Geophysical Research Letters

South Asian dust episode in June 2006: Aerosol observations in the central Himalayas



[1] A dust storm blew through the Thar Desert on 12 June 2006, which has significantly influenced aerosol physical and optical properties over the central Himalayas (Nainital, 29.4°N; 79.5°E, 1958 m amsl) on 13 June 2006. Aerosol number concentrations in the coarse and giant modes on 13 June 2006 are found to be five (26 × 106 m−3) and ten (17.2 × 103 m−3) times higher compared to their respective monthly mean values. Aerosol optical depth values also showed two to four times increase, particularly at longer wavelengths suggesting increase in the concentrations of coarse and giant particles. This is supported by three to five times increase in Ångström turbidity coefficient (β) and significant reduction in Ångström wavelength exponent (α). Absence of enhancements in black carbon and accumulation mode particles suggests negligible changes in the influences of anthropogenic activities at the site during the study period.

1. Introduction

[2] Dust storms have been recognized as having a wide range of impacts on climate and environment [e.g., Liu et al., 2004]. Desert dust constitutes a large fraction (∼50%) of the naturally occurring tropospheric aerosols [Gobbi et al., 2000]. These particles are responsible for an additional cooling effect in the lower troposphere, while in the middle and upper troposphere these particles lead to an additional warming [e.g., Gobbi et al., 2000, and references therein]. Dust particles, which are generally of size greater than 1μm, mainly consist of quartz, mica, and clay minerals and are produced by the natural weathering that breaks the rocks [Duce et al., 1991]. Desertification is another potential source of dust production. However, natural weathering has greater influence on Asian dust emissions and associated dust storm occurrences when compared to the desertification processes [Zhang et al., 2003]. Further, dust from East Asia is typically of mineral type [e.g., Park et al., 2005], while limited studies suggest it of mix type over South Asia [Satheesh and Moorthy, 2005].

[3] Globally, there have been different studies on dust storms using satellite and ground based observations [e.g., Prospero, 1999; Washington et al., 2003]. Such studies in Asian region are largely focused on dust storms originating from East Asia during spring, mainly in April, [e.g., Husar et al., 2001]. However, studies on dust storms are sparse over South Asia, where dust storms are observed in May and June with frequency of 8 to 10 per month [e.g., Dey et al., 2004]. The Great Indian Desert (Thar Desert), along the eastern Pakistan and western India, is one of the major sources of atmospheric dust for this region. Despite of being large sources of dust storms in South Asia, studies on long-range transport of dust from this region are very few and are limited to Indo-Gangetic basin [e.g., Dey et al., 2004; Deepshikha et al., 2005; Moorthy et al., 2007]. Better understanding of dust storms over South Asia and quantification of their influences on radiation budget and environmental changes over this part of world is very essential, particularly for studies on precipitation [Intergovernmental Panel on Climate Change, 2007]. Natural aerosols forcing, mainly induced by dusts, is shown to be about 1.5 times higher when compare to that of anthropogenic aerosols forcing over South Asia [Satheesh and Srinivasan, 2002]. Here we report large changes in aerosols physical and optical properties at a high altitude site in the central Himalayan region during a dust storm, which blew through the Thar Desert in Pakistan and India on 12 June 2006. Though, such dust storms will have important influence on radiation budget, its estimate is not attempted in this paper.

2. Observation Site and Techniques

[4] Observation site, Nainital (29.4°N; 79.5°E, 1958 m amsl; Figure 1), is located in the Central Himalayas. Sharply peaking mountains of Himalayan ranges are due North and Northeast of the site, while very low elevation (<500 m amsl) regions are due South. Observation site is at a mountain top and is about 1 km away from Nainital city, which has population of about 0.5 million. There is no industry in Nainital and some small-scale industries are located in nearby towns (Haldwani and Rudrapur about 30 km away) at lower altitude (<500 m amsl).

Figure 1.

Six days air back-trajectories (at 2000 m) (a) during dust episode periods (black, 12 June 16 GMT; red, 13 June 4 GMT; green, 13 June 8 GMT; blue, 13 June 16 GMT) and (b) after dust episodes (black, 14 June 4 GMT; red, 14 June 8 GMT; green, 15 June 8 GMT).

[5] An optical particle counter (OPC, model 1.018 of Grimm Aerosol Technik, GmbH, Germany) is used for observations of number concentration of composite aerosols in 15 size ranges (from > 0.15 μm to > 10 μm radius) and an Aethalometer (model AE-42, Magee Scientific, USA) is used for observations of black carbon mass concentration. Five minutes averaged data are collected from both these systems. A high volume air sampler (Envirotech Inc., model APM 430) is employed to measure mass loading of composite aerosols at fortnightly intervals. Detailed descriptions of all these instruments and methodology of data archival have been given by Pant et al. [2006] (also see auxiliary material). Columnar aerosol optical depths (AODs) and columnar water vapour are measured using Microtops II Sunphotometer (Solar Light Co., USA) at the central wavelengths of 380, 440, 500, 675, 870 and 1020 nm with a full-width-of-half-minimum of 6-10 nm during unobscured solar sky conditions. AOD observations are made six to eight times in a day from about 0900 hour LT at hourly intervals. The performance and methodology of such Microtops II Sunphotometers have been described by Ichoku et al. [2002]. Standard meteorological sensors (Dynalab, India; Campbell Scientific Inc., Canada) are used for observations of meteorological parameters.

