Altitude profiles of CCN characteristics across the Indo-Gangetic Plain prior 5 to the onset of the Indian summer monsoon

Abstract. Concurrent measurements of the altitude profiles of cloud condensation nuclei (CCN) concentration, as a function of supersaturation (ranging from 0.2 % to 1.0 %), and aerosol optical properties (scattering and absorption coefficients) were carried out aboard an instrumented aircraft across the Indo-Gangetic Plain (IGP) covering coastal, urban and arid environments, just prior to the onset of the Indian summer monsoon (ISM) of 2016, under the aegis of the SWAAMI - RAWEX campaign. In general, the CCN concentration has been highest in the Central IGP, decreasing spatially from east to west above the planetary boundary layer (PBL), which is ~ 1.5 km for the IGP during pre-monsoon. Despite of this, the CCN activation efficiency at 0.4 % supersaturation has been, interestingly, the highest over the eastern IGP (~ 72 %), followed by the west (~ 61 %), and has been the least over the central IGP (~ 24 %) within the PBL. In general, higher activation efficiency is noticed above the PBL than below it. The Central IGP showed remarkably low CCN activation efficiency at all the heights, which appears to be associated with high black carbon (BC) mass concentration there, indicating the role of anthropogenic sources in suppressing the CCN efficiency. First ever CCN measurements over the western IGP, encompassing "The Great Indian desert", show high CCN efficiency, ~ 61 % at 0.4 % supersaturation, indicating hygroscopic nature of the dust. The vertical structure of CCN properties is found to be airmass-dependent; with higher activation efficiency even over the central IGP during the prevalence of marine airmass. Precipitation episodes seem to reduce the CCN activation efficiency below cloud level. An empirical relation has emerged between the CCN concentration and the scattering aerosol index (AI), which would facilitate prediction of CCN from aerosol optical properties.



