**4. Background aerosol optical and physical properties over Thar desert**

## **4.1. Aerosol optical depth**

The seasonal variation of AOD spectrum at the hilltop station over western India is shown in Figure 4. Vertical lines represent ±1*σ* variation about the mean AOD. The solid lines are the OPAC model fitted AOD spectrum and the shaded regions indicate the variation of simulated AOD within that season. At all wavelengths, AODs are maximum during premonsoon followed by postmonsoon and monsoon and are minimum during winter. At 0.5 *μ*m, the AODs are 0.20±0.08, 0.18±0.04, 0.10±0.02, and 0.08±0.03 during premonsoon, postmonsoon, monsoon and winter, respectively. The reasons could be as follows. Firstly, there is a large variation of boundary layer height. In an earlier research work carried out over a tropical Indian station, Gadanki (13.5*o*N, 79.2*o*E), Krishnan & Kunhikrishnan [37] studied the annual boundary layer height variation and observed a minimum during winter and maximum during premonsoon. In the present study, during winter the boundary layer height is lower than the observation altitude and hence the observation site is in the free troposphere region. As a result, AOD is minimum during this season. During premonsoon the observation site is within the boundary layer and hence AOD increases. In addition, there is significant amount of dust transported from arid region which results in maximum AOD. On the other hand, during monsoon though the boundary layer height is significantly high, AOD is low due to wash out of aerosols from the atmosphere by the heavy monsoonal rain events. Monsoon rain has a major role to wash out the aerosol loading from the atmosphere causing significant decrease of AOD. A case study over tropical Indian station reported about 64% decrease of AOD due to heavy rain [69]. In the present study, there is no significant decrease of AOD during monsoon. This is because of the presence of a very stable aerosol layer of about 1.5

**Figure 3.** Picture of land surface over western India during premonsoon (top) and potmonsoon (middle) seasons, obtained from MODIS-Terra satellite. Dark black, gray and green colors represent oceanic surface, arid bare land, forest regions, respectively. (Bottom) Monthly variation of surface reflectance at 1.64 *μ*m wavelength during 2007 and the vertical bars represents ±1*σ* deviation about the monthly mean. Note that the surface reflectance increases by 70% over study area during premonsoon.

Natural vs Anthropogenic Background Aerosol Contribution to the Radiation Budget over Indian Thar Desert 91

90 Atmospheric Aerosols – Regional Characteristics – Chemistry and Physics Natural vs Anthropogenic Background Aerosol Contribution to the Radiation Budget over Indian Thar Desert <sup>11</sup> Natural vs Anthropogenic Background Aerosol Contribution to the Radiation Budget over Indian Thar Desert 91

10 Will-be-set-by-IN-TECH

during postmonsoon due to green vegetation born during monsoon rain. As a result, surface

In the present study, MODIS derived surface reflectance data over Mt. Abu is used in the estimations of radiative forcing. It is obtained from Nadir BRDF-Adjusted Reflectance 16-Day L3 Global 0.5 km SIN Grid product which is derived at the mean solar zenith angle of Terra overpasses for every successive 16-day period, calculating surface reflectance as if every pixel in the grid was viewed from nadir direction. Surface reflectance data available in seven wavelength bands of MODIS centered around 0.47, 0.56, 0.65, 0.86, 1.24, 1.64 and 2.13 *μ*m are used to reproduce the spectral dependence of surface reflectances for the entire SW range using a combination of three different surface types, namely, vegetation, sand and water. The monthly variation of surface reflectance at 1.64 *μ*m during 2007 is shown in Figure 3c. Vertical lines in this Figure represent ±1*σ* variation about the monthly mean values. Average surface reflectance is found to be high at about 0.35 during premonsoon (Apr-May) and low at about 0.20 during postmonsoon (Sep-Nov) and winter (Dec-Feb). Space-borne observations show that the land over western India increases its brightness by about 75% during premonsoon season. This could be due to bare surface and deposited dust that is transported from arid region. Model simulations to fit the surface reflectance combining the three surfaces suggest that during premonsoon sand surface contributes a maximum of about 70% and during postmonsoon and winter it contributes a minimum of about 20% while vegetation surface contributes 15% and 60%, respectively. These varying land properties are also considered in

**4. Background aerosol optical and physical properties over Thar desert**

The seasonal variation of AOD spectrum at the hilltop station over western India is shown in Figure 4. Vertical lines represent ±1*σ* variation about the mean AOD. The solid lines are the OPAC model fitted AOD spectrum and the shaded regions indicate the variation of simulated AOD within that season. At all wavelengths, AODs are maximum during premonsoon followed by postmonsoon and monsoon and are minimum during winter. At 0.5 *μ*m, the AODs are 0.20±0.08, 0.18±0.04, 0.10±0.02, and 0.08±0.03 during premonsoon, postmonsoon, monsoon and winter, respectively. The reasons could be as follows. Firstly, there is a large variation of boundary layer height. In an earlier research work carried out over a tropical Indian station, Gadanki (13.5*o*N, 79.2*o*E), Krishnan & Kunhikrishnan [37] studied the annual boundary layer height variation and observed a minimum during winter and maximum during premonsoon. In the present study, during winter the boundary layer height is lower than the observation altitude and hence the observation site is in the free troposphere region. As a result, AOD is minimum during this season. During premonsoon the observation site is within the boundary layer and hence AOD increases. In addition, there is significant amount of dust transported from arid region which results in maximum AOD. On the other hand, during monsoon though the boundary layer height is significantly high, AOD is low due to wash out of aerosols from the atmosphere by the heavy monsoonal rain events. Monsoon rain has a major role to wash out the aerosol loading from the atmosphere causing significant decrease of AOD. A case study over tropical Indian station reported about 64% decrease of AOD due to heavy rain [69]. In the present study, there is no significant decrease of AOD during monsoon. This is because of the presence of a very stable aerosol layer of about 1.5

reflectance is maximum during premonsoon and minimum during postmonsoon.

the radiative forcing calculations.

