Introduction

Carbon dioxide is the greenhouse gas that absorbs solar energy and increases atmospheric temperature by re-emitting infrared energy. Surface active constituents influencing the CO2 in the atmosphere could come from oceans, landmasses, vegetation, industries, and volcanic eruptions that change the composition geographically and even seasonally. Weathering process also alters the composition of atmospheric CO2. The exploitation of natural resources, excess CO2 emissions and ecological hazards (Leakey 2009) facilitate the additional heat trapped in the earth’s atmosphere and increase the global temperature. This increase in temperature leads to melting ice caps, rise in ocean levels and amplifies the acidic character of oceans, rainstorms and intense floods. The Intergovernmental Panel on Climate Change (IPCC) has predicted that the global surface temperature will likely increase by 2.6–4.8 °C by 2100 (IPCC 2013). Mansour et al. (2020) projected that by the end of the twenty-first century, Northern Arabian Peninsula would be warmer than Southern Arabian Peninsula. Higher temperatures are expected in Kuwait since it is located in the northeast of the Arabian Peninsula. The distribution of CO2 in the mid-troposphere, a part of the atmosphere, was observed by Atmospheric Infrared Sounder (AIRS) instrument. It looks patchy with high concentrations in the northern hemisphere and low in the southern hemisphere. The CO2 gas transport and distribution were controlled by the jet stream, large weather systems, and large-scale atmospheric circulations (nasa.gov 2013). About 1.5 g of CO2 is produced annually in the USA has an annual uptake is 0.5 tons. The remaining contributes to the increase in atmospheric CO2 and climate changes (USGA 2020). For decades, CO2 emission has been accelerating. In the 1960s, the annual growth was 0.8 ppm per year, which reached 1.5 ppm in 198090 s, but now it has been exceeded above 2.0 ppm of annual rise (NOAA 2020). The average concentration of CO2 was 414.7 ppm observed at NOAA’s Mauna Loa Atmospheric Baseline Observatory in Hawaii (Voiland 2019), reflecting the consistent increase since pre-industrial times and earlier in May 2013 observation reported 400 ppm (Monastersky 2013). This scenario also affects economic growth, primarily due to carbon emissions (Baltimoore 2010). The driving force of climate change is due to enhanced CO2 levels in the atmosphere affecting crop productivity in conjunction with increased water stress and temperature (Hoffman et al. 1986; Stinner et al. 1988; Rosegrant et al. 2006). Studies have also reported that increasing CO2 levels affect plants physiology and growth rate (Ziska 2008).

The atmospheric carbon contains two different forms known as dissolved inorganic carbon (DIC) and dissolved organic carbon (DOC). The organic carbon is generally oxidized to carbon monoxide and inorganic carbon (Hallquist et al. 2009), subsequently removed from the atmosphere through precipitation (Goldstein and Galbally 2007). The process of oligomerization reactions transforms less soluble higher molecular masses, also facilitated by molecules oxidation. The oxidation process is more soluble, incorporated in nuclei condensation and removed from the atmosphere through rainfall (Hallquist et al. 2009). The role of organic carbon in the global carbon cycle models is also limited (Jurado et al. 2008). Significant gaps exist in assessing the global carbon cycles due to the influence of anthropogenic activities. The difference in emission and uptake obtains the atmospheric CO2 increase (N). The emission is mainly due to fossil fuel (EF) and the land-use change (EL). The uptake of CO2 is predominantly the oceanic (VO) and the terrestrial (VT) sinks. Then the net increase in Atmospheric CO2 is derived as follows:

$$N = \left( {E_{{\text{F}}} + E_{{\text{L}}} } \right){-}\left( {V_{{\text{O}}} + V_{{\text{T}}} } \right).$$
(1)

The net increase in atmospheric CO2 concentration was estimated by Melinikov and Neill (2006) as 3.2 PgC/a. Thus, the photosynthetic uptake by aquatic biota, global water cycle, and atmospheric CO2 combines to play a key role in CO2 dynamics (Gombert 2002; Lerman and Mackenzie 2005; Iglesias-Rodriguez et al. 2008).

