Skip to main content
European Commission logo
français français
CORDIS - Résultats de la recherche de l’UE
CORDIS
CORDIS Web 30th anniversary CORDIS Web 30th anniversary
Contenu archivé le 2024-06-18

Stratospheric and upper tropospheric processes for better climate predictions

Final Report Summary - STRATOCLIM (Stratospheric and upper tropospheric processes for better climate predictions)

Executive Summary:
Anthropogenic emissions of greenhouse gases (GHGs), aerosols and their precursors are all drivers of changes in Earth’s climate. Growing evidence indicates that changes in the chemical composition of the Upper Troposphere and Stratosphere (UTS) also play a key role in affecting surface climate.

StratoClim planned, organized, and conducted an Aircraft Field Campaign (AFC) and interprets the results in order to characterize major processes which dominate particle and trace gas transport from one of the most polluted regions of the world into the upper troposphere and lower stratosphere (UTLS), within and above the Asian Summer Monsoon Anticyclone (ASMA).
The AFC has supplied the very first detailed and almost complete data set on atmospheric parameters of the ASMA 2017, an atmospheric domain that had previously not been characterized in any detail but is of high relevance to the global climate development. The data has been quality checked and is now available to all partners from the project database and starting from summer 2020 also to the public. First studies based on the interpretation of the aircraft data are published already and currently more than 30 publications are in preparation by the StratoClim community alone. These will appear in the peer reviewed literature until end of 2020, most of them in a dedicated special issue in Atmos. Chem. Phys.. Apart from that, the unique dataset will provide material for many more studies once released to the public.

Further, StratoClim established a new Atmospheric Observatory in Palau, in the tropical West Pacific (TWP) warm pool at 7.3°N, 134.5°E. The Palau Atmospheric Observatory has been operational since January 2016 with major addition of a second lab container in 2018, hosting a comprehensive instrumental setup for O3, H2O, NO2 and OCS. The measurements in Palau have been complemented with ballooning campaigns in Nainital in India, Dhulikhel in Nepal and Bhola island in Bangladesh. The balloon measurements yield unique datasets for the annual cycle of especially O3, and are included a cooperative effort with the US NASA/NOAA campaign POSIDON in October 2016, which was awarded with a NASA group achievement award.

A variety of new data have been integrated to the regional and global models in order to enhance the process understanding of various trace gases in the atmosphere. New in-situ data has been used to calculate and model entry pathways of specific air masses into the stratosphere and model results together with aircraft data and satellite observations have been used to improve our understanding of the processes controlling the role of stratospheric aerosol, water vapour and ozone on recent climate and future climate change.


Project Context and Objectives:
Anthropogenic emissions of greenhouse gases (GHGs), aerosols and their precursors are all drivers of changes in Earth’s climate. Growing evidence indicates that changes in the chemical composition of the Upper Troposphere and Stratosphere (UTS) also play a key role in affecting surface climate. On a decadal timescale, there is strong evidence that variability in the UTS affects the troposphere and surface climate. The impact of changes in the UTS on surface climate is inextricably linked to many other components of the climate system. However, the representation of these components in current Chemical Climate Models (CCMs) and Earth System Models (ESMs) is still largely inadequate.

The overall goals of StratoClim are (a) to quantitatively assess the role of the UTS in climate change, and (b) to improve climate projections by developing and including within ESMs new, interactive modules for stratospheric aerosols and ozone and by improving our understanding of UTS water vapour variations and the representation of upper tropospheric clouds.

To reach these overarching goals the objectives of StratoClim are:
(1) to develop the scientific basis for including the climate relevant components of the UTS as interactive modules in ESMs,
(2) to construct and implementing such modules,
(3) to assess the UTS’s role in climate,
(4) to produce new and better climate model projections.

To achieve these objectives StratoClim combines a large-scale tropical aircraft campaign, longer-term operation of a tropical measurement station, satellite data analysis including development of new satellite data products, process and regional modelling, global modelling with CCMs and ESMs, studies of the socioeconomic Climate Change implications of the findings and stakeholder fora, expert panels and public outreach programs.

Work will focus on two areas: (a) improving the understanding of the chemical and dynamical processes that determine the composition of the UTS and the formation, loss and redistribution of ozone, aerosol, water vapour and clouds, and how these processes will alter under climate change; (b) developing model tools that allow to fully include the interactive feedbacks from UTS ozone and aerosol on surface climate and use of these models to produce improved climate predictions.

The project will integrate data from in situ measurements, from space-based observations and modelling to improve our understanding of key atmospheric processes, and how these processes are initiated or affected by natural and anthropogenic emissions, which interact, also through feedback mechanisms, with ecosystems and climate. Results will directly improve the representation of these processes in relevant models and the assessments of socio-economic implications, contributing to evidence-based decision-making in EU policies for environment and climate and support EU climate actions.

The project long-term database will be populated gradually, as final versions of the data become available. All data will be freely available via the HALO database at https://halo-db.pa.op.dlr.de/mission/101 by the summer 2020 at the latest.

Project Results:
Work Package 1 (WP1) of StratoClim planned, organized, and conducted the Aircraft Field Campaign (AFC) and (together with WP4 and WP5) interprets the results in order to characterize major processes which dominate particle and trace gas transport from one of the most polluted regions of the world into the upper troposphere and lower stratosphere (UTLS) within and above the Asian Summer Monsoon Anticyclone (ASMA). Focal processes include:
- Distribution of chemical species and aerosols in the ASMA and the surrounding atmospheric domains and the relevant exchange processes,
- Vertical transport to and in the ASMA by large scale ascent vs. convective transport,
- Chemical transformation of trace gases upon transport (Processing),
- Formation of particles and clouds from trace gases and their reaction products (New Particle Formation, NPF),
Due to their direct radiative effects and their coupling with e.g. ozone chemistry these processes are key for understanding climate feedback mechanisms of this atmospheric domain and need to be represented in global models to improve the predictability of the global climate evolution and its local consequences.
However, before StratoClim the ASMA was one of the most under-researched regions world-wide and beneath heavily averaged satellite and some balloon data no suitable complete atmospheric datasets existed to test our process understanding. With the AFC carried out in July/August 2017 from Kathmandu, Nepal, this situation has been improved dramatically.
In order to probe the complete vertical domain of interest from around 12 up to 20 km altitude the established and unique high-flying research aircraft M55-Geophysica was employed as the platform of choice. The aircraft has been substantially overhauled before the campaign with considerable support of the StratoClim partner institutions. However, to enable measurements of all relevant gas-phase and particle parameters the instrument payload onboard Geophysica had to be upgraded considerably, especially with respect to:
- Chemical composition as well as phase-state and shape of atmospheric particles (solid and liquid) of all relevant sizes (<10nm to >1mm).
- Sulfur containing gases (OCS, SO2, H2SO4) that play an important role as aerosol precursors locally (SO2, H2SO4) and after further transport to the stratosphere (OCS)
- Isotopic composition of water vapor in order to probe condensation and evaporation processes of the most important climate gas.
- Combination of fast and spatially highly resolved in-situ measurements with remote-sensing of atmospheric composition giving 2- and even 3-dimensional distributions.

Aircraft Instrument Upgrades
To achieve these capabilities four new gas-phase as well as seven new particle instruments were developed or adapted, characterized, and integrated onboard the aircraft. Two more instruments were upgraded considerably. All these activities were covered from institutional funds and using synergies in particular with the ERC Advanced Grant project EXCATRO of MPI Mainz (for most of the particle instruments) and by inclusion of associate partners like the Univ. Chicago for the water isotopologues instrument. For almost complete particles characterization for the first time a 3-channel single particle and bulk aerosol mass spectrometer was flown. In-situ aerosol size distributions were measured down to 65 nm particle size and cloud particle shapes were for the first time measured by a holographic cloud particle imaging instrument in the ASMA and at such altitudes. The new instruments were tested and improved during and after a combined scientific and test campaign from Kalamata, Greece, in August/September 2016. This was an important first step for the later successful deployment from Kathmandu. Besides that, there hardly is any other operational aircraft platform rivalling the unique capabilities of M55 Geophysica in the analysis of atmospheric chemistry, dynamics, and microphysics, now.

Campaign Organization and Execution
As expected the realization of the AFC on the Indian subcontinent was a highly challenging task to the organization and logistics team. A strong local cooperation had been established with the Indian Institute of Tropical Meteorology (IITM) in Pune starting even before StratoClim and support from its head organization the Ministry of Earth Sciences (MOES) in Delhi had been secured. However, the first attempt to execute this campaign from Nagpur, India, during the Asian Summer Monsoon (ASM) in 2016 failed due to missing clearances by the Indian Ministry of Defense. Therefore, a phase-one campaign was organized from Kalamata, Greece, and successfully completed to measure AM anticyclone outflow features over the Mediterranean and in parallel performing instrument tests (see above) during a total of three flights. After positive signs of logistical recovery from the 2015 earthquake from the highly supportive German Embassy in Nepal the main campaign in 2017 was relocated to Kathmandu International Airport (TIA) in Nepal. With help of an influential local agent (Lt. Col. D. K. Karki) all clearances by the Nepalese authorities were obtained based on the scientific cooperation established with the Nepalese Academy of Science and Technology (NAST), Lalitpur, Kathmandu. A Memorandum of Understanding as well as an Implementation Agreement were concluded between AWI and NAST. Other cooperation partners in Nepal were the International Centre for Integrated Mountain Development (ICIMOD), Kathmandu, with a strong atmospheric boundary layer research program and the Department of Hydrology and Meteorology (DHM), Kathmandu, for local meteorological support. For Bangladesh our associate partner Dhaka University supported us to get clearances. With the help of the Civil Aviation Authority of Nepal flight permissions for India could be obtained on a single flight basis giving us access to most of the Indian subcontinent. All important logistic prerequisites could be put in place by our subcontractors Enviscope and FinkCAS GmbH supported by D. K. Karki.
For efficient and smooth conduction of the main campaign several dedicated tools operated by members of the WP1, WP4 and WP5 teams were in place locally at the Kathmandu base or operated remotely during the campaign in order to predict and alert on the scientifically most interesting situations. Real-time information of the MLS-AURA microwave limb sounder were provided by M. Santee et al., JPL. Local flight planning meetings were held every day to analyze the meteorological development and identify new flight options. Detailed flight planning was done employing the "Mission Support System" (MSS) an open source software project that had been heavily updated for the Geophysica deployment. Its local client used offline webmap servers operated by Jülich and LMD in order to reduce the local internet load.
As a result of the efforts detailed above eight local science flights covering all main science objectives could be successfully completed during the StratoClim ASMA campaign. Very good overall instrument performance in an extreme environment (up to 40°C during flight preparation on the ground and below 80°C around the tropopause) provided a wealth of data on all relevant processes. Regarding the extremely challenging circumstances this exemplifies an outstanding performance of the complete campaign group. Data has been quality checked and is now available in a restricted data base exemplifying the very first detailed data set on the ASMA. Wherever possible aircraft data have been compared internally or with parallel satellite or balloon-borne data with good results (e. g. CO, water vapour, ambient temperature). Data will become available to the general science community and public until summer 2020 on a long-term data base hosted by DLR (https://halo-db.pa.op.dlr.de/mission/101).