5
Cloud nucleating ability of aerosols is fundamental in understanding the aerosol-cloud interactions (ACI) and associated feedback processes, which are complex in nature and pose a major challenge in quantifying the indirect climate forcing of aerosols (Boucher et al., 2013;IPCC 2013). Cloud Condensation Nuclei (CCN) form a sub-set of atmospheric aerosols (also known as Condensation Nuclei, CN) and take part in cloud processes, accelerate the condensation of water vapour leading 10 to the formation of liquid cloud droplets and modify the microphysical properties of clouds depending on the number size distribution, chemical composition, and mixing state of aerosols (Dusek et al., 2006;Farmer et al., 2015;Zhang et al., 2017). Several investigators have examined temporal and spatial distribution of the CCN properties and their processing by non-precipitating clouds over both continental and marine environments (Hoppel et al., 1973, Hudson and Xie, 1999, 15 Jurányi et al., 2011, Paramonov et al., 2015, Schmale et al., 2018. Significant variability in the CCN activation efficiency has also been reported over regions influenced by urban (Sotiropoulou et al., 2007) and industrial emissions (Asa-Awuku et al., 2011). Efforts have also been made to infer or predict CCN properties based on aerosol concentration and optical properties (for example, Jefferson, 2010;Liu and Li, 2014). However, due to the region-specific and heterogeneous nature 20 of the composition of aerosols, their chemical interactions, vertical mixing and advection to long distances, significant uncertainties still persist in characterising the CCN activation efficiency, especially its region-specific nature and altitude variation in the realistic atmosphere (Zhang et al., 2017). The information on the vertical distribution of the CCN number concentration, CCN efficiency and its variation with supersaturation are some of the vital parameters needed in 25 quantifying the ACI. In-situ measurements of the vertical distribution of the CCN activity especially over polluted regions are very important in accounting for the ACI in climate models (Li et al., 2016).
In the above context, the importance of South Asian region is unequivocal. Aerosol 5 physicochemical properties show large spatio-temporal variation over this region owing to the diverse source influence, both natural and anthropogenic, which show large seasonality (Jethwa et al., 2005) and dependence on large-scale meteorology (Lawrence and Lelieveld, 2010;Babu et al., 2013;Nair et al., 2016). Even within South Asia, the Indo-Gangetic Plains (IGP) fall under those regions in the globe where very high aerosol loading persists almost throughout the year (Di 10 Girolamo et al., 2004) and also depict a steady increasing trend in the Aerosol Optical Depth (AOD) (Dey and Girolamo, 2011;Babu et al., 2013), increasing surface dimming (Padmakumari et al., 2007;Badrinath et al., 2010), and enhanced mid tropospheric warming (Satheesh et al., 2008). Through modelling efforts, Vinoj et al., (2014) have shown possible linkages of West Asian dust loading over the Arabian Sea with the Indian summer monsoon (ISM). The competing roles 15 of natural (mostly mineral dust and marine aerosols) and anthropogenic aerosols over this region and their high seasonality, aided by the large-scale industrial and agricultural activities in this region and its particular orography makes the IGP one of the best natural laboratories for investigating the complex aerosol impacts on clouds and precipitation .
Despite these, characterisation of the vertical structure and the spatial variability of the CCN 20 characteristics across the IGP remains quite limited, except for some recent efforts using instrumented aircraft during the summer monsoon season under the Cloud Aerosol Interaction and Precipitation Enhancement Experiment (CAIPEEX) (Prabha et al., 2012;Konwar et al., 2014;Padmakumari et al., 2017). A few ground-based measurements also exist scattered across the subcontinent (Bhattu and Tripathy, 2014;Gogoi et al., 2015;Jayachandran et al., 2017;Singla et al., 25 2017).
In light of the above, and with a view to understand the ACI and its linkage to the ISM, an experimental campaign was undertaken under the aegis of SWAAMI (South-West Asian Aerosol -Monsoon Interactions) and RAWEX (Regional Aerosol Warming Experiment), executed jointly by the Indian Space Research Organisation (ISRO) and the Ministry of Earth Sciences (MoES) of 5 India, and the Natural Environment Research Council (NERC) of the UK. Under this, concurrent and collocated airborne measurements of the vertical structure of the CCN characteristics and aerosol scattering and absorption coefficients were carried out across the IGP, just prior to the onset of the ISM. The campaign was planned to quantify the vertical distribution of total aerosols (CN) and CCN concentrations at different supersaturations and its spatial variation across the IGP, 10 just prior to the onset of ISM, when different aerosol types are known to co-exist over this region.
The data are analysed to understand the altitude distribution of CCN characteristics, its activation efficiency and its relationship with scattering and absorption properties of aerosols, and the variation of those from west to east across the IGP. The campaign details along with the measurement protocols are given below, followed by the results and discussions.  1.3 ± 0.5 km for BBR, VNS, and JDR, respectively (Vaishya et al., 2018). Due to the unpressurised 15 mode of operation of the aircraft, the ceiling altitude of airborne measurements was ~ 4 km a.m.s.l.
In all, 14 sorties were made, 5 from each base station, except from JDR where only 4 sorties were made. Each sortie was for a period of ~ 3.5 hours, during which, the measurements were made at six altitude levels -~ 500, 1000, 1500, 2000, 2500, and 3000 m above the ground level (a.g.l.), following the 'staircase pattern' shown in Figure 2b (Babu et al., 2016). Accordingly, after takeoff, the aircraft climbed to the first level (500 m a.g.l.), stabilized the attitude and flew at that level for 5 ~ 30 minutes during which it covered a horizontal distance of ~ 150 km; before climbing up to the next higher level and retracing the path. This procedure was repeated until the highest level (ceiling altitude) was reached, after which the aircraft descended to the base. The sorties were repeated on consecutive days, except that on each day the aircraft proceeded to a different radial direction from the base, as shown in Figure 2a, so that the five sorties together provided a gross picture of the

Measurements
Ambient air was aspirated to the instruments using a solid diffuser inlet (University of Hawaii) maintained at isokinetic flow conditions, as detailed in Babu et al., (2016), with a volumetric flow rate of 70 LPM (litres per minute), for the average cruising speed of 300 km h -1 of the aircraft. The efficacy of the inlet to sample aerosols below 4 μm, under such conditions, has been demonstrated during the DC-8 Inlet Characterization Experiment (McNaughton et al., 2007). Further details of 5 the experiment setup is explained in Babu et al., (2016) and Vaishya et al., (2018). The air, aspirated through this inlet, is then fed to different instruments through a manifold. Aerosol instruments onboard were calibrated prior to and after the campaign to ensure consistency in the measurements. Concurrent time and space coordinates were logged continuously using a highresolution global positioning system (GPS).

