A partial National Park Service Interagency Monitoring of Protected Visual Environments (IMPROVE) site was established at Mauna Loa Observatory (MLO) in 1989 to measure the mass and composition of fine particles (i.e., D_p <2.5 m diameter). While this site has always produced data on the more common trace elemental species such as zinc (see previous CMDL summary reports), analytical advances in early 1993 greatly increased sensitivity and allowed a much wider suite of trace elemental species to be measured. While the IMPROVE site at MLO quickly confirmed prior findings of regular Asian dust transport each spring, it also demonstrated that elevated levels of trace elements also occur during the dust episodes. In this summary, we will compare the present results to recently published aerosol summaries at MLO, examine the intense Asian dust episodes that occurred during the spring of 1994, and outline intensive measurements being made at MLO during the spring of 1996.

Comparisons with other long-term aerosol records became possible when Zieman et al. [1995] summarized their high volume air sampler (Hi-Vol) data from MLO. The Hi-Vol sampler was operated for a period of 1 week during nighttime/downslope winds and collected particles <30 m in diameter. For comparison purposes, we have summed the two weekly IMPROVE sampling periods (i.e., Wednesday through Friday and Saturday through Tuesday) to give mean weekly values during downslope winds.

The excellent agreement for sulfates in Table 1 indicates that these particles are predominantly <2.5 m in diameter. The results for soil, however, are typically different. For example, Hi-Vols normally collect about four times more soil mass than the IMPROVE PM_2.5 air sampler. Thus, one might expect the Hi-Vol to collect ~2.9 g m^-3 of soil during the spring Asian dust episodes and ~0.4 g m^-3 during the entire year. Table 1 shows that less than half of this soil mass was observed during the dust season. This implies that the soil at MLO during the spring is much finer than typical local soils, which is consistent with transport. Braaten and Cahill [1986] found a mass median diameter for Asian soils of ~1.2 m diameter, but did not measure particles above ~11 m. Thus, while it is likely that the Hi-Vol data from Zieman et al. [1995] contains some local soils, most of the soil is probably transported from distant sources during the dust episodes. The data for zinc and selenium implies that most of the mass of these species is <2.5 m in diameter, which is also consistent with transport from Asia.

TABLE 1. Comparison of Downslope Aerosol Sampling at MLO

	Mass	RCMA*	SO4	Organics	Soil	b (abs)	Zn	Se
	(g m-3)	(g m-3)	(g m-3)	(g m-3)	(g m-3)	10-4 m-1	(ng m-3)	(ng m-3)
Dust Season:
UC Davis	1.94	1.62	0.37	0.36	0.72	1.92	0.47	0.029
Zieman et al.,  1995			0.37		1.27		0.51	0.038
Overall:
UC Davis	0.84	0.68	0.29	0.22	0.10	0.62	0.18	0.021
Zieman et al.,  1995			0.26		0.55		0.33	0.031

Note: All UC Davis data is from 1993 and 1994.

*Mass is gravimetric, reconstructed mass (RCMA) is the sum of measured species. The difference is mainly water.

The second topic involves the intense dust storms that occurred during the spring of 1994. Figure 1 shows the concentration of fine soil, soot (b(abs), measured optically), organic matter (inferred from hydrogen), and arsenic during this time period. Although the soil concentration on April 9 is about twice the maximum in most prior years, it is the other elements that make this episode interesting. For example, the maximum arsenic concentrations at MLO during the spring of 1994 were roughly equivalent to the median of the peak arsenic concentrations at the other 70 or so IMPROVE sites at mainland parks and monuments. The high arsenic concentrations at MLO were accompanied by other trace elements similar to those produced from non-ferrous smelter activity (Table 2). Ten-day backward isenotropic air trajectories for this episode pass directly over northern China, site of most Chinese smelting activity.
 

Fig. 1. Aerosol concentrations at Mauna Loa Observatory, spring 1994. All concentrations are given in nanograms m^-3. For details of the parameters, see Malm et al. [1994].
 

