Stutz, Jochen, Werner, Bodo, Spolaor, Max, Scalone, Lisa, Festa, James, Tsai, Catalina, Cheung, Ross, Colosimo, Santo F., Tricoli, Ugo, Raecke, Rasmus, Hossaini, Ryan, Chipperfield, Martyn P., Feng, Wuhu, Gao, Ru-Shan, Hintsa, Eric J., Elkins, JamesW., Moore, Fred L., Daube, Bruce, Pittman, Jasna, Wofsy, Steven, Pfeilsticker, Klaus, Stutz, Jochen, Werner, Bodo, Spolaor, Max, Scalone, Lisa, Festa, James, Tsai, Catalina, Cheung, Ross, Colosimo, Santo F., Tricoli, Ugo, Raecke, Rasmus, Hossaini, Ryan, Chipperfield, Martyn P., Feng, Wuhu, Gao, Ru-Shan, Hintsa, Eric J., Elkins, JamesW., Moore, Fred L., Daube, Bruce, Pittman, Jasna, Wofsy, Steven, and Pfeilsticker, Klaus
Observations of atmospheric trace gases in the tropical upper troposphere (UT), tropical tropopause layer (TTL), and lower stratosphere (LS) require dedicated measurement platforms and instrumentation. Here we present a new limb-scanning Differential Optical Absorption Spectroscopy (DOAS) instrument developed for NASA's Global Hawk (GH) unmanned aerial system and deployed during the Airborne Tropical TRopopause EXperiment (ATTREX). The mini-DOAS system is designed for automatic operation under unpressurized and unheated conditions at 14-18 km altitude, collecting scattered sunlight in three wavelength windows: UV (301-387 nm), visible (410-525 nm), and near infrared (900-1700 nm). A telescope scanning unit allows selection of a viewing angle around the limb, as well as realtime correction of the aircraft pitch. Due to the high altitude, solar reference spectra are measured using diffusors and direct sunlight. The DOAS approach allows retrieval of slant column densities (SCDs) of O-3, O-4, NO2, and BrO with relative errors similar to other aircraft DOAS systems. Radiative transfer considerations show that the retrieval of trace gas mixing ratios from the observed SCD based on O-4 ob-servations, the most common approach for DOAS measurements, is inadequate for high-altitude observations. This is due to the frequent presence of low-altitude clouds, which shift the sensitivity of the O-4 SCD into the lower atmosphere and make it highly dependent on cloud coverage. A newly developed technique that constrains the radiative transfer by comparing in situ and DOAS O-3 observations overcomes this issue. Extensive sensitivity calculations show that the novel O-3-scaling technique allows the retrieval of BrO and NO2 mixing ratios at high accuracies of 0.5 and 15 ppt, respectively. The BrO and NO2 mixing ratios and vertical profiles observed during ATTREX thus provide new insights into ozone and halogen chemistry in the UT, TTL, and LS.