Atmospheric acid deposition is an environmental concern worldwide, and the determination of its impacts in remote areas can be problematic [1]. Acidification is primarily related to the emissions of SO₂ and NO_x, since these gases are the precursors of major acids like H₂SO₄ and HNO₃. On the other hand, the neutralization of acidity in rainwater can be achieved either by CaCO₃ in airborne dust [2,3] and or by ammonia released from industrial, agricultural, and other natural sources [4].

The chemical composition of rainfall depends strongly on the atmosphere. The concentrations of chemical species in precipitation vary widely in relation to a multiplicity of factors, such as the type and distribution of aerosol sources, transport of aerosols, chemical species, and the scavenging processes of species. Meteorological factors are often considered to be the predominant factor [5-7] in relation to rainfall; the two main processes [8, 9] are described. The first process refers to the washout of the below-cloud atmosphere during precipitation events by raindrops which scavenge and dissolve particles and gases as they fall; the washout components are mainly of local/regional origin. The second process, called "rainout," corresponds to the condensation of water vapor on aerosol particles during the formation of cloud droplets and incorporation of gases surrounding the droplets by aqueous-phase reactions. This process corresponds mainly to the incorporation of long-range transported components which may be traced back to their origins by air masses trajectories.

The chemical composition of rainfall depends strongly on the atmosphere. The concentrations of chemical species in precipitation vary widely in relation to a multiplicity of factors, such as the type and distribution of aerosol sources, transport of aerosols, chemical species, and the scavenging processes of species. Meteorological factors are often considered to be the predominant factor [5-7] in relation to rainfall; the two main processes [8, 9] are described. The first process refers to the washout of the below-cloud atmosphere during precipitation events by raindrops which scavenge and dissolve particles and gases as they fall; the washout components are mainly of local/regional origin. The second process, called "rainout," corresponds to the condensation of water vapor on aerosol particles during the formation of cloud droplets and incorporation of gases surrounding the droplets by aqueous-phase reactions. This process corresponds mainly to the incorporation of long-range transported components which may be traced back to their origins by air masses trajectories.

2.1. Sampling Site

Rainwater samples were collected on a daily basis for one year (January-December 2008) using an automated wet only collector. Samples were collected at 9 AM local time and dispatched to the laboratory for analysis within a week. The sampling bucket was cleaned every day after sample collection using de-ionized water with conductivity <0.15 mS/m.

The rainwater samples were weighed to determine the amount of water and were measured for electro-conductivity (EC) and pH under a controlled temperature of 25.0℃ using conductivity and pH meters, respectively. The samples were filtered through 0.45 μm cellulose acetate filter paper and stored in the dark at 4℃ until ion analysis was performed. Inorganic cations (Na⁺, NH₄⁺, K⁺, Ca²⁺, and Mg²⁺) and anions (HCOO^-, CH₃COO^-, Cl^-, NO₃^-, and SO₄^2-) in the rainwater samples were analyzed by Ion Chromatograph (Metrohm, Herisau, Switzerland). Hydrogen ion (H⁺) concentrations were derived from pH values [16].

In this study, principal component analysis (PCA) was used to determine the various sources of ion composition of the wet depositions. A varimax rotation with Kaiser Normalization of Principle Components Analysis by SPSS program (version 14; SPSS Inc., Chicago, IL, USA) was used for the determination of factors underlying the inter-correlations between the measured species. Pearson correlation (r) was used for data analysis to identify relationships between ionic species in rainwater.

The total number of collected samples was 132 and precipitation volume was 1,286.7 mm. The frequency distribution of the rainwater pH values (n = 123 with precipitation >;4 mm) is illustrated in Fig. 2.

The total number of collected samples was 132 and precipitation volume was 1,286.7 mm. The frequency distribution of the rainwater pH values (n = 123 with precipitation >;4 mm) is illustrated in Fig. 2.

