Understanding the Chemistry and Sources of Precipitation Ions in the Mid-brahmaputra Valley of Northeastern India

The chemistry of rainwater over mid-Brahmaputra Valley was studied for three consecutive years (2012–2014; n = 285). The samples were analyzed for major chemical parameters viz. pH, electrical conductivity (EC), and ions (SO4, NO3, Cl–, F–, Br–, Ca2+, NH4, Mg2+, Na+, K+, and Li+), organic acids (HCOO– and CH3COO) and dissolved organic carbon (DOC). The mean pH for the entire study period was found to be 5.66, which ranged from 4.51 to 7.68, and the volume weighted (VW) mean pH was found to be 5.16. Over 55% of the samples showed pH between 5 and 6, and a few samples had pH<5. Ionic concentration followed the order NH4 > Ca2+ > SO4 > NO3 > Cl– > Na+ > K+ > Mg2+ > H+ > HCO3 > Br– > F– > Li+, indicating dominance of alkaline ions over acidic ions such that 94% of mineral acid was neutralized. The secondary ions, NH4, SO4, and NO3, showed high wet deposition fluxes. The chemistry exhibits explicit seasonality. The airmass clusters of monsoon and non-monsoon seasons, and the associated chemistry varied, which showed influence of long-range transport. The interspecies correlations varied between the monsoon and non-monsoon time samples meaning variation in the source strengths of the contribution sources of the chemical species of the rainwater. Positive Matrix Factorization (PMF) was applied to the data which extracted six factors that explained the sources and chemistry of the rainwater constituents which are of sea, agriculture, coal burning, biomass burning, and secondary origin.


INTRODUCTION
For many decades, rainwater chemistry has drawn significant attention because of its acidity and 34 the associated consequences on ecosystems. Researchers from various parts of the world have 35 extensively studied rainwater chemistry to gain understanding of this issue (Eriksson, 1952; 36 Charlson and Rodhe, 1982; Zeng and Hopke, 1989;Galloway et al., 1982;Likens et al., 1976;   ground) and the collector height was maintained at 1m above the resting surface to avert 94 contamination of samples from splashes as per Rastogi and Sarin (2005). 95 Measurements of pH, electrical conductivity (EC), and volume were taken immediately in the 96 laboratory after each rain event. It was then filtered to remove any insoluble materials. Two 97 aliquots were stored for further analyses. One was used for the determination of anions and the 98 other was for cations. It was treated with chloroform to minimize microbial degradation of ions 99 such as ammonium and stored at 4°C (GAW, 2004; Keene et al., 1983).

A C C E P T E D M A N U S C R I P T
5 For the analysis of carbon, filtered samples were stored in brown vials at 4 o C to avert degradation 101 of organics till analysis (Willey et al., 2000;Campos et al., 2007;Kieber et al., 2002). This 102 campaign included 285 rain events (2012= 116, 2013=102, and 2014=67).

A C C E P T E D M A N U S C R I P T
6 Quality assurance/quality control (QA/QC) procedures were followed for the precipitation 124 chemistry monitoring. Samples contaminated with direct leaf fall or bird droppings were excluded.

125
The bottles and funnels were pre-washed with ultrapure water (Ultrapure Type 1, Simplicity, 126 Millipore, resistivity 18.2MΩ.cm) before sampling. The collectors were deployed just before the 127 start of rain and removed immediately after the rain stopped. Field blanks were evaluated 128 frequently and analyzed for all the parameters by the same procedures as the rain samples. Field 129 blanks were taken by pouring ultra-pure water into the sampler and following the procedures as 130 previously described. The sample values were adjusted by subtracting the field blank values.

131
Replicate measurements were also conducted for the samples that showed RSD≤ 5%. In case of 132 pH, the difference between replicate measurements were found to be less than 0.05 while for EC, 133 they were less than 1µS/cm.

134
The ion balance was evaluated by linear regression analysis between sum of anions and sum of 135 cations. The result showed a good correlation between total anions and cations (r =0.876), which 136 was significant at p<0.01 (Fig. 2). The slope (m=1.546) deviated to the higher side of ideal 1:1 137 relationship, which would mean that more anions were missed from the analyses, e.g. phosphate, 138 organic acids, etc. Also, in the assessment of ion balance, bicarbonate (HCO3 -) and hydrogen ion 139 (H + ) concentrations were also considered by indirectly estimating their concentrations from the 140 pH values. HCO3was estimated using the relationship between pH and HCO3 - (Granat, 1972)

A C C E P T E D M A N U S C R I P T
7 pHVWM = -log ( Where pHVWM is the volume weighted average pH, and pHi and Vi are the value of pH and volume 148 of the sample i, respectively. The volume weighted mean concentrations (VWM) of ionic 149 constituents were calculated using Eq 3: Where Ci is the ionic concentration of each ion in µeqL -1 , Pi is the volume of each rainy event in The arithmetic and volume weighted (VW) mean pH and EC of rainwater observed during the 177 study period are given in Supplementary Table S1. The mean pH for the entire study period was been illustrated in Fig.3 showing that most rain events were in the pH range of 5.5 to 6.0, which 182 pointed at near the alkaline nature of the rainwater.

