Trans-Pacific dust events observed at Whistler, British Columbia during INTEX-B

The meteorology and physico-chemical characteristics of aerosol associated with two new cases of long range dust transport affecting western Canada during spring 2006 are described. Each event showed enhancements of both sulfate aerosol and crustal material of Asian origin. However, the events were of quite different character and demonstrate the highly variable nature of such events. The April event was a significant dust event with considerable enhancement of fine particle sulfate while the May event was a weaker dust event, also with significant fine particle sulfate enhancement. The latter event was notable in the sense that it was of short duration and was quickly followed by a large increase of organic material likely of regional origin. Comparison of these two events with other documented cases extending back to 1993, suggests that all dust events show coincident enhancements of sulfate and crustal aerosol. However, events vary across a wide continuum based on the magnitude of aerosol enhancements and their sulfate to calcium ratios. At one extreme, events are dominated by highly significant crustal enhancements (e.g. the well-documented 1998 and 2001 "dust" events) while at the other are events with some dust transport, but where sulfate enhancements are of very high magnitude (e.g. the 1993 event at Crater Lake and the 15 May 2006 event at Whistler). Other events represent a "mix". It is likely that this variability is a function of the comparative strengths of the dust and anthropogenic SO 2 sources, the transport pathway and in particular the extent to which dust is transported across industrial SO 2 sources, and finally, meteorological and chemical processes.


Introduction
During April-May 2006, INTEX-B (the Intercontinental Chemical Transport Experiment) was focused on the North Pacific with the goal of providing detailed chemical analysis of tropospheric air following a trans-Pacific pathway to the North American continent. This was prompted by over a decade of observational and modelling studies demonstrating the relatively rapid (especially during springtime) trans-Pacific transport of both crustal material (predominately desert dusts) and anthropogenic pollutants from sources in Eurasia to North America (e.g. Jaffe et al., 1999Jaffe et al., , 2003Husar et al., 2001;Holzer et al., 2003;Jacob et al., 1999). In this context, the mobilisation and transport of min-5 eral dust from the arid regions of the world is considered of major significance due to the role airborne crustal material plays in the global radiation balance (and hence climate forcing), cloud processes, atmospheric chemistry, oceanic and terrestrial biogeochemical processes (e.g. dust is a major source of iron and calcium), as vectors for microbes, and as a factor influencing local air quality in both "source" and "sink" regions (Prospero et al., 2002). Furthermore, recent studies suggest that dust participates in important interactions with anthropogenic pollutants such as sulfur dioxide (SO 2 -a source of sulfate aerosol). Such pollutants are often mixed into, and interact with, the dust plumes as they pass over urban/industrial sources (Li-Jones and Prospero, 1998;Heald et al., 2006). 15 Studies of trans-Pacific dust transport suggest a strong linkage between anthropogenic sulfate aerosol and crustal dust transport from Asia. For example, VanCuren's (2003) analysis of aerosol data at Crater Lake, (Oregon) and Mt. Lassen (California) shows that a mixture of dust and combustion products dominate Asian aerosol that arrives in North America. Major fine particle (<2.5 µm diameter) constituent fractions in western North America. This is particularly relevant when set against a background of burgeoning emissions of SO 2 from Asia, where not only have SO 2 emissions increased by 119% between 1980119% between -2003119% between (Ohara et al., 2007, but also Asian sulfate influx to North America has increased 2. 4-3.4 fold between 19854-3.4 fold between -20064-3.4 fold between (van Donkelaar et al., 2008. Streets and Waldhoff (2000) indicate that Asian SO 2 emissions are 5 projected to increase from 25.2 mt in 1995 to 30.6 mt in 2020 (assuming emission controls are implemented on large power plants) and possibly to 60.7 mt without emission controls.
Beginning with the well-documented 1998 dust episode, a growing inventory of trans-Pacific dust events affecting North America has been established. Most of these events 10 have involved sources in the Gobi and Takla Makan deserts Asia (Husar et al., 2001;Jaffe et al., 2003;Thulasiraman, 2002), although recently, a case of Saharan dust transport across Asia and the Pacific to North America has been documented (McKendry et al., 2007). In addition, modelling studies have identified many aspects of the climatology, inter-annual variability and pathways of dust transport and deposition 15 (Holzer et al., 2003(Holzer et al., , 2005Gong et al., 2006;Zhao et al., 2006). The meteorological mechanisms associated with boundary layer -free tropospheric (BL-FT) exchange of dust (and anthropogenic pollutants) have also largely been identified. BL-FT exchange processes most relevant to dust source areas include the warm conveyer belt (WCB) mechanism (Cooper et al., 2004;Stohl et al., 2002) and frontal lifting, while in recep- 20 tor regions (such as western North America) subsidence, and mountain wave activity are important in bringing mid-tropospheric dust layers in range of BL entrainment processes Hacker et al., 2001).
Against this background of improved understanding of many aspects of trans-Pacific dust transport, the intensive observations afforded by INTEX-B provide a further op-25 portunity to extend the inventory of documented dust events and to examine the commonalities emerging from such a catalog of events. In this context, our goals in this paper are to: -Describe two separate dust events identified during INTEX B. 10278 -Compare and contrast these events with previously documented cases, particularly with respect to their magnitude and chemical signature. For the latter, sulfate aerosol measurements during INTEX-B provide a rare opportunity to examine the linkage between sulfate and mineral aerosol in these cases.