The warmer temperatures due to CO2 emissions enhance evapotranspiration and reflect an unbalanced hydrological cycle on the earth’s surface. Apart from the terrestrial sources, the ocean serves as the atmospheric sink of CO2. Other significant sinks of CO2 are rainwater and soil water (Liu et al. 2010). The rainwater is highly sensitive to atmospheric changes and dissolves CO2 with other components like major ions, gases, minerals, trace elements, etc. (Hutchinson 1957; Carroll 1962). A numerical study in Nigeria observed that the decrease in pH significantly showed the increasing levels of CO2 of rainwater. Robert et al. predicted that the pH of the rainwater would be reduced to 5.49 by 2100, considering pCO2 at 25 °C and 1 atmosphere (Bogan et al. 2009). The acidity of rainwater was predicted to increase from 5.3% in 2000 to 93.7% by 2050 (Nwaeze and Ehiri 2017). The CO2 emissions into the atmosphere also depend on the exchange of pCO2 with the aquatic environment (Chung et al. 2018). An understanding of CO2 export from the atmosphere through rainwater or gaseous exchange through sea air fluxes provides a solid base for predicting the trends of CO2 in the atmosphere (Jones et al. 2003; Liu et al. 2010; Ashton et al. 2016). Therefore, the increase of atmospheric CO2 can be estimated through rain deposition onto the surface.

CO2 in water is at equilibrium with carbonic acid, bicarbonates and carbonates. They all represent in the form of dissolved inorganic carbon (DIC), governed by pH of the solution. H2CO3 is predominant in water with pH < 6; HCO3− at the pH range of 7 to 9 and CO32− at a pH > 10.5 (Qian and Li 2011, 2012).

$${\text{H}}_{{2}} {\text{O}} + {\text{CO}}_{{2}} \rightleftharpoons {\text{H}}_{{2}} {\text{CO}}_{{3}} \rightleftharpoons {\text{HCO}}_{{3}}^{ - } + {\text{H}}^{ + } \rightleftharpoons {\text{CO}}_{{3}}^{{{2} - }} + {\text{2H}}^{ + } .$$
(2)

The increase in DIC reflects the contribution of inorganic carbon as HCO3− either through carbonate dust or inorganic HCO3− derived from the oxidation of organic carbon in the atmosphere. Studies report that the terrestrial and aquatic environment absorbs approximately half of the CO2 derived by the fossil fuel emission and the remaining CO2 is supplemented to the atmospheric sink (Sabine et al. 2004; Reay et al. 2007).

Climate projection models show significant warming by the end of the twenty-first century in the Middle East and North Africa (Bucchignani et al. 2018). Kuwait is one of the largest oil producers globally, hence studies have predicted pollution due to CO2 released by the operations in the oil sector (Al-Salem 2015; Al-Mutairi et al. 2017). Kuwait’s long-term historical carbon dioxide emission record (World Bank Open data) reflects a steady rise in atmospheric CO2. The data indicate climatic changes in Kuwait, with warmer summers extending till October. An unusual short intense rainfall event was observed during mid-November 2018, and relatively more frequent rain events were observed up to March 2019. The seawater surface and the atmospheric interface are very complex, and the exchange of CO2 is studied and modeled by several researchers (Zappa et al. 2009; Turk et al. 2010). A significant buffering along the sea surface resulting from the deposition of calcareous planktons along the ocean floor (Ridgwell and Zeebe 2005) controls the sea surface carbon dynamics. The ocean surface is very dynamic, and rainfall facilitates the exchange of air-sea CO2 flux, enhancing the uptake of oceanic CO2 (Turk et al. 2010). Observation on long-term records reported a seasonal variation, leading to the change in the phytoplankton population. In recent decades, the plankton population has been influenced by sewage inflow, temperature variations, salinity in Kuwait Bay and the open sea environment (Devlin et al. 2019). Thus, geographically the terrestrial landscape adjoining the Bay and the open sea has experienced drastic growth in urbanization, population, and oil-based industries. This change in land use has resulted in variation of CO2 in the atmosphere relating to the change in rainwater CO2.