10
CCN concentration at different supersaturations was measured at every second using a continuous flow CCN counter (CCN-100 model by Droplet Measurement Technologies), by feeding the aspirated air continuously to the cylindrical column of the counter at a constant flow rate of 0.5 LPM, where it is exposed to desired supersaturations. Details of the principle of operation of the CCN counter are available elsewhere (Roberts and Nenes, 2005;Lance et al., 2006). Aerosols, 15 according to their composition and size, having a critical supersaturation less than the effective supersaturation inside the column, will spontaneously grow into a droplet as they exit the column.
These droplets are counted with an optical counter using a laser of 650 nm wavelength. During each set of measurements, the supersaturation was varied through 0.2, 0.3, 0.4, 0.7, and 1.0% over a cycle of 30 minutes, and the cycle is repeated at each altitude level so that a complete CCN 20 spectra (of CCN vs supersaturation) is available at every altitude level. In the present study, the CCN concentrations never exceeded 5000 cm -3 , and hence the correction for water vapour depletion (Lathem et al., 2011) is not applied. Pressure correction was done to the set supersaturation at each altitude layer depending upon the change in pressure between ambient and calibration pressure (Lance et al., 2009). Data points during supersaturation transition are excluded 25 due to the inherent ambiguity in the stability of the attained supersaturation. The measured CCN concentration has a maximum uncertainty of 10 % (Rose et al.,2008).

Total aerosol number (CN) concentration was measured using an Ultrafine Condensation Particle
Counter (Model 3776, TSI), developed by Stolzenburg and McMurry, (1991). It measures CN of diameter 2.5 nm and above, with a time base of 1 minute. The aspirated air is continuously fed at 5 1.5 LPM, mixed with clean sheath air, which is saturated with butanol vapour while passing through a saturator. The resultant flow is passed through a condenser where a sudden cooling result in the condensation of butanol vapour onto aerosols due to supersaturation and the droplets are counted using a counter working with a laser diode at 650 nm. Further details of the instrument and its adaptability for aircraft-based experiments are explained by Takegawa et al., (2017). 10 Aerosol absorption measurements at 7 different wavelengths (370, 470, 520, 590, 660, 880, and 950 nm) were carried out using a dual spot Aethalometer (AE 33 model of Magee Scientific) (Drinovec et al., 2015) which works on the principle of filter-based optical attenuation technique (Hansen et al., 1984). Filter loading artifact of the instrument is corrected in real time as explained by Drinovec et al., (2015). Absorption measurements were corrected for change in flow rate at 15 high altitudes following Moorthy et al., (2004). Optical attenuation at 880 nm is used to estimate the black carbon (BC) mass concentration using the specific absorption cross section value (7.77 m 2 g -1 ). The Integrating Nephelometer (3563 model of TSI) measured the scattering coefficient (σsca) at 450, 550, and 700 nm wavelengths. Scattering measurements were corrected for nonlinearity in the angular truncation error following Anderson et al., (1998). 20 For the CCN data analysis, initial five minutes of data at each altitude level were discarded considering the stability of the measurements and the data was averaged for every minute. Hence a minimum of 20 minutes of usable data comprising 5 supersaturations is available for each altitude level. CN, spectral scattering, and spectral absorption measurements were also synchronized to the 1-minute averaged CCN data. Thus, for each region (East, west, and central IGP), 5 vertical 25 profiles of CCN and CN concentrations, and scattering and absorption coefficients were obtained.