TABLE 2. Comparison of Trace Elements From Potential Sources of Arsenic

	1000 Se/SO4	1000 Cu/SO4	1000 Zn/SO4	1000 As/SO4
Volcanoes:
Kilauea,	1.9	0.39	0.37	0.32
Nov. 21, 1992
Copper Smelters:
Average Arizona	1.7	20.0	28.0	6.0
"Smelter No. 5"	0.02	1.0	3.8	2.3
Mauna Loa
(downslope)
April 9, 1994	<0.08	0.9	3.8	2.4

Note: All values have been normalized to sulfate.

While we expected to find traces of Asian industrial activity in the aerosol record at MLO, we did not expect such high concentrations of arsenic. Prior research had shown that significant dilution of soil mass occurred between the Asian sources and MLO. Such high arsenic concentrations could only be achieved if either the concentrations were very high at mainland sites, or the trace elements did not suffer as much dilution as the soil dust. The former hypothesis seems unlikely. For example, if we use typical soil dilution factors to calculate the ambient levels of arsenic and other toxic species at the source regions in Asia, we find that the concentrations would need to be in the microgram m^-3 range over a wide area of the mainland during periods of strong ventilation and dilution associated with the dust storms. In non dust-storm periods, one would then expect arsenic levels unprecedented in western experience, levels that would likely be lethal in a relatively few years. As a result, the second explanation is probably the dominant factor. For example, using data from dust storms that occurred during the spring of 1981, we found that the dilution factor for very fine (D_p <0.5 m) particles was only a factor on the order of 3 (Table 3). If we use this small particle dilution factor to calculate the arsenic levels at Asian sources, the resulting arsenic concentrations are high, but not unprecedented. Thus, it is possible that other materials in the smallest size ranges (including many anthropogenic pollutants) could also be efficiently transported to MLO.

TABLE 3. Comparison of Fine Soil Dust Transport Efficiencies from China (April 19, 1981) to MLO (May 12, 1981)*

	Beijing	Mauna Loa	Ratio
Dp<0.5 m	0.6 g m-3	0.17 g m-3	31%
0.5<Dp<3.0 m	80 g m-3	0.22 g m-3	0.3%
3.0<Dp<16 m	136 g m-3	0.09 g m-3	0.07%

*Note that these are two different dust storms, so that the proportionality is only very approximate, but the ratios of coarse to fine dust should be reasonably similar.

The final topic involves experiments to be conducted during the spring of 1996 to test some of the hypotheses raised by the prior work. The experiment will take place during April and May of 1996 and include the following: (a) direct measurement of organic particles by carbon combustion from quartz filters and by GC/MS from quartz filters, (b) establishment of nitrate levels via direct collection of denuded nylon filters, (c) size/time resolved sampling of optically efficient aerosol (three sizes <2.5 m), and (d) ¹⁴carbon dating of organic aerosols.

These measurements will be complemented by the IMPROVE Channel A sampler and the full suite of MLO observables, including carbon soot via aethelometer and isentropic trajectories every 12 hours from CMDL.

Acknowledgments. This work would not have been possible without the capable support of the staff at MLO who have delivered samples biweekly with an efficiency that is among the very best in the entire 70-site network. We also wish to thank Joyce Harris for both her insights and her trajectories.

Braaten, D.A. and T.A. Cahill. Size and composition of Asian dust transported to Hawaii. Atmos. Environ., 20, 1105-1109, 1986.

Malm, W.C., J.F. Sisler, D. Huffman, R.A. Eldred, and T.A. Cahill, Spatial and seasonal trends in particle concentration and optical extinction in the United States, J. Geophys. Res., 99, 1347-1370, 1994.

Zieman, J.J., J.L. Holmes, D. Connor, C.R. Jensen, W.H. Zoller, D.M. Hermann, J.R. Parrington, and G.E. Gordon, Atmospheric aerosol trace element chemistry at Mauna Loa Observatory, 1, 1979-1985, J. Geophys. Res., 100, 25,979-25,994, 1995.

[back] [contents] [next]