**4.1. Aerosol optical depth**

**Figure 3.** Picture of land surface over western India during premonsoon (top) and potmonsoon (middle) seasons, obtained from MODIS-Terra satellite. Dark black, gray and green colors represent oceanic surface, arid bare land, forest regions, respectively. (Bottom) Monthly variation of surface reflectance at 1.64 *μ*m wavelength during 2007 and the vertical bars represents ±1*σ* deviation about the monthly mean. Note that the surface reflectance increases by 70% over study area during premonsoon.

#### 12 Will-be-set-by-IN-TECH 92 Atmospheric Aerosols – Regional Characteristics – Chemistry and Physics Natural vs Anthropogenic Background Aerosol Contribution to the Radiation Budget over Indian Thar Desert <sup>13</sup>

km thickness over the inversion layer during monsoon in the western India, as reported by Ganguly et al. [19].

The spectral dependence of AOD is parameterized through Ångstr*o*¨m exponent (*α*) which is the slope of the logarithm of AOD versus the logarithm of wavelength (in micron) and provides the basic information about the columnar particle size distribution [66]. *α* is higher for relatively higher number of smaller particles and as the number of bigger particles increases *α* decreases. It can even reduce to ∼0 for very large number of coarse-mode soil particles [51, 77, 81, etc]. In the present study, *α* is obtained from Microtops measured AOD for the entire wavelength (0.380 - 1.020 *μ*m) and is given in Figure 4 along with the variation in the parenthesis. It varies from 0.2 to 0.6. During monsoon *α* is minimum at 0.2±0.15 indicating dominance of bigger aerosols. It is due to the presence of bigger water soluble aerosols which increase in size due to accumulation and coagulation processes in high relative humidity conditions. During premonsoon also, when RH is low and the atmosphere is dry and warm, *α* is low at about 0.3±0.25 indicating the dominance of bigger aerosols. These are the soil born dust aerosols produced by the frequently occurring dust storms in the Thar desert and transported to other parts of India [17, 55] including Mt. Abu. *α* is found to be maximum at about 0.6±0.01 during postmonsoon indicating dominance of smaller aerosols. This is probably due to dominance of fine seasalt aerosols transported from Arabian Sea [19, 64] . During winter also *α* is found to be high showing dominance of smaller aerosols which could be due to anthropogenic aerosols coming from burning tree branches and dry leaves by the poor villagers living in the surrounding hill areas to keep themselves warm during cold mornings and evenings.

## **4.2. Aerosol mass concentration**

Aerosol mass concentration measured separately in ten different sizes by Quartz Crystal Microbalance (QCM) cascade impactor has been classified into three different categories, viz., nucleation (radius*<*0.1 *μ*m), accumulation (0.1*μ*m≤radius≥1.0 *μ*m) and coarse (radius*>*1.0 *μ*m) mode particles. Nucleation mode aerosols represent total aerosols collected in stages 9-10, accumulation mode aerosols are the total aerosols in stages 5-8 and coarse mode aerosols are the total aerosols collected in stages 2-4. Aerosols collected in stage 1 are not considered in the calculations because all aerosols whose radius is greater than 12.5 *μ*m are collected in this stage and thus, there is no definite aerosol radius representing this stage.

**Figure 4.** Seasonal variation of observed AOD spectrum at Mt. Abu. Vertical bars represent ±1*σ* variation about the mean. Solid line is the OPAC model derived AOD spectrum and the shaded regions are the variation in the simulated AOD. Ångstr*o*¨m exponent (*α*) is also given along with the standard

enhances the nucleation mode aerosols which explains the maximum mass of 6.4±1.1 *<sup>μ</sup>*g.m−<sup>3</sup> observed during this season. During winter the nucleation mass concentration decreased to 3.2±0.1 *<sup>μ</sup>*g.m−<sup>3</sup> as the boundary layer height decreased and the measurement site was in free troposphere. The nucleation aerosol mass was 2.9±1.8 and 3.6±0.7 *<sup>μ</sup>*g.m−<sup>3</sup> during premonsoon and monsoon, respectively. During premonsoon, the boundary layer height was maximum which gives more room for these fine aerosols to dilute and high temperature with low RH are not favorable for gas-to-particle conversion processes. In addition, strong wind also helps in removing the aerosols from the measurement site during this season and makes mass of the nucleation mode aerosols minimum. Monsoon also experiences high boundary layer height and strong wind condition, however, nucleation mode aerosols are significant compared to premonsoon. This could be due to the transport of seasalt coming from Arabian

Natural vs Anthropogenic Background Aerosol Contribution to the Radiation Budget over Indian Thar Desert 93

The seasonal variation observed in the accumulation aerosols is similar to the nucleation aerosols. The accumulation aerosol mass concentration was minimum at 8.4±2.8 *<sup>μ</sup>*g.m−<sup>3</sup> during premonsoon and maximum at 12.6±0.6 *<sup>μ</sup>*g.m−<sup>3</sup> during postmonsoon followed by monsoon (10.0±1.0 *<sup>μ</sup>*g.m−3) and winter (9.6±1.6 *<sup>μ</sup>*g.m−3). Accumulation aerosols are mainly produced by the condensation growth and coagulation of nucleation aerosols. During

deviations in parenthesis.

Sea.

Figure 5 shows the seasonal variation of aerosol mass concentration (*μ*g.m−3) of all three modes, viz., nucleation, accumulation and coarse modes at the hill top region, Mt. Abu from January 2006 to December 2007. Total aerosol mass concentration observed was minimum at 16.5±1.5 *<sup>μ</sup>*g.m−<sup>3</sup> during winter and maximum at 25.8±2.7 *<sup>μ</sup>*g.m−<sup>3</sup> during postmonsoon followed by premonsoon (19.9±5.6 *<sup>μ</sup>*g.m−3) and monsoon (16.7±6.0 *<sup>μ</sup>*g.m−3). This variation is similar to that observed in columnar AOD at Mt. Abu. The accumulation aerosol mass was contributing maximum to the total aerosol mass during all seasons and the coarse mode aerosol mass was contributing equivalently only during premonsoon. This is due to large transportation of dust aerosols from Thar desert during this season that enhanced the coarse mode aerosol mass.