Major Scientific Results
Interpretation of the unique data set provided from the AFC has yielded several important early results which are presented below and have partly been published already.
Based on the tracer measurements inside the ASMA a number of transport processes could be identified. Profiles of CO2, CO, N2O, and CH4 (Volk et al., 2019, von Hobe et al., 2019, S. Viciani et al., 2018) reveal clear signatures of fast convective transport of polluted boundary layer air up to at least 370 K (ca. 13 km) which is the level of main outflow and still clearly below the local cold point tropopause (380-390 K, ca. 17 km). From the main outflow level, air can be transported out of the ASMA isentropically, e. g. by eddy-shedding. Above that the profiles indicate continuous slow radiative upwelling through the cold point tropopause to around 400-410 K (ca. 19 km) within about 2 months consistent with vertical velocities and heating rates from reanalysis data. This proceeds in relative isolation from the older air of the extratropical stratosphere. Above 410 K horizontal mixing with older air rapidly dilutes the upwelling anticyclonic air. The age of air increases to about 2.5 years at altitudes around 460 K (ca. 21 km). Many signatures of convective input of tropospheric air masses have been observed around the CPT level

Signs of overshooting convection at altitudes in excess of 400 K have been detected within the water vapour measurements (gas-phase as well as total water) as well as in cloud particle data by the presence of thin cirrus clouds. It must be mentioned here that these high reaching overshoots are not reflected clearly in the CO data discussed above, which is an issue under discussion. However, water vapour is a much more sensitive tracer for convectively injected air than CO due to its abundance in the troposphere being 2 to 3 orders of magnitude higher than in the UTLS. On the other hand, it is much harder to interpret due to its phase conversions. A high-resolution convective model study for the origin of a moistening signature at 400 K (18.5 km) altitude in flight 7 originating from south China has been conducted and submitted for publication (Lee et al., 2018). It is able to reproduce the measurement quite well. Based on water vapour two meteorological regimes have been identified: a wet and warm phase (in the tropopause region) during the first four flights with much less high reaching convective activity and a cold and dry period towards the end of the campaign with much more convective activity. During the last period frequent cirrus particles at altitudes up to more than 400 K (ca. 18.5 km) above the cold point tropopause along with reduced values of gas-phase water were observed indicating some influence of higher reaching convection. Record values of ice water content (2500 ppmv) and ice particle size (0.6 mm) were measured during flight 8 (10.08.2017) within active overshoots at tropopause level (385 K, ca. 17 km). According to meteorological situation convective overshoots may cause moistening or even strong dehydration down to the saturation level depending on whether the injected ice particles sublimate or sediment (after growing by collecting water vapour). Both situations have been observed and are analyzed (Khaykin et al., 2019 and 2020, Krämer et al., 2019 and 2020). A high-resolution convective model study for the origin of a moistening signature at 400 K (18.5 km) altitude in flight 7 originating from south China has been conducted and submitted (Lee et al., 2018).

Carbonyl sulfide (OCS), the most abundant atmospheric sulfur gas (~500 ppt typical), was significantly enhanced in the ASMA with mixing ratios around 600 ppt up to 17 km indicating strong sources in the ASM region. The vertical gradient above the CPT was smaller than expected (Kloss et al., 2020). SO2 mixing ratios range between 30-50 pptv and show distinct enhancements of up to 300 ppt at the main convective outflow level between 12-14 km altitude. Interestingly, SO2 signatures of convective outflow were also found in the vicinity of the cold point tropopause (16.5-18 km). In the lower stratosphere also enhanced H2SO4 concentrations were detected (compared to balloon observations at mid-latitudes) in line with the high SO2 mixing ratios (Schlager et al., 2019).

Analysis and interpretation of the data obtained by the unique GLORIA limb-imaging infrared spectrometer along with the new ERICA aerosol mass spectrometer has brought major progress in the understanding of the Asian Tropopause Aerosol Layer (ATAL) which was first identified in 2008 but had remained unexplained since (Vernier et al., 2011). A spectral band observed by GLORIA had also been observed in the same region and season before by the CRISTA satellite experiment (Cryogenic Infrared Spectrometers and Telescopes for the Atmosphere, Offermann et al. 1999). Through intensive laboratory measurements at the AIDA cloud and aerosol chamber at KIT, this signature could be identified as a deformation band of solid ammonium nitrate (NH4NO3, “AN”) particles. The AIDA experiments also allowed to develop a quantitative scheme for the retrieval of vertical profiles of ammonium nitrate mass density from infrared limb-sounding observations. The derived AN profiles show clear enhancements over the altitude regime of the ATAL layer (15-17 km) in good agreement with the in-situ aerosol observations by ERICA. The ERICA observations furthermore indicate that most of the nitrate particles observed also contained sulfate which is a prerequisite for AN crystallization as also shown in the AIDA experiments. The GLORIA observations also clearly identified an extended cloud of strongly enhanced Ammonia (NH3) of up to 1 ppbv in mixing ratio located about 1 km below the AN layer. The ammonia cloud itself exhibited enhanced AN around its boundary. Ammonia is the precursor gas for formation of ammonia aerosols and was detected in the ASMA for the first time in MIPAS satellite data, however, at much lower abundance (Höpfner et al., 2016). Trajectory calculations indicate that the region of origin of the observed ammonia cloud are convective events over northwest India and northeast Pakistan, a region known for high ammonia emissions from agriculture. This highly interesting study, together with the analysis of CRISTA and MIPAS satellite data which show that the ATAL has been present since at least 1997, has already been published in (Höpfner et al., 2019).

Summarizing WP1 of StratoClim has supplied the very first detailed and almost complete data set on atmospheric parameters of the ASMA 2017, an atmospheric domain that had previously not been characterized in any detail but is of high relevance to the global climate development. The data has been quality checked and is now available to all partners from the project database and starting from summer 2020 also to the public. First studies based on the interpretation of the aircraft data are published already mostly in close cooperation with WP4 and WP5 and currently more than 30 publications are in preparation by the StratoClim community alone. These will appear in the peer reviewed literature until end of 2020, most of them in a dedicated special issue in Atmos. Chem. Phys.. Apart from that the unique dataset will provide material for many more studies once released to the public.



The WP2 established the Palau Atmospheric Observatory (PAO), located in the tropical West Pacific (TWP) warm pool at the Palau Community College in Koror/Palau at 7.3°N, 134.5°E. The PAO has been operational since January 2016 with the major addition of a second lab container in 2018, hosting a comprehensive instrumental setup. Observations were performed continuously with local support and in campaigns with visiting scientists and technicians and yield a unique dataset, including a cooperative effort with the US NASA/NOAA campaign POSIDON in October 2016, which was awarded with a NASA group achievement award.

O3 and H2O balloon activities
The unique 3-year-dataset of tropospheric O3 balloon measurements from the PAO gives important insights into the seasonal variability of this trace gas in the warm pool region. The observations captured the significant seasonal signals of the annual O3 cycles in the TTL (maximum in August) and in the mid-troposphere (minimum between August and October). Taking into account the close coupling between tropospheric O3 and the oxidizing capacity (OH) we can conclude, that the transport in the TWP to the stratosphere with longest lifetimes for different chemical species occurs in late boreal summer/fall. The impact of the strong El Nino in spring 2016 is reflected in higher O3 volume mixing ratios (VMR) in the upper troposphere and the driest and ozone-richest mid-troposphere throughout the record. We are currently analyzing the relation between O3 and H2O in mid-tropospheric air filaments. Our hypothesis on five different ways of transport related to these filaments will be tested using trajectory analysis in the near future and will contribute to resolving a current scientific dispute (e.g. Pan et al., 2015, Anderson et al., 2016, Randel et al., 2017).

Trace gases by FTIR
Measurements of atmospheric trace gases using the solar absorption spectrometry in the infrared started in December 2015. The observations require the direct sunlight and have been performed remotely in the late evening or early morning remotely from Bremen. The analysis so far concentrated on OCS and O3. Data for OCS in Palau have been compared to other sites of the NDACC. A first interpretation reveals enhanced values of OCS in the TTL, indicating an enhanced uplift in the warm-pool region. The retrieval of tropospheric O3 yields very low values, in agreement with the balloon sonde data.

Aerosol and cloud Lidar
The Italian lidar system MuLid measured 3650 profiles during one month in spring 2016 before instrumental failure. Optically thick and thin clouds were detected in two preferred regions in the TTL. The additional analysis of temperature fluctuations showed the influence of large-scale equatorial wave activity, which is able to modulate cloud presence and thickness. Further studies of cloud occurrence with respect to temperature lapse rate will assess the possible influence of gravity waves with the convective history of clouds.
The ComCAL run by AWI to detect aerosol and subvisible clouds in the UTLS region above Palau is able to work even under daylight conditions. Operations are automatized to 3-hour measurements at night time and monitored remotely. Installed in Oct 2018, a technical problem related to polarization measurements has compromised data analysis so far, but repairs are planned for fall 2019.

MaxDOAS instrument PANDORA
The MaxDOAS instrument Pandora 2-S was installed in Palau by AWI in 2016 as part of an ESA-funded initiative of instruments at remote locations within the Pandonia network (www.pandonia.net). Data are submitted in near-real-time to the Pandonia database and will be used for satellite validations by ESA. Comparison of the O3, NO2 and AOD (Aerosol Optical Depth) retrievals from the instrument with coinciding balloon soundings, FTIR and ComCAL measurements is aspired in the future.

Bangladesh: O3 balloon sondes
AWI performed 14 balloon sonde launches from Bhola Island, Bangladesh, in cooperation with the University of Dhaka as part of the StratoClim campaign. Launches were performed during the Asian summer monsoon season in July and August 2016 and July and August 2017.

In-Situ instrument MICA (Bangladesh/Palau)
The MICA instrument by FZJ (a ground-based version of the AMICA instrument that was deployed during the two StratoClim aircraft campaigns) measures the trace gases carbonyl sulfide (OCS), carbon monoxide (CO), carbon dioxide (CO2) and water vapour (H2O) by integrated cavity output spectroscopy. The instrument was deployed at Bhola Island alongside the ballooning campaign in 2016 and in Palau in September 2017. The Palau observations revealed background OCS values (440-500 ppt) not indicating a dominant local/regional source or sink. In Bangladesh rather moderate air pollution was observed (CO2 < 420 ppm, OCS < 500 ppt, CO mostly < 200 ppb), also supported by the tropospheric ozone sounding data. The prevailing wind direction was South-East, i.e. air was coming from the sea. Thus, the continental convection seems to be more important than the convection over the Bay of Bengal, which may not carry large amounts of pollutants, at least in terms of pollutant transport into the anticyclone and to the stratosphere (see Bergman et al., JGR, 2013). From closer examination of the correlations between the different trace gases we conclude that the local CO2 and OCS variability is predominantly of biogenic, not anthropogenic origin.

