Atmos. Chem. Phys., 16, 8389–8403, 2016 www.atmos-chem-phys.net/16/8389/2016/ doi:10.5194/acp-16-8389-2016 © Author(s) 2016. CC Attribution 3.0 License. Observations of PAN and its confinement in the Asian summer monsoon anticyclone in high spatial resolution Jörn Ungermann, Mandfred Ern, Martin Kaufmann, Rolf Müller, Reinhold Spang, Felix Ploeger, Bärbel Vogel, and Martin Riese Institut für Energie- und Klimaforschung – Stratosphäre (IEK-7), Forschungszentrum Jülich GmbH, Jülich, Germany Correspondence to: Jörn Ungermann ([email protected]) Received: 14 January 2016 – Published in Atmos. Chem. Phys. Discuss.: 1 February 2016 Revised: 30 May 2016 – Accepted: 15 June 2016 – Published: 12 July 2016 Abstract. This paper presents an analysis of trace gases in 1 Introduction the Asian summer monsoon (ASM) region on the basis of observations by the CRISTA infrared limb sounder taken The Asian summer monsoon (ASM) anticyclone is the dom- in low-earth orbit in August 1997. The spatially highly re- inant circulation pattern in the summertime Northern Hemi- solved measurements of peroxyacetyl nitrate (PAN) and O3 sphere in the upper troposphere/lower stratosphere (UTLS). allow a detailed analysis of an eddy-shedding event of the It has a major impact on stratosphere–troposphere exchange ASM anticyclone. We identify enhanced PAN volume mix- (STE) and thus on the trace gas composition of the north- ing ratios (VMRs) within the main anticyclone and within ern lowermost stratosphere and tropical upper troposphere. It the eddy, which are suitable as a tracer for polluted air origi- is caused by persistent thermal heating during summer with nating in India and China. Plotting the retrieved PAN VMRs a thermal low at lower altitudes and a strong anticyclone in against potential vorticity (PV) and potential temperature re- the upper troposphere (e.g. Krishnamurti and Bhalme, 1976). veals that the PV value at which the PAN VMRs exhibit the The ASM can be found in the larger vicinity of the Tibetan strongest decrease with respect to PV increases with poten- Plateau and is enclosed by the subtropical westerly jet on tial temperature. These PV values might be used to identify its northern side and the tropical easterly jet on its south- the extent of the ASM. Using temperature values also de- ern side (e.g. Randel and Park, 2006). These jets act as a rived from CRISTA measurements, we also computed the strong transport barrier and prevent the air inside the anti- location of the thermal tropopause according to the WMO cyclone from mixing along isentropic levels into the extrat- criterion and find that it confines the PAN anomaly vertically ropical lower stratosphere or the equatorial tropics (e.g. Park within the main ASM anticyclone. In contrast, the shed eddy et al., 2008). Previous studies using satellite data have thus exhibits enhanced PAN VMRs for 1 to 2 km above the ther- found an enhancement of carbon monoxide (e.g. Li et al., mal tropopause. Using the relationship between PAN as a tro- 2005; Park et al., 2007) and other trace gases in the ASM, pospheric tracer and O3 as a stratospheric tracer to identify indicating that polluted air is trapped within the anticyclone mixed air parcels, we further found the anticyclone to contain below the thermal tropopause (Park et al., 2008). The ASM few such air parcels, whereas the region between the anticy- is largely responsible for the moistening of the lowermost clone and the eddy as well as the eddy itself contains many stratosphere in summer (e.g. Gettelman et al., 2004; Ploeger mixed air parcels. In combination, this implies that while the et al., 2013). It is, however, not fully clear how much of the anticyclone confines polluted air masses well, eddy shedding STE occurs isentropically, by eddy shedding, or vertically provides a very rapid horizontal transport pathway of Asian across the tropopause (e.g. Hsu and Plumb, 2000; Randel pollution into the extratropical lowermost stratosphere with et al., 2010; Vogel et al., 2015; Garny and Randel, 2016). a timescale of only a few days. The CRyogenic Infrared Spectrometers and Telescopes for the Atmosphere (CRISTA; e.g. Offermann et al., 1999) ex- periment was flown on two Space Shuttle missions (STS66 and STS85) in November 1994 and August 1997 respec- Published by Copernicus Publications on behalf of the European Geosciences Union. 