The study on the rainwater chemistry in Kuwait with few samples was carried out in 2005 (Tariq Rashid 2008) and subsequently on trace elements in rainwater during 2020 (Samayamanthula et al. 2021). Isotopic studies in rainwater have helped derive the Kuwait Meteoric water line and characterize the sources of groundwater recharge (Hadi et al. 2016). However, these studies focused only on the major ion concentration, local meteoric water line derivation, and groundwater rainwater interaction, but CO2 levels in rainwater were not determined and discussed. pCO2 plays a significant role in the carbonate chemistry of rainwater and governs their dissolution capacities. Therefore, the geochemical process and its relation to the saturation index of the carbonate minerals would be understood from pCO2 values of water (Chidambaram et al. 2011). Therefore, this is the first baseline attempt in Kuwait to study rainwater CO2 and compare it with the values of groundwater and seawater in Kuwait. The study aims to determine the dissolved inorganic components in rainwater, Total CO2, DIC, pCO2 and the spatial variations of these components in rainwater to industrial and residential areas.

Study area

Kuwait is an arid country with an annual rainfall of 110, -120 mm (Samayamanthula et al. 2021). The rainfall is expected from October to December, and January to March. Atmospheric temperature varies from 45 to 50 °C (Chidambaram et al. 2020), with maximum and minimum values reported from June to August and December to February, respectively. Frequent dust storms are reported during the shift of seasons (Yassin et al. 2018). The annual evaporation rates vary from 2500 to 4500 mm in the coastal region and inland. Northwest winds, also known as Shamal winds, are the primary moisture source for rainfall in Kuwait. Observation on long-term data indicates that after 2005, a sudden change in wind direction was observed, mainly from the North. The variation in the atmospheric pressure that makes air move from high to low pressure is mainly due to a change in wind direction. Frequent sandstorms and variations in the wind are the sources for inputs of ions into the atmosphere. Meteorological parameters, such as temperature, rainfall, wind speed, wind direction, and humidity, also influence rainwater quality indirectly (Samayamanthula et al. 2021). The study area falls between 28.43° and 30.27° N Latitude, 46.57° and 48.58° E Longitude covering the sampling regions.

Methodology

Thirty-two rainwater samples from residential regions and 18 from the industrial areas accounting for a total of fifty were collected from twelve different locations in Kuwait (Fig. 1) during the rainy season, between November 2018 and December 2019, by adopting standard procedures (Peden et al. 1979; IAEA/GNIP 2014). The samples collected from industrial regions represented KISR, Shuwaikh, and Shuaiba. The samples were filtered by 0.45 µm Millipore filter paper, immediately transferred to air-tight sterile polyethylene bottles and preserved in a cool dry place to enable high accuracy during the analyses of samples in the Water Research Center’s (WRC) laboratories of the Kuwait Institute for Scientific Research (KISR). The physical, chemical, and microbiological parameters, such as alkalinity, pH, electrical conductivity, total dissolved solids, major cations and anions, total coliform, and E. coli, were determined using the standard methods (SMEWW 2017; ASTM 2009). The QC/QA procedures were adapted to the analytes as per the standard methods such as duplication, standard checks, and QC sample (Supplementary Table 1). Also, the samples were verified by the ion balance error equation for accuracy. The analytical results of the rainwater samples were considered as inputs in the PHREEQC software to determine total CO2. The study also focused on pCO2 and DIC, and these values were obtained using WATEQ4F (Version 4.0). The sample locations and spatial variations of the present study pH, DIC, pCO2 and total CO2 were mapped using Map Info professional software (Version V17.0.3). The groundwater chemistry of the 113 samples analyzed in the previous study (Sabarathinam et al. 2020b) was considered to calculate the total CO2, pCO2, and DIC. Similarly, 20 samples collected in Kuwait Bay and the open sea environment by Bhandary et al.(2018) were utilized for seawater chemistry. These were used to calculate the above parameters (Bhandary et al. 2018).

Fig. 1
figure 1

Rainwater sampling locations of residential and industrial regions

Full size image

Results and discussion

The microbial analysis helps recognize carbon derived from microbes in rainwater as the microbes can capture carbon (Roger et al. 2018). All the filtered rainwater samples were analyzed for total coliform and E. coli. There were no detectable coliforms in the samples. A few samples were selected randomly and analyzed for the microbial parameters after filtration using different sized filters. It was observed that unfiltered samples contained total coliform and E. coli, whereas the samples filtered using 0.45 micron filter under suction ensured the complete removal of microbes. This indicates that atmospheric dust influences the microbes and enhances the dissolution of CO2.