Nainital/India and Dhulikhel/Nepal O3, H2O and aerosol balloon campaigns
The region of the southern slopes of the Himalayas is known as a hot-spot for the transport of water vapor, aerosols and pollutants to Asian summer monsoon anticyclone (ASMA) (e.g. Pan et al., 2016). For this reason, three balloon campaigns were conducted throughout the years 2016-2017 in this region, carried out from Nainital (India) and Dhulikhel (Nepal). The campaigns were organized by ETH Zürich (Switzerland) in collaboration with the Indian Institute of Tropical Meteorology (India) and the German Meteorological Service/GRUAN Lead Centre (Germany). The Nainital campaign was conducted from the Aryabhatta Institute of Observational Sciences (ARIES) (29.35°N, 79.46°E) and consisted of two operating periods, one in August 2016 and one in November 2016. The Dhulikhel campaign was hosted by the Himalayan Cryosphere Climate and Disaster Research Center (HiC-CDRC) at Kathmandu University (KU) (27.62°N, 79.46°E) and took place in July-August 2017, simultaneously with the Geophysica-M55 aircraft campaign in Kathmandu (Nepal). All soundings employed Vaisala RS41 meteorological radiosondes (Vaisala, Finland) for pressure and temperature measurements, Electrochemical Concentration Cell (ECC) (En-Sci, USA) for ozone (O3) mixing ratio, Cryogenic Frost-point Hygrometer (CFH) (En-Sci, USA) for water vapor (H2O) mixing ratio, and Compact Optical Backscatter Aerosol Detector (COBALD) (MyLab, Switzerland) for aerosol backscatter at two wavelengths (455 nm and 940 nm). An overview of all the measurements is given by Brunamonti et al. (2018).
This dataset provides unprecedented insights into the structure and composition of ASMA above the southern slopes of the Himalayas, leading to the proposal of a new framework for the vertical thermodynamic structure of the ASMA (see Brunamonti et al. 2018 for a detailed description).The average lapse rate minimum (LRM) and cold-point tropopause (CPT) levels were found at higher altitude in Dhulikhel compared to Nainital (400-600 m), but due to lower temperatures in Dhulikhel, the shift in potential temperature space is small (1.5-2 K). At both locations, high H2O and low O3 were found in the Asian Tropopause Transition Layer (ATTL) (altitudes between LRM and CPT) and confined lower stratosphere (CLS) (between CPT and the top of confinement, TOC), which is the signature of deep convection, extending up to 1.5-2 km above the CPT. An isolated H2O maximum in the CLS was observed in Dhulikhel, which is most likely due to overshooting convection injecting ice crystals directly above the CPT (Brunamonti et al., 2018). The high H2O observed in the CLS is particularly interesting because of its potential implications for stratospheric moistening. Cloud-filtering of the Nainital aerosol backscatter measurements reveals the signature of the Asian tropopause aerosol layer (ATAL) extending from the LRM to the TOC.

Arctic H2O frost point sondes
In the Arctic, stratospheric water vapor profiles have been retrieved regularly on a bi-monthly basis by balloon-borne cryogenic frostpoint hygrometers (CFH) at the AWIPEV research base in Ny-Ålesund, Svalbard. The Ny-Ålesund CFH sounding program has been accepted to the Network for the Detection of Atmospheric Composition Changes (NDACC), and all quality-checked profiles obtained within the frame of StratoClim are now included to the NDACC database, available with open access for further use for e. g. model parameterization and validation beyond the StratoClim community.

O3 Arctic Match campaign
Since the winter 2013/14 the climatic conditions in the Arctic stratosphere had been monitored for 6 winters (until winter 2018/19). In three winters, i. e. 2013/14, 2015/16 and 2017/18, international ozonesonde Match campaigns started in early January due to low stratospheric temperatures which are favorable for ozone depletion. During the other three winters the temperatures were too high throughout the winter and consequently no Match had been coordinated. During the winters 2013/14 and 2015/16 the Arctic stratosphere showed a similar behavior as it stayed cold into the first half of March with a short interruption in February. The warming in each March stopped the ozone depletion and shortly afterwards the Match campaigns. Both winters showed moderate ozone losses compared to the ozone loss record in the decades before. ATLAS model calculations complemented the observations during the Match campaigns.



The overall task of WP3 was the combined analysis of key species (SO2, COS), upper troposphere and stratosphere (UTS) sulfate aerosol distribution and properties, and UTS cloud properties from various satellite instruments together with chemistry-transport modelling and respective chemistry climate model (CCM) simulations. This has allowed studying the sulfur budget of the lower stratosphere, the transport pathways from the troposphere into the stratosphere, and the transformation into aerosols on a global scale. The work was split into 4 sub-tasks.

Analysis of the processes that regulate global sulfur distributions from MIPAS data and CTM modelling
Single profile SO2 and OCS data for the UTLS region (Höpfner et al., 2015; Glatthor et al., 2015; 2017) have been derived from MIPAS spectra for the full mission lifetime (2002 – 2012) within StratoClim. Climatologies have been generated, their variability has been studied, links to all relevant volcanic eruptions have been identified, and the lifetime of SO2 in the UTLS has been derived. The variability of OCS has been studied. Sink regions over the tropical rain forests (South-America and Africa) that require stronger sinks than assumed in today’s models have been identified; these enhanced sink regions must be compensated by enhanced oceanic emissions in models to realistically reproduce measured distributions. Both SO2 and OCS gridded data have been made publicly available and for use within StratoClim, and a data description report has been delivered (D3.1). Further, a climatology of ammonia (NH3) in the upper troposphere showing enhanced values within the region of the Asian summer monsoon has been derived from MIPAS/Envisat measurements in the time period 2002-2012 (Höpfner et al., 2016a, 2016b). The composition of aerosol particles in the ATAL is influenced by NH3, indicating that at least a fraction are composed of ammonium, like ammonium sulfate or ammonium nitrate. Finally, retrieval of sulfate aerosol (in terms of volume densities) from MIPAS spectral data was developed and a comprehensive data set of sulfate aerosol volume densities covering the 10 years of the MIPAS mission was derived.
A comprehensive treatment of stratospheric sulfur chemistry was implemented into an isentropic chemical transport model. The model results together with the MIPAS observations were used to improve our understanding of the processes controlling the budget and variability of sulfur in the stratosphere over the past decade, in particular for the volcanic eruptions of Kasatochi in August 2008 and Sarychev in June 2009. For the case studies of the two volcanic eruptions, we have shown that the MIPAS sulfate aerosol and SO2 data are qualitatively and quantitatively consistent with each other. Further, we demonstrated that the lifetime of SO2 is explained well by its oxidation by hydroxyl radicals (OH). While the sedimentation of sulfate aerosol plays a role, we found that the long-term decay of stratospheric sulfur after these volcanic eruptions in midlatitudes is mainly controlled by transport via the Brewer–Dobson circulation. Sulfur emitted by the two midlatitude volcanoes resides mostly north of 30°N at altitudes of ~10–16 km, while at higher altitudes (~18–22 km) part of the volcanic sulfur is transported towards the equator where it is lifted into the stratospheric overworld and can further be transported into both hemispheres. For these two eruptions it is demonstrated that sulfur that enters the stratosphere at 50°N can impact the sulfur loading of the entire hemisphere, and even the southern hemisphere. These findings regarding residence time and transport of sulfur, which show the potentially global effect of sulfur that is injected in the northern mid-latitudes, must be considered in potential proposed climate-engineering schemes using sulfate aerosol.
Regarding the stratospheric sulfur burden, simulations in- and excluding volcanic sulfur show that measured sulfur masses for the different species are in agreement, when considering input from OCS and volcanic SO2 with a relatively high frequency of small volcanic eruptions (98 eruption days in ~10 years). MIPAS OCS suggests a strong increase of stratospheric OCS from 2006–2008 (+ 13–32 Gg yr-1), while the stratospheric OCS burden remained at a rather constant level before and after this period. One possible reason for this increase in stratospheric OCS might be caused by a decrease in tropopause heights during this period. Time series of sulfate aerosol mass further suggest an increase during periods of relatively weak volcanic influence. Additionally, comparisons between measured and simulated SO2 and liquid-phase H2SO4 emphasize the importance of volcanic contributions to reach the levels of observed stratospheric sulfur masses.
The new MIPAS data set of sulfate aerosol is a step forward to close the sulfur budget for 2002–2012, and can be used to further constrain models, to quantify volcanic emissions, and to help improve our understanding of stratospheric sulfur and its effects on climate. Model simulations showed agreement between the MIPAS data sets of SO2 and sulfate aerosol after two volcanic eruptions, in terms of mass and transport patterns, as well as for MIPAS data sets of OCS, SO2 and liquid-phase H2SO4 under so-called “background” (non-volcanic) conditions. It has been demonstrated by model simulations that in the period 2006–2011 the sulfur released from OCS only accounted for about 7–25 % of the stratospheric sulfur mass contained in SO2 and for 30–55 % of the stratospheric sulfur mass contained in sulfate aerosol, in both cases regarding periods of relatively low volcanic influence. The presented findings help to better understand the transformation between the three sulfur species, and prove the importance of volcanic sulfur for the stratospheric sulfur loading.

Characterization of the global sulfate aerosols distribution and properties from high spectral resolution infrared sounders
This task focuses on the retrieval of sulfate and other aerosols from nadir-looking infrared sounders like IASI, TANSO-FTS, and SEVIRI. We have developed a comprehensive theoretical basis for the observation of UTLS secondary sulfate aerosols (SSA) from TIR nadir satellite instruments; the main result of these analyses is that broad-band radiometers like SEVIRI cannot retrieve SO2 and stratospheric sulfate aerosol total masses as independent quantities, while high-spectral-resolution sounders, like IASI, can retrieve both with limited theoretical uncertainties.
The detection of sulfate aerosol precursors (SO2, OCS and H2S) emitted by volcanic eruptions from high spectral infrared sounders IASI and TANSO-FTS has been developed. Spectral signatures suitable to retrieve SO2 have been identified in both IASI and TANSO-FTS spectra. OCS can be retrieved from IASI spectra, however, TANSO-FTS does not cover the spectral region where OCS signatures are. OCS emissions during volcanic eruptions are expected to be of minor relevance, however. The spectral signature of H2S has been identified as well, however, volcanic emissions far beyond the expected ones would need to happen that H2S could be detected by IASI or TANSO-FTS.
The sensitivity of satellite nadir TIR observations to secondary sulfate aerosols (SSA) has been analyzed (Sellitto and Legras, 2016). Both high-spectral resolution (HR) and broad-band radiometers (BB) have been simulated, basing on IASI, MODIS and SEVIRI instruments. A new algorithm has been developed to estimate the Aerosol Optical Depth (AOD) at 10 µm from the Metop-A and Metop-B IASI spectra. This algorithm combines the advantages of being very fast (applicable in near real time) and exploiting all the information contained in the measurement. The technical description of the algorithm and the product format description is contained in D3.6. The IASI dataset has been put under production for a 2-year period of IASI B (CNRS, D3.6). With this algorithm (AEROIASI-Sulphates), we have characterized a moderate volcanic eruption of Mount Etna volcano (starting on 18/03/2012). After a few hours from the start of the eruption, we have found sulfate aerosol AOD reaching values as big as 0.15. AEROIASI-Sulphates applied to IASI observations has identified the co-existence of two different plumes for this event, one at higher altitudes (8 to 13 km) and one at lower altitudes (6-7 km), probably produced by wind sheers at the injection altitude.
Geostationary satellite constellation can provide continuous (very high temporal resolution) and global monitoring. Based on the outcomes described above, we have developed an SSA detection algorithm for SEVIRI as a test instrument (Sellitto et al., 2016). While SEVIRI is insensitive to background SSA, it is in principle capable to detect SSA from severe volcanic conditions (Sellitto and Legras, AMT, 2016). This algorithm allows for the first time the identification of regions with presence of SSA particles from geostationary observations and can readily be coupled with existing cirrus clouds products from SEVIRI (or other similar geostationary platforms) for studying the glaciogenic impact of such UTLS aerosol emissions. The algorithm has been tested on the build-up and Asian monsoon anticyclone entrainment of sulfur emissions following the Nabro volcano eruption of 2011 (Sellitto et al., 2017). This product, being very fast, may be used in NRT volcanic plume dispersions analyses following major volcanic eruptions. It has been recently proposed as a test algorithm during the SPARC SsiRC VolRes activities foreseen for the possible imminent Agung volcano eruption. To this aim, the transfer of the existing SEVIRI algorithm to the Himawari observing system is presently ongoing. Further, studies are ongoing to extend the SSA retrieval theoretical basis of Sellitto and Legras (2016) to other non-sulfate UTLS aerosols, like secondary organic aerosols, black carbon or others. These aerosols types are expected to be present in the ATAL, where a purely SSA retrieval scheme may fail.