Vertical distribution of CN and CCN
Vertical profiles of CN and CCN concentrations (at 0.4% supersaturation) for the three sub-regions 5 of the IGP are shown in Figure 3. Each profile is an average of all the sorties carried out from the base station. Significant differences are seen below ~ 1.5 km, which represents the well-mixed region within the PBL, and are attributed to the sub-regional scale emissions. As such, the CN concentrations are up by nearly a factor of 2 at the Central IGP (VNS) compared to the eastern or western ends of the IGP; owing to the large-scale anthropogenic activities in the central IGP. 10 Beyond ~ 2 km altitude, the CN concentrations remain quite comparable in magnitude, across the entire IGP with similar vertical variations.
In contrast to this, there is a significant difference in the aerosol type across the IGP (attributable to the source-heterogeneity), as revealed by the CCN concentration in the right panel of the same  Irrespective of all these, the CCN concentrations remain high (1000 to >2000 cm -3 at 0.4% 10 supersaturation), even at 3 km altitude, which is above the base of monsoon clouds (Das et al., 2017). This will have strong implications in cloud modification, as has been established elsewhere (Andreae et al., 2004;Rosenfeld et al., 2008); however, their influence on the monsoon rainfall over the study region has not yet been quantified. Based on the aircraft observations during the CAIPEEX, over Hyderabad (17.45° N, 78.38° E) in southern India, Padmakumari et al., (2017) 15 reported the suppression of warm rain process due to the presence of high CCN concentration.
During the collaborative Regional Aerosol Warming Experiment (RAWEX) and the Ganges Valley Aerosol Experiment (GVAX), Gogoi et al., (2015) have reported CN and CCN (at 0.46% supersaturation) concentrations of ~ 2500 and ~ 1100 cm -3 , respectively, for June 2011, from a 5 high-altitude station (~ 2 km a.m.s.l.), Nainital in Central Himalayas. The high CN and CCN concentrations observed in this study is in line with values reported from Nainital, which is an optimal high-altitude site to study regional (IGP) as well as transported aerosol characteristics over the IGP. In another study over the Loess plateau in China during July-August months, Li et al., (2015) have reported high concentrations of CN and CCN; peaking within the PBL and decreasing decreasing spatially from the eastern IGP to the western IGP especially above the PBL.

Altitudinal dependence of CCN -CN association
Aerosol number-size distribution and composition are known to show vertical variations (Zhang et al., 2011;Li et al., 2015). Hence it is imperative to examine altitudinal dependency of CCN on 25 CN and its region-specific nature. In Figure 4,   To further investigate the above hypothesis of the role of local emissions in weakening the relationship between CCN and CN over Central IGP, the variation of CCN number concentration at 0.4% supersaturation with BC mass concentration is examined in Figure 5. For this, the concurrent BC mass concentration measurements carried out from the same platform is used.
Central IGP showed highest absorption coefficient (column averaged) of 26 ± 9 Mm -1 , followed 5 by the west (16 ± 2 Mm -1 ) and east (15 ± 3 Mm -1 ) IGP (Vaishya et al., 2018). It is interesting to note that the linear relationship is maintained for low to moderate concentrations of BC (up to around 1000 ng m -3 , which occurs mostly above PBL), while significant scatter occurs for higher values of BC (exceeding 2000 ng m -3 ), which occurs mostly in the lower altitudes, supporting the hypothesis. Similar deviations in CCN -CN relationship with respect to altitude has also been 10 reported by Srivastava et al., (2013) over the central IGP region, using aircraft measurements, where they attributed it to the impact of local anthropogenic emissions.