In general, nucleation aerosols contribute least to the total aerosol mass concentration. This contribution was maximum during postmonsoon when the wind speed was almost calm and RH was relatively high. This atmospheric condition helps in gas-to-particle conversion and 12 Will-be-set-by-IN-TECH

km thickness over the inversion layer during monsoon in the western India, as reported by

The spectral dependence of AOD is parameterized through Ångstr*o*¨m exponent (*α*) which is the slope of the logarithm of AOD versus the logarithm of wavelength (in micron) and provides the basic information about the columnar particle size distribution [66]. *α* is higher for relatively higher number of smaller particles and as the number of bigger particles increases *α* decreases. It can even reduce to ∼0 for very large number of coarse-mode soil particles [51, 77, 81, etc]. In the present study, *α* is obtained from Microtops measured AOD for the entire wavelength (0.380 - 1.020 *μ*m) and is given in Figure 4 along with the variation in the parenthesis. It varies from 0.2 to 0.6. During monsoon *α* is minimum at 0.2±0.15 indicating dominance of bigger aerosols. It is due to the presence of bigger water soluble aerosols which increase in size due to accumulation and coagulation processes in high relative humidity conditions. During premonsoon also, when RH is low and the atmosphere is dry and warm, *α* is low at about 0.3±0.25 indicating the dominance of bigger aerosols. These are the soil born dust aerosols produced by the frequently occurring dust storms in the Thar desert and transported to other parts of India [17, 55] including Mt. Abu. *α* is found to be maximum at about 0.6±0.01 during postmonsoon indicating dominance of smaller aerosols. This is probably due to dominance of fine seasalt aerosols transported from Arabian Sea [19, 64] . During winter also *α* is found to be high showing dominance of smaller aerosols which could be due to anthropogenic aerosols coming from burning tree branches and dry leaves by the poor villagers living in the surrounding hill areas to keep themselves warm during cold

Aerosol mass concentration measured separately in ten different sizes by Quartz Crystal Microbalance (QCM) cascade impactor has been classified into three different categories, viz., nucleation (radius*<*0.1 *μ*m), accumulation (0.1*μ*m≤radius≥1.0 *μ*m) and coarse (radius*>*1.0 *μ*m) mode particles. Nucleation mode aerosols represent total aerosols collected in stages 9-10, accumulation mode aerosols are the total aerosols in stages 5-8 and coarse mode aerosols are the total aerosols collected in stages 2-4. Aerosols collected in stage 1 are not considered in the calculations because all aerosols whose radius is greater than 12.5 *μ*m are collected in this

Figure 5 shows the seasonal variation of aerosol mass concentration (*μ*g.m−3) of all three modes, viz., nucleation, accumulation and coarse modes at the hill top region, Mt. Abu from January 2006 to December 2007. Total aerosol mass concentration observed was minimum at 16.5±1.5 *<sup>μ</sup>*g.m−<sup>3</sup> during winter and maximum at 25.8±2.7 *<sup>μ</sup>*g.m−<sup>3</sup> during postmonsoon followed by premonsoon (19.9±5.6 *<sup>μ</sup>*g.m−3) and monsoon (16.7±6.0 *<sup>μ</sup>*g.m−3). This variation is similar to that observed in columnar AOD at Mt. Abu. The accumulation aerosol mass was contributing maximum to the total aerosol mass during all seasons and the coarse mode aerosol mass was contributing equivalently only during premonsoon. This is due to large transportation of dust aerosols from Thar desert during this season that enhanced the coarse

In general, nucleation aerosols contribute least to the total aerosol mass concentration. This contribution was maximum during postmonsoon when the wind speed was almost calm and RH was relatively high. This atmospheric condition helps in gas-to-particle conversion and

stage and thus, there is no definite aerosol radius representing this stage.

Ganguly et al. [19].

mornings and evenings.

mode aerosol mass.

**4.2. Aerosol mass concentration**

**Figure 4.** Seasonal variation of observed AOD spectrum at Mt. Abu. Vertical bars represent ±1*σ* variation about the mean. Solid line is the OPAC model derived AOD spectrum and the shaded regions are the variation in the simulated AOD. Ångstr*o*¨m exponent (*α*) is also given along with the standard deviations in parenthesis.

enhances the nucleation mode aerosols which explains the maximum mass of 6.4±1.1 *<sup>μ</sup>*g.m−<sup>3</sup> observed during this season. During winter the nucleation mass concentration decreased to 3.2±0.1 *<sup>μ</sup>*g.m−<sup>3</sup> as the boundary layer height decreased and the measurement site was in free troposphere. The nucleation aerosol mass was 2.9±1.8 and 3.6±0.7 *<sup>μ</sup>*g.m−<sup>3</sup> during premonsoon and monsoon, respectively. During premonsoon, the boundary layer height was maximum which gives more room for these fine aerosols to dilute and high temperature with low RH are not favorable for gas-to-particle conversion processes. In addition, strong wind also helps in removing the aerosols from the measurement site during this season and makes mass of the nucleation mode aerosols minimum. Monsoon also experiences high boundary layer height and strong wind condition, however, nucleation mode aerosols are significant compared to premonsoon. This could be due to the transport of seasalt coming from Arabian Sea.

The seasonal variation observed in the accumulation aerosols is similar to the nucleation aerosols. The accumulation aerosol mass concentration was minimum at 8.4±2.8 *<sup>μ</sup>*g.m−<sup>3</sup> during premonsoon and maximum at 12.6±0.6 *<sup>μ</sup>*g.m−<sup>3</sup> during postmonsoon followed by monsoon (10.0±1.0 *<sup>μ</sup>*g.m−3) and winter (9.6±1.6 *<sup>μ</sup>*g.m−3). Accumulation aerosols are mainly produced by the condensation growth and coagulation of nucleation aerosols. During

the typical aerosol size distributions for the four seasons. The vertical bars represent ±1*σ* variation about the monthly mean number concentration of different sizes of aerosols. In all seasons, the size distribution showed tri-modal distribution and each mode could be fitted