Sulfate aerosols from spaceborne lidar combined with Lagrangian modelling
Unfortunately, the CALIOP lidar we wanted to use for this study has been affected by technical issues after more than 11 years of operation, which prevented it from acquiring data during the StratoClim campaign. We therefore focused on high-altitude lidar observations that were available to us in the Indian Ocean region. The report (D3.4) summarizes the results of our efforts aiming to better characterize the processes responsible for the formation, development and dissipation of high tropospheric clouds in the Tropics. LMD-CNRS has conducted an extensive study based on the calculation of backward and forward diabatic Lagrangian trajectories across the TTL between the tropical tropopause (defined as the θ=380 K surface) and the top of convective cloud. We use winds and heating rates from three different meteorological reanalysis and cloud top observations from geostationary satellites. The region of the Sea of China and Sea of Philippines was found to be the first contributor of the convective sources to the mass flux of air entering the stratosphere, during the Asian monsoon, followed in terms of mass flux by India and the Bay of Bengal. Parcels released at convective top with brightness temperature less than 230 K over the Tibetan plateau have the highest probability, compared to other regions, to cross the 380 K surface. This is consistent with previous findings of a vertical conduit over the continent. The Tibetan plateau remains, however, a minor contributor to the total mass flux across the 380 K surface. The cloud products from the EUMETSAT Satellite Application Facility (SAF) on Nowcasting (NWC) and Very Short-Range Forecasting has been used, in combination with Lagrangian transport analysis, for the study of tropospheric-stratospheric exchanges in the Asian Monsoon region. The cloud data from the geostationary satellites proved to be of great relevance for the identification of such intense convective events, thanks to their high temporal and spatial resolution. In addition, the capability of the satellites to detect the convective events was in turn tested by comparison with the tropospheric trace gases variations measured in the UTLS by the aircraft.

Clouds in the UTS, determination of the altitude of anvils and overshoots from geostationary satellite data
The full dataset of the Meteosat 10 and Himawari 8 satellites in the extended monsoon region (0-50N, 10W-160E) have been processed using an improved version of the operational algorithm of the Eumetsat NWC SAF supporting nowcasting and short-range forecast (Derrien & Le Gléau, 2010 and http://www.nwcsaf.org/AemetWebContents/ScientificDocumentation/Documentation/GEO/v2016/NWC-CDOP2-GEO-MFL-SCI-ATBD-Cloud_v1.1.pdf). The improvement consists in using the ERA5 reanalysis with hourly data at high vertical resolution to better describe the atmospheric profile and the radiative transfers, in particular in the tropopause regions. The processing has been made from May to September 2017, covering the whole monsoon season. The product contains a classification of clouds and an estimate of the cloud top pressure. It is produced at the same resolution and frequency as the satellite images. The entire region of interest was covered by merging the two satellites at 90°E of longitude, where the field of view of the two satellites overlap. Even if the highest cloud tops can be found predominantly over maritime regions (especially over the Pacific Ocean, where several typhoon events took place), convection over North India and in the southern flank of the Himalayan Plateau can also contribute to very high cloud formation. Overall, July appears to be the month with the deepest and most frequent convective activity, both on the continental and maritime regions.
As campaign support, satellite data have been provided in real time during the campaign with merged products from MSG1 and Himawari 8 (IR, WV, cloud classification, cloud top altitudes) that were presented to the flight planning team using a specially tailored interface. This was combined with weather forecast products described in WP3. The dataset will be freely distributed and accessible from the StratoClim database. It will be used as a basis for the final version of the convective source products developed in WP4. A new version of the SAFNWC cloud product being available in 2019, a new processing will be made using this version that will be the final StratoClim product.



An important contribution from the work in WP4 to StratoClim is the selection of the campaign locations and the planning of the flights of the Geophysica both during the Kalamata (2016) and the Kathmandu (2017) campaign. While the choice of the location was dominated by logistic concerns (see WP1) the design of the scientifically most valuable flight plan was a contribution of the WP4 teams present during the campaigns. Often different alternative flight plans had to be designed that had to be chosen and adjusted according to the constraints of the campaign. Of course, the discussion of the flight plan was conducted in a larger group including the experimental teams. The actual design of the flight plan that was communicated to the aircraft (Geophysica) team was then prepared employing the MSS software tool. An overall achievement of the work in WP4 is the general pattern of the transport of air masses from convection (in the region of the Asian monsoon) to the stratosphere. The StratoClim studies show that this transport occurs in three distinct steps: first, very rapid uplift within the convective range up to about 360 K within the Asian monsoon anticyclone and outside in the tropical adjacent regions (within a few days); second, uplift above 360 K (up to about 460 K) within the upward spiralling range (within a few months); and third, transport within the tropical pipe to altitudes higher than 460 K associated with the large-scale Brewer–Dobson circulation (within about a year). The thermal tropopause does not constitute a transport barrier regarding the slow, diabatically driven ascent. The tropical pipe, which largely isolates tropical air masses from isentropic mixing with mid-latitude air is fed indirectly with monsoon lifted air by export of air quasi horizontally from the anticyclone; the tropical pipe is not situated just above the monsoon anticyclone.

Moreover, based on backward trajectories, the distribution of convective sources was established in WP4 to determine the origin of air (and the CO concentrations) sampled during the StratoClim campaign in Kathmandu. It has been found, using CO tracer measurements that the best trajectories are provided by using the new ERA5 high resolution reanalysis (0.25° resolution, hourly sampling, 137 levels) with radiative heating rates used to determine the vertical motion. Satellite observations were used for the identification of convective sources. In spite of the positive results regarding ERA5, a problem was identified: the ERA5 simulations tend to exhibit excessive penetrative convection over the Tibetan plateau.

Höpfner et al. (2019) have studied the possible origin of the elevated NH3 concentrations observed by GLORIA in the upper troposphere by trajectory analyses in combination with satellite data of total column amounts of NH3. During the days prior to the aircraft measurements, enhanced column amounts of NH3 have been observed at lower atmospheric levels in the region of NW-India and NE-Pakistan, a region known as a hot-spot for NH3 emissions. Trajectory simulations suggest that the high amounts of NH3 observed by the GLORIA instrument have initially been transported convectively to altitudes of 12–14 km consistent with the transport pattern in the Asian monsoon discussed above. Subsequently, the NH3 was advected by the anticyclonic upper tropospheric monsoon circulation to the location of the airborne observations within a few days. These measurements prove that NH3 reaches the upper troposphere in amounts sufficient to explain the mass density of ammonium nitrate observed. The relevant processes reducing the impact of washout of NH3 during convection might involve low acidity of convective rain droplets, or a release of NH3 during the freezing process of cloud particles.



Most of the highlights of the WP5 presented in the following chapters have also been published in high-ranking journals. The large amount of peer-reviewed papers is demonstrating the outstanding success of research of the global modelling activities in StratoClim.

StratoClim Campaign Support Document
The aim of this detailed document was to provide suitable information from the StratoClim modelling group in order to support the planning of the StratoClim flight campaign within the monsoon area in its first stage (Nützel et al., 2015). The model used primarily within this study was the global CCM EMAC. Additional model contributions from CLaMS and UM-UKCA were included. Observational and reanalysis data complement the model studies. The main aspects focused on are: 1. Asian Summer Monsoon (anticyclone) variability in location and strength with respect to interannual and intraseasonal time scales. 2. Trace gas signatures within the anticyclone. 3. Identification of target periods and areas.

Interannual variability of the boreal summer tropical UTLS in observations and CCMVal-2 simulations
During boreal summer the upper troposphere/lower stratosphere (UTLS) in the Northern hemisphere shows a distinct maximum in water vapor (H2O) mixing ratios and a coincident minimum in ozone (O3) mixing ratios, both confined within the Asian monsoon anticyclone (AMA). This well-known feature has been related to transport processes emerging above the convective systems during the Asian summer monsoon (ASM), further modified by the dynamics of the AMA. In this study, the ability of chemistry–climate models (CCMs) to reproduce the climatological characteristics and variability of H2O, O3, and temperature in the UTLS during the boreal summer was compared with MIPAS satellite observations and the ERA-Interim reanalysis (Kunze et al., 2016). By using a multiple linear regression model the main driving factors, the strength of the ASM, the quasi-biennial oscillation (QBO), and the El Niño–Southern Oscillation (ENSO), were separated. The regression patterns related to ENSO showed a coherent, zonally asymmetric signal for temperatures and H2O mixing ratios for ERA-Interim and the CCMs, and suggested a weakening of the ASM during ENSO warm events. The QBO modulation of the lower-stratospheric temperature near the equator is well represented as a zonally symmetric pattern in the CCMs.
Changes in H2O and O3 mixing ratios are consistent with the QBO-induced temperature and circulation anomalies but less zonally symmetric than the temperature pattern. Regarding the ASM, the results of the regression analysis show for ERA-Interim and the CCMs enhanced H2O and reduced O3 mixing ratios within the AMA for stronger ASM seasons. The CCM results can further confirmed earlier studies which emphasize the importance of the Tibetan Plateau/southern slope of the Himalayas as the main source region for H2O in the AMA. The results suggested that H2O is transported towards higher latitudes at the north-eastern edge of the AMA rather than towards low equatorial latitudes to be fed into the tropical pipe.

Investigation of the dynamic variability of the Asian monsoon anticyclone
The Asian monsoon anticyclone shows considerable variability (Garny and Randel, 2013). In the study from DLR the variability of the Asian monsoon anticyclone and its causes was investigated (Nützel et al., 2016). For the analyses (partly) data from multiple reanalysis were used. Special attention was brought to the subject of the bimodality of the South Asia High (SAH), i.e. the preferred location of the SAH center over the Tibetan and Iranian Plateaus (Zhang et al., 2002). With respect to the bimodality it was found, that only an (older) reanalysis was showing a pronounced bimodal structure for the SAH center location. More recent reanalysis rather showed a broad spread of the center location. Analyses based on this publication will be part of the TTL (Tropical Tropopause Layer) Chapter 8 of the S-RIP (SPARC-Reanalysis Intercomparison Project) report (https://s-rip.ees.hokudai.ac.jp/ report/structure.html). The SRIP report is currently finalized and is planned to be submitted for review in the near future.