CCN spectra and parameterisation for different altitudes
Using the measurements of CCN number concentration as a function of supersaturation the mean CCN spectra are constructed, for different sub-regions of IGP, and is shown in Figure 6 for different altitudes. In addition to the regional distinctiveness in the CCN number concentrations seen in Figure 3, it is interesting to note the rapidly levelling off of the spectra with increasing 5 supersaturation, at the eastern IGP (represented by BBR, blue lines in Figure 6), especially above 1 km; in contrast to the other two regions, where the CCN concentrations keep on increasing with increasing supersaturation at all heights. This clearly demonstrates a change in the hygroscopicity of aerosols across the IGP, especially in the free-troposphere. To quantify this, the CCN spectra are parameterized by evolving a least square fit with Twomey's relation (Twomey 1959), where CCN (ss) is the number concentration of CCN at a particular supersaturation (ss), C and k are empirical coefficients. Lower k values are reported more frequent for marine airmass compared to continental airmass (Twomey and Wojciechowski, 1969;Khain, 2009). The fine mode anthropogenic aerosols exhibit high k values, while hygroscopic and larger aerosols like seasalt 15 have low k values (Hegg et al., 1991;Jefferson et al., 2010). The shape of the CCN spectra, represented by the 'k' values, showed significant altitudinal variations. The altitude variations in the CCN spectra, which can be due to the variations in aerosol number size distribution, will have an impact on the droplet size distribution of the warm cloud formation (Raga and Jonas., 1995).      ~ 20 % to 65% and ~ 0.2 to 0.7, respectively. Similar inverse association between CCN efficiency and k is reported by Hegg et al., (1991) and Jayachandran et al., (2017). High k values are due to the dominant presence of small or less soluble particles in the aerosol system, which in turn reduce the CCN efficiency. However, over central IGP, very low CCN efficiency (<20 %) were observed with low k values (~ 0.4), which is not in-line with the general inverse relationship. These cases 25 were observed within the PBL, indicating a CCN-inactive aerosol system even at high (>0.8 %) supersaturations. At high altitudes (>3 km) over the IGP, Srivastava et al., (2013) have reported aerosol size distribution peaking below ~ 40 nm due to new particle formation (NPF) events and cloud processing. Rose et al., (2017) have reported the significant role of NPF in CCN activation study, the role of cloud processing or in-cloud scavenging for low CCN efficiency and flat CCN spectra (low k) at cloud forming heights cannot be neglected. Based on measurements at the mean sea level and at 1 km above ground level,  have shown the vertical heterogeneity existing in CCN efficiency and CCN spectra during the ISM at the south coast of India. Li et al., (2015) have shown that the anthropogenic 15 influences can cause a strong variation in CCN efficiency from 10% to 70% from near ground level to about 4.5 km over China during Asian summer monsoon season. More than 50% of the aerosols are CCN active over the regions other than central IGP, which indicates the dominant role of natural aerosols in warm cloud droplet activation over the sub-continent region just prior to the ISM season. The airmass traversing through the polluted-continental regions is responsible for the 5 lowering of CCN activation efficiency at the free-troposphere heights over east IGP. The back trajectory analysis of airmass reaching at 500 m and 3000 m over BBR (figure not included) clearly showed that the particles reaching 3000 m have pure continental history of passing across the IGP from the arid regions of western India and West Asia, whereas those reaching at 500 m pass over oceanic region of Bay of Bengal before arriving at the location. This distinctiveness in the airmass 10 history at higher altitudes are also causing the scatter in CCN-CN association as seen in Figure 4. The significant influence of the nature of airmass on CCN activation over the Indian region is illustrated by the closure studies carried out by Srivastava et al., (2013) at various altitudes. Jayachandran et al., (2017) have reported higher CCN activation efficiency for marine airmass than continental from ground-based observations from peninsular India during the ISM. Within 15 the PBL including near to the ground level, CCN efficiency is very high over the east IGP (coast) which will support the cloud droplet formation with a sharp droplet size distribution.
One of the striking features emerging from this study is the high CCN efficiency over the arid region of Western IGP, which is reported for the first time. This region is known for its dust dominance (both locally generated and advected from the Middle East and Eastern Africa). 20 Though pure dust is water inactive, its CCN efficiency will enhance when coated or mixed with soluble salts like sulphates and nitrates. (Zhang et al., 2006;Kelly et al., 2007). Though Feingold et al., (1999) have shown that coarse mode dust aerosols can act as giant CCN and initiate drizzle formation, their number concentration is far less numerous, especially at high altitudes (Padmakumari et al., 2013). Thus, the observations of moderately high CCN activation efficiency, 25 lower values of k and higher concentration of CCN are interesting and need discussions. Figure 9 shows airmass back trajectories for five days and arriving at 500 m, 1500 m, and 3000 m a.m.s.l above (a) east IGP -BBR, (b) central IGP -VNS, and (c) west IGP -JDR. From Panel (c), it can be seen that the airmass reaching JDR (conducive for dust-advection) has significant history over the northwestern Arabian Sea, and hence would also carry significant moisture. It is known that 5 the presence of hygroscopic salt aerosols can catalyse the reaction of dust with acidic gases (Tobo et al., 2010), changing its hygroscopicity. Thus, the airmass reaching the desert region, having a strong marine component could enhance the activation efficiency of the aerosols. Strong convection at the lower atmosphere will also take salt aerosols to the atmosphere from the regional dry salt lakes. Begue et al., (2015) have reported CCN efficiency of ~70% for 0.2% supersaturation 10 over the Netherlands during a dust transport event due to the accumulation of solute particles on dust. The present study shows that about 66% of the total aerosols in the PBL of western IGP -JDR were activated as CCN at 1% supersaturation. At BBR, there have been two episodes of pre-monsoon precipitation on 4, 5 June 2018 (much before the sortie timings), with accumulated rainfall of 58 and 8 mm. The vertical profile of CCN activation efficiency over BBR averaged for measurements before and after rainfall is shown in bold and dotted lines, respectively, in Figure 10(b). There is a decrease (though weak) in the 15 activation efficiency (especially below the cloud level, 2 km), after the precipitation, probably due to removal of hygroscopic aerosols by the precipitation. Even though the CCN efficiency found to be slightly reduced below 2 km, the CCN activation efficiency is found to be higher above 2 km compared to that of observations before the rainfall. Near the ground level, CCN concentration (mean ± standard deviation) reduced from 3431 ± 572 to 1320 ± 454 cm -3 and from 1755 ± 105 to 20 460± 209 cm -3 at ~3 km a.g.l. After the rainfall, a reduction (<10%) is seen in the CCN efficiency over BBR, meanwhile, there is a large diminution in the number concentration of CN and CCN. The theoretical framework of wet scavenging process accounts for nucleation, gravitational and inertial impactions, and turbulence scavenging mechanisms (Pruppacher and Klett, 1997). 10 However, uncertainties and difficulties still exist in attributing the observational evidences of wet scavenging of aerosols to different scavenging mechanisms, especially in the case of moving air parcels. Efficiency of below-cloud scavenging (wash out) mainly depends on the number size distribution of both aerosols and raindrops, while the in-cloud scavenging (rain out) depends mainly on the solubility of the aerosols (Garrette et al., 2006). The decrease in CCN concentration 15 over BBR after the rainfall, and the high CCN efficiency seen in the present study indicates the highly soluble nature of aerosol system prevailing over the region. The difference in CCN activation efficiency at different altitude levels before and after rainfall reinstates the difference in the aerosol types at different altitudes. One of the possibilities for the observed CCN efficiency is 5 that the rainfall has removed coarser and hygroscopic particles by wet scavenging, resulting in the reduction of the CCN activation efficiency below 2 km. Cloud processing broadening the aerosol distribution as reported by Flossmann et al., (1987) may be enhancing the CCN activation efficiency above 2 km. However, the effect of cloud formation and further rainfall on CCN characteristics needs further investigation. The modification in CCN efficiency over VNS and 10 BBR underlines the role of type of airmass and rainfall in determining the vertical structure of CCN activation in a short duration.