Natural vs Anthropogenic Background Aerosol Contribution to the Radiation Budget over Indian Thar Desert 95

exp ⎡ <sup>⎣</sup><sup>−</sup> log2

where *N* is the number concentration (cm−3), *σ<sup>m</sup>* is the width of the distribution and *rm* is the mode radius for a particular mode. The three modal parameters for all the seasons are given in Table 1. At Mt. Abu the number concentrations (*N*) of nucleation and accumulation modes are lower by an order of magnitude than that at other urban region in western India, Ahmedabad while for coarse mode it is comparable [19]. Since Mt. Abu is far from anthropogenic activity, the anthropogenically influenced modes (nucleation and accumulation) have smaller number concentrations. However, the proximity to Thar desert and similarity of the surface conditions of Mt. Abu make the coarse mode number concentrations comparable. The radii of nucleation mode lie in the range 0.018-0.020 *μ*m and number concentration for this mode is found to be maximum during postmonsoon and minimum during premonsoon and monsoon. Similarly the radii for corresponding accumulation and coarse modes lie in the range of 0.12-0.19 *μ*m

Season *N* r*<sup>m</sup> σ N* r*<sup>m</sup> σ N* r*<sup>m</sup> σ*

Winter 12000 0.019 2.0 18 0.14 2.0 0.02 1.4 1.9 Premonsoon 10000 0.018 1.9 22 0.19 1.8 0.01 2.2 1.8 Monsoon 15000 0.018 1.9 50 0.13 1.9 0.01 1.7 1.8 Postmonsoon 17000 0.020 1.9 60 0.12 1.9 0.08 1.1 1.8 **Table 1.** Average values of size distribution parameters obtained by fitting lognormal curves to the

Accumulation aerosols are mainly produced by the condensation growth and coagulation of nucleation aerosols. During winter accumulation mode aerosols number concentration (*N*) was minimum at 18 cm−3. During premonsoon, the anthropogenic activities were maximum at Mt. Abu which increased *N* of accumulation mode to 22 cm−3. During monsoon, it further increased to 50 cm−3. It is due to the wind coming from Arabian sea (Figure 2) that carried large amount of sea salt and enriched the sea salt aerosols at the hill top region [64]. During high RH conditions these sea salt aerosols belong to the accumulation mode. Later during postmonsoon, wind was south-easterly and the transported sea salt reduced. However, burning of biomass like garbage and fallen leaves increased and hence, BC particle concentration was enhanced. Therefore, high production, shallow boundary layer height and low wind speed made the accumulation mode aerosol number concentration reach a

During premonsoon, there is large transportation of mineral dust aerosols from Thar Desert which enhanced the abundance of coarse mode aerosols at the hill top area. The coarse mode radius was maximum at 2.2 *μ*m. During monsoon, rain washes out these dust aerosols from the atmosphere and reduces their number and mode radius. However, the abundance

� *<sup>r</sup> rm* �

2

⎤

⎦ (2)

2 (log *σm*)

Nucleation Accumulation Coarse

cm−<sup>3</sup> *μ*m *μ*m cm−<sup>3</sup> *μ*m *μ*m cm−<sup>3</sup> *μ*m *μ*m

using using three lognormal modes of the following form.

*dn* (*r*)

and 1.1-2.1 *μ*m, respectively.

measured aerosol number distribution over Mt. Abu

maximum at 60 cm−<sup>3</sup> during this season.

*dr* <sup>=</sup> *<sup>N</sup>*

<sup>√</sup>2*<sup>π</sup>* log *<sup>σ</sup><sup>m</sup>*

**Figure 5.** Seasonal variation of aerosol mass concentrations in nucleation (radius < 0.1 micron), accumulation (0.1≥radius≤1.0 micron) and coarse (radius >1.0 micron) modes. The vertical bars represent ±1*σ* variation about the mean.

premonsoon these processes are slowed down due to low RH and hence the low accumulation aerosol mass. During monsoon seasalt aerosols are transported from the Arabian sea and high RH maintains the accumulation mode aerosols, increasing the mass concentration. During postmonsoon, minimum wind speed results in further increase resulting in the observed maximum. And during winter it is minimum as the the measurement site is in the free troposphere.

The coarse mode aerosols show a slightly different seasonal behaviour at Mt. Abu. During premonsoon, they mainly consist of dust aerosols transported from Thar desert and the mass concentration is maximum at 8.6±0.4 *<sup>μ</sup>*g.m−3. It is minimum at 3.1±0.5 *<sup>μ</sup>*g.m−<sup>3</sup> during monsoon due to wash out of the dust aerosols by heavy rains. During postmonsoon, the coarse aerosol mass concentration was slightly enhanced to 6.8±1.0 *<sup>μ</sup>*g.m−<sup>3</sup> as the accumulation aerosols, which mainly consist of seasalt particles, swell up by absorbing water vapor at high RH conditions and become coarse mode particles. During winter, their mass concentration becomes 3.6±0.6 *<sup>μ</sup>*g.m−<sup>3</sup> when low boundary layer height helps to keep them low at the hill-top region.

### **4.3. Aerosol number concentration**

Aerosol number concentration is also obtained from the observed aerosol mass concentration from QCM observations for the hilltop area using appropriate mass density valid for semi-arid background atmosphere and prevailing relative humidity conditions [13, 26]. Figure 6 shows the typical aerosol size distributions for the four seasons. The vertical bars represent ±1*σ* variation about the monthly mean number concentration of different sizes of aerosols. In all seasons, the size distribution showed tri-modal distribution and each mode could be fitted using using three lognormal modes of the following form.

14 Will-be-set-by-IN-TECH

**Figure 5.** Seasonal variation of aerosol mass concentrations in nucleation (radius < 0.1 micron), accumulation (0.1≥radius≤1.0 micron) and coarse (radius >1.0 micron) modes. The vertical bars

premonsoon these processes are slowed down due to low RH and hence the low accumulation aerosol mass. During monsoon seasalt aerosols are transported from the Arabian sea and high RH maintains the accumulation mode aerosols, increasing the mass concentration. During postmonsoon, minimum wind speed results in further increase resulting in the observed maximum. And during winter it is minimum as the the measurement site is in the free

The coarse mode aerosols show a slightly different seasonal behaviour at Mt. Abu. During premonsoon, they mainly consist of dust aerosols transported from Thar desert and the mass concentration is maximum at 8.6±0.4 *<sup>μ</sup>*g.m−3. It is minimum at 3.1±0.5 *<sup>μ</sup>*g.m−<sup>3</sup> during monsoon due to wash out of the dust aerosols by heavy rains. During postmonsoon, the coarse aerosol mass concentration was slightly enhanced to 6.8±1.0 *<sup>μ</sup>*g.m−<sup>3</sup> as the accumulation aerosols, which mainly consist of seasalt particles, swell up by absorbing water vapor at high RH conditions and become coarse mode particles. During winter, their mass concentration becomes 3.6±0.6 *<sup>μ</sup>*g.m−<sup>3</sup> when low boundary layer height helps to keep them

Aerosol number concentration is also obtained from the observed aerosol mass concentration from QCM observations for the hilltop area using appropriate mass density valid for semi-arid background atmosphere and prevailing relative humidity conditions [13, 26]. Figure 6 shows

represent ±1*σ* variation about the mean.

troposphere.

low at the hill-top region.