The Asian monsoon and its influence on the upper troposphere and lower stratosphere
The development and successful completion of the PhD thesis of Matthias Nützel was mostly financially supported by the StratoClim project (Nützel, 2018).

Assessing water vapor transport from the Asian monsoon to the stratosphere
In a joint project between DLR and FZJ (including university affiliations) the CTM CLaMS (e.g. McKenna et al., 2002) was used to investigate the transport pathways of water vapor from the Asian monsoon region to the NH extratropical lower stratosphere and to the deep tropics (Nützel et al., 2019 – accepted for publication). The presented results for the Asian monsoon region are set into context by comparing them with other source regions such as the tropics during winter and summer and the warm pool region as well as the North American monsoon region. This study complements previous findings by Ploeger et al. (2017), which were focused on the mass transport from the Asian monsoon anticyclone to the stratosphere. The results presented in Nützel et al. (2019) suggest that the Asian monsoon region is (when judged by its size) especially efficient in transporting water vapor and mass to the stratosphere. Only the warm pool region during NH winter shows a comparable (higher) efficiency for water vapor (mass) transport to the deep tropics.

On the representation of major stratospheric warmings in reanalysis
Major sudden stratospheric warmings (SSWs) represent one of the most abrupt phenomena of the boreal wintertime stratospheric variability, and constitute the clearest example of coupling between the stratosphere and the troposphere. A good representation of SSWs in climate models is required to reduce their biases and uncertainties in future projections of stratospheric variability. The ability of models to reproduce these phenomena is usually assessed with just one reanalysis. However, the number of reanalysis has increased in the last decade and their own biases may affect the model evaluation. In this study the representation of the main aspects of SSWs across reanalysis was compared (Ayarzagüena et al., 2019). The examination of their main characteristics in the pre- and post-satellite periods reveals that reanalysis behave very similarly in both periods. However, discrepancies are larger in the pre-satellite period than afterwards, particularly for the NCEP/NCAR reanalysis. All datasets reproduce similarly the specific features of wavenumber-1 and wavenumber-2 SSWs. A good agreement among reanalysis is also found for triggering mechanisms, tropospheric precursors and surface fingerprint. In particular, differences in blocking precursor activity of SSWs across reanalysis are much smaller than between blocking definitions.

No robust evidence of future changes in major stratospheric sudden warmings: a multi-model assessment from CCMI
Major mid-winter stratospheric sudden warmings (SSWs) are the largest instance of wintertime variability in the Arctic stratosphere. Because SSWs are able to cause significant surface weather anomalies on intra-seasonal timescales, several previous studies have focused on their potential future change, as might be induced by anthropogenic forcings. However, a wide range of results have been reported, from a future increase in the frequency of SSWs to an actual decrease. Several factors might explain these contradictory results, notably the use of different metrics for the identification of SSWs and the impact of large climatological biases in single-model studies. To bring some clarity, the question of future SSW changes was revisited, using an identical set of metrics applied consistently across 12 different models participating in the Chemistry–Climate Model Initiative (Ayarzagüena et al., 2018). The analysis revealed that no statistically significant change in the frequency of SSWs will occur over the 21st century, irrespective of the metric used for the identification of the event. Changes in other SSW characteristics – such as their duration, deceleration of the polar night jet, and the tropospheric forcing – were also assessed: again, no evidence of future changes over the 21st century was found).

Drivers in the impact of stratospheric final warming variability on surface climate
Springtime stratospheric final warming (SFW) variability has been suggested to be linked to the tropospheric circulation, particularly over the North Atlantic sector. These findings, however, are based on reanalysis data that cover a rather short period of time (1979 to present). The present work aims to improve the understanding of drivers, trends and surface impact of dynamical variability of boreal SFWs using chemistry-climate models. We use multi-decadal integrations of the fully coupled chemistry-climate model Community Earth System Model (CESM) version 1 (Whole Atmosphere Community Climate Model). Four sensitivity experiments are analyzed to assess the impact of external factors; namely, the quasi-biennial oscillation, sea surface temperature (SST) variability, and anthropogenic emissions. SFWs are classified into two types with respect to their vertical development; that is, events which occur first in the mid stratosphere (10-hPa first SFWs) or first in the upper stratosphere (1-hPa first SFWs). Our results confirm previous reanalysis results regarding the differences in the time evolution of stratospheric conditions and near-surface circulation between 10 and 1-hPa first SFWs. Additionally, a tripolar SST pattern is, for the first time, identified over the North Atlantic in spring months related to the SFW variability. Our analysis of the influence of remote modulators on SFWs revealed that the occurrence of major warmings in the previous winter favors the occurrence of 10-hPa first SFWs later on. We further found that quasi-biennial oscillation and SST variability significantly affect the ratio between 1-hPa first and 10-hPa first SFWs. Finally, our results suggest that ozone recovery may impact the timing of the occurrence of 1-hPa first SFWs.

Dynamics of the monsoon anticyclone and its variability
The StratoClim project is motivated by the potential importance of the Asian monsoon system in transporting chemical species from lower troposphere to upper troposphere and from upper troposphere to stratosphere. Many studies have shown how they are episodically transported in filaments which are drawn out of the upper tropospheric monsoon anticyclone and eventually mix with the stratospheric air mass to the north.
In order to build confidence in the ability of general circulation models and chemistry-climate models to reproduce the monsoon anticyclone in current and future climates it is important to have clear understanding of the role of different physical processes. The PhD project of Philip Rupp working in DAMTP, University of Cambridge, was primarily supported by StratoClim and had the aim of understanding the particular role of different dynamical processes in determining the structure and time variation of the monsoon anticyclone, which controls the troposphere-to-stratosphere transport described above.
The approach taken in this work was to use a simple general circulation model (SGCM) which includes a good representation of large-scale dynamics, but a highly simplified representation of physical processes such as radiation and tropospheric convection. Numerical simulations with this model have shown: (i) A time-independent localized forcing, representing the effect of convection over northern India and the Bay of Bengal, can lead to a realistic monsoon anticyclone which fluctuates in time. These fluctuations are determined entirely by the upper-level dynamics and can be regarded as an ‘eddy-shedding’ process in which a low potential vorticity region grows and extends to the west of the convective forcing region, forming a sequence of anticyclonic eddies.
The relative strength of the two anticyclonic circulations to the west of the forcing region varies in time. At one day the eastern circulation is stronger whereas eight days later it is the western circulation that is stronger. This eddy-shedding phenomenon may account for the ‘bimodality’ in the monsoon anticyclone which is sometimes reported in observational studies.
This behavior is typically seen in configurations of the SGCM where the midlatitude eastward jet is relatively weak. When the eastward jet is stronger a different behavior is often seen, where low potential vorticity air is drawn out of the monsoon anticyclone to the northeast and forms an anticyclonic circulation to the east of the convective forcing region.
This behavior appears relevant to the seasonal evolution of the ‘Bonin high’ anticyclone which can have a substantial effect on weather in Japan in late summer. Previous work has interpreted this as a Rossby wave response to forcing located in SW Asia. In the SGCM there is no such forcing and the anticyclonic circulation results from an interaction between synoptic-scale eddies which continuously disturb the mid-latitude jet and the primary monsoon anticyclone.
Other work in this project included examination of the vertical structure of the monsoon anticyclone and the transitions in the type of westward eddy shedding that occurred as external parameters were varied.
The results are reported in Rupp (2018) (PhD thesis, University of Cambridge, ‘On the structure and dynamics of the Asian Monsoon Anticyclone’) and two journal papers are in preparation.

Response of the stratospheric quasi-biennial oscillation to enhanced CO2 concentrations
Investigations regarding the response of the stratospheric quasi-biennial oscillation (QBO) to a doubling and quadrupling of CO2 have carried out by the Met Office. A joint paper on the results (including contributions from ISAMAR-CNR) has been submitted for publication in a Special Section of QJRMS (Richter et al., 2019). Key finds of this study were:
• No consistency in the QBO period response.
• For the multi-model mean QBO, amplitudes decreased by 36% and 51% for a doubling and quadrupling of CO2 amounts, respectively.
• Differences in period response across the models were found to be most strongly related to how the upward momentum flux due parameterized gravity waves, and the equatorial residual vertical, responded to the increased CO2 concentration.
• The robust amplitude decreases in the changing climate correlated well across the models with the corresponding increases in tropical upwelling and also to the changes in the gravity wave momentum fluxes and to a lesser extent the resolved upward wave flux.
For 4 x CO2 the QBO signal becomes less discernible, especially in the lower stratosphere. The results were also represented at the SPARC General Assembly in Kyoto, Japan, in October 2018.

Coupling between the ENSO and the QBO
The main part of DMIs work in StratoClim has been related to the coupling between the ENSO and the QBO. We have also worked on the definition of stratospheric sudden warmings and the mechanism of the downward propagation on high-latitudes. Together with Dr. Serva (CNR) we have made modifications to EC-Earth to store the diagnostics relevant for the DynVarMip under CMIP6. A preliminary analysis of a 10-years test shows that in general the transformed Eulerian mean momentum-budget is closed to a good degree.
In Christiansen et al. (2016) we noted that in observations the QBO in the 3 to 4 years after the three warm ENSO events in 1982, 1997, and 2015 was aligned with the ENSO. We investigated this indicated relationship with a version of the EC-Earth climate model which includes non-orographic gravity waves. We analyzed the modelled QBO in ensembles consisting of 10 AMIP-type experiments forced with climatological SSTs and 10 experiments forced with observed daily SSTs. In the ensemble with observed SSTs we found a strong and significant alignment of the ensemble members in the equatorial stratospheric zonal winds in the 2 to 4 years after the strong ENSO events 1997. This alignment also includes the observed QBO. No such alignment was found in the ensemble with climatological SSTs. This model result, together with the observed alignment after the warmest ENSO events, indicates that strong warm ENSO events can lock the phase of the QBO.
In Serva et al. (2019) the phase alignment of the QBO after strong warm ENSO events were confirmed in a larger ensemble of uncoupled experiments. In this work we also analyzed atmosphere-only and ocean-atmosphere coupled simulations from a large multi-model ensemble. We found that the annual cycles of the QBO amplitude and QBO descent rate are well represented in both coupled and uncoupled models. In Christiansen et al. (2016) we showed that the previously reported observation that the phase-speed of the QBO is faster under warm ENSO conditions that under cold conditions is well represented in the EC-Earth experiments with observed SSTs. However, in Serva et al. (2019) we found that for the coupled models a relatively high horizontal resolution is necessary to reproduce this modulation.