Monthly mean pH values are shown in Fig. 3. Acidic precipitation (pH < 5.6) was found during the months of March to April. Monthly mean pH values were lowest in March (pH = 4.78) and highest in November (pH = 6.50). In the dry season, high amounts of pollutants accumulated in the atmosphere due to lower levels of precipitation. Therefore, rain samples collected during the dry season contain high ionic concentrations, which presented low pH values.

The number of rainwater samples measured for EC was 123. EC values ranged from 0.18 ? 6.82 mS/m and the volume weight mean EC value was 0.62 mS/m. The month with the highest mean EC value was March, whereas the month with the lowest value was November (Fig. 4).

The EC value is positively correlated with the level of ionic contamination in precipitation. Therefore, samples with high EC values contain high ionic concentrations. Based on the dilution effect, the amount of precipitation is inversely related to the EC values.

The rainwater shows remarkable seasonal variations. Monthly average anion and cation concentrations, together with precipitation amount, are illustrated in Fig. 5.

The highest ion concentrations were found in March (dry season) due to the low amount of rain precipitation and long period of non-precipitation. On the other hand, low ion concentrations were found in the rainy season.

The order of anion and cation equivalents were NH₄⁺ (28%) >; Ca²⁺ (18%) >; SO₄^2- (14%) >; NO₃^- (13%) >; Cl^- (6%) >; CH₃COO^- (5%) = Mg²⁺ (5%) >; K⁺ (4%) = Na⁺ (4%) >; HCOO^- (3%). The cation and anion species with the highest percentage of their respective categories were NH₄⁺ and SO₄^2-. The ratio between total anions and total cations was 43:57 with units of μeq/L.

In order to get high precision and correct data, it is necessary to do quality control. Data quality of rainwater samples was checked by ionic balance (R₁) and conductivity balance (R₂). The data was rejected if it did not meet the quality criteria. R₁ and R₂ values illustrate the accuracy of ion analysis and EC measurements, respectively. The principle of electro-neutrality in precipitation water requires that total anion equivalents be equal to total cation equivalents. Ion balance in a precipitation samples can be calculated by the following equations [18];

Where C represents anion equivalents, C_ci is the concentration of i-th cation in μmol/L, Vi is the valence of the given ion.

Where A represents cation equivalent, C_Ai is the concentration of i-th anion in μmol/L.

This is a simplified from of the corresponding equation used by the US Environmental Protection Agency where the denominator is the average of the two sums.

For dilution solutions (e.g. below 10^-3 M), the total conductivity can be calculated in mS/m from the molar concentrations and molar conductivity (at infinite dilution) of the individual ions. The calculation is as follows:

Where λ_calc denotes the calculated conductivity of the solution (in mS/m), C_i represents the ionic concentration of the i-th ion (in μmol/L), and λ_i⁰ represents the molar conductivity (in S cm²/ mol) at infinite dilution and 25.0℃.

The calculation of conductivity values can then be compared to the observed value for precipitation samples with the relationship described in the equation below;

The total number of precipitation samples with complete measurements was 123 (100%). The number of samples that met the quality control criteria in terms of R₁ and R₂ were 62 (50.4%) and 96 (78.0%) respectively. Only 58 samples (47.2%) were qualified based on R₁ and R₂ combination criteria and therefore were used for statistical analysis.

Some base ions found in precipitation (e.g., Ca²⁺ and NH₄⁺) act as buffers for the acidity of rainwater. To estimate the neutralization capacity of each alkaline compound, the neutralization factor (NF) was calculated by the following equation [19]: NF_Xi = [Xi]/([SO₄^2-] + [NO₃^-]), where Xi is the chemical component of interest, with all the ions expressed in μeq/L. The corresponding values for the major cations are given in Table 1. NH₄⁺ was found to have the highest neutralization effect, with Ca²⁺ having the second highest, while the contribution of Na⁺ and Mg²⁺ to the overall neutralization process was very low.

