*Also at Department of Hydrology and Meteorology, Nepal.

+Also at Department of Earth Sciences, University of Now Hampshire.

Abstract - Aerosol and surface snow samples were collected from Hidden Valley in the Dhaulagiri region of western Nepal during the summer monsoon of 1994. Temporal variations of major ion (Na⁺, NH₄⁺, K⁺, Mg²⁺,Ca²⁺, Cl^-, N0₃^- and SO₄^2- ) concentrations in the aerosol samples are clearly related to the influx of monsoon air masses. Snow was enriched in NH₄⁺, and NO₃^-, while Na⁺/Cl^- ratios were lower in the snow compared to the aerosol. A large part of this is explained by the difference in the air masses represented by aerosol and snow chemistry. Snow chemistry in general represented stronger southerly monsoon circulation, which resulted in precipitation events in Hidden Valley, whereas aerosol chemistry represented weaker monsoon or local circulation as the sampling was not conducted during rainy and foggy weather. Enrichment of NH₄⁺ and NO₃^- in snow is attributed to their biogenic and agricultural sources from villages to the south and east of Hidden Valley. In addition, scavenging of HN0₃ present in the air could also have contributed to the enrichment of N0₃^- in the snow. A lower Na⁺/Cl^- ratio in snow is attributed to scavenging of HCl present locally and/or due to less fractionation of monsoon air masses during more intense circulation and shorter travel time. The observed differences in the chemistry of the two media due to the influence of monsoon versus local air masses supports the concept of using glaciochemical records from that region to interpret monsoon variations in the past. Although the aerosol samples show excess cations, our data suggest the presence of acidic gases in the air locally. The overall major soluble ion concentrations of the aerosol are comparable or lower than those measured at several other remote tropospheric sites. Our results further support the concept that high elevation mountain sites in the Himalayas can be used to investigate the composition and the evolution of the remote continental troposphere. © 1997 Elsevier Science Ltd

Key word index: Aerosol, air snow relationship, Hidden Valley, Nepal, monsoon air mass, water soluble inorganic ions.

Introduction
The Site
Experimental
Results
Discussion
Conclusions
References
 

INTRODUCTION

Investigation of aerosol and precipitation chemistry in the Himalayas to date has been limited in both temporal and spatial scales, mainly due to difficult logistics. However, as the Himalayas are far removed from highly industrialized centers they provide suitable locations to monitor the chemistry of the remote troposphere and to study the evolution of atmospheric composition (e.g. Cunningham and Zoller, 1981). A limited number of studies conducted in the past, including short-term aerosol chemistry investigations on the southern slopes of Nepal Himalaya (Ikegami et al., 1978; Davidson, 1981; Wake et al., 1994), in north western India (Kapoor and Paul,1980), as well as surface snow studies (Mayewski et al., 1983; Wake, 1989), have indeed shown that the concentrations of pollution related Species Such as nitrate and sulfate are comparable to, or lower than, those measured in several other remote regions of the world.

Glaciochemical records have proven to be an excellent tool to investigate the atmospheric composition of the past (e.g. Mayewski et al., 1983, 1990; Wake et al., 1990). The Himalayan region contains several sub-tropical high elevation glaciers from which ice core records spanning decades to several centuries of atmospheric chemistry variations could be collected (e.g. Mayewski et al., 1984; Lyons et al., 1991; Wake and Mayewski, 1993; Wake et al., 1993). Due to its high elevation, the Himalayan range acts as a boundary limiting the northern extent of the Indian summer monsoon, and therefore ice core records from such sites could provide insight into the variations of the monsoon in the past. However, understanding of the chemical composition of the air, its temporal and spatial variability, and the relationships between the composition of snow and the air it forms in are essential for an improved interpretation of glaciochemical records. The main purpose of this study was to obtain an improved perception of aerosol composition in the high Himalayan region and investigate relationships between aerosol and snow chemistry. A secondary goal was investigating the influence of atmospheric circulation patterns on aerosol and snow chemistry. Hidden Valley in western Nepal was selected as the site for the study of aerosol and snow chemistry mainly because it is located to the north of main Himalayan range, almost at the edge of monsoon influence. Since only stronger monsoon circulation can overcome the orographic barrier and reach the site, relatively small variations in monsoon circulation could have significant impact on the chemistry of local atmosphere.

Aerosol samples were collected at 5050 m in Hidden Valley, during the summer monsoon season of 1994. These samples were analyzed for the water soluble major inorganic ions (Na⁺, NH₄⁺, K⁺, Mg²⁺, Ca ²⁺ ,Cl^-, N0₃^-, S0₄^2-). Similarly, fresh and surface snow samples collected from different locations in the valley were also analyzed for major ions. Although sampling only covered 13 d, this study is the most detailed of its kind in this remote region.