3. Results and Discussion

[6] A dust storm blew through the Thar Desert on 12 June, 2006 and reached to the mountain region in northeast India later on (also see This has been observed by Moderate Resolution Imaging Spectroradiometer (MODIS) and Ozone Monitoring Instrument (OMI), which shows very high aerosols index value (3.5–4) over northwest Indian region during 12–14 June 2006. Six days air back-trajectories show transport of air masses from Thar Desert to the Central Himalayas during this period (Figure 1a).

3.1. Surface Meteorology

[7] The prevailing meteorology during June at Nainital is characterized by north-westerly wind. Wind passes through arid regions of western India and generally brings dry air masses from Southwest Asia. Daily average temperature was highest on 11–12 June 2006 at this site (Figure 2a) and temperature was also higher over north India in comparison to the average temperature in June 2006 over this region. Generally, relative humidity remains higher during the month of June due to the preset of southwest monsoon. But prior to dust event, moisture content was lowest (∼25–30%) on 11–12 June (Figure 2a). The columnar water vapour content was also lowest (0.52 ± 0.11 cm) during this period, which probably suggests vertical downward transport of dry air mass. During the onset of dust episode, wind speed was maximum (7.6 ± 2.3 m s−1) (Figure 2a), while average speed was 3.3 ± 1.8 m s−1 in rest of June month. Wind direction was consistently from northwest for about a week, with very less variability during dust episode period (see Table S1 of the auxiliary material).

Figure 2.

Daily average variations in (a) wind speed, temperature, and relative humidity, (b) aerosol number concentrations for OPC data in coarse and giant mode, and (c) accumulation mode particles and mass concentration for aethalometer in June 2006. One hourly average data of aerosols number concentrations are also shown for three sizes. Scales of aerosol number concentrations are different in Figures 2b and 2c. Vertical bars are one standard deviation. Observations from OPC were not available on 6 and 10 June 2006. Shaded area shows the period of dust storm.

3.2. Observations with the Optical Particle Counter

[8] Average variations (daily and hourly) in aerosols number concentrations in giant (>10.0 μm), coarse (1.0 μm < r < 10.0 μm) and accumulation (0.15 μm < r < 1.0 μm) modes, during June 2006 are shown in Figures 2b and 2c. Daily average aerosol number concentrations of coarse and giant mode particles show a gradual increase after 9 June and reached the maximum value of 16.4 ± 3.9 × 106 and 7.9 ± 3.0 × 103 m−3, respectively on 13 June (Figure 2b). Thereafter their concentrations show a decrease. Nainital is 800–1000 km away from the source region of dust storm and therefore, considering the wind speed of 8–10 m sec−1, maximum influence of dust storm on next day (i.e. 13 June) over Nainital is justified.

[9] Abrupt changes in concentrations started at about 2000 hour LT on 12 June. On the next day, 13 June, large variability was observed in coarse and giant mode particles with increase in number concentrations up to 26 × 106 m−3 and 17.2 × 103 m−3, respectively. Average concentrations in these two modes are observed to be 12.8 ± 4.6 × 106 m−3 and 6.0 ± 3.1 × 103 m−3 during 12–14 June, respectively, while their corresponding monthly mean values in June are only 5.5 ± 3.8 × 106 m−3 and 1.7 ± 2.1 × 103 m−3, respectively (Table S1 of the auxiliary material). Interestingly, variabilities are observed to be greater in giant mode particles (52%) when compared with coarse mode particles (36%). Further, increase in giant mode particles is as high as 10 times, while increase in coarse mode is only about 5 times. Concentrations in these two modes showed a gradual decrease from 14 June onwards. During this period the air masses arrive from north India and/or Arabian Sea (Figure 1b).

[10] A comparison of the number size distribution of aerosols from the OPC data between dust episode days (12–14 June) and rest of the observational days in June 2006, shows a very clear enhancement in normalized (with their sizes) aerosol number concentrations (dn/dr) for radii above 1.0 μm (Figure 3). This enhancement is as high as ∼50 times for sizes greater than 5.0 μm radius, while it is only ∼4 times for sizes greater than 10.0 μm radius. The observed enhancement is appearing to be due to mixing of two air masses with different aerosol populations. Smaller sizes do not show the enhancement, in–fact their value is lesser. As mentioned earlier, variations in these sizes are basically due to local processes. Estimated effective radius during dust episode days is found to be 1.47 μm, while it is only 0.28 μm during rest of the days in June 2006.

Figure 3.

Normalized (with their sizes) aerosol number concentration size distribution during dust episode (June 12–14) and rest of the observational days in June 2006 (days excluding June 12–14, 2006). Vertical bars are one standard deviation.