Troposphere-stratosphere coupling at seasonal and interannual timescales
The scientific understanding of troposphere-stratosphere coupling at seasonal and interannual timescales has been improved, in particular regarding Stratospheric Sudden Warmings (SSWs), the ENSO phenomenon and the relationship between both through e.g. blocking (Barriopedro and Calvo, 2014; Iza and Calvo, 2015; Iza et al., 2016; Calvo et al., 2017; Palmeiro et al., 2017; Ayarzagüena et al., 2018; Ayarzagüena et al., 2019). We have shed light on the controversy on the Northern Hemisphere polar stratospheric impact of the different flavors of ENSO (Central Pacific vs Eastern Pacific) in both reanalysis data and general circulation models (Iza and Calvo, 2015; Calvo et al., 2017). Our research showed that SSWs played a prominent role in separating both signals in the Northern Hemisphere polar stratosphere. Iza et al. (2016) showed for the first time a robust stratospheric pathway for La Nina in reanalysis datasets, with implications for seasonal forecasting. The influence of SSWs and the polar stratospheric ENSO signal over the North Atlantic European (NAE) region has been addressed in several studies, showing that only Eastern Pacific El Nino events have an impact on NAE region through the stratosphere (e.g. Calvo et al., 2017; Palmeiro et al., 2017; Iza et al., 2016). Ayarzagüena et al. (2019) found decadal variability in the ENSO/NAE teleconnection through the stratosphere.
In addition, we have deepened our understanding of polar ozone variability at different timescales, from seasonal to multidecadal. De la Camara et al. (2018) provided understanding of the dynamics that control changes of Arctic ozone during the life cycle of SSWs quantifying advective transport and mixing. Calvo et al. (2015) and Ivy et al. (2017) found modeling and observational evidences of the impact of interannual Arctic stratospheric ozone changes on NH tropospheric climate in spring. On the other hand, the impact of long-term ozone trends on the stratospheric mean meridional circulation has been addressed in Abalos et al. (2019) and Palmeiro et al. (2017) and the mechanisms for the long-term changes in the stratospheric mean meridional circulation in Calvo et al. (2017) and Palmeiro et al (2014).

The role of stratospheric ozone for Arctic-mid-latitude linkages
The rate at which Arctic temperature has risen in recent decades is more than twice the rate of lower latitudes (e.g. Cohen et al., 2014). The extent of Arctic sea-ice has declined rapidly (Stroeve et al., 2012) and a decrease of its thickness could be observed (Kwok & Rothrock, 2009). In addition, observations and some model studies suggest a negative phase shift in wintertime Arctic Oscillation (AO) which can be related to the loss of Arctic sea-ice (e.g. Nakamura et al., 2015). However, a lot of climate models are not able to clearly reproduce this important feature of Arctic-mid-latitude linkages, which is strongly modulated by upward propagation of planetary waves into the stratosphere and subsequent interactions with the stratospheric polar vortex, which then itself impacts the tropospheric circulation in the following months.
To allow for a more accurate representation of stratospheric processes, we coupled the computationally fast interactive stratospheric chemistry module SWIFT (Wohltmann et al., 2017) with the atmospheric general circulation model ECHAM6 (Stevens et al., 2013). SWIFT is based on a set of coupled differential equations, which simulate the polar vortex-averaged mixing ratios of the key species involved in polar ozone depletion.
To investigate the impact of decreasing Arctic sea-ice on the atmospheric circulation we performed model simulations with ECHAM6 and the newly coupled ECHAM6-SWIFT using different sea-ice conditions as the model’s boundary condition, computing a total of 480 model years (Romanowsky et al., 2019). The interactive chemistry enhances the model’s dynamical response to Arctic sea ice reduction, and the simulated dynamical characteristics are in much better agreement with ERA-Interim reanalysis data. Also, a clear shift of the AO to a negative phase in late winter and spring could be reproduced with ECHAM6-SWIFT. The ECHAM6 standalone model showed neither a statistically significant response of the Arctic stratosphere nor a change in the AO. This shows that for an improved understanding of Arctic-mid-latitude linkages the coupling between dynamics and stratospheric ozone chemistry should be considered.

Stratospheric ozone changes in future
Researchers at University of Cambridge have used the UM-UKCA model to understand how total column ozone may change in the future, and examined the drivers of these changes, a key objective of WP5. Total-column ozone (TCO) has a direct effect on human health by preventing harmful ultraviolet (UV) radiation from reaching the surface. It is therefore important to gain a quantitative understanding of how TCO values may evolve over the 21st century.
Keeble et al. (2017) investigated the drivers of past and future changes in tropical averaged total column ozone using the UM-UKCA model. Ozone concentrations in the lower and upper stratosphere are governed by different processes and thus show distinct behaviors that combine to determine the overall evolution of total column ozone. In the upper stratosphere, where the chemical lifetime of ozone is short (~ 1 day), future reductions in ODS concentrations and stratospheric cooling from increased GHG concentrations both lead to increased upper stratospheric partial column ozone. Conversely in the lower stratosphere, where the chemical lifetime of ozone is typically >1 month, the partial column ozone values are predominantly controlled by changes to transport. Projected increases in GHGs lead to an acceleration of the Brewer-Dobson circulation (BDC), which is associated with increased transport of relatively ozone poor air masses into the tropical lower stratosphere, thereby decreasing ozone mixing ratios and the partial lower stratospheric column ozone, despite reductions in ODS concentrations. This highlights that future projections of tropical stratospheric column ozone are the result of a complex interplay between drivers of ozone trends in the lower and upper stratosphere.
Keeble et al. (2018) used and ensemble of UM-UKCA simulations to investigate different definitions of progress on the road to ozone recovery. The impacts of modelled internal atmospheric variability are accounted for by applying a multiple linear regression model to modelled total column ozone values, and ozone trend analysis is performed on the resulting ozone residuals. Three definitions of recovery are investigated: (i) a slowed rate of decline and the date of minimum column ozone, (ii) the identification of significant positive trends and (iii) a return to historic values. Trends for the 2000–2017 period are positive at most latitudes and are statistically significant in the mid-latitudes in both hemispheres when natural cycles are accounted for. Significant trends cannot be identified by 2017 at the highest latitudes, due to the large interannual variability in the data, nor in the tropics, due to the small trend magnitude, although it is projected that significant trends may be identified in these regions soon thereafter. While significant positive trends in total column ozone could be identified at all latitudes by ~2030, column ozone values which are lower than the 1980 annual mean can occur in the mid-latitudes until ~2050, and in the tropics and high latitudes deep into the second half of the 21st century.
In addition to these key findings, researchers at Cambridge have also assessed the impacts of ozone zonal asymmetries on the Arctic vortex, and developed a computationally inexpensive technique for imposing dynamically consistent ozone asymmetries on prescribed 2D fields; explored the novel use of machine learning techniques to make seasonal predictions of the North Atlantic Oscillation; and performed nudged simulations with the UM-UCKA model for comparison with the aircraft observations made during the campaign period and other WP5 models.

Radiative forcing by stratospheric aerosol
The group of MPIC performed transient simulations with the chemistry-climate model EMAC with fully interactive stratospheric and tropospheric aerosol for the period 1990 to 2017 including the 1991 Pinatubo eruption as contribution to SPARC-SSIRC (Stratosphere-troposphere Processes And their Role in Climate/ Stratospheric Sulfur and Its Role in Climate, Timmreck et al., 2018). The resolution is T63/L90 (1.9° up to 1 Pa), the one the convection scheme of the GCM was developed for. The lower boundary conditions are similar to Jöckel et al. (2016). Tropospheric meteorology is nudged to ERA-Interim of ECMWF and the Quasi Biennial Oscillation (QBO) to the observations compiled by Free University of Berlin. 3D-SO2 plumes of about 460 explosive volcanic eruptions were derived from limb satellite observations (SAGE II, MIPAS, GOMOS, OSIRIS) and added to the modelled SO2 at the eruption times (data for the ENVISAT period available under doi 10.1594/WDCC/SSIRC_1). The fast increase in anthropogenic SO2 emissions in India (Lelieveld et al., 2018) is included.
In Brühl et al. (2018) it is demonstrated that Asian desert dust lifted up by the Asian Summer Monsoon contributes significantly to the seasonal cycle of stratospheric aerosol optical depth especially in the subtropics of the Northern Hemisphere, a feature also seen in the global radiative forcing by stratospheric aerosol. Radiative forcing by medium strength tropical volcanic eruptions is largest in 2006, 2011 and 2015 with about -0.23 W/m2 while the major eruption of Pinatubo caused about -5 W/m2. Observations of a large nonvolatile fraction of lower stratospheric aerosol by COPAS on the GEOPHYSICA point to the presence of non-sulfate particles. The modelled size distribution with altitude but also the profiles of composition in the UTLS appear to be consistent with the campaign data (COPAS, UHSAS, ERICA). A PhD thesis entitled: “Stratospheric aerosol, budgets, chemistry and radiative transfer based on a complex chemistry climate model and satellite and field campaign data” will be published soon at the Johannes Gutenberg University in Mainz (by Schallock, 2019).

Contribution to SSIRC
C. Timmreck (MPI-M) has co-led the WCRP’s activity SPARC/SSIRC (Stratosphere-troposphere Processes And their Role in Climate /Stratospheric Sulfur and Its Role in Climate SSiRC and initiated within SSiRC the ISA-MIP model intercomparison project (see next point below). MPI-M was therefore mainly involved in the SSIRC compilation of the first comprehensive review of stratospheric aerosol since 2006 (Kremser et al., 2016). The main achievements since 2006 are that differences between in situ and space-based inferences of stratospheric aerosol properties have been resolved. Improved understanding of the role of minor volcanic eruptions and non-sulfate aerosols on the stratospheric aerosol layer has been achieved. In addition, chemistry-climate models have substantially increased in quantity and sophistication. Still, changes in stratospheric aerosol levels less than 20% cannot be confidently quantified. The volcanic signals tend to mask any non-volcanically driven change, making them difficult to understand. While the role of carbonyl sulfide as a substantial and relatively constant source of stratospheric sulfur has been confirmed by new observations and model simulations, large uncertainties remain with respect to the contribution from anthropogenic sulfur dioxide emissions. (See also Timmreck, 2018.)

Interactive stratospheric model Intercomparison (ISA-MIP)
Large inter-model differences have been found for the global stratospheric aerosol models which participated in the Tambora precursor study of the CMIP6 Model Intercomparison Project on the climatic response to volcanic forcing (VolMIP; Zanchettin et al., 2016; Marshall et al., 2018). Dedicated multi-model comparison projects with process-oriented comparisons, will therefore be imperative to disentangling the reasons for model differences in AOD and corresponding radiative forcing and sulfur deposition. Hence, partners in StratoClim initiated among others the SSiRC international stratospheric aerosol model intercomparison project ISA-MIP (Timmreck et al., 2018) ISA-MIP aims to address existing uncertainties and differences among the models with respect to aerosol radiative forcing and its climate response. The experiments have been designed to investigate key processes which influence the formation and temporal development of stratospheric aerosol in different time periods of the observational record. The “Background” (BG) experiment will concentrate on microphysics and transport processes under volcanically quiescent conditions, where the stratospheric aerosol is only modulated by seasonal changes and interannual variability. The “Transient Aerosol Record” (TAR) experiment will explore the role of small- to moderate-magnitude volcanic eruptions and transport processes over the period 1998-2012 and its possible role for the warm hiatus period. Two further experiments will investigate the stratospheric sulfate aerosol size distribution under the influence of large volcanic eruptions. The “Historical Eruptions SO2 Emission Assessment” (HErSEA) experiment will focus on the uncertainty in the initial emission of recent large volcanic eruptions, while the “Pinatubo Emulation in Multiple models” (PoEMS) experiment will provide a full uncertainty analysis of the radiative forcing of the Mt. Pinatubo eruption. First results are expected in 2019.