THE SITE

The site selected for this study is located at 28^o50' N latitude and 83^o35' E longitude in Hidden Valley, immediately north of main crest of the Himalaya in western Nepal (Fig. 1). The elevation of the valley floor is 5050 m a.s.l. Hidden Valley is surrounded by mountains ranging from 6000 m to over 8000 m a.s.l. including Dhaulagiri (8167 m). Hidden Valley possesses unique physiographical and climatological characteristics different from the rest of the country, mainly because the main crest of Himalayan range restricts the entry of the summer monsoon into these areas. The main pathways for monsoon moisture to Hidden Valley are Dhampus pass in the southeast and French pass in the south of the valley (Shrestha et al.,1976). The valley floor is covered by light vegetation, whereas the moderately sloped mountain sides are not vegetated and are severely weathered.

EXPERIMENTAL

Aerosol samples

Aerosol samples were collected over a 13 d period (from 14 to the 27 of August 1994) on 2 µm pore size, 90 min diameter Zefluor^TM Teflon filters (Gelman Sciences) mounted in a protective polyethylene cover.

The air was drawn by a 24 V high-volume pump powered by a combination of photovoltaic cells and batteries. The volume of air sampled was measured by an in-line flow meter. Corrections for ambient temperature and pressure allowed conversion of the measured volumes to standard cubic meters (scm). The mean flow rate was 4.35 scm h^-1 and the mean volume of samples was 41.2 scm. Zefluor^TM filters have been found to have minimum positive artifact with regard to sorption of gaseous species like HN0₃ and S0₂ (Spicer and Schumacher, 1979; Appel et al., 1984). In contrast there might have been some evaporation of N0₃^- due to extended sampling, therefore the N0₃^- values reported here should be regarded as lower limit of true airborne concentration. For the mean flow rate of 4.35 scm h^-1 the velocity at the face of the filter was sufficiently high such that the efficiency of collection for particles as small as 0.3 µm is greater than 99% (Liu et al., 1984). Aerosol samples were collected in filters facing downwards, placed in a cylindrical protective cover. The flow velocity at the opening of the protective cover was calculated to be 0.04 m s^-1, according to which the cut-off for large particle, as given by sedimentation velocity of particles is estimated at about 10 µm (Davidson, 1987; Warneck, 1988). Filters were changed twice daily representing approximately daytime and nighttime samples, but samples were not collected during rainy/foggy periods.

Care was taken to minimize contamination both in the laboratory and in the field. The filter cartridges were loaded in a Class 100 clean lab, packed and transported in clean plastic bags. After sampling the filters, still in their cartridges, were returned to the original clean plastic bags. Loading and unloading of samples was carried out wearing a nonparticulating clean suit, hood, face mask and plastic gloves. Blank filters were handled in the same manner as the samples.

In order to analyze the major ion concentration the sample and blank filters were wetted with 0.5 ml ultra pure methanol. The soluble components were then extracted with three 5 ml aliquots of deionized Milli-Q water. Major ion concentrations in the aqueous extracts were determined by ion chromatography using Dionex model 4000 ion chromatograph. Cations were analyzed using a CS12 column, 22 mM MSA eluent and CSRS suppresser and anions were analyzed using an AS11 column, 6.6 mM NaOH eluent and MMS suppresser. The sample loop was 250 µl and the sample running time was approximately 9 min.

Five field blanks and two lab blanks were collected. We define the detection limit for major inorganic ions as the standard deviation of all the blanks divided by the mean volume of all the samples (after Talbot et al., 1986). The detection limits for aerosol species were (in neq scm Na⁺ (0.143), NH₄⁺ (0.028), K⁺ (0.033), Mg²⁺ (0.003), Ca²⁺ (0.018), Cl^- (0.048), N0₃^- (0.025), SO₄^2- Blank values were variable, especially for Na⁺, K⁺ and Cl^-. Mean blank values were subtracted from the ample concentrations, which resulted in below detection limits (bd) values for Na⁺, K⁺ and Cl^- in some of the samples.

The uncertainty in aerosol major inorganic ion concentrations associated with variability in the blank values and recision of ion chromatograph analysis were determined based on the propagation of errors (Miller and Miller, 1988). The overall mean uncertainty for samples were (in neq scm^-1): Na⁺ (0.153), NH₄⁺ (0.048), K⁺ (0.034), Mg²⁺ (0.004), Ca²⁺ (0.038), Cl^- (0.050), N0₃^- (0.065), IS0₄^2- (0.080).