**4.3. Aerosol number concentration**

$$\frac{dn\,(r)}{dr} = \frac{N}{\sqrt{2\pi}\log\sigma\_m} \exp\left[-\frac{\log^2\left(\frac{r}{r\_m}\right)}{2\left(\log\sigma\_m\right)^2}\right] \tag{2}$$

where *N* is the number concentration (cm−3), *σ<sup>m</sup>* is the width of the distribution and *rm* is the mode radius for a particular mode. The three modal parameters for all the seasons are given in Table 1. At Mt. Abu the number concentrations (*N*) of nucleation and accumulation modes are lower by an order of magnitude than that at other urban region in western India, Ahmedabad while for coarse mode it is comparable [19]. Since Mt. Abu is far from anthropogenic activity, the anthropogenically influenced modes (nucleation and accumulation) have smaller number concentrations. However, the proximity to Thar desert and similarity of the surface conditions of Mt. Abu make the coarse mode number concentrations comparable. The radii of nucleation mode lie in the range 0.018-0.020 *μ*m and number concentration for this mode is found to be maximum during postmonsoon and minimum during premonsoon and monsoon. Similarly the radii for corresponding accumulation and coarse modes lie in the range of 0.12-0.19 *μ*m and 1.1-2.1 *μ*m, respectively.


**Table 1.** Average values of size distribution parameters obtained by fitting lognormal curves to the measured aerosol number distribution over Mt. Abu

Accumulation aerosols are mainly produced by the condensation growth and coagulation of nucleation aerosols. During winter accumulation mode aerosols number concentration (*N*) was minimum at 18 cm−3. During premonsoon, the anthropogenic activities were maximum at Mt. Abu which increased *N* of accumulation mode to 22 cm−3. During monsoon, it further increased to 50 cm−3. It is due to the wind coming from Arabian sea (Figure 2) that carried large amount of sea salt and enriched the sea salt aerosols at the hill top region [64]. During high RH conditions these sea salt aerosols belong to the accumulation mode. Later during postmonsoon, wind was south-easterly and the transported sea salt reduced. However, burning of biomass like garbage and fallen leaves increased and hence, BC particle concentration was enhanced. Therefore, high production, shallow boundary layer height and low wind speed made the accumulation mode aerosol number concentration reach a maximum at 60 cm−<sup>3</sup> during this season.

During premonsoon, there is large transportation of mineral dust aerosols from Thar Desert which enhanced the abundance of coarse mode aerosols at the hill top area. The coarse mode radius was maximum at 2.2 *μ*m. During monsoon, rain washes out these dust aerosols from the atmosphere and reduces their number and mode radius. However, the abundance

after CO2 and has a larger impact on direct radiative forcing than that of methane. As a result, in populated countries like China and India, the large production of BC aerosols has a large impact on the hydrological cycle and precipitation pattern [46, 61, 71]. In India the fraction of BC production from fossil fuel burning, open burning and biofuel combustion to the global emission is significantly large and hence, it is necessary to estimate radiative impact

Natural vs Anthropogenic Background Aerosol Contribution to the Radiation Budget over Indian Thar Desert 97

**Figure 7.** Diurnal variation of BC mass concentration during each month in 2007. White region indicates

In recent years, global climate has received considerable attention due to increase in the percentage contribution of anthropogenic aerosols on the Earth's radiation budget [23]. BC particles exist mainly in the accumulation mode and can be transported over long distances [12] from source regions to far off pristine environment and perturb the climate of the latter, like that of Mt. Abu. The diurnal variation of BC mass concentration during different months over Mt. Abu is shown in Figure 7. Observations were not possible in July and August due to heavy rain. Minimum BC concentration was observed during monsoon (0.428±0.128 *<sup>μ</sup>*g.m−3) and maximum was observed during premonsoon (0.665±0.478 *<sup>μ</sup>*g.m−3) followed by winter (0.608±0.246 *<sup>μ</sup>*g.m−3) and postmonsoon (0.620±0.158 *<sup>μ</sup>*g.m−3). The annual mean BC mass concentration was 0.580±0.104 *<sup>μ</sup>*g.m−3. At Mt. Abu the BC concentration is an order of magnitude less than that at any other urban region in India. BC during April is found to be as high as 1.00±0.170 *<sup>μ</sup>*g.m−<sup>3</sup> which is a factor of 2 higher than the previous month. Backtrajectory analysis indicates wind coming from IGP which increases the BC concentration. In another study at a high altitude station, Nainital (29.4*o*N, 79.5*o*E, 1950m

no data during Jul-Aug due to heavy monsoonal rain.

of different kinds of BC not only on global scale but also in the regional scale.

**Figure 6.** Seasonal variation of aerosol number distribution. Vertical bars represent ±1*σ* variation about the mean. The solid lines are the best-fitted curves representing nucleation, accumulation and coarse modes.

of coarse aerosols is found to be maximum with minimum mode radius of about 1.1 *μ*m during postmonsoon which indicates the transfer of aerosols from accumulation mode due to hygroscopic and coagulation growth of particles at high RH conditions.

### **4.4. Black carbon mass concentration**

Black carbon (BC) produced due to incomplete combustion of carbon-based fuels [3, 31, 53, 86, etc] is the most efficient light absorbing aerosol component in the atmosphere. BC has major contribution to alter the radiative balance by absorbing the solar radiation in the visible spectrum. As a result, it cools the surface and warms the atmosphere [24, 38]. A recent study of BC contribution to radiative forcing by Jacobson [31] showed that BC has a great contribution towards global warming and is the second most important component of global warming after CO2 and has a larger impact on direct radiative forcing than that of methane. As a result, in populated countries like China and India, the large production of BC aerosols has a large impact on the hydrological cycle and precipitation pattern [46, 61, 71]. In India the fraction of BC production from fossil fuel burning, open burning and biofuel combustion to the global emission is significantly large and hence, it is necessary to estimate radiative impact of different kinds of BC not only on global scale but also in the regional scale.