The chemistry of rainwater plays a major role in determining the amount of CO2 carried by rain from the atmosphere. It varies with the area due to natural and anthropogenic activities. Studies reveal that anthropogenic input and subsequent rainwater infiltration affect groundwater quality (Wu et al. 2020; Ren et al. 2021). The electrical conductivity (EC) of the samples ranged from 16.08 to 1974 µS/cm. Bicarbonate and sulphate ions were abundant in rainwater. The average concentration of ions in the residential regions reflected the following order of dominance HCO3− > SO42− > Ca2+  > Cl− > NO3− > Na+  > K+  > NH4+  > Mg2+  > NO2− > F− > PO43− > Br− and in industrial region SO42− was dominant followed by Ca2+ and CO32−. The other parameters of industrial regions showed a similar relationship to that of the residential regions (Fig. 2). The alkalinity of the samples was mainly due to bicarbonates which ranged from 1.1 to 83.1 mg/L and an average of 28.80 mg/L. Samples from the KISR site in 2018 contained lower bicarbonate than samples collected in 2019 January, February, and March events. It was also noted that samples in Fahaheel contain maximum alkalinity. The pH of the rainwater samples in residential and industrial regions ranged from 5.69 to 8.21, with an average of 7.12. The samples collected during November were more acidic than samples of December, January, February, and March months. Alkaline pH was also observed in many locations (Fig. 3a). The sample from Shuaibha contains an acidic pH (5.69) due to the influence of anthropogenic sources from industrial activities (Kattan, 2020).

Fig. 2
figure 2

Rainwater chemistry data with minimum, maximum and average values for residential and industrial areas

Full size image
Fig. 3
figure 3

a, b, c and d Spatial distribution for the average values of pH, Total CO2, DIC and pCO2 in rainwater

Full size image

Spatial distribution of pH, total CO2,DIC and pCO2

The mean analytical results pH, total CO2, DIC, and pCO2 of the study areas concerning the residential and industrial region were distributed spatially to understand the areas enriched with CO2 (Fig. 3a–d). Shuaiba, being an industrial region, observed an acidic pH (5.69) with the highest pCO2 (388.73 ppmV). Ishbiliya was reported to contain the highest DIC as 1147.92 µmole/L and total CO2 as 50.41 mg/L due to alkaline pH 7.59. pH above 7.0 usually tends to increase the DIC. Jahra and Riggae areas were noticed with low concentrations of total CO2 (< 10 mg/L). The rainwater samples were categorized into the residential and industrial regions to distinguish the variation in the concentration of total CO2. The output PHREEQC was classified as residential regions and industrial (Table 1) with the pH, alkalinity, and total CO2. In general, uncontaminated rainwater is acidic, but the neutralizing ions present in the atmosphere increase the pH and become alkaline. The chemistry of precipitation waters in Lhasa, Tibet was alkaline with an average of pH 8.36 in 1987–1988 (Zhang et al. 2003). As pH increased, the alkalinity increased, and total CO2 also increased in rainwater samples. The alkalinity of rainwater samples (Table 1) reflect the influence of carbonate dust present in large amounts in the atmosphere (Kattan 2020), as the acidic nature is generally neutralized by atmospheric dust (Balasubramanian et al. 2001; Al-Momani et al. 2008; Al-Khashman 2009; Rao et al. 2016). The total CO2 in residential region samples was comparatively lower than industrial region samples, but few samples showed higher, especially in the Fahaheel area close to the Industrial sector (Fig. 3b). In general, total CO2 was higher in 2019 compared to 2018 events in the industrial region. This higher trend continued until October 2019, followed by a reduction of total CO2 in December rain events. Dissolved CO2 is the key component governing the pH of rainwater. The rainwater pH is mainly controlled by greenhouse gases (Menz and Seip 2004; Senanayake et al. 2005; Wang and Wang 2006). The change in pH values of rainwater due to the rise in CO2 levels affected the ecosystem, reported by earlier studies (Zhang et al. 2003; Hendershot et al. 1993).