Snow samples

Surface snow samples were collected in Hidden Valley during the summer monsoon of 1994. Since post-depositional processes might alter the composition of snow chemistry, only fresh surface snow samples (28 samples) collected in different parts of the valley during the period of aerosol sampling are included in this study. All snow samples were collected within 12 h of deposition. Table 1 and Fig. 1 provide information on the site, date, and number of snow samples collected. Samples were collected using pre-cleaned plastic scrapers and were placed in sealed plastic "Whirl-Pak" bags for melting. Snow samples were allowed to melt at the base camp and were poured into laboratory cleaned vials. Care was taken during sample collection and transfers to avoid outside contamination. Samples were kept frozen in the field using a gas powered freezer, but thawed during transport and were refrozen only when they arrived in New Hampshire. The samples were analyzed for major inorganic ions using Dionex model 2010i ion chromatograph.

Blanks were prepared by filling Whirl-Pak bags from the same lot as those used for samples, which had been transported to the site and back, with 50 ml of Milli-Q water. They were sealed and agitated for about 15 s. The bags were then cut open by a razor and Milli-Q water was transferred to laboratory cleaned 40 ml vials.

The blank values were variable for Na⁺ and NH₄⁺, but were relatively constant for other major inorganic ions. Detection limits of snow samples were taken as one standard deviation of all blanks. The detection limits of species in snow were in (µeq kg^-1): Na⁺ (0.139), NH₄⁺ (0.198), K⁺ (0.054), Mg²⁺ (0.017), Ca²⁺ (0.040), Cl^- (0.062), N0₃^- (0.006) S0₄^2- (0.021).

Meteorological data

A temporary meteorological station was established in the valley at an elevation of 5050 m a.s.l. Temperature and relative humidity were measured continuously by a thermohydrograph, which had been set up in a Stevenson screen. Precipitation was measured by a Hellmann type rain gauge daily at 8:00 AM. Barometric pressure was measured every 4 h using an altimeterbarometer.

RESULTS

Water soluble inorganic ions in the aerosol

Concentrations of water soluble inorganic ions in the aerosol (Na⁺, NH₄⁺, K⁺, Mg²⁺, Ca²⁺, Cl^-, NO₃^-  and S0₄^2- ) in neq scm^-1 are presented in Table 2. The average total aerosol load (total anion + total cation)was 7.36 neq scm^-1.  S0₄^2-, NH₄⁺, and Ca²⁺ are the dominant ions with concentrations above 1 neq cm^-1 and account on average for more than 80% of the total ion burden. The sum of cations generally exceeded the sum of anions in our samples. The ion budget was in the range of - 0.32 to + 1.54 with a mean value of + 0.75 (negative sign indicates excess anion and positive sign indicates excess cation). Note that H⁺, C0₃^2- and HCO₃^- were not measured in our samples.

So far only one study conducted in the Nepal Himalaya reported the same suite of chemical species as this study. Figure 2 compares the average concentrations of major inorganic ions in the aerosol in our samples with samples from Ngozumpa Glacier, Khumbu Himal, eastern Nepal (Wake et al., 1994). The concentrations of aerosol related to secondary aerosol such as S0₄^2-  in Hidden Valley are comparable to those in Ngozumpa, whereas NO₃^- and NH₄⁺ are even lower in Hidden Valley. Wake et al. (1994) pointed out that NO₃^- and S0₄^2- concentrations in Ngozumpa samples were comparable to concentrations measured in the remote free troposphere. On the other hand, Na⁺, Cl^- and Ca²⁺ are higher in Hidden Valley.

All the water soluble inorganic ions, in general, show similar temporal variation. Nevertheless there are certain differences, which facilitates dividing them into three groups: (1) Na⁺ and Cl^-, (2) Ca ²⁺, Mg²⁺ and NO₃^-, and (3) S0₄^2- and NH₄⁺, presented in Fig. 3a, b, and c, respectively. For comparison, precipitation and pressure anomalies are presented in Fig. 3d. The concentrations of all the ions are low in the beginning of the sampling period. Due to large detection limits inter-sample variability in the first group (Na⁺ and Cl^-) cannot be resolved with much certainty. Nevertheless, general trends in concentration over the sampling period is obvious. For example, peaks on 20 and 22 August are obvious in Fig. 3a. The temporal variations in the second group (Ca ²⁺, Mg²⁺ and NO₃^-) show three peaks. An abrupt increase occurred on 16/17 August, after which the concentrations gradually decreased, but peaked again on 20 August. The third peak occurred around 23 August. Species in the third group (S0₄^2- and NH₄⁺) display temporal variations similar to those for species in the second group, although they lack the peak on 20 August. In this group the 23 August peak was greatly subordinate to the one on 16-17 August (Fig. 3c). The temporal distribution of K⁺ is somewhat different from others and follows partly Na⁺ and Cl^- (after 20 August), and partly the second group (before 20 August). Concentrations of all the ions were low at the end of the sampling period (after 25 August).