5
Measurements of particles and trace gases are made by Environment Canada at a high elevation site in Whistler, BC, approximately 100 km north of Vancouver (Fig. 1). The site is located at the top of Whistler Peak (2182 m above sea level). There are no continuous combustion sources at the peak and influences from snowmobiles have been identified and removed from the data set (Macdonald et al., 2006). During the period 10 of interest, particle size distributions were measured both with a TSI Scanning Mobility Particle Sizer (SMPS) Model 3934 and a Grimm optical particle counter (OPC) (Model 1.108). The SMPS was used to measure mobility diameter in the 0.01 µm to 0.4 µm size range while the OPC provided optical diameter from 0.3 µm to 20 µm. Particle Chemistry is from bulk filter packs, collected every 48 h and analyzed for inorganic ions 15 by ion chromatography. All particle sampling takes place through a heated stainless steel manifold and filter packs are preceded by a cyclone operated at approximately a 2 µm size cut. Particle sampling is suspended in the presence of fog to ensure no contamination of samples by cloud and that the sampling inlet remains dry and free of rime. In addition, A high mass resolution time-of-flight aerosol mass spectrometer Measurements of O 3 (TECO 49), CO (Aerolaser), size distributions of aerosol particles (CPC, PCASP100X, FSSP300) and the chemical size distribution (Quad-AMS) were conducted from a Cessna 207. Flights were approximately two hours long, and most flights consisted of ascent and descent profiles at Whistler to an altitude of about 5.3 km. Data were averaged every second, except for the AMS that was averaged at 5 one minute intervals in order to try to optimize temporal resolution and sufficient signal for detection. Further details of the instrumentation and comparisons are discussed by Leaitch et al. (2008) 2 .
Back-trajectories from Whistler were calculated using the HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory) model (www.arl.noaa.gov/ready/hysplit4. 10 html). The model is the newest version of a complete online system for computing simple air parcel trajectories to complex dispersion and deposition simulations for any location and date (depending on data availability) using a variety of standard data input products (e.g. the NCEP Reanalysis 1948-present). Further details and validation of the model can be found in Draxler and Hess (1998).

Dust events at Whistler
Observations at Whistler Peak revealed at least four distinct aerosol episodes during April-May 2006 (Fig. 2). This is consistent with observations of sand/dust storm events (SDS) in Northeast Asia (Zhou et al., 2007), Aeronet observations from Saturna Island 20 ( Fig. 1) and modeling results described by Zhao et al. (2007). Of these episodes, two are particularly noteworthy in the context of this study. The first, of highest magnitude, Introduction  2) and was associated with material with modal particle diameter in the range 2-3 µm. The second occurred from 15-16 May (Fig. 2), was of somewhat lesser magnitude and was notable in the sense that there appeared to be a shift late in the event to a particle size distribution dominated by fine particles (∼0.2 µm). Both events were in broad agreement in terms of magnitude 5 and timing with NAAPS forecasts for Cheeka Peak (available at www.nrlmry.navy.mil/ aerosol/), a nearby atmospheric chemistry monitoring site.