CCN and aerosol optical properties
Concurrent measurement of aerosol scattering and absorption coefficients during the campaign provided an opportunity to examine possible links between CCN and the optical properties of 15 aerosols. Li, (2014), andJefferson (2010) have illustrated the potential of using aerosol optical properties as a proxy and prognostic variable for studying the CCN properties. Liu and Li, (2014) have used the scattering aerosol index (AI), which is the product of scattering coefficient  A scatter between the Extinction aerosol index and CCN concentration at 0.4% supersaturation is generated and shown in Figure 12. If absorption contributed insignificantly to the extinction, then this plot would not differ significantly from Figure 11. However, it can be seen in Figure 12 that 20 there is a significant reduction in the slope over western and Central IGP (JDR and VNS). This indicates the reduction in CCN activation due to absorbing aerosols, probably dust. However, there is no remarkable change in the slope over BBR, which might be due to the reduced concentration 5 of dust (as most of it get removed as dust is advected across the IGP and also due to mixing of dust with other more hygroscopic aerosol species as it gets aged in the atmosphere). There is an increase in correlation coefficient over east IGP when we consider aerosol absorption also, which might be indicative of contribution of these aerosols to e to CCN activation; probably due to co-emitted or co-existing soluble inorganic particles. Examining Figure 11 along with the CN profile shown in Figure 3( (Vaishya et al., 2018). However, the slope at BBR is nearly half of that seen at JDR, despite it having the highest activation efficiency. On the similar lines, it appears that the size distribution of aerosols over VNS has more fine particles (higher Angstrom exponent, but lesser 5 activation efficiency). Thus, the size distribution and chemistry of the aerosol influence the relationship between scattering aerosol index and CCN concentration. This dependency is useful in developing empirical relationship connecting CCN and light scattering properties at least in a region-specific scale. The number concentration of Aitken mode aerosols, especially the aerosols at 60-100 nm range and its composition is the main factor in governing the variability in CCN 10 properties, while the relative dominance of accumulation mode aerosols will be determining the scattering properties. Figure 11 demonstrates the strong relationship that exists between the aerosol scattering properties and CCN concentration in the vertical column over the IGP. The relationship between CCN and aerosol optical properties further implied the use of satellite-retrieved AOD products in the region, which are now matured and fairly accurate, and model-generated aerosol 15 profile aided by ground / space-based lidar, in predicting CCN..