Tropical deep convection impact on southern winter stationary waves
The impact of tropical deep convection on southern winter stationary waves and its modulation by the quasi-biennial oscillation (QBO) have been investigated in a long (210 years) climate model experiment and in ERA-Interim reanalysis data for the period 1979-2018 (Pena-Ortiz et al., in review 2019). Model results reveal that tropical deep convection over the region of its climatological maximum modulates high latitude stationary planetary waves in the southern winter hemisphere, corroborating the dominant role of tropical thermal forcing in the generation of these waves. In the tropics, deep convection enhancement leads to wavenumber 1 eddy anomalies that reinforce the climatological Rossby-Kelvin wave couplet. The Rossby wave propagates towards the extratropical southern winter hemisphere and upward through the winter stratosphere reinforcing wavenumber 1 climatological eddies. As a consequence, stronger tropical deep convection is related to greater upward wave propagation and, consequently, to a stronger Brewer Dobson circulation and a warmer polar winter stratosphere. This linkage between tropical deep convection and SH winter polar vortex is also found in the ERA-Interim reanalysis. Furthermore, model results indicate that the enhancement of deep convection observed during the easterly phase of the QBO (E-QBO) gives rise to a similar modulation of the southern winter extratropical stratosphere, which suggests that the QBO modulation of convection plays a fundamental role in the transmission of the QBO signature to the southern stratosphere during the austral winter revealing a new pathway for the QBO-SH polar vortex connection. ERA-Interim corroborates a QBO modulation of deep convection, however the shorter data record does not allow to assess its possible impact on the SH polar vortex.

Contribution to CMIP6
MPI-M was actively involved in (co-leading) the set-up and the finalization of the experimental design for three CMIP6 (Coupled Model Intercomparison Project, Phase 6, Eyring et al., 2016) activities: the Model Intercomparison Project on the climatic response to Volcanic forcing (VolMIP), the CMIP6 diagnostic Model Intercomparison Project on Dynamics and Variability of the Stratosphere-Troposphere system (DynVarMIP), and the geo-engineering model intercomparison initiative GeoMIP (Kravitz et al, 2015). These activities will substantially contribute to an understanding of the microphysical, chemical and dynamical processes in the UTS and their role in climate. VolMIP focuses on a multi-model assessment of climate models' performance under strong volcanic forcing conditions (Zanchettin et al., 2016). DynVarMIP provides coordination for dynamically oriented diagnostics relevant for the assessment and understanding of momentum transfer in the atmosphere (Gerber and Manzini, 2016). VolMIP simulations have not been submitted to the CMIP6 Archive during the StratoClim project time but MPI-M has worked on the interactions between forced and unforced variability in the early 19th century one of the core periods for VolMIP (Zanchettin et al., 2019, Schurer et al. 2019 submitted). First DynVar results have been presented at the CMIP6 Model Analysis Workshop in March 2019. Analysis will be extended and presented at the DynVarMIP & S2S workshop on the Atmospheric Circulation in a Changing Climate, in Madrid in October 2019. First GeoMIP simulation were performed but are not submitted to the CMIP6 archive jet.

Revisiting the radiative forcing of the 1963 eruption of Mt. Agung
In 1963 a series of eruptions of Mt. Agung, Indonesia, resulted in the 3rd largest eruption of the 20th century and claimed about 1900 lives (Self and Rampino, 2012). Two eruptions of this series injected SO2 into the stratosphere, a requirement to get a long-lasting stratospheric sulfate layer. The first eruption on March 17th injected roughly twice as much SO2 into the stratosphere as the second eruption on May, 16th (Self and Kiang, 1996). In recent volcanic emission data sets these eruption phases are merged together to one large eruption phase for Mt. Agung in March 1963 with an injection rate of 7 Tg SO2. The injected sulfur forms a sulfate layer in the stratosphere. The evolution of sulfur is non-linear and depends on the injection rate and aerosol background conditions. To test if there is a significant difference when simulating two medium eruptions instead of a single large one only, we performed ensembles of two model experiments, one with a single eruption of 7 Tg SO2 (AGUNG1) and a second one with two eruptions (AGUNG2) of 4.7 Tg SO2 and 2.3 Tg SO2 (Niemeier et al., 2019). The two smaller eruptions result in a lower burden, smaller particles and 0.1 to 0.3 W/m2 (10 - 20%) lower radiative forcing in monthly mean global average compared to the individual eruption experiment. The differences are the consequence of slightly stronger meridional transport due to different seasons of the eruptions, lower injection height of the second eruption and the resulting different aerosol evolution. The differences between the two experiments are significant but smaller than the variance of the individual ensemble means. Overall, the evolution of the volcanic clouds is different in case of two eruptions than with a single eruption only. We conclude that there is no justification to use one eruption only and both climatic eruptions should be taken into account in future emission datasets.

Antarctic total ozone trend reversal
Time series of Antarctic total ozone for signatures of trend reversal based on level-crossing statistical techniques were analyzed. In contrast to earlier studies, we did not remove components of total ozone variability (e.g. QBO or volcanic activity) from the ozone time series, and we did not prescribe a turning point for Antarctic total ozone in the analysis. For total ozone at 79.5°S, we found a clear trend reversal with a turning point around the year 2000. For the lower section of the difference time series (difference between 79.5°S and two lower latitudes 59.5°S and 58.5°S), which coincides with data from the September–November period, we likewise found clear turning points. These turning points are more stable than those derived from the raw record of polar TO (i.e. less dependent on the required initial guess values) and lie in the early and mid-1990s, depending on the statistical technique employed. We concluded that a robust statistical signal of a trend reversal is present in the Antarctic total ozone record 1978–2011. However, an accurate determination of the year of the turning point is not possible because of the statistical uncertainties of the record (Várai et al., 2015).

Global trend analysis of the drought severity index
A global trend analysis of the remotely sensed drought severity index (DSI) time series were performed (Orvos et al., 2014; 2015). Special attention is paid to testing the statistical significance of local linear trends by data. The main result is that 17.34 % of the land area exhibits significant trends of either signs (12.03 % drying and 5.31 % wetting), and most of these locations form large, geographically connected areas. We emphasize that the usual field significance tests cannot give more reliable estimates, because a DSI value as defined provides a fully local characterization, and the same numerical value can be related to very different local circumstances. The relatively short period of 12 years hinders linking the trends to global climate change; however, we think that the observations might reveal a slow (decadal) mode of natural climate variability. Correlations with other atmospheric and oceanic variables are found at various (statistically insignificant) levels; therefore, at the moment we cannot prove any causal relationship or propose a solid explanation of the observations.

Lightning density analyses
The most recent version of the ECMWF forecasting system (IFS Cycle 45r1, 5 June 2018) includes lightning density as a new forecast parameter. The new parameterization of total lightning flash densities involves quantities that are diagnosed from the ECMWF convection scheme: the convective available potential energy, the vertical profile of the frozen precipitation convective flux, the profile of cloud condensate amount within the convective updraft, and the convective cloud-base height (Lopez, 2016). It is demonstrated that the new parameterization gives better results overall than the main existing lightning parameterizations designed for global models. Nevertheless, it is remarkable that not any aerosol related parameter is incorporated into the new module. Our main goal was to construct and test a prediction scheme where parameters depend on the geographic location, and optimal to have a minimal set of explanatory (input) variables. As a first step, we compiled data of lightning activities focusing on the geographic area of the South Asian Monsoon which turns large parts of India from a kind of semi- desert into green lands. This area is an optimal testbed for lightning prediction because of the complex surface orography. Data are evaluated from the ground base World Wide Lightning Location Network (14 years). We have focused on a five-year period (2013-2017), when the WWLLN observing network had a stable detection efficiency around 9-12 % (no changes in ground stations).
An attempt is made to correlate the spatiotemporal structure of lightning patterns with the following atmospheric parameters: convective available potential energy (CAPE), cloud base height, cloud base height gradient, convective precipitation, convective rain rate, total column cloud ice water (TIWC), salt aerosol concentration (three size fractions), dust aerosol concentration (3 size fractions), hydrophobic and hydrophilic organic matter aerosol and black carbon aerosol. Aerosol concentrations are considered on 16 pressure levels from 1000 to 70 hPa. In this way, the number of potential explanatory parameters is 166, thus traditional approaches (e.g. based on cross correlation functions, etc.) are not feasible. Instead, we implemented a machine learning algorithm (Orthogonal Matching Pursuit from the Python Scikit-learn module), which provides an optimal linear fit with an initially fixed number of parameters. Preliminary conclusions suggest that the main explanatory variables are convective precipitation (m), CAPE (J/kg) and TIWC (kg/m2), and the incorporation of aerosol variables only slightly improve the fits. Seemingly the most important aerosol fraction is hydrophilic organic matter aerosol in the upper troposphere (500-250 hPa). Those grid-cells which have larger explained variance with aerosol variables belong to extended continuous geographic clusters. Work is in progress, and publication is expected in autumn 2019.

During the last StratoClim annual meeting, it was decided that all StratoClim data will be hosted for long-term within the HALO database (https://halo-db.pa.op.dlr.de/mission/101) at DLR (Deutsches Zentrum für Luft- und Raumfahrt).
The HALO database is a long-term archiving system tailored to low-volume data of aircraft observations. Currently there are listed 80 missions with over 1000 separate flights and 4000 datasets. The current userbase is around 500 individuals and gives thus a good access to the StratoClim data in a venue already known and used by the UTLS research community.
The results from model studies using data acquired during StratoClim will be stored in relevant international databases, while metadata and links to the original datasets will be provided via the HALO database. Most of the model data sets are also part of the IGAC/SPARC activity CCMI (Chemistry Climate Model Initiative). Among others, such simulations have been performed for the upcoming ozone assessment (UNEP-WMO) and therefore these data are stored at the BADC (http://data.ceda.ac.uk/badc/wcrp-ccmi/data/CCMI-1/output/ ), where they are already publicly available (open access).
The project HALO database will be populated gradually, as final versions of the data become available. However, all data will be available via the HALO database by the summer 2020 at the latest.



Potential Impact:
From the campaign activities within StratoClim WP1 provides a unique data set as well as first process studies focusing on the ASMA, which is one of the major input regions of tropospheric air into the upper atmospheric layers and the most important region for input of anthropogenically highly polluted air masses from the South-East Asian boundary layer. These products mark the starting point of a strategically planned chain within StratoClim with a final goal to improve existing chemical climate models (CCM) and earth system models (ESM) in order to enable more accurate predictions of the global as well as local climate development and possibly identify hitherto unknown coupling mechanisms which can reduce or enhance global warming or local rainfall etc. The importance of such processes in the UTLS on the global climate has been underestimated for a long time but clearly shown by e.g. (Solomon et al., 2010, Solomon et al., 2011, Maycock et al., 2013). The further StratoClim strategy employs the analysis of process data to develop a detailed dynamical and/or chemical-physical understanding (WP4) and mathematical parametrizations that enable inclusion of the relevant processes into the CCMs and later ESMs (WP4 and WP5). Then the impact on reduced or enhanced emissions in SE Asian on global and local climate can be much more reliably determined and possible measures can be developed if deemed necessary. The completion of these steps will naturally happen only years after the end of the project in 2019.
An important and obvious example of such a process is the formation of the Asian Tropopause Aerosol Layer (ATAL) (Vernier et al., 2011) which has been discussed since due to its potentially high radiative and therefore climatological impact and now has been shown by StratoClim to be caused primarily through ammonium nitrate formation from ammonia emissions most probably due to agricultural activity (cattle farming and fertilization) (Höpfner et al., 2019). By the introduction of this process into global models the impact of further high or reduced emissions of ammonia and their main emission regions can be modelled. However, in order to reliably model such scenarios also the dynamical as well as other chemical-physical processes need to be checked and improved or even introduced into the models. StratoClim WP1 has also delivered many relevant and detailed data to for the first time enable such important checks and upgrades. As indicated above this will not just result in better global climate prognostics but also (in the longer term) improve local weather forecasts in order to provide reliable early warnings for storm and flooding events (causing landslides, etc.) or upcoming drought periods by more reliable long-term forecasts. Therefore, the results are not just important for the general global but also for the local public sectors in SE Asia.

