on stray light.