Conclusions
Extensive characterisation of the altitude distribution of CCN and its spatial variation across the IGP has been carried out, for the first time, using in-situ measurements aboard an instrumented aircraft just prior to the onset of the Indian summer monsoon (ISM). The results concluded below 20 form a significant step towards characterisation/understanding the ACI during the Indian Summer Monsoon, though the impact on cloud microphysics needs further investigation.
• Spatial heterogeneity in total aerosol concentration exist over the IGP with high concentrations (>13000 cm -3 ) over the central IGP (near to the ground level) and the least over the western IGP while its vertical variation remain the same above the planetary 25 boundary layer (PBL) at all regions.
• High CCN concentration (above 1000 cm -3 at 0.4% supersaturation) is observed up to 2.5 km across the IGP, indicating significant possibility of aerosol indirect effects.
• Central IGP shows higher CCN activation efficiency above the PBL(>1.5 km), than within, 5 despite the latter having high CN and CCN concentrations indicating activation of aerosols as CCN is suppressed by freshly emitted aerosols, mostly from anthropogenic sources.
• High CCN activation efficiency, ~61% at 0.4% supersaturation, at ~1.5 km above the ground level is observed over the dust dominated western IGP. This high CCN activation efficiency of dust aerosols can modify the cloud microphysics over the region, hence 10 affecting the precipitation pattern as well as the regional radiation balance.
• It is seen that while precipitation reduces the CCN activation efficiency below cloud level, advection of marine airmass enhances CCN efficiency, even over arid regions.
• An empirical relationship between the CCN activation and optical properties of aerosols has further implied the use of satellite-retrieved AOD products and model-generated 15 aerosol profile aided by ground / space-based lidar, in predicting CCN over the region.

Data availability
Data are available upon request from the contact author, S. Suresh Babu (s_sureshbabu@vssc.gov.in). 20 The authors declare that they have no conflict of interest.

Author contributions
SSB, SKS and KKM conceptualized the experiment and finalized the methodology. SSB, VJ, AV and MMG were responsible for the data collection onboard aircraft. VJ carried out the scientific analysis of the data supported by SSB, VSN and AV. VJ drafted the manuscript. 25 SSB, KKM and SKS carried out the review and editing of the manuscript.

Acknowledgments
Interaction -Regional Aerosol Warming Experiment) campaign. We thank Director, National