16 Will-be-set-by-IN-TECH

**Figure 6.** Seasonal variation of aerosol number distribution. Vertical bars represent ±1*σ* variation about the mean. The solid lines are the best-fitted curves representing nucleation, accumulation and coarse

of coarse aerosols is found to be maximum with minimum mode radius of about 1.1 *μ*m during postmonsoon which indicates the transfer of aerosols from accumulation mode due

Black carbon (BC) produced due to incomplete combustion of carbon-based fuels [3, 31, 53, 86, etc] is the most efficient light absorbing aerosol component in the atmosphere. BC has major contribution to alter the radiative balance by absorbing the solar radiation in the visible spectrum. As a result, it cools the surface and warms the atmosphere [24, 38]. A recent study of BC contribution to radiative forcing by Jacobson [31] showed that BC has a great contribution towards global warming and is the second most important component of global warming

to hygroscopic and coagulation growth of particles at high RH conditions.

**4.4. Black carbon mass concentration**

modes.

**Figure 7.** Diurnal variation of BC mass concentration during each month in 2007. White region indicates no data during Jul-Aug due to heavy monsoonal rain.

In recent years, global climate has received considerable attention due to increase in the percentage contribution of anthropogenic aerosols on the Earth's radiation budget [23]. BC particles exist mainly in the accumulation mode and can be transported over long distances [12] from source regions to far off pristine environment and perturb the climate of the latter, like that of Mt. Abu. The diurnal variation of BC mass concentration during different months over Mt. Abu is shown in Figure 7. Observations were not possible in July and August due to heavy rain. Minimum BC concentration was observed during monsoon (0.428±0.128 *<sup>μ</sup>*g.m−3) and maximum was observed during premonsoon (0.665±0.478 *<sup>μ</sup>*g.m−3) followed by winter (0.608±0.246 *<sup>μ</sup>*g.m−3) and postmonsoon (0.620±0.158 *<sup>μ</sup>*g.m−3). The annual mean BC mass concentration was 0.580±0.104 *<sup>μ</sup>*g.m−3. At Mt. Abu the BC concentration is an order of magnitude less than that at any other urban region in India. BC during April is found to be as high as 1.00±0.170 *<sup>μ</sup>*g.m−<sup>3</sup> which is a factor of 2 higher than the previous month. Backtrajectory analysis indicates wind coming from IGP which increases the BC concentration. In another study at a high altitude station, Nainital (29.4*o*N, 79.5*o*E, 1950m asl), in central Himalayas, mean BC was observed to be 1.36±0.99 *<sup>μ</sup>*g.m−<sup>3</sup> during December 2004 [55]. This shows that Mt. Abu is less affected by anthropogenic activities.

of seasalt aerosols from Arabian Sea. During Oct-Feb, high values of *α* are found indicating enhancement of anthropogenic aerosols. Ground-based observations show high abundance of BC on the hill-top region during winter. All these observations suggest that dust is dominating during premonsoon, anthropogenic aerosols during winter and natural seasalt are present in

Natural vs Anthropogenic Background Aerosol Contribution to the Radiation Budget over Indian Thar Desert 99

**Figure 8.** Space-borne daily observations of AOD, Ångstr*o*¨m exponent, small mode fraction (SMF) obtained from MODIS onboard Terra and Aqua satellites and aerosol index obtained from OMI onboard

Seasonal variation of aerosol vertical profiles over the study region is obtained from CALIPSO observations. Figure 9 shows the seasonal variation of aerosol extinction coefficient (km−1). The horizontal dotted line at 1.7 km represents the height of Mt. Abu. The extinction coefficient is directly proportional to the total aerosol loading. It is clearly seen from the figure that aerosol loading over Mt. Abu is minimum during winter and higher during other seasons. There is a peak found near 2.2 km altitude during monsoon which becomes weak during postmonsoon. Ganguly et al. [19] reported that this peak is due to seasalt aerosols transported from Arabian sea and chemical analysis also supports this result showing significantly high amount of seasalt present over Mt. Abu during monsoon [64]. During premonsoon, there is a peak at 4.2 km which is due to the transported dust layer. MODIS and

OMI observations also indicate significant amount of dust present in the atmosphere.

Near surface region also shows high extinction coefficient values. This could be due to locally produced anthropogenic aerosols. In the present study, the properties of aerosols at the

the atmosphere during monsoon season.

Aura satellite during 2006 and 2007.

**5.2. Aerosol vertical profile**

The diurnal variation of BC mass concentration does not show any significant morning and nocturnal peaks like other urban regions. However, increased BC was observed during the noon hours except during November and December. The reason for such an increase is during the day time the thermal convection becomes stronger and as a result, the pollutants at the foothill area rise up to the hilltop region and enhance the BC concentration. This day time enhancement was prominent during winter and postmonsoon because during these seasons there is a large difference between the day and night time temperatures. During November and December the night time BC concentration was larger by a factor of two. During these months the nearby villagers burn wood and fallen leaves to keep themselves warm thereby increasing the BC mass concentration. During January this nocturnal enhancement was not observed. The reason is that the boundary layer height is less than the station altitude and the night time BC that is produced cannot reach the hill top region due to weak thermal convection. During this period hill top region becomes pollutant free region.