Additionally, during StratoClim the high-altitude research platform M55-Geophysica has been considerably upgraded to a truly unique atmospheric research platform for such important process studies. It is now fit to aim for the exploration of other important atmospheric domains where we lack suitable understanding to check and improve our global models or even for further more detailed exploration of the ASMA. As a side effect of the intensive instrument development for the StratoClim campaigns the new innovative sensors are available also for integration on other atmospheric measurement platforms which has happened already or is ongoing for several key instruments.

The interest in the outcomes of the StratoClim studies has manifested in many local cooperations that where established with renowned institutions in the countries affected directly by the Asian Summer Monsoon and where better local forecast are vital for the lives of the population as well as economic evolution but also will be affected dramatically by global climate change should weather patterns change or just by rising sea levels along their coast lines. StratoClim has several Associate Partners (AP) as well as more Strategic Partners (without official association):
• University of Chicago, USA, Prof. Elizabeth Moyer, the group integrated and operated CHIWIS onboard Geophysica, a very important instrument to measure water vapour isotopic signatures
• Jet Propulsion Laboratory (JPL), Pasadena Ca., USA, Dr. Michelle Santee, the group supplied near-real time satellite data (MLS onboard AURA) on several important species for flight planning and data interpretation, the group took part in the ASMA campaign
• Dhaka University, Bangladesh, Physical-Chemistry Institute, Prof. Abdus Salam (AP)
• Chinese Academy of Sciences, Beijing China, Institute of Atmospheric Physics, LAGEO, Prof. Jianchun BIAN (AP)
• University of Cochin, India, Pr Kesavapillai Mohanakumar (AP)
• NASA Langley Research Centre, USA, Chemistry and Dynamics Branch, Science Directorate, T. Duncan Fairlie (AP)
• Indian Institute for Tropical Meteorology (IITM), Pune, India, IITM (with its head office Ministry of Earth Sciences) was a very important strategic partner for StratoClim WP1 when the AM aircraft campaign was still planned in India, however there are still close cooperation in data interpretation and modelling., Dr. Suvarna Fadnavis.
• Nepal Academy of Science and Technology (NAST), Lalitpur, Kathmandu, Nepal, Prof. Pokharel (former vice chancellor)
• International Centre for Integrated Mountain Development, Kathmandu, Nepal, (ICIMOD), Prof. A.K. Panday
• Department of Hydrology and Meteorology (DHM), Kathmandu, Nepal, Dr. Jagadiswor Karmacharya (Deputy Dir. General)
Years before the Kathmandu campaign contacts were established with the Nepalese organizations, in order to raise the interest in the local science community but also to get some political support for the process of flight and measurement permission approval with the administration (however, interrupted by the earthquake in 2015 and re-established in late 2016). Three months before the campaign a public hearing was organized to inform the public and the local media and all cooperating science institutes. Two common stakeholder meetings to inform about early and more elaborate results of the campaign were held during the campaign and in early 2017 in Kathmandu. The campaign activities were well covered by the local newspapers with several highly visible articles. It must be mentioned that all Nepalese civil and military authorities were very helpful during the campaign approval process that was heavily supported by the German Embassy at Nepal and our local representative Lt. Col. Karki.

A new atmospheric observation site was setup up at the Palau Community College in Koror/Palau. Besides PCC further cooperation partners on Palau are Coral Reef Research Foundation (CRRF) and CTSI logistics. The collaboration with these local partners were established successfully and are stable. Capacity building activities in Palau were pursued from the opening ceremony of the PAO onwards with growing success. A number of lectures by visiting scientists was held at the hosting Palau Community College (PCC) during the project (M. Rex, AWI (2x), J. Notholt, Uni Bremen, K. Müller, AWI (2x), F. Cairo, CNR). With the support of the German Honorary Consul of Palau these events were addressing not only college students and staff, but also invited Palauan policy makers and diplomates, the media (print and TV) and the interested public. The lectures were used to inform the audience and wider public about the scientific progress of the project as well as ongoing measurements and future perspectives of the PAO.
A poster series at the PCC presents the principle background of the instrumentation at the PAO and interested students and staff are more and more frequently attending the balloon launches in the Olympic Stadium. Additional presentations were frequently given to diplomatic delegations and at the local Rotary Club. The PCC is the largest body of education between Manila and Hawaii, with a large international studentship from Pacific island states. Hence, our educational messages on our current research objectives and results as well as general education related to the local climate and climate change are carried into the whole area.
In spring 2017 a fruitful collaboration with the Palau High School was started with direct support of the Palauan Minister of Education, which could be extended throughout the following years. The program for the students includes basic atmospheric science lectures on the relevance of the Stratosphere in the TWP and hands-on ozone sonde preparations and launches (by K. Müller (2x), AWI and J. Tradowsky (6x)). Also, in 2017, an ozone sonde presentation and launch was performed in front of a full stadium twice, reaching all high school students and staff, i. e. approximately 750 people each time. Recently the environmental science curriculum has been extended and our contributions are very welcome. In the future the education of teachers as multipliers could be a new focus, which was already asked for by the Minister of Education.
Especially our observations’ attestation of the clean air environment in Palau is well received and might even be marketed within the tourism industry, which is the strongest economic branch in Palau. On a political dimension, the Alliance of Small Island States (AOSIS) with Palau as a member benefits from its role as a host of a climate research station. The standing of the AOSIS in the UN during negotiation for climate change mitigation will be strengthened.
Our balloon campaigns in Bangladesh, Nepal and India were also accompanied by outreach activities. During the balloon campaign in Nainital, scientists and students from IITM and ARIES were trained to the deployment of the balloon-borne payloads, so that after the end of the monsoon season, 5 extra balloon soundings were performed by local personnel in November 2016. In January 2018 Ralph Lehmann (AWI Potsdam) stayed at the University of Dhaka for two weeks, in order to give lectures on stratospheric chemistry to master students of the Department of Chemistry. From November 2018 to January 2019 a master student from the University of Dhaka stayed at AWI Potsdam for 3 months.
As part of our Arctic Match campaign the public was informed via a press release because the analysis of the meteorological situation in February 2016 suggested a risk for a strong depletion later on. A second press release was published after the stratospheric warming when it became evident that a situation like in the winter 2010/11 (with record ozone losses) was not happening again in that winter.

In WP3, mission-long 10-year data sets of Envisat/MIPAS SO2, OCS, sulfate aerosol and NH3 with global coverage and altitude coverage from the middle troposphere (cloud top) to the lower stratosphere have been derived and have been made publicly available for further scientific studies. The sources, transport, and chemical conversion of sulfur in the atmosphere has been analyzed on basis of the MIPAS data sets with a help of a chemical transport model. The lifetime of SO2 is explained well by its oxidation by hydroxyl radicals. The numerous minor volcanic eruptions in the years 2002 to 2012 have contributed to more than 50 % of the total sulfur mass of the atmosphere. The studies indicated that sulfate aerosols injected into the Northern mid-latitudinal stratosphere will be distributed over the full globe, also over the southern hemisphere. This has to be taken into account for potential proposed climate-engineering schemes using sulfate aerosol. These results have been published in (Höpfner et al., 2015; Glatthor et al., 2015; 2017; Höpfner et al., 2016a, 2016b; Günther et al., 2017, 2018).
Detection schemes for sulfate aerosol precursors (SO2, OCS, H2S) have been developed for nadir-looking infrared sounders with high spectral resolution (IASI, TANSO-FTS). A new very fast algorithm (applicable in near-real time) has been developed to estimate the sulfate Aerosol Optical Depth (AOD) at 10 µm from the Metop-A and Metop-B IASI spectra. A two-years data set from IASI will be made publicly available. While the geostationary instrument SEVIRI is insensitive to background secondary sulfate aerosol, it is in principle capable to detect sulfate aerosol from severe volcanic condition. An algorithm developed for SEVIRI allows for the first time the identification of regions with presence of secondary sulfate aerosol particles from geostationary observations and can readily be coupled with existing cirrus clouds products from SEVIRI (or other similar geostationary platforms) for studying the glaciogenic impact of such UTLS aerosol emissions. This product, being very fast, may be used in NRT volcanic plume dispersions analyses following major volcanic eruptions. It has been recently proposed as a test algorithm during the SPARC SsiRC VolRes activities foreseen for the possible imminent Agung volcano eruption. To this aim, the transfer of the existing SEVIRI algorithm to the Himawari observing system is presently ongoing. These results have been published in (Sellitto and Legras, 2016; Sellitto, 2016, 2017).
For classification of clouds and detection of cloud top pressures, the full dataset of the Meteosat 10 and Himawari 8 satellites in the extended monsoon region (0-50N, 10W-160E) for May to September 2017 have been processed using an improved version of the operational algorithm of the Eumetsat NWC SAF supporting nowcasting and short-range forecast. The final version of this data set will be made available to the public.

In the European part of WP6, a methodology has been developed in which the severity of cold weather can be quantified and compared across different locations and periods of time. This allows, for example, the severity of cold weather in the future (based on models) to be compared to previous years. This allows local authorities to plan appropriately for predicted future winters, balancing the negative consequences of underpreparing and the wasted resources of over-preparing. At national or EU level, it could allow for checking the preparedness of cities or countries, and potentially the allocation of resources to areas expected to be hard hit by extreme weather.
On the Asian monsoon side of WP6, an analysis was completed quantifying the benefit to the Indian economy of reducing SO2 emissions (from India) via the increase in rainfall from the reduced SO2(the effects of SO2 on crop and human health are not captured in this analysis). This showed that the benefit of SO2 emissions mitigation is between around US$200-400 per ton. At around $400 per ton, around half of India’s total SO2 emissions could be mitigated cost-neutral or better. This information could be used to justify action to mitigate SO2 emissions by the Indian government, which might include income redistribution to disadvantaged groups, reduction of SO2 generating activities (especially coal-fired electricity generation) and/or removal of SO2 from e.g. power-plant exhausts (which would have crop and human health benefits in addition to those mentioned above).


List of Websites:
stratoclim.org
Project Coordinator, Prof. Markus Rex, Alfred Wegener Institute, Markus.Rex@awi.de
Aircraft campaign Coordination, Dr. Fred Stroh, Jülich GmbH, f.stroh@fz-juelich.de
Modelling studies, Prof. Martin Dameris, DLR, Martin.Dameris@dlr.de
Socio-economic impacts, Dr. Neil Harris, Cranfield University, neil.harris@cranfield.ac.uk