Altitude of atmospheric tracer transport towards Antarctica inpresent and glacial climate

https://doi.org/10.1016/j.quascirev.2009.06.020Get rights and content

Abstract

The preferential altitude of transport of continental tracers towards Antarctica under present and Last Glacial Maximum (21 kyr BP) conditions is analysed using an atmospheric general circulation model with idealized tracers which are emitted at the surface of Australia and South America. It is shown that the difference between the preferential transport altitude of Australian and South American tracers is similar in glacial and interglacial climates. Australian tracers arriving in Antarctica are consistently transported at higher altitudes than tracers emitted in South America. The frequency of low-level transport is stronger at the LGM than at present, reflecting a more vigorous atmospheric circulation at the LGM as a consequence of increased baroclinicity. While the spatial patterns of the total tracer concentrations at the Antarctic surface differ for Australian and South American tracers, with the regions of maximum surface concentration being located to the south-east of the respective tracer sources, the spatial distribution of the part advected via upper atmospheric levels is very similar for the Australian and South American tracers, with a maximum over Queen Maud Land. The simulated changes in transport characteristics cannot explain observed glacial to interglacial variations of dust size spectra which have been interpreted as indicators of the relative intensity of upper and lower level atmospheric dust transport to Antarctica.

Introduction

Aerosols and other species such as trace gases found in central Antarctic ice cores are in most cases clearly correlated with major climate shifts on orbital time scales (e.g., Petit et al., 1999, Lambert et al., 2008). For example, glacial ice core dust concentration at Vostok (see Fig. 1 for geographical setting) exceeds interglacial values by a factor of about 25–50 (e.g., Petit et al., 1981, Petit et al., 1999), and the link between millennial-scale variations of climate and dust deposition tightens during cold periods (Lambert et al., 2008). Changes in aerosol deposition in central Antarctica can schematically be reduced to changes in (1) source intensity; (2) atmospheric aerosol lifetime; (3) atmospheric pathway; and (4) deposition processes. Measured aerosol or ice core particle concentrations will inevitably be determined by all of these factors. Therefore it is necessary to disentangle the influence of these individual factors in order to be able to correctly interpret measured variations of ice core data. This study focuses on glacial to interglacial changes of atmospheric transport, and particularly on the altitudinal characteristics of this transport, between the mid-latitude continents and Antarctica.

Mass convergence in the middle troposphere above Antarctica, and subsequent subsidence, are forced by a persistent and almost exclusively unidirectional horizontal divergence flow in the boundary layer (James, 1989). This near-surface wind pattern is a drainage flow induced by the topography of the ice surface and the persistent, continental-scale surface inversion (e.g., Ball, 1956, Phillpot and Zillman, 1970, Parish and Waight, 1987, Parish and Bromwich, 1991). Egger (1985) and James (1989) related the Antarctic drainage flow (and the mid-tropospheric convergence and subsidence) to characteristics of the atmospheric flow at lower latitudes. James (1989) concluded that the Antarctic drainage flow and its associated mid-tropospheric circulation patterns aloft are probably maintained through the export of cyclonic vorticity by interactions between the polar vortex and decaying mid-latitude systems, without which the drainage flow would cease within a few days. Multiple factors of a local and regional nature thus determine the strength and localization of the subsidence of middle tropospheric air over Antarctica. It is obvious that these factors can vary during climatic transitions. Therefore the intensity of this direct polar cell might also vary across terminations. The intensity and frequency of lower level advection events are also subject to variations during climatic transitions, because they are linked to the presence of cyclonic systems off the Antarctic coast (Krinner and Genthon, 2003), the density of which varies with climate (Krinner and Genthon, 1998). Very schematically, there are thus two types of tropospheric tracer transport towards the interior of the Antarctic continent. The first type is fast, low-level advection enhanced by cyclonic systems off the Antarctic coast. The second type is advection via mass convergence in the middle troposphere above Antarctica. The relative intensity and efficiency of these types of transport depend of course on the transported species itself, but they certainly also vary with climate.