## **5. Satellite observed aerosol properties over Mt. Abu region**

## **5.1. Aerosol optical and physical properties**

In the current satellite era, large databases are available to study aerosol properties from space, both in the regional and the global scale, that are essentially demanding. For the present study, Terra and a series of satellite sensors flying on the A-train platform provide the required data. MODIS on board Terra and Aqua provide aerosol parameters in the morning and afternoon. OMI on board Aura satellite provides AI. The joint information of AOD, Ångstr*o*¨m exponent (*α*) and small mode fraction (SMF) retrieved from MODIS and AI retrieved from OMI can be utilized to estimate the optical properties of aerosols with their size and type. In addition, the aerosol vertical distribution obtained from CALIPSO fulfills the requirement for the regional climate change study. In the present study, AOD, *α*, SMF and AI obtained from above multi-satellite observations are considered to distinguish the dominant natural and anthropogenic aerosols during different seasons. Figure 8 shows the multi-satellite observed AOD, *α*, SMF and AI over the study area during 2006–2007. Open circles represent the parameters obtained from MODIS-Aqua and filled circles are MODIS-Terra observations. AI gives information about the dust aerosols while SMF provides information about the anthropogenic and natural aerosols. Low SMF with low *α* indicates the presence of natural aerosols and the reverse represents the dominance of anthropogenic aerosols. AI has large values during Mar-Jul, whereas, *α* and SMF have low values. These combined observations suggest the abundance of coarse dust aerosols during these periods. OMI captures many dust storms over Thar Desert during premonsoon season in the AI images and enhancement of AI is due to transport of the dust plume from the desert region. During Aug-Feb, SMF is found to be high and AI is found to be very low indicating less abundance of dust aerosols in the atmosphere. In addition, *α* is also found to be very low during Aug-Sep but AOD is significantly high. Earlier studies from chemical composition of aerosol samples collected at Mt. Abu reported the enhancement of seasalt aerosols transported from Arabian sea during these periods and ground-based lidar observations at Ahmedabad, located 300 km to the south of Mt. Abu, reported the existence of a layer of seasalt aerosol in between 2–4 km [19, 64]. It can thus be inferred that the increase of monsoonal AOD is due to transport of seasalt aerosols from Arabian Sea. During Oct-Feb, high values of *α* are found indicating enhancement of anthropogenic aerosols. Ground-based observations show high abundance of BC on the hill-top region during winter. All these observations suggest that dust is dominating during premonsoon, anthropogenic aerosols during winter and natural seasalt are present in the atmosphere during monsoon season.

**Figure 8.** Space-borne daily observations of AOD, Ångstr*o*¨m exponent, small mode fraction (SMF) obtained from MODIS onboard Terra and Aqua satellites and aerosol index obtained from OMI onboard Aura satellite during 2006 and 2007.

## **5.2. Aerosol vertical profile**

18 Will-be-set-by-IN-TECH

asl), in central Himalayas, mean BC was observed to be 1.36±0.99 *<sup>μ</sup>*g.m−<sup>3</sup> during December

The diurnal variation of BC mass concentration does not show any significant morning and nocturnal peaks like other urban regions. However, increased BC was observed during the noon hours except during November and December. The reason for such an increase is during the day time the thermal convection becomes stronger and as a result, the pollutants at the foothill area rise up to the hilltop region and enhance the BC concentration. This day time enhancement was prominent during winter and postmonsoon because during these seasons there is a large difference between the day and night time temperatures. During November and December the night time BC concentration was larger by a factor of two. During these months the nearby villagers burn wood and fallen leaves to keep themselves warm thereby increasing the BC mass concentration. During January this nocturnal enhancement was not observed. The reason is that the boundary layer height is less than the station altitude and the night time BC that is produced cannot reach the hill top region due to weak thermal

In the current satellite era, large databases are available to study aerosol properties from space, both in the regional and the global scale, that are essentially demanding. For the present study, Terra and a series of satellite sensors flying on the A-train platform provide the required data. MODIS on board Terra and Aqua provide aerosol parameters in the morning and afternoon. OMI on board Aura satellite provides AI. The joint information of AOD, Ångstr*o*¨m exponent (*α*) and small mode fraction (SMF) retrieved from MODIS and AI retrieved from OMI can be utilized to estimate the optical properties of aerosols with their size and type. In addition, the aerosol vertical distribution obtained from CALIPSO fulfills the requirement for the regional climate change study. In the present study, AOD, *α*, SMF and AI obtained from above multi-satellite observations are considered to distinguish the dominant natural and anthropogenic aerosols during different seasons. Figure 8 shows the multi-satellite observed AOD, *α*, SMF and AI over the study area during 2006–2007. Open circles represent the parameters obtained from MODIS-Aqua and filled circles are MODIS-Terra observations. AI gives information about the dust aerosols while SMF provides information about the anthropogenic and natural aerosols. Low SMF with low *α* indicates the presence of natural aerosols and the reverse represents the dominance of anthropogenic aerosols. AI has large values during Mar-Jul, whereas, *α* and SMF have low values. These combined observations suggest the abundance of coarse dust aerosols during these periods. OMI captures many dust storms over Thar Desert during premonsoon season in the AI images and enhancement of AI is due to transport of the dust plume from the desert region. During Aug-Feb, SMF is found to be high and AI is found to be very low indicating less abundance of dust aerosols in the atmosphere. In addition, *α* is also found to be very low during Aug-Sep but AOD is significantly high. Earlier studies from chemical composition of aerosol samples collected at Mt. Abu reported the enhancement of seasalt aerosols transported from Arabian sea during these periods and ground-based lidar observations at Ahmedabad, located 300 km to the south of Mt. Abu, reported the existence of a layer of seasalt aerosol in between 2–4 km [19, 64]. It can thus be inferred that the increase of monsoonal AOD is due to transport

2004 [55]. This shows that Mt. Abu is less affected by anthropogenic activities.

convection. During this period hill top region becomes pollutant free region.

**5. Satellite observed aerosol properties over Mt. Abu region**

**5.1. Aerosol optical and physical properties**

Seasonal variation of aerosol vertical profiles over the study region is obtained from CALIPSO observations. Figure 9 shows the seasonal variation of aerosol extinction coefficient (km−1). The horizontal dotted line at 1.7 km represents the height of Mt. Abu. The extinction coefficient is directly proportional to the total aerosol loading. It is clearly seen from the figure that aerosol loading over Mt. Abu is minimum during winter and higher during other seasons. There is a peak found near 2.2 km altitude during monsoon which becomes weak during postmonsoon. Ganguly et al. [19] reported that this peak is due to seasalt aerosols transported from Arabian sea and chemical analysis also supports this result showing significantly high amount of seasalt present over Mt. Abu during monsoon [64]. During premonsoon, there is a peak at 4.2 km which is due to the transported dust layer. MODIS and OMI observations also indicate significant amount of dust present in the atmosphere.