Linear regression on all water soluble species in the aerosol showed that Na⁺ and Cl^- are highly correlated (r = 0.91). Similarly, Ca ²⁺ correlates well with Mg²⁺ (r = 0.80), while S0₄^2- is highly correlated with NH₄⁺ (r = 0.99). In addition, Ca ²⁺ correlates well with N0₃^- (r = 0.90) and S0₄^2- (r = 0.74). All of these correlation coefficients are significant at p = 0.01. Correlation coefficients between all species are shown in Table 3.

During the first half of the sampling period, the temporal variations in the aerosol ion concentrations show a relationship with meteorological conditions, especially precipitation. The concentration levels were low during the precipitation of 14-17 August, and sharply increased as the intensity of precipitation diminished. However, the scenario is different during the second precipitation event of 23-27 August as concentrations of the aerosol species, especially Ca ²⁺ remained moderately high for the first two days of precipitation. By the night of 24, all species showed low concentrations.

Despite the high correlation between Na⁺ and Cl^- there is a deficiency in Cl^- in the aerosol compared to sea-water Na⁺/Cl^- ratio (Table 2). Cl^- -deficit shows an inverse relationship with S0₄^2- in our samples (with the exception of the 22 August sample). This relationship improves when the amount of S0₄^2- necessary to fully neutralize NH₄⁺ is deducted from the total S0₄^2- (Fig. 4). The correlation coefficient between Cl^- -deficit and S0₄^2- minus NH₄⁺ is - 0.56 (excluding 22 August sample), which is statistically significant (p = 0.05).

The 22 August sample shows the highest Cl^- -deficit ( - 0.71, Fig. 4), but also the highest Na⁺ and second highest Clconcentrations of all the samples (Table 2). The high value of Na⁺ in this sample, responsible for the high Cl^- deficit (negative excess), is difficult to explain, as the Na⁺ was not likely of crustal origin since concentrations of all other crustal species were low in this sample.

Another interesting result is the striking inverse relationship between Ca ²⁺/S0₄^2- ratio and NH₄⁺/S0₄^2- ratio in our samples (Fig. 5). The correlation between these ratios is 0.75 (p=0.01). For comparison (Ca ²⁺ + NH₄⁺)/S0₄^2- ratio is also plotted in Fig. 5. Seven samples show the ratio greater than one. The same number of samples have the ratio less than one, while three samples show the ratio close to unity.
 

It has been suggested that acidic aerosols can be contaminated by ammonia in ambient air while processing filter in laboratory (Hayes et al., 1980; Silvente and Legrand, 1993). The ammonium concentrations can be suspect if the high correlation between S0₄^2- and NH₄⁺ is due to laboratory contamination of our samples. It is unlikely that such reaction took place in our samples, as S0₄^2- is not fully neutralized in any of our samples, since NH₄⁺ to S0₄^2- equivalence ratios are less than unity (Table 3). Furthermore, the study of Hayes et al. (1980) included highly acidic stratospheric aerosol samples, whereas our samples are generally basic or only slightly acidic.
 

Snow chemistry

Samples of surface snow chemistry collected during the aerosol sampling program were compared with aerosol chemistry. Although the number of snow samples collected within the period of aerosol sampling is limited, they facilitate comparison of overall trends in these two media. Snow samples were not collected on 23 and 24 August, the last two days of the four-day precipitation event.

It has to be noted that snow samples included in this study were collected from several different locations in the valley. It is therefore possible that the temporal variations observed in the snow chemistry could be largely due to variations between the sampling locations.

Logistical considerations resulted in the collection of fresh snow at more than one location only for the 23 August event. For this one case samples were collected from one site in Rika Samba glacier and at North Glacier (Table 1). The Student's t-test shows that for that event the mean concentrations of all soluble ions were statistically not different (p = 0.05) at the two sites. Assuming this result is valid for other events as well, we suggest that the variability in snow chemistry is dominated by temporal variability rather than variability between sites.

Crustal species, S0₄^2- and to some extent sea salt species, in the snow show temporal variations in general similar to those in the aerosol (Fig. 6). Other species on the other hand do not show such similarity. The most remarkable difference is observed in the temporal variations of NH₄⁺ and N0₃^- in the aerosol and in the snow (Fig. 6). In addition, unlike in the aerosol, NH₄⁺ in the snow is not correlated with S0₄^2-. The NH₄⁺/S0₄^2- ratio in the aerosol is never higher than 1, whereas this ratio in snow in all with the exception of one event, is always higher than 1 and even exceeds 25 on 21 August (Fig. 7a). On the other hand the ratio NH₄⁺/(S0₄^2- + N0₃^- ) in snow is much lower, ranging from 0.5 to 1.5.