Meteorology and Asian dust storms
GEOS-Chem v7-04-09 at 2 • ×2.5 • (www.as.harvard.edu/chemistry/trop/geos/) was used to estimate dust emissions and transport from Asia to North America. It is 10 driven using assimilated meteorological data from the Goddard Earth Observing System (GEOS-4) at the NASA Global Modeling Assimilation Office (GMAO). The dust module is described in detail by Fairlie et al. (2007) and includes the effects of gravitational settling as well as wet and dry deposition.
The top panel of Fig , which extended to about 3 km during late afternoon, the particles were composed of about 60% organic material, 30% sulfate, and much of the balance was made up by nitrate and ammonium. Above 3 km, the fine particle aerosol was almost completely dominated by sulfate. The number concentrations of particles <1 µm (i.e. PCASP and 7610) were higher in the BL and much lower above 3 km, consistent 15 with the transport of dust from Asia. The presence of a predominantly sulfate aerosol is also consistent with past observations (e.g. Brock et al., 2004) and other observations during Intex-B (e.g. Peltier et al., 2008;Dunlea et al., 2008 3 ;van Donkelaar et al., 2008), and suggest that the aerosol was formed by the oxidation of SO 2 during transport, in both cloud and the gas phase. The fine particles above 3 km were also larger 20 than those in the boundary layer. This is evident in the PCASP size distributions (not shown) and from Fig. 7; the reduction in particle number concentrations from the BL to above 3 km is about a factor of four whereas the reduction in the fine particle mass is a factor of two or less. An increase in the size of the fine particles is an indication of a longer lived aerosol. Ozone is relatively high in the BL, and it increases from about 65 ppbv at 3 km to about 70 ppbv at 4.7 km. It increases another 10 ppbv in the plume between 4.7 and 5.2 km, evidence that photochemical production of ozone accompanied the production of sulfate. The INTEX-B observations were interpreted with a global chemical transport model (GEOS-Chem) to estimate that Asian anthropogenic 5 emissions during the period increased the mean profiles observed over Whistler for fine particle sulfate by 0.3-0.5 µgm −3 (van Donkelaar et al., 2008) and for ozone by 6-8 ppbV (Walker et al., 2008 4 ).
As discussed by Zhang et al. (2008), a predominantly sulfate plume impacted the Peak site on the morning of 15 May (Fig. 5b). The profile data show an increase in 10 sulfate between 2 km and 2.5 km at an elevation coincident with Whistler Peak. Accompanying this was a modest increase in O 3 and a substantial increase in CO (Leaitch et al., 2008).
Lidar imagery from the morning of 15 May 2006 (Fig. 6) confirms the presence of an aerosol layer at approximately the elevation of the Peak station (2182 m a.s.l.) and is in 15 agreement with the profile data at this time (Fig. 5b). During the course of the May 15 event, lidar suggests that subsidence was a significant process bringing the sulfate-rich aerosol layer to mountain ridge level. The importance of subsidence in bringing dust layers in reach of the planetary boundary-layer over mountainous regions of western North America was identified by Hacker et al. (2001). 20 The situation changed dramatically by the afternoon of 15 May (Fig. 5c). While there was still a thin sulfate plume similar to that during the morning, the fine particle aerosol was clearly dominated by organics between 2 km up to about 4.4 km. This abrupt change was also seen in the Peak data (Fig. 7,  for this time indicate subsiding air at the 2-5 km level, with origin from the south to south west over the previous one to two days (Fig. 8). Longer range trajectories come from the mid-Pacific and northern China (Fig. 7). Sun et al. (2008) attribute the origin of the organic plume (Case 2, Fig. 7) to regional sources with likely contributions from the urbanized regions extending from southern Puget Sound 5 to Georgia Strait (Seattle -Bellingham-Vancouver) with possible contributions from biomass burning that was evident along the California-Oregon border during the period (map.ngdc.noaa.gov/website/firedetects/viewer.htm). In addition, given elevated tree emissions in the Whistler valley during the time period, biogenic secondary organic aerosol might also have contributed to the observed enhancement of organic 10 aerosol mass.