Near surface region also shows high extinction coefficient values. This could be due to locally produced anthropogenic aerosols. In the present study, the properties of aerosols at the

are coming from factories and vehicular emissions. However, there is significant contribution from marine sources as di-methyl sulphate. On the other hand, BC is mainly anthropogenic, but it becomes natural when produced during natural forest fires. In the present study, dust and seasalt are considered as natural aerosols and BC, sulphate and nitrates as anthropogenic aerosols. BC is obtained from ground-based measurements using Aethalometer. Other aerosols like dust, sulphate and nitrates are obtained from the chemical analysis of aerosols samples collected over this hill-top region [39, 40]. These chemical compositions are used as input to the OPAC model to obtained aerosol optical properties and compared with measured values. OPAC model is also used to distinguish the natural and anthropogenic aerosols by separating the natural and anthropogenic components. A scatter plot of monthly averaged AODs obtained from Microtops observations and OPAC model is shown in Figure 10. The solid line represents the 1:1 line. Model derived and observed AODs are linearly varying with a slope of 0.90 and very close to the 1:1 line which indicates that the model derived AOD are very close to the observed values. However, the model is underestimating the AOD by about 10%. This is due to the cut-off radius of aerosols at 7.5 micron considered by the model, but in reality, aerosols are larger, especially over semi-arid regions, though their residence period is

Natural vs Anthropogenic Background Aerosol Contribution to the Radiation Budget over Indian Thar Desert 101

only for a few hours and their contribution towards optical depth is small.

**6.2. Source identification of natural and anthropogenic aerosols**

region during premonsoon increasing natural dust aerosols.

**7.1. Seasonal variation of aerosol radiative forcing**

**7. Natural vs anthropogenic background aerosol radiative forcing**

Aerosol radiative forcing is estmated using SBDART model considering aerosol optical properties obtained from OPAC, aerosol vertical profile from CALIPSO and MODIS surface reflectance. Aerosol radiative forcings in different seasons are given in table 2. Aerosol radiative forcing is found to vary from -3.2 to +0.2 Wm−<sup>2</sup> at TOA and from 6.1 to 23.6 Wm−<sup>2</sup> within the atmosphere. Aerosol radiative forcing at TOA is found to be maximum of about 0.2±2.5 Wm−<sup>2</sup> during premonsoon, followed by -1.3±0.5, -2.7±1.6, and -3.1±1.3 Wm−<sup>2</sup> during monsoon, winter and postmonsoon, respectively. Forcing within the atmosphere is maximum of about 23.6±5.5 Wm−<sup>2</sup> during premonsoon, followed by 12.5±3.9, 7.4±1.8, and 6.1±1.8 Wm−<sup>2</sup> during monsoon, postmonsoon and winter, respectively. Annual mean aerosol forcing at Mt. Abu is found to be 8.7±3.4 Wm−<sup>2</sup> which is lower than other urban

Seven days air parcel back trajectories are considered to identify the possible source regions of the natural and anthropogenic aerosols at Mt. Abu. The back-trajectories during premonsoon and winter are shown in Figure 11(a) and (b), respectively. Air parcels are mainly coming from IGP during winter and the heights of the trajectories are within 2 km. Ground based observations show that BC values at Mt. Abu are higher during winter and it is also clearly seen that there is long-range transportation of anthropogenic aerosols like BC from IGP within the boundary layer height. On the other side, air parcels are direction during premonsoon season. The heights are also greater than 3 km. Earlier chemical analyses report that dust concentration during this season is maximum of about 80% (in mass) of the total aerosols [39]. Therefore, one can easily conclude by these trajectories that the source of these dust aerosols is the nearby desert region. The back-trajectory analysis indicates that there is significant contribution of IGP during winter enhancing anthropogenic aerosols and that by nearby arid

**Figure 9.** Seasonal variation of vertical distribution of aerosol extinction coefficient obtained from CALIPSO observations. The horizontal dashed line at 1.7 km is the altitude of Mt. Abu.

hill-top region are considered and defined as the 'background aerosols'. The vertical profile of aerosols indicate that these background aerosols are less influenced from these locally produced anthropogenic aerosols. Therefore, the aerosol properties observed over Mt. Abu are assumed to represent those of the background aerosols over semi-arid region of western India.

## **6. Natural and anthropogenic background aerosol properties**

## **6.1. Estimation of natural and anthropogenic aerosols**

The estimation of natural and anthropogenic aerosols over this background site is a challenging task because many aerosol compositions have both origins. For example, sulphates are mainly considered as anthropogenic components over urban regions as they 100 Atmospheric Aerosols – Regional Characteristics – Chemistry and Physics Natural vs Anthropogenic Background Aerosol Contribution to the Radiation Budget over Indian Thar Desert <sup>21</sup> Natural vs Anthropogenic Background Aerosol Contribution to the Radiation Budget over Indian Thar Desert 101

> are coming from factories and vehicular emissions. However, there is significant contribution from marine sources as di-methyl sulphate. On the other hand, BC is mainly anthropogenic, but it becomes natural when produced during natural forest fires. In the present study, dust and seasalt are considered as natural aerosols and BC, sulphate and nitrates as anthropogenic aerosols. BC is obtained from ground-based measurements using Aethalometer. Other aerosols like dust, sulphate and nitrates are obtained from the chemical analysis of aerosols samples collected over this hill-top region [39, 40]. These chemical compositions are used as input to the OPAC model to obtained aerosol optical properties and compared with measured values. OPAC model is also used to distinguish the natural and anthropogenic aerosols by separating the natural and anthropogenic components. A scatter plot of monthly averaged AODs obtained from Microtops observations and OPAC model is shown in Figure 10. The solid line represents the 1:1 line. Model derived and observed AODs are linearly varying with a slope of 0.90 and very close to the 1:1 line which indicates that the model derived AOD are very close to the observed values. However, the model is underestimating the AOD by about 10%. This is due to the cut-off radius of aerosols at 7.5 micron considered by the model, but in reality, aerosols are larger, especially over semi-arid regions, though their residence period is only for a few hours and their contribution towards optical depth is small.