Similarly, temporal patterns of ratios of Na⁺ and Cl^- also display conspicuous contrasts in the two media. The Na⁺/Cl^- ratio is much lower in snow compared to aerosol (Fig. 7b). The Na⁺/Cl^- ratio in the aerosol ranges from 1.25 to 7.24 (i.e. Cl^- deficit) whereas, the ratio in the snow ranges from 0.41 to 1.16 indicating Cl^- excess in most (5 out of 6) samples.

DISCUSSION

Temporal variations in the aerosol and its relationship with meteorological condition

Wet deposition is an effective mechanism for the removal of aerosols, especially in the size range of 0.1 to 10 µm (Barrie, 1985; Warneck, 1988; McGann and Jennings, 1991). It is therefore expected that low concentration of atmospheric constituents will occur during and following precipitation events. This simple relation is not always observed in our study. During the 14-16 August precipitation event, a high amount of precipitation fell on the first day but diminished rapidly in the following days (Fig. 3d). In contrast, during the 23-27 August event precipitation was maintained at a moderate rate for 4 d. Study of upper air synoptic maps at 500 mb level shows that the precipitation of 14-16 and 23-27 August had a similar cause: formation of high pressure cell over Bay of Bengal. The difference between these two events was that the high pressure cell in the first event was of higher magnitude (592 mb) but lasted for only a short period, whereas in the second event the magnitude was not as high (~588 mb) although it lasted for a longer period.

The patterns shown by synoptic maps are consistent with the results of the atmospheric pressure measurements made in Hidden Valley (Fig. 3d). The first precipitation event coincides with sharp but short drop in pressure, whereas the pressure pattern during the second event shows more gradual but long lasting decrease. Consequently, precipitation in the first event was of high intensity in the first day, which efficiently scavenged the atmospheric constituents. The circulation in the following days (15-17 August) was not as strong, the precipitation was also low, which possibly was not efficient to scavenge atmospheric constituents. Hence the levels of most species increased considerably during those days. The circulation in the second event was of moderate strength resulting in moderate intensity precipitation, which likely was not as effective in removing atmospheric constituents. In addition, the continued moderate circulation brought in atmospheric constituents in the following days. The dry period between event one and two was relatively free from monsoon circulation and more dominated by local circulation, therefore concentrations of sea salt species were low and concentrations of crustal species were high. K⁺ is better correlated with Mg²⁺ and Ca ²⁺ during this period. Conversely during periods of stronger monsoonal influence such as after 20 August, K⁺ correlates better with sea salt species. Although airborne concentrations depend on many processes, including upwind chemical processes, removal by wet and dry deposition, and variability in source regions, our results suggest an important role due to the influx on monsoonal air masses and precipitation intensity.

Relationship between snow and aerosol chemistry

It is common to observe enrichment of S0₄^2- relative to NH₄⁺ in the snow compared to the aerosol due to S0₂ oxidation in clouds (e.g. Calvert et al., 1985).On the contrary, our samples display huge enrichment of NH₄⁺ relative to S0₄^2- in the snow.  NH₄⁺/S0₄^2- ratio shows that only a small fraction of NH₄⁺ in snow is neutralized by S0₄^2- (Fig. 7a). Unlike the aerosol, a large fraction of NH₄⁺ is balanced by N0₃^- in snow. Fractionation of NH₃ cannot explain such a large difference as the NH₄⁺/S0₄^2- ratio in the air is always less than 1 (Table 2). The only plausible explanation for such a contrast between aerosol and snow chemistry is that different air masses are represented by the aerosol and snow samples. The aerosol sampling was not conducted during intensive precipitation and fog, therefore the filter samples were strongly biased towards air masses representing weaker monsoon circulation or local circulation. On the other hand, fresh snow chemistry explicitly reflects monsoon air mass with more southerly origin. Monsoon air masses, before entering Hidden Valley, travel over low lying valleys and mountain slopes dominated by cultivated land, forests, and vegetation. There are also several small villages, where animal husbandry and fertilization of fields with manure is a common practice, and burning of firewood is the main source of domestic energy. The high proportion of NH₄⁺ in snow is therefore suggested to be due to enrichment of NH₃ while the air mass traveled over these areas. As there are no industrial activities in the region, S0₄^2- enrichment is low compared to NH₃. As a result NH₃ in these air masses was not fully neutralized by S0₄^2-, and NH₃ was therefore available to react with HN0₃ resulting in a higher proportion of N0₃^- in snow. However, some of the samples show NH₄⁺/(S0₄^2- + N0₃^-) ratio less than 1, suggesting additional scavenging of HN0₃  in the air by snow.