The regression coefficient of the relationship of the sum of the concentrations of the cations to the sum of the concentration of the anions was 0.849 (Fig. 6). From this relatively high correlation, we can estimate that nss-SO₄^2- and NO₃^- are in neutralized forms. Such neutralization is frequently reported and attributed to NH₄⁺ and/or CaCO₃ [20]. It is expected that the sum of the concentrations of the three main cations ([H⁺] + [NH₄⁺] + [Ca²⁺]) correlated with sum of the concentrations of the major anions ([nss-SO₄^2-] + [NO₃^-]) if the acidity of the precipitation is mainly neutralized by typical bases in the atmosphere, i.e. NH₃ and soil dust.

3.6.1 Correlation Analysis of Ion Composition

Correlation analysis is a useful technique to characterize relationships among the ions present in rainwater. In order to find the associations among ions in precipitation as well as the possible sources of pollutants, correlations among ions in precipitation are calculated and listed in Table 2. A correlation was seen between SO₄^2- and NO₃^- (R² = 0.70) indicating they originated from similar sources. This was due to the similarity in their behavior in precipitation and the result of the emissions of their precursors, SO₂ and NO_x. Similarly, a strong correlation was seen between Ca²⁺ and Mg²⁺ (R² = 0.82), suggesting the common source of these ions was a natural source (e.g. crustal origin). Other relatively good correlations were observed between NH₄⁺ and NO₃^-, Mg²⁺ and Cl^-, Ca²⁺ and NO₃^-, and NH₄⁺ and SO₄^2-.

Most of these well-correlated pairs have common occurrences in precipitation as a result of atmospheric chemical reactions likely from the reaction of the acid in the atmosphere, such as HNO₃ and H₂SO₄, with alkaline compounds rich in carbonate materials, which were carried into the atmosphere by wind-blown dust. This shows that the wind-carried dust and soil play important roles in the chemistry of rainwater. Ammonium was correlated with NO₃^- (R² = 0.88). Other relatively good correlations were observed between Mg²⁺ and Cl^-, between Ca²⁺ and NO₃^-, and between NH₄⁺ and SO₄^2-, with correlation values of R² = 0.73, 0.72, and 0.68, respectively. The ammonium compounds applied to soil can escape into the atmosphere by means of gaseous NH₃ or as NH₄NO₃ and (NH₄)₂SO₄ particles. When NH₄NO₃ and (NH₄)₂SO₄ particles are incorporated into rain, they change the NO₃^-and SO₄^2- concentrations, but do not affect the pH. However, when ammonium is incorporated into rain, it can neutralize the acidity of rainwater [3].

3.6.2. Source Analysis of Chemical Components

Statistical analysis was performed using SPSS for Windows version 14.0 (SPSS Inc.). Data was log-transformed to achieve a normal distribution. Table 3 shows the varimax rotated principal component patterns for individual precipitation events. Three factors contributed to wet deposition during the sampling. Components 1, 2, and 3 were contributed 64.6%, 12.9 and 7.5%, respectively. Based on observations, only factor loads of higher than 0.5 have been deemed to be statistically significant [21]. The first principal component consists of high loads of SO₄^2-, NO₃^- and NH₄⁺. SO₄^2- and NO₃^- were derived from fuel combustion [22] and NH₄⁺ mainly originated from anthropogenic activities such as agricultural activity [23]. The second component shows high loads of Na⁺, Ca²⁺, and Mg²⁺, which indicated that they originated from soil resuspension (Mg²⁺ and Ca²⁺) and marine sources (Na⁺) [24]. The last principal component had a high load of K⁺ came from biomass burning [25].

This study of the chemical composition of rainwater was carried out in the northern part of Thailand from January to December 2008. NH₄⁺ and SO₄^2- were the dominant cation and anion found in the wet deposition in this area. Ion concentrations were high in the dry season and low in the rainy season.

The PCA illuminated the sources of the pollutants detected in precipitation, including fuel combustion (SO₄^2- and NO₃^-) and agricultural activity (NH₄⁺) in the first component. The second component suggested the contribution of soil resuspension (Mg²⁺ and Ca²⁺) and marine sources (Na⁺). The third component was associated with biomass burning (K⁺).

Financial support from the Pollution Control Department (PCD) and the Higher Education Commission of Thailand are gratefully acknowledged.