8390 J. Ungermann et al.: Observations of PAN in the ASM tively. During free-flying periods of about 10 days, global 2 Measurements and model data observations of temperature and numerous trace gases were made with unprecedented and so far unmatched spatial reso- The measurements presented in this paper were made aboard lution (Riese et al., 1997, 1999). It was designed to examine the Space Shuttle experiment CRISTA. The limb sounder dynamical processes primarily in the stratosphere and meso- CRISTA was integrated into the astronomical Shuttle Pallet sphere, but it also measured the UTLS region. System (ASTRO-SPAS; Wattenbach and Moritz, 1997) and By using three viewing directions simultaneously, a hor- launched into orbit by the NASA space shuttle for two mis- izontal across-track sampling of 600 km was achieved. The sions: once in November 1994 (CRISTA-1) and once in Au- along-track sampling was 200 to 400 km, depending on the gust 1997 (CRISTA-2). The instrument employs three Her- measurement mode. The vertical sampling was 1.5 km in the schel telescopes (pointing towards −18, 0, and 18◦ in ref- first mission and 2 km in the second mission (Grossmann erence to the viewing direction, which varied, depending on et al., 2002). The spectral resolution was sufficient to deter- the latitude, between 108 and 252◦ with respect to flight di- mine 25 trace gas species (Offermann et al., 1999). While rection Grossmann et al., 2002), four Ebert–Fastie spectrom- more recent satellites offer marginally better vertical reso- eters (Fastie, 1991), and 24 mid-infrared detectors ranging lution in the UTLS (especially solar occultation instruments from 155 to 2390 cm−1 and covered an altitude range from such as ACE-FTS; Bernath et al., 2005) or a higher spectral ≈ 9 to 150 km altitude in the limb. Here, we focus on the resolution (such as MIPAS; Fischer et al., 2008), no other in- detector channels covering the spectral range from 777 to strument since has offered such a spatial measurement den- 862 cm−1 (with a spectral resolution of λ/1λ ≈ 500) more sity. relevant to this work and the altitude range between 9 and Here, we will examine the confinement of polluted air 45 km. The field of view has a size of 3 arcmin × 30 arcmin, within the ASM by means of trace gases retrieved from which corresponds to about 1.7 km × 17 km at 18 km tan- CRISTA-2 measurements (that is measurements taken dur- gent height, whereby spectra were usually measured in ing the second CRISTA mission) taken in August 1997. We 2 km altitude steps. The typical along-track sampling for focus on peroxyacetyl nitrate (PAN) measurements, which the three measurement tracks and the discussed measure- offer the highest contrast between tropospheric air within the ments is ≈ 250 km, while the typical across-track sampling anticyclone and clean air outside compared to other tracers is ≈ 600 km. As the detectors and the optics were cooled available from CRISTA measurements. by cryogenic helium, a high measurement speed could be PAN is a tracer of tropospheric pollution. It is a sec- achieved in combination with a good signal-to-noise ratio – ondary pollutant that forms in the troposphere from precur- recording one spectrum took only 1.2 s, allowing the mea- sors whose main sources are biomass burning and anthro- surement of full profiles in less than 1 min with four spec- pogenic pollution (e.g. Stephens, 1969; Penketi et al., 1975; trometers operating in parallel. The vertical sampling is one Singh et al., 1986; Wunderli and Gehrig, 1991; Singh et al., spectrum every 2 km, which, taking field of view and pro- 2007). Its lifetime is highly temperature dependent and it cessing into account, translates to a vertical resolution of may be only seconds in the boundary layer but, close to the retrieved products of ≈ 2 km for temperature and O3 and the cold-point tropopause, PAN can accumulate for weeks ≈ 3 km for PAN in the UTLS. As the retrieval grid is aligned or months (Roberts, 1990). The combination of the relatively with the tangent altitudes of the measured spectra, there is long lifetime of PAN at tropopause altitudes together with no loss of resolution due to interpolation. No limb sounder the confinement of air in the ASM anticyclone is expected to since has achieved a comparable spatial measurement den- lead to an accumulation of PAN in the ASM circulation. Fur- sity at this vertical resolution. ther sources of PAN are thunderstorms in convective systems Previous work on CRISTA concentrated mainly on the ex- that produce NOx by lightning. The typical sinks are thermal ploration of the middle atmosphere, with only few studies decomposition at lower altitudes and photolysis in the strato- looking on the UTLS region (e.g. Spang et al., 2015). Here, sphere. Due to the low temperatures prevalent in the ASM an- we examine the ASM, which was observed for 8 consecu- ticyclone, PAN is well suited to identify polluted air masses tive days from 8 to 16 August 1997 during the CRISTA-2 that have been uplifted in the current season. For example, mission. PAN was also used in model studies to determine the influ- The calibrated spectra were available from previous stud- ence of regional emissions on the air within the ASM (e.g.