Much lower Na⁺/Cl^- ratio in the snow compared to the aerosol (Fig. 7b) could be due to two factors: scavenging of gaseous HCl present in the air, or relatively low fractionation of sea salt particle in the air mass represented by the snow. Since the snow chemistry represents stronger monsoon circulation with shorter travel time, it is possible that the fractionation of Cl^- with respect to Na⁺ is less in such air masses compared to air masses representing weaker circulation. However, most of the snow events show excess Cl^-, which cannot be achieved by this process alone. Therefore scavenging of HCI is possibly the major cause of the observed excess Cl^- in snow.

The difference in NH₄⁺ and the ratio of sea salt species in air and snow indicates that the sensitivity towards monsoon strength is well characterized by snow chemistry. Only stronger monsoonal circulation can overcome the large orographic barrier and reach the site, while weaker circulation cannot, so local circulation dominate during such periods. Such sensitivity towards monsoon strength suggest that the ice core chemistry can provide a strong record of monsoon variations in the past.

Water soluble inorganic ions in the aerosol and their sources

The concentration distributions during the measurement period were distinct for sea salt aerosols (Na⁺ and Cl^-), continental aerosols (Ca ²⁺ and Mg²⁺) and secondary aerosols (S0₄^2- and NH₄⁺). The high correlation between Na⁺ and Cl^- suggests their common origin. We know of no evaporite deposits in the region and there is generally a poor correlation with Ca ²⁺ and Mg²⁺. Both of these factors argue against a crustal source of Na⁺ and Cl^- and, despite the distance from the ocean, we believe Na⁺ and Cl^- were likely of marine origin. The deficiency in Cl^-( ~ 30%) in the aerosol compared to Na⁺/Cl^- ratio in sea water, and the inverse relationship between excess Cl^- and S0₄^2- (Fig. 4), suggests that Cl^- was depleted due to reaction between sea salt particle and H₂SO₄ in the air (Hitchcock, 1980; Ohta and Okita, 1990). This result further suggests that the enrichment of Cl^- in the snow was partially due to scavenging of HCl in the air.

The fairly high correlation between crustal species and S0₄^2- and N0₃^- and the inverse relationship between Ca ²⁺/S0₄^2- and NH₄⁺/S0₄^2- ratios (Fig. 5) can also be explained by the affinity of H₂SO₄ and HN0₃ to be absorbed on the surface of mineral particles and reacting to form salt (Wolff, 1984; Mamane and Gottlieb, 1989,1992; Wu and Okada, 1994). While gaseous H₂SO₄ has a very short lifetime against condensation onto existing aerosol or formation of new aerosol, S0₂ can enter aqueous aerosols or films and oxidize to S0₃^- and then to H₂SO₄. The samples with (Ca ²⁺ + NH₄⁺)/S0₄^2- ratio less than one suggest the existence of H₂SO₄ aerosols and/or S0₂ in the air masses sampled. The H₂SO₄ and/or S0₂ in the ambient air not neutralized by NH₃ would be available to react with the surface of crustal particles, possibly CaC03 releasing C0₂. The covariance between crustal species and N0₃^- in our samples is also believed to be due to reaction between N0₂ and HN0₃ and crustal particles after all the S0₄^2- was neutralized.

The excellent correlation between NH₄⁺ and S0₄^2- suggests that these species are present in the same particles. The mean value of NH₄⁺/S0₄^2- equivalence ratio is 0.69 (Table 2), nearly in the middle of the range 0.5 and I corresponding to NH₄HSO₄ and (NH₄)₂SO₄ respectively. Sulfate, on the other hand, was also correlated with crustal species and Na⁺, indicating its association with these species. We suggest that the NH₄⁺ existed mainly as NH₄HSO₄, while S0₄^2- existed also in the form of Na₂SO₄ and CaS0₄.

Earlier studies suggested that S0₄^2- in the Himalayan region did not represent anthropogenic emissions from more southerly location but reflected the influx of crustal species in the form of CaS0₄ (Mayewski et al., 1983). This study was conducted in the western Himalaya, where the influence of dust from Thar desert in northwest India is significantly greater compared to the eastern Himalaya (Wake et al., 1993). In our samples from Hidden Valley evidence of reactions between atmospheric acids and sea salt, as well as locally derived crustal particles, suggests the presence of acids (although quantitatively low) such as H₂SO₄, HN0₃, and HCI in the air despite excess cations observed in the aerosol in this part of Himalaya. In addition, the widespread biomass burning (firewood) for domestic energy in Nepal cannot be ignored as a source of N0₃^- and S0₄^2- (Davidson et al., 1986).