For present-day conditions, a model-based investigation by Li et al. (2008) showed that different dust sources (Patagonia, Australia, northern hemisphere) produce distinctive patterns of horizontal and vertical transport. In particular, they show that over Antarctica, dust originating from Patagonia is concentrated in the lower troposphere, while dust from Australia has its highest concentration in the upper troposphere (above 400 hPa). Changes of transport altitude on orbital time scales would certainly leave their imprint on the dust particle size spectra or on the fine particle fraction (defined as the proportion of total dust mass made up by particles below a given threshold size), smaller particles being typically associated with long-range, high-altitude pathways including subsidence over the Antarctic Plateau (Delmonte et al., 2004). Mahowald et al. (2006) state that their climate model, including a prognostic mineral dust cycle using four grain size bins, simulates an increase in the fraction of particles in the smallest bin (<2.5 μm). Their model therefore seems to reproduce the tendency of glacial to interglacial changes of dust grain sizes at Dome C (Antarctica, see Fig. 1) reported by Delmonte et al. (2004). Mahowald et al. (2006) conclude that their model results, indicating smaller dust grain size at Dome C during the Last Glacial Maximum, are consistent with the underlying mechanism proposed by Delmonte et al. (2004), that is, a longer atmospheric pathway. It is noteworthy, however, that Mahowald et al. (2006) do not indicate whether their model reproduces the opposite glacial to interglacial changes that Delmonte et al. (2004) reported for ice cores at a few hundred kilometres distance (see Fig. 1). Lunt and Valdes (2001) analysed back trajectories of air masses arriving at Dome C using modelled glacial and modern wind fields. Concerning the vertical characteristic of atmospheric transport, they mention that the trajectories generally decrease in height as they move northwards away from the pole.

Here we use idealized tracers in an atmospheric general circulation model to address the following questions: Do glacial to interglacial atmospheric circulation changes induce changes in the preferential altitude of tracer transport towards the Antarctic? If yes, are these changes different for tracers originating from Australia and Patagonia? And, if existent, do these changes exhibit spatial patterns on sub-regional scales (∼100 to ∼1000 km), and do they depend on the tracer lifetime? Our results are discussed with respect to ice core dust data which were interpreted as indications of strong regional-scale differences in the vertical characteristics of dust advection (Delmonte et al., 2004). We address the question whether our simulations support such interpretations.

Section snippets

Methods

We use the LMDZ3.3 atmospheric general circulation model with specific modifications improving the model performance in polar regions (Krinner et al., 1997). The model is run with 96 (longitude) × 72 (latitude) × 19 (vertical) grid points, irregularly spaced in latitude to obtain a horizontal resolution of about 100 km over Antarctica. Two simulations, lasting 20 years (plus 1 year of spin-up) each, were carried out: one for present-day climate and one under Last Glacial Maximum (LGM, 21 kyr BP)

Results

The annual mean vertical and zonal distribution of TS,3 and TA,3 above the Antarctic (defined here as the region polewards of 75°S to facilitate comparison with the study by Li et al. (2008)) is broadly similar at present and at the LGM (Fig. 3). At both periods, the concentration of the South American tracers TS,3 reaches high values in lower levels between about 30°W to 0°W, that is, in the Atlantic sector of the Antarctic (Filchner-Ronne Ice Shelf and Queen Maud Land). The concentration of

Discussion

The results above can be summarized as follows:

  • For tracers with atmospheric lifetimes between 3 and 10 days, transport from South America and Australia to Antarctica preferentially occurs via levels below 400 hPa both at present and at the LGM. Upper-level advection is more frequent for Australian than for south American tracers, in agreement with previous modelling results (Li et al., 2008).

  • For both source regions, the fraction of tracers of given lifetime advected via the upper-level pathways

Conclusion

Li et al. (2008) showed that under present climatic conditions, dust transport from Australia to Antarctica preferentially occurs at high altitude (above 5000 m), while transport of South American dust occurs at lower levels. Our analysis shows that a similar distinction can also be made for idealized, radon-like tracers, and that this difference between Australian and South American tracers also holds for the Last Glacial Maximum. However, in glacial climate, the frequency of fast low-level

Acknowledgements

This work is a contribution to the European Project for Ice Coring in Antarctica (EPICA), a joint ESF (European Science Foundation)/EC scientific programme, funded by the European Commission and by national contributions from Belgium, Denmark, France, Germany, Italy, the Netherlands, Norway, Sweden, Switzerland and the United Kingdom. Complementary analyses were carried out with the French project ANR PICC. The simulations were carried out on the Mirage platform (UJF Grenoble) and at

References (35)

  • M. Frezzotti et al.