198
The monthly variations in pH against volume rainwater are shown in Fig. 4. The months that 199 experienced very low rainfall volume and/or rain events that came after a long spell of dry days  The VW-mean EC was 16.22 µS cm -1 and the arithmetic mean was 29.16 µS cm -1 . The monthly 207 variation of EC against volume of rainwater is shown in Fig. 4. The EC of the rain events decreased 208 with the increasing rainfall volume. Thus, after continuous, heavy rainfall, the particles and gases 209 contributing to the EC of rainwater decreased because of 'washing and cleaning' the atmosphere. The equivalence ratio, sum of the anions to sum of the cations (∑Anion/∑Cation) and the indicator 213 of the completeness of measured major constituents (Wang and Han, 2011), was found to be 0.78, 214 suggesting that most of the major anions and cations were measured. It was observed that the 215 arithmetic means were greater than the VW mean concentrations, which might mean that higher 216 concentrations are usually associated with lower volume of precipitation as reported by (Xiao,217 2016). On average, the ionic representation of rain samples followed the order of 218 NH4 + >Ca 2+ >SO4 2->NO3 ->Cl -> Na + >K + >Mg 2+ >H + >HCO3 -> Br ->F ->Li + .

219
A comparison of the concentrations with the reported studies from elsewhere in India has been put 220 up in Table S1. It was observed that the VWM concentration of acid buffering ions (i.e. Ca 2+ , Mg 2+ , to the current results. The present study found that alkaline ions (Ca 2+ , NH4 + ) were dominant during 227 the entire study period. Ca 2+ and NH4 + contributed 18% and 17%, respectively, while SO4 2and 228 NO3provided 12% and 11%, respectively, to the total ion budget (Fig. 2). So, SO4 2acted as a 229 dominant acidic ion in the study region followed by NO3 -. This result would suggest that most of 230 the associations between these alkaline and acidic species might be in the form of neutral salts 231 such as (NH4)2SO4, NH4NO3, CaSO4, and Ca(NO3)2. The presence of Ca 2+ could be soil derived 232 from construction activities, and/or might be due to the long-range transport from more arid 233 regions. Cland Na + , markers of sea salt, also contributed averages of 18% and 8%, respectively, 234 to the total ion balance of rainwater during the entire study period.

11
To appreciate the influence of season in rainwater composition, the relationship between measured 236 Na + and Clions of each event with reference to seawater Cl -/Na + have been plotted in 237 supplemental Fig. S1. It was observed that the relationship was poor, except for the year 2014. or formation in the atmosphere during long-range transport could account for their presence (Deka 256 and Hoque, 2014b).

A C C E P T E D M A N U S C R I P T
Seasonal variability 259 Seasonal variations in the ionic composition were observed (Fig. S2). The seasonal variability in the composition of rainwater may also be related to long-range transport.

273
There is an explicit shift of directions from where the air mass trajectories that reached Tezpur 274 originate in different seasons. Trajectory clusters reaching Tezpur of Brahmaputra Valley at 500m 275 height above ground level for monsoon and non-monsoon period illustrated in Fig. 1

A C C E P T E D M A N U S C R I P T
13 monsoon period and those rains were associated with the clusters that originated from and/or 281 crossed the polluted regions like the IGP.

282
The calculated anthropogenic contribution of the chemistry was found to be much higher during 283 the non-monsoon period. The clusters 1 and 5 represent airmasses arriving from the polluted IGP 284 region and, therefore, the associated anthropogenic contributions were found to be maximum. It is 285 also evident that the contribution of dust was higher during the non-monsoon period. So, the 286 chemistry suggests that both local and trans-boundary movement of pollutants with the rain 287 bearing airmasses did influence the rainwater chemistry of the study area. The annual wet deposition along with the standard deviations and annual total rainfall collected 305 during the study period over the mid-Brahmaputra valley are given in Table S2. It has been 306 observed that SO4 2and NO3among the anions, and Ca 2+ and NH4 + among the cations were

A C C E P T E D M A N U S C R I P T
15 acidity caused by mineral acids was neutralized by alkaline species. There is a 6% gap in terms of 327 species that caused the rainwater acidification that could be due to the species other than NO3and 328 SO4 2-.

A C C E P T E D M A N U S C R I P T
17 and Fappears only in this factor. Thus, this factor can be attributed to emission from coal  Bris suspected to be of agricultural origin and methyl bromide (CH3Br) is the likely precursor. implies that most of the soil is not taking up acidic NO3and SO4 2-, but that process represents a 483 major mechanism for getting NO3and SO4 2into the precipitation.