Measurements of halogen oxides by scattered sunlight differential optical absorption spectroscopy

This work describes measurements of the tropospheric halogen species iodine monoxide (IO) and bromine monoxide (BrO) with multi-axis differential optical absorption spectroscopy (MAX-DOAS) at various locations: Svalbard (79 deg N), Andoya (69 deg N), List (55 deg N), Crete (35 deg N), and Maldives (5 deg N). Furthermore, the radiative transfer model SCIATRAN was used to investigate the sensitivity of the MAX-DOAS technique towards tropospheric absorbers as well as to estimate concentrations of IO and upper limits only for BrO. IO was detected at Svalbard, List, and Maldives with concentrations of 0.4 ppt, 2.2 ppt, and 2.8 ppt, respectively, but was below the average detection limits of 0.6 ppt and 1.3 ppt for Andoya and Crete, respectively. Uncertainties are estimated to be in the range of 2-3 ppt. These concentrations are averaged horizontally as well as vertically assuming a well-mixed surface layer of 500 m thickness or in the case of Svalbard and List 200 m. On a local scale, higher concentrations can occur. Tropospheric BrO could not be measured at any location outside the polar spring. Minimum detection limits for BrO were 2.2 ppt. Ground-based zenith-sky DOAS measurements of stratospheric chlorine dioxide have been performed at the Arctic sites Svalbard and Summit (73 deg N) as well as at the mid-latitudinal site Bremen (53 deg N) and used to validate SCIAMACHY OClO data for the exceptionally cold stratospheric spring 2005. OClO was also derived with a chemical stacked box model. The agreement of all three data sets is excellent for the time of overpass of the satellite instrument, i.e. 10 AM. However, the ground-based measurements could not be reproduced with the model simulations for large solar zenith angles as well as for large concentrations. Sensitivity studies have been performed with the chemistry model and this exercise demonstrated that the measured OClO columns cannot be explained within the known uncertainties of the model parameters including the involved photochemical data.