Deep-convective influence on the upper troposphere-lower stratosphere composition in the Asian monsoon anticyclone region: 2017 StratoClim campaign results

The StratoClim stratospheric aircraft campaign took place in summer 2017 in Nepal (27 July 10 August) and provided for the first time a wide dataset of observations of air composition inside the Asian monsoon anticyclone (AMA). In the framework of this project, with the purpose of modelling the injection of pollutants and natural compounds into the stratosphere, we performed a series of diffusive back trajectory runs along the flights' tracks. The availability of in situ measurements of trace gases has been exploited to evaluate the capability of the trajectory system to reproduce the transport in the upper troposphere lower stratosphere (UTLS) region. The diagnostics of the convective sources and mixing in the air parcel samples have been derived by integrating the trajectory output with highresolution observations of cloud tops from the Meteosat Second Generation (MSG1) and Himawari geostationary satellites. Back trajectories have been calculated using meteorological fields from European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis (ERA-Interim and ERA5) at 3 and 1 h resolution, using both kinematic and diabatic vertical motion. The comparison among the different trajectory runs shows, in general, a higher consistency with observed data as well as a better agreement between the diabatic and kinematic version when using ERA5-based runs with respect to ERA-Interim. Overall, a better capacity in reproducing the pollution features is finally found in the diabatic version of the ERA5 runs. We therefore adopt this setting to analyse the convective influence in the UTLS starting from the StratoClim observations. A large variety of transport conditions have been individuated during the eight flights of the campaign. The larger influence by convective injections is found from the continental sources of China and India. Only a small contribution appears to be originated from maritime regions, in particular the South Pacific and the Bay of Bengal, which, unexpectedly, was not particularly active during the period of the campaign. In addition, a mass of clean air injected from a typhoon has also been detected at around 18 km. Thin filamentary structures of polluted air, characterized by peaks in CO, are observed, mostly associated with young convective air (age less than a few days) and with a predominant South China origin. The analysis revealed a case of direct injection of highly polluted air close to the level of the tropopause (anomalies of around 80 ppbv injected at 16 km) that then kept rising inside the anticyclonic circulation. Due to the location of the campaign, air from continental India, in contrast, has been only observed to be linked to air masses that recirculated within the anticyclone for 10 to 20 d, resulting in a lower concentration of the trace gas. The analysis of a flight overpassing an intense convective system close to the southern Nepalese border revealed the injection of very young air (few hours of age) directly in the tropopause region (∼ 18 km), visible in the trace gases as an enhancement in CO and a depletion in the O3 one. From the whole campaign, a vertical stratification in the age of air is observed: up to 15 km, the age is less than 3 d, and these fresh air masses constitute almost the totality of the air composition. A transition layer is then individuated between 15 and 17 km, where the convective contribution is still dominant, and the ages vary between 1 and 2 weeks. Above this level, the mean age of the air sampled by the aircraft is estimated to be 20 d. There, the convective contribution rapidly decreases with height and finally becomes negligible around 20 km.