    New estimations of precipitation and surface sublimation in East Antarctica from snow accumulation measurements

    Clim. Dyn

    (2004)
  • J.M. Gregory et al.

    Ice-sheet contributions to future sea-level change

    Philos. Trans. R. Soc. A

    (2006)
  • F. Hourdin et al.

    The use of finite-volume methods for atmospheric advection of trace species. Part I: Test of various formulations in a general circulation model

    Monthly Weather Rev.

    (1999)
  • I.N. James

    The Antarctic drainage flow: implications for hemispheric flow on the Southern Hemisphere

    Antarctic Sci.

    (1989)
  • Joussaume, S., Taylor, K.E. 1995. Status of the Paleoclimate Modeling Intercomparison Project (PMIP), In: Proceedings...
  • G. Krinner et al.

    GCM simulations of the Last Glacial Maximum surface climate of Greenland and Antarctica

    Clim. Dyn

    (1998)
  • G. Krinner et al.

    Tropospheric transport of continental tracers towards Antarctica under varying climatic conditions

    Tellus

    (2003)
  • Cited by (29)

    • Seasonality of Radon-222 near the surface at King Sejong Station (62°S), Antarctic Peninsula, and the role of atmospheric circulation based on observations and CAM-Chem model

      2022, Environmental Research
      Citation Excerpt :

      It has been used in atmospheric transport studies to examine vertical mixing in the atmospheric boundary layer at the local scale (Lee and Larsen, 1997; Lupu and Cuculeanu, 2001) and analyzing the transportation from its origin at the regional scale (Arnold et al., 2010). Recently, in the development and improvement of global atmospheric chemistry models for research on climate change and its impact, radon has been widely used as an atmospheric tracer to evaluate the performance of these models (e.g., Krinner et al., 2010; Locatelli et al., 2015; van Noije et al., 2014; Zhang et al., 2021, 2011, 2008). The large uncertainty of the general circulation models (GCM) over the Antarctic region makes continuous, high-quality radon data a valuable evaluation metric.

    • Late Holocene dust provenance at Siple Dome, Antarctica

      2021, Quaternary Science Reviews
      Citation Excerpt :

      Interglacial dust samples from Vostok and Dome C show higher εNd (∼−10 to −4), but a similar range of 87Sr/86Sr ratios (Fig. 2). Located at elevations >3500 m, the Vostok and Dome C ice cores sample far-traveled dust transported at altitudes of ≥4000 m, and are less influenced by local sources (Delmonte et al., 2004, 2013; Krinner et al., 2010). Surprisingly, the Dome A (∼4000 m) snow pit data have lower εNd (mostly −18 to −10, but as low as −27), more similar to Siple Dome, likely indicating the input of Antarctic dust (Du et al., 2018).

    • Stable isotopic composition in snowpack along the traverse from a coastal location to Dome A (East Antarctica): Results from observations and numerical modeling

      2020, Polar Science
      Citation Excerpt :

      The offshore region (50–60°S) near the coast also contributes moisture to ZS, but the total fractions are less at SP6 and SP12. This difference could be associated with the coastal cyclonic systems, which drive the evaporated vapor off the Antarctic coast to be transported to a limited distance (Krinner et al., 2010). The longitude of dominant moisture source region lies in the westward sector with respect to precipitation sites (~77°E, Fig. 4).

    • Causes of dust size variability in central East Antarctica (Dome B): Atmospheric transport from expanded South American sources during Marine Isotope Stage 2

      2017, Quaternary Science Reviews
      Citation Excerpt :

      Dust emissions from both sources likely increased during MIS 2, although in different proportions (e.g. Albani et al., 2012). It has been established that the fine component of dust from these sources can travel long-range to the high southern latitudes of Antarctica, although the preferential altitude of transport is different (Li et al., 2008; Krinner et al., 2010; Albani et al., 2012). Independently from climate conditions, Australian dust is transported consistently at higher elevation than dust from South America.

    View all citing articles on Scopus
    View full text