Table 1 pH, Alkalinity and total CO2 of samples collected during individual rain events from residential and industrial region
Full size table

The samples collected on 24 November 2018 in the residential and industrial region were observed to contain high amounts of total CO2 except for two samples from the residential area, Riggae. On observing the total CO2, it was noticed that the residential region of Fahaheel had a maximum concentration (64.6 mg/L) for the sample collected on 31 December 2018. The two samples from Isbhiliya also showed higher values of CO2. The residential samples collected adjacent to the coast had high pH, thus varying the total CO2 concentrations in rainwater. The biological driven carbonate deposition and nutrient loadings to the ocean or surface water bodies also significantly reduce CO2 diffusing from water bodies to the atmosphere (Wang and Wang 2006). An increase in phytoplankton species would increase inorganic carbon into the oceans, contributing to CO2 emissions and results through sea spray (Rousseaux and Gregg 2012). Samples near the coast were neutralized due to the influence of sea spray and the neutralizing ions from anthropogenic dust (Chidambaram et al. 2014). Thus a relatively higher amount of CO2 was noted in the residential regions (Fig. 3b). The contribution of CO2 from the industrial region is mainly due to fuel combustion.

In contrast, in the residential region, CO2 is contributed from automobile exhaust, baking at homes, construction material, and air cooler (Turk et al. 2010; Al-Salem 2015). The direction of wind flow associated with sandstorms enhances the concentration of CO2 (Czikowsky et al. 2018). Other impurities carried into the atmosphere by the effect of wind flow consequently alter rainwater's nature and vary the pH. Thus it is inferred that rainfall acts as the carbon sink by absorption of CO2 (Bharti et al. 2017).

On comparing the average values of pH, alkalinity, and total CO2 concerning time series, it was observed that the alkalinity and total CO2 increased constantly from 2005 to 2019 in both industrial and residential regions whereas pH showed slight variation in the industrial region (Fig. 4). When the total CO2 values were compared with the industrial and residential areas on a temporal basis, residential regions were observed to have higher CO2 than industrial (Fig. 3b). The change in lifestyle and increase in population, and other factors have contributed to more gases in residential regions and increased total CO2, as observed with the rise in pH and alkalinity.

Fig. 4
figure 4

Trends of pH, Alkalinity and Total CO2 in rainwater on temporal basis

Full size image

Dissolved inorganic carbon

The rain events were studied for the variation of DIC in the samples (Fig. 5). There was a drastic increase in the DIC of November–December 2018 events. The samples represented different spells on the same day in locations, especially in KISR and Riggae. The values indicated that the first sample collected in the day had a higher value and was subsequently reduced. This phenomenon was also clearly witnessed in the 2019 samples collected at KISR since the first rain event after the summer sandstorm event removed the predominant dust from the atmosphere. In Kuwait, dust storms are more frequent before the seasonal shift from summer to winter (Yassin et al. 2018). The higher values of DIC were observed at KISR located in the southern boundary of Kuwait Bay, and a few samples of Farwaniya, Ishbiliya, and Salmiya represented the western region of the study area (Fig. 3c). Atmospheric dust primarily from the inland desert influenced the DIC in the rain shower of the region. Other factors that influence the natural CO2 dissolution in rainwater are wind speed and direction, the atmospheric temperature during the rain event, and the sandstorms prior to the event. So, the interplay of these events during 24 November 2018, 10 and 30 December 2018 has increased DIC concentration (Fig. 5) in rainwater.

Fig. 5
figure 5

DIC of rain events during 2018 and 2019, Comparison of study area with DIC of CO2–H2O at 18 °C and DIC of CaCO3–CO2–H2O system at 15 °C and 18 °C (Liu et al. 2010)

Full size image

Studies have indicated that high wind speed, usually greater than 20 m/s and precipitation above 15 mm, scavenges the CO2 (Pathakoti et al. 2018). In contrast, in the study region, the wind speed was less than 20 m/s and precipitation was noticed in a few rain events with less than 10 mm. So the wind speed and rainfall amount have contributed to higher CO2. The rain events show that SSE and ESE direction of the wind with speed ranging from 3.9 to 4.4 m/s was predominant during 14–16 November 2019. Subsequently, from 22 to 24 November 2018 the prevailing direction was from SSE and NW with an average wind speed of 6.7–7.5 m/s, but the rain events during 10 and 21 December 2019 were from NE and had an average speed of 1.1 m/s. The lesser wind speed has facilitated the reaction time of the dust with the rain droplets, thereby increasing the DIC concentration in rainwater during this period. In general, the value of DIC is greater than 800 µmoles/L were noted in western parts of the study area and the South of Kuwait Bay.