Comparison with previously documented cases
In order to compare the dust events observed at Whistler during April-May 2006 with other documented dust events, IMPROVE data from the closest high altitude sites, with a sufficiently long record to encompass the Spring 1993 event, were examined 15 (Crater Lake, and Mt Hood, Oregon). Crater Lake is considered to be a pristine high altitude site that has been used in previous studies to examine trans-Pacific transport (Jaffe et al., 2005;Van Curen and Cahill, 2002;Zhao et al., 2007) Fig. 9a context of the longer term record from Crater Lake (Fig. 9), the dust events pale in comparison to the large events of 1998 and 2000 when calcium concentrations were enhanced by factors of 30-40 over mean background levels. However, the dust event of 23-26 April 2006 observed at Whistler (and apparent at Crater Lake) appears to be one of the largest 4-5 events observed over the thirteen year period. The May 2006   5 dust event on the other hand, is of low magnitude with concentrations comparable to the Saharan dust event observed in 2005(McKendry et al., 2007. For all events listed in Table 1, aerosol sulfate concentrations at Crater Lake were enhanced over annual mean background concentrations by factors in the range 2-5.6. This is smaller than the range of enhancements shown for crustal dust and likely re-10 flects the relatively constant anthropogenic source strength for Asian SO 2 compared to the highly variable dust sources. Of particular note is the large magnitude 15 May 2006 sulfate aerosol event that produced concentrations at Whistler peak (1.8 µgm −3 ) exceeding the maximum sulfate aerosol associated with documented crustal dust events observed at Crater Lake (1.74 µgm −3 on 28 April, 1993).

15
Ratios of SO 4 to Ca fine aerosol at Crater lake (Table 1) indicate some measure of the "mix" of aerosol for the different events, and together with enhancement factors, provide the basis for a somewhat subjective characterization of "events" with respect to the relative mix and magnitude of the sulfate vs mineral dust components (column two of Table 1). On this basis it is possible to define major dust events (29 April, 1998, 16 20 April 2001, 23-26 April 2006 in which there was low to moderate sulfate enhancement. In other cases, the event incorporated dust but was a significant sulfate event (28 April 1993, 15

Discussion and conclusions
During INTEX B, two significant aerosol events incorporating Asian dust were observed at Whistler Peak and in nearby aircraft profiles over southwestern British Columbia. Both events showed enhancements of both sulfate aerosol and crustal material of Asian origin. However, the events were of quite different character and demonstrate 5 the highly variable nature of such events. The April event was a significant dust event with moderate sulfate enhancement while the May event was a weak dust event with very significant sulfate enhancement. The latter event was interesting in the sense that it was of short duration and was quickly followed by significant enhancement of organic material likely of regional origin.

10
In terms of the meteorology of the two dust events observed during INTEX B, both were traceable to typical spring dust storms in China and shared the common pattern evident in previously documented cases (e.g. Husar et al., 2001;McKendry et al., 2005) of subsidence over western North America. Trajectories in the eastern Pacific were quite different in the two cases and were likely responsible for the differences in 15 chemical "signature" of the two events. In contrast to the northerly approach to western North America in the April case, the May 2006 episode trajectory approached British Columbia from the south. This particular trajectory was likely responsible for the switch to high organic concentrations late on the 15 May, likely a response to regional sources over western North America and to the south of Whistler. 20 In summarizing the documented Asian dust events in Western North America, it is apparent that, in agreement with Heald et al. (2006), all dust events appear to show coincident enhancements of sulfate and crustal aerosol. However, events appear to vary across a wide continuum based on the magnitude of aerosol enhancements and their ratios. At one endpoint are events dominated by highly significant crustal enhance-25 ments (e.g. the well documented 1998 and 2001 "dust" events) while at the other end of the spectrum are events with some dust transport, but where sulfate enhancements are of very high magnitude (e.g. the 1993 event at Crater Lake and the 15 May 2006 event at Whistler). Other events lie somewhere between. It is likely that this variability is a function of: 1. the comparative strengths of the dust and anthropogenic SO 2 sources; 2. the transport pathway and in particular the extent to which dust is transported across industrial SO 2 sources 5 3. meteorological and chemical processes.
The coincident transport of sulfate and mineral aerosol evident in these events has potential implications for air quality compliance and visibility in western North America, and is likely evolving due to changes in land-use and anthropogenic emissions in Eurasian source regions. Furthermore, transport of sulfate and dust from Asia may 10 have significant impacts on climate due to changes in particle size distributions and increased residence times of particles in the atmosphere. Introduction  2000RG000095, 2002. Stohl, A., Eckhardt, S., Forster, C., James, P., andSpichtinger, N.: On the pathways and timescales of intercontinental air pollution transport, J. Geophys. Res., 107, 4684, doi:10.1029/2001JD001396, 2002 Present and future emissions of air pollutants in China: SO 2 , 5 NO x , and CO, Atmos. Environ., 34, 363-374, 2000. Thulasiraman, S., O'Neill, N. T., Royer, A., Holben, B. N