3.3. Variations in Concentrations of Accumulation Mode Particle and Black Carbon

[11] In contrast to coarse and giant mode, accumulation mode particles show a different variation and do not show higher value on 13 June (Figure 2c). Instead, concentrations of these particles show a gradual increase from 6 June itself with a maximum value on 9 June. Aerosol number concentrations of accumulation mode particles also show higher values on two other occasions (16 and 22 June) and confirming that variations in these particles are different than variations in coarse and giant mode particles.

[12] Observations of black carbon (BC) show variations nearly similar to that in accumulation mode particles with higher values on 6–9 June, 16 June, and 22 June 2006. Lower value of BC during dust episode period confirms the negligible contribution from local sources. The BC mass fraction was also found to be minimum (<0.5%) during the dust events as compared to dust free days (∼6%). Air back-trajectory also shows that air mass transport during dust episode basically does not coincide with regions of local anthropogenic emissions, which confirms the least contribution of local anthropogenic sources during dust episode. Therefore, we suggest that variations in the accumulation mode particles and BC are largely due to local processes, while variations in larger particle sizes are due to long-range transport.

3.4. Spectral Aerosol Optical Depths (AODs)

[13] Figure 4 shows average variations in spectral AOD during June 2006. AOD shows a gradual increase from 11 June with a maximum value of 0.63–0.67 at all six wavelengths on 13 June. Normally, spectral variations of AOD at this site show higher values at lower wavelengths [Sagar et al., 2004]. However, this spectral dependency in AOD reduces significantly on 13 June when dust episode was observed at this site. AOD at longer wavelengths shows larger increase in values, when compare to the increase in lower wavelengths during the dust episode period. This is attributed to the increase in number of coarse and giant particles. Observations made during a Saharan Dust Experiment [Tanré et al., 2003], at Islands in the Indian Ocean [Satheesh and Srinivasan, 2002], and in the desert region [Moorthy et al., 2007] showed increase in AOD at higher wavelengths during dust episodes.

Figure 4.

Variations in aerosol optical depth at different wavelengths in June 2006. Box plot at right side shows statistics of AOD in June, 2006, at 1020 nm. The upper and lower whiskers represent 95 and 5 percentiles, respectively. The box's upper and lower limits are 75 and 25 percentiles; straight and dotted line inside the box are median and mean, respectively. Shaded area shows the period of dust storm.

[14] Ångström wavelength exponent (α) is estimated to be very low (−0.03 ± 0.01) on 13 June, suggesting contribution of desert dust and the reduced abundance of fine particles during peak dust episode (Table S1). Additionally, the higher value of Ångström turbidity (β) (0.65 ± 0.03) on 13 June indicates abundance of coarse aerosols.

4. Discussions and Conclusions

[15] Ground based observations at a high altitude site in Nainital have provided a good evidence of influences of dust storm over the Himalayan region. The dust storm blew through the Thar Desert on 12 June 2006. Aerosol number concentrations in coarse and giant mode were observed to be 5 and 10 times higher respectively, during dust episode days in comparison to their monthly mean values of June 2006. Apart from larger increase in concentrations of giant mode particles, temporal variabilities were also larger in these particles in comparison to those in coarse mode particles.

[16] The spectral aerosol optical depth values showed two to four times increase, particularly at longer wavelengths, suggesting the increase in the columnar loading of coarse mode particles during the dust episode. This is also confirmed by higher values of turbidity coefficient and near-zero value of Ångström wavelength exponent. Arrival of air masses from dry areas of the Thar Desert regions is seen during the dust episode using air back-trajectories. In contrast to the bigger particles, number concentrations of accumulation mode particles and black carbon do not show enhancement in their values during dust episode, suggesting insignificant contributions from local sources.

[17] Large changes in aerosols optical and physical properties over the central Himalayan region during dust episode will have significant impact on regional radiative budget and environmental changes. East Asian dust episode during 19–23 March 2002 resulted to add total suspended particle of ∼400 μg m−3 over Korea, which has led to −11 Wm−2 mean radiative forcing at the surface and −6 Wm−2 at top of atmosphere [Park et al., 2005]. In a similar way, enhancement in total suspended particle by nearly seven times (282 μg m−3) of average summertime value (Figure S1) during present dust episode could perturb the radiation budget for this region.

[18] The Thar Desert is one of the major sources of atmospheric dust in South Asia. Therefore, it is highly essential to quantify the influences of such events that will help in reducing the uncertainties in radiation budget and improve the air quality prediction over South Asia. Different studies have also shown that dust from Asian continental sources could reach as far as over the Pacific and have influences there. Thereby, it is desirable to make more detailed observations, with a chemical composition analysis, and evaluating radiative forcing for better understanding of such dust episode over South Asia and its impact on global climate.


[19] This work is carried out as a part of ISRO-Geosphere Biosphere Program project. We would like to thank A. Taori for help in meteorological observations. We greatly appreciate the constructive and useful suggestions of anonymous reviewers, which significantly improved the contents of the paper.