It is calculated that 380ppmV is the global mean atmospheric CO2 (GMAP) at global mean annual surface temperature (GMAST) 15 °C has 20 µmoles/L of DIC in CO2 -H2O system (Dreybrodt 2012) at equilibrium. There are two probable reactions of Ca with CO2, either as (CaCO3–CO2–H2O) or with Ca– silicate. In carbonate system, the dissolution of CaCO3 consumes less CO2, as one CO2 is dissolved for each Ca to form two HCO3−. In this system at equilibrium condition, the DIC is estimated to be 1231 µmoles/L at GMAP and GMAST, which is 66 times greater than the DIC of the CO2-H2O system. In the silicate system, weathering of the silicate minerals also consumes relatively more CO2 (Dupré et al. 2003; Mortatti and Probst 2003). The reactions of the Ca-carbonate and Ca-silicate system with CO2–H2O can be expressed as follows (Li et al. 2013, 2016).

$${\text{H}}_{{2}} {\text{O}} + {\text{CO}}_{{2}} + {\text{CaCO}}_{{3}} \rightleftharpoons {\text{2HCO}}_{{3}}^{ - } + {\text{Ca}}^{{{2} + }} .$$
(3)
$${\text{3H}}_{{2}} {\text{O}} + {\text{2CO}}_{{2}} + {\text{2CaSiO}}_{{4}} \rightleftharpoons {\text{2HCO}}_{{3}}^{ - } + {\text{H}}_{{4}} {\text{SiO}}_{{4}} + {\text{Ca}}^{{{2} + }} .$$
(4)

It is clear from the above reactions that half of the HCO3− in the carbonate system and total HCO3− in the silicate system is derived from the atmospheric CO2. Thus, atmospheric sequestration is mainly influenced by the above processes. The silicate minerals are generally reported to have a slow weathering rate and very slow solubility compared to carbonates (Singh et al. 2005; Wu et al. 2005, 2008; Moon et al. 2007). The contribution of CO2 to the atmospheric sink is less significant due to silicate weathering because of low reaction kinetics. But the carbonate system is very dynamic and plays a key role in the pCO2 and DIC of rainwater. In Kuwait, the carbonate dissolution is facilitated in rainwater chiefly by the carbonate dust attributed to the dolomitic formations.

The rainwater DIC flux was determined as the product of the mean DIC and total rainwater (Shiklomanov 1993; Arsene et al. 2007; Ladouche et al. 2009; Celle-Jeanton et al. 2009). The total annual rainfall during 2018 and 2019 was reported as 343.8 mm and 69.3 mm, respectively. Therefore, the average DIC derived from rainwater in 2018 and 2019 was 540 µmoles/L and 568 µmoles/L, respectively. Hence, the rainwater flux during these years was estimated to be 1,85,652 µmoles/L and 39,362 µmoles/L during the years 2018 and 2019.