The comparable or lower concentrations of chemical species in the aerosol in Hidden Valley with respect to Ngozumpa glacier (Fig. 2) further supports the possibility of using high Himalayan sites to monitor the global remote troposphere. Relatively higher amount of sea salt species in Hidden Valley is likely due to marine influence on air mass during the monsoon season, although the air has traveled a considerable distance from marine environments to reach the site, while the samples from Ngozumpa glaciers were collected during the post-monsoon season. Higher concentrations of crustal species suggests local generation of dust. There are more local sources of dust in Hidden Valley, which has more weathered, moderately sloped terrain compared to steep unweathered rock faces above treeline or vegetation covered slopes below treeline in Ngozumpa. In addition Hidden Valley receives much less precipitation than Ngozumpa glacier.

CONCLUSIONS

Samples collected from Hidden Valley during the monsoon of 1994 show that short-term variation in monsoonal circulation was reflected in aerosol chemistry such that temporal variations in the concentrations of aerosol species were related to the pattern and intensity of atmospheric circulations. Differences between the temporal variations in chemical species in the air and in the snow is likely due to different air masses represented by these media. A higher proportion of NH₄⁺ in snow indicates that snow chemistry reflects stronger monsoon circulation, advecting air masses rich in NH₃ of agricultural and biogenic origin in low lying areas in the vicinity of Hidden Valley. Aerosol sampling biased against periods of precipitation and fog conditions, and therefore, reflect weaker circulation and the dominance of local air masses. This finding suggests that isolated high Himalayan sites are sensitive towards monsoon strength and this sensitivity is translated to aerosol and snow chemistry. Glaciochemical records from these sites should therefore provide a valuable record of monsoon variations of the past.

Our study shows the presence of acidic species in air such as H₂SO₄, HN0₃ and HCl. The result of this study provides the most detailed data so far on soluble species in atmospheric aerosol in the high Himalaya during the summer monsoon season. This data series is longer, and therefore adds confidence to the results of Wake et al. (1994), reinforcing the possibility of using high elevation sites in the Himalaya as a platform to investigate the chemistry of the remote troposphere.

The variations in the composition of snow and aerosols, besides changes in atmospheric circulation, depend on many processes within and before the sampling region. Therefore the concept of chemical variation due to changes in atmospheric circulation should be put to a more rigorous test by extending the sampling to a longer period covering different seasons and conducting synchronous sampling of air and snow media. Parallel sampling at stations in progressively lower elevation extending to the southern, lower elevation regions of the country will enhance the knowledge of variation in chemical composition of air masses as they move northwards.

Acknowledgements - The authors thank David S. Averill for collecting the aerosol samples and providing the snow chemistry data, and Sallie Whitlow for analyzing the aerosol samples. Thanks are also due to Department of Hydrology and Meteorology, Nepal, for providing meteorological instruments and synoptic maps. We also wish to thank Mr Sher Bahadur for collecting the meteorological data and Mr Dawa Nurbu for helping logistic support.
 

REFERENCES

Appel, B. R., Tokiwa, Y., Haik, M. and Kathny, E. L. (1984) Artifact particulate sulfate and nitrate formation on filter media. Atmospheric Environment 18, 409-416.

Barrie, L. A. (1985) Atmospheric particles: their physical and chemical characteristics, and deposition processes relevant to the chemical composition of glaciers. Ann. Glaciol. 7, 100-108.

Calvert, J. K., Lazrus, A., Kok, G. L., Heikes, B. G., Walega, J. H., Lind, J. and Cantrell, C. A. (1985) Chemical mechanisms of acid generation in the troposphere. Nature 317, 27-35.

Cunningham, W. C. and Zoller, W. H. (1981) The chemical composition of remote area aerosols. J. Aerosol Sci. 12, 367-384.

Davidson, C. 1. (1981) Airborne lead and other elements derived from local fires in the Himalayas. Science 214, 1344-1346.

Davidson, C. 1. (1987) Mechanisms of Wet and Dry Deposition of Atmospheric Contaminants to Snow Surfaces. Dahlem Workshop on the Environmental Record in Glaciers.

Davidson, C. I., Lin, S., Osborn, J. F., Pandey, M. R., Rasmussen, R. A. and Khalil, M. A. K. (1986) Indoor and Outdoor Air Pollution in The Himalayas. Envir. Sci. Technol. 20, 561-567.

Hayes, D., Snetsinger, K., Ferry, G., Oberbeck, V. and Farlow, N. (1980) Reactivity of stratospheric aerosols to small amounts of ammonia in the laboratory environment. Geophys. Res. Lett. 7, 974-976.

Hitchcock, D. R. (1980) Sulphuric acid aerosols and HCI release in coastal atmospheres: evidence of rapid formation of sulphuric acid particulates. Atmospheric Environment 14, 165-182.

Ikegami, K., Inoue, J., Higuchi, K. and Ono, A. (1978) Atmospheric aerosol particles observed in high altitude Himalayas. Seppyo 40(special issue), 50-55.

Kapoor, R. K. and Paul, S. K. (1980) A study of the chemical components of aerosols and snow in the Kashmir region. Tellus 32, 33-40.

Liu, B. Y. H., Pui, D. Y. H. and Rubow, K. L. (1984) Characteristics of air sampling filter media. In Aerosols in the Mining and Industrial Work Environments, Vol. 3, Instrumentation, (edited by Marple V. A. and Liu B. Y. H.). Ann Arbor Science, Michigan.

Lyons, W. B., Wake, C. and Mayewski, P. A. (1991) Chemistry of snow from high altitude, mid/low latitude glaciers. In Seasonal Snowpacks, Vol. G 28 (edited by Davies T. D. et al.), pp. 359-383. Springer, Berlin.

Mamare, Y. and Gottlieb, J. (1989) Heterogeneous reactions of mineral with sulphur and nitrogen oxides. J. Aerosol Sci. 20, 303-311.

Mamare, Y. and Gottlieb, J. (1992) Nitrate formation on sea-salt and mineral particles - a single particle approach. Atmospheric Environment 26A, 1763-1769.

Mayewski, P. A., Lyons, W. B. and Ahmad, N. (1983) Chemical composition of a high altitude fresh snowfall in the Ladakh Himalayas. Geophys. Res. Lett. 10, 105-108.

Mayewski, P. A., Lyons, W. B., Ahmad, N., Smith, G. and Pourchet, M. (1984) Interpretation of the chemical and physical time-series retrieved from Sentik Glacier, Ladakh Himalaya, India. J. Glaciol. 30, 66-76.

Mayewski, P. A., Spencer, M. J., Twickler, M. S. and Whitlow, S. (1990) A glaciochemical survey of the summit region, Greenland. Ann. Glaciol. 14, 186-190.

McGann, B. T. and Jennings, S. G. (1991) The efficiency with which drizzle and precipitation sized drops collide with aerosol particles. Atmospheric Environment 25A, 791-799.

Miller, J. C. and Miller, J. N. (1988) Statistics for Analytical Chemistry. Eltis Horwood, New York.

Nakawo, M., Fujii, Y. and Shrestha, M. L. (1976) Flow of glaciers in Hidden Valley, Mukut Himal. Seppyo 38 (Special Issue), 39-43.

Ohta, S. and Okita, T. (1990) A chemical characterization of atmospheric aerosol in Sapporo. Atmospheric Environment 24A, 815-822.

Shrestha, M. L., Fujii, Y. and Nakawo, M. (1976) Climate of Hidden Valley, Mukut Himal during the Monsoon 1974. Seppyo 38 (special issue), 105-108.

Silvente, E. and Legrand, M. (1993) Ammonium to sulphate ratio in aerosol and snow of Greenland and Antarctic regions. Geophys. Res. Lett. 20, 687-690.

Spicer, C. W. and Schumacher, P. M. (1979) Particulate nitrate: laboratory and field studies of major sampling interferences. Atmospheric Environment 13, 543-552.

Talbot, R. W., Harris, R. C., Browell, E. V., Gregory, G. L., Sebacher, D. I. and Beck, S. M. (1986) Distribution and geochemistry of aerosols in the tropical North Atlantic troposphere: relationship to Saharan dust. J. geophys. Res. 91, 5173-5182.

Wake, C. P. (1989) Glaciochemical investigations as a tool for determining the spatial and seasonal variation of snow accumulation in the Central Karakoram, Northern Pakistan. Ann. Glaciol. 13, 279-284.

Wake, C. and Mayewski, P. A. (1993) The spatial variation of Asian dust and marine aerosol contributions to glaciochemical signals in central Asia. In Proc. Int. Symp. on Snow and Glacier Hydrology, 16-21 November, 1992, Kathmandu, Nepal, IASH, 218.

Wake, C. P., Mayewski, P. A., Zichu, X., Ping, W. and Zhongqin, L. (1993) Regional distribution of monsoon and desert dust signals recorded in Asian glaciers. Geophys. Res. Lett. 20, 1411-1414.

Wake, C. P., Dibb, J. E., Mayewski, P. A., Zhongqin, L. and Zichu, X. (1994) The chemical composition of aerosols over the Eastern Himalayas and Tibetan Plateau during low dust periods. Atmospheric Environment 28, 695-704.

Warneck, (1988) Chemistry of Natural Atmosphere. Academic Press, San Diego.

Wolf, G. T. (1984) On the nature of nitrate in coarse continental aerosols. Atmospheric Environment 18, 977-981.

Wu, P. and Okada, K. (1994) Nature of coarse nitrate particles in the atmosphere - a single particle approach. Atmospheric Environment 28, 2053-2060.