Comparison of DIC with pCO2

The DIC was compared to the pCO2 in rainwater to determine the influence of carbonate dust. The samples of 2018 and 2019 were compared to the equilibrium values of pCO2 of CO2–H2O (Longinelli et al. 2005; IPCC 2013) equilibrium values of CO2 for CaCO3–CO2–H2O system and the projected global mean value at 18 °C along with the observed values during 2005. There was excess rain shower during 2018 than normal years. The pCO2 value ranges from 6.78 to 227.9 ppmV during 2018, while 14.02 to 388 ppmV in 2019 (Fig. 6). Similarly, DIC varies from 39.7 to 1469.6 µmoles/L in 2018 and 134.7–1127.46 µmoles/L during 2019. The samples of 2018 had a wide range of DIC compared to 2019, greater than equilibrium values of the CO2-H2O system but less than CaCO3–CO2–H2O at 18 °C. The global mean DIC in rainwater is 82.8 µmoles/L, which is significantly lesser than the equilibrium value of 1231 µmoles/L, due to the short interaction time of rainwater with the carbonate dust < 104 s (Dreybrodt 2012). They do not attain equilibrium instantly and are believed to have a value lesser than the equilibrium value. The DIC value also changes with atmospheric temperature. A definite trend was observed between pCO2 and DIC except for positive linearity at equilibrium condition of CO2–H2O and CaCO3–CO2–H2O system. A wide range of pCO2 was observed in 2019 samples than the 2018 and 2005 samples. All the samples except one are below the equilibrium value of pCO2 reported at different temperatures for both the CO2–H2O system as well as CaCO3–CO2–H2O. The above equilibrium values may be because several organic carbon species like polycarboxylic acid substances are surface-active and form nuclei for condensation (Orlović-Leko et al. 2010). Further, Shulman et al. (1996) inferred that these compounds facilitate nucleation by decreasing the supersaturation level, the surface tension and hygroscopicity of the inorganic suspension in air mass (Saxena et al. 1995), which also influences the association of major ions to CO2 in rainwater and thus the pCO2 values.

Fig. 6
figure 6

Comparison of pCO2 study area with Equilibrium value of CO2–H2O and CaCO3–CO2–H2O, Global projected CO2–H2O and CaCO3–CO2–H2O (Liu et al. 2010), Open Sea (Bhandary et al. 2018, Kuwait Bay (Ali and Chidambaram 2020) and Groundwater (Sabarathinam et al. 2020a, b)

Full size image

DIC and pCO2 comparison of rainwater with groundwater

The DIC in the Kuwait aquifers were compared with the DIC of rainwater samples of the current study (Fig. 6). The average DIC value was derived from the analytical results of the groundwater samples collected for previous studies in the region (Sabarathinam et al. 2020b). The groundwater of the region is mostly saline, and the DIC concentration is on an average of 239 mg/L, with a pCO2 value of 501 ppmV. The lithology of the older Dammam aquifer in Kuwait is mainly dolomitic and that upper Kuwait group aquifer is predominantly arenaceous, with intercalations of Chert and Gypsum (Mukhopadhyay et al. 1996). The behaviours of the Karst aquifer and their associated pH were studied in different parts of the world (Liu et al. 2007; Macpherson et al. 2008). The pCO2 relationship to pH in the coastal groundwater (Rajendiran et al. 2019), continental silicate aquifers (Panda et al. 2019), and the alluvial aquifers (Chidambaram et al. 2011) proved that a negative association exists between the two variables. The reforestation and temperatures are some of the significant factors reported to increase the soil pCO2 (Liu and Zhao 2000). This phenomenon was reflected with the increased HCO3− value in groundwater during summer. The high pCO2 favours the dissolution of carbonates and thus enhances the HCO3− values in groundwater, but the groundwaters of both the aquifers in Kuwait are saline, and the predominant ions are Na+ and Cl− (Chidambaram et al. 2020). The CO2 generated from the soil facilitates the dissolution of dolomite and calcite in the Karstic aquifer (Macpherson et al. 2008). The increase of CO2 in groundwater indicates that this component of the hydrological cycle is an excellent sink for CO2 through the dissolution of carbonate rocks. The saturation state of calcite is observed to be near zero in Dammam aquifers and much negative in non-carbonate Kuwait group aquifer (Sabarathinam et al. 2020a). Similar results were observed globally (Millot et al. 2003; Liu et al. 2010; Devaraj et al. 2018). The greater volume of groundwater and lesser possibility of return flux (except in the discharge region) serve as an excellent carbon store compared to atmospheric CO2 sink and the surface water bodies (Kempe 1979). Since CO2 production in soil increases with temperature, the release of CO2 from the soil matrix is significantly reduced due to the complete or partial saturation of the pore spaces. Such variations are also dependent on the frequency and duration of rain event (Kabwe et al. 2006).

DIC and pCO2 of the Bay and open Seawater

The earlier studies on sea-air CO2 flux exchange have not considered the CO2 transfer from hot deposition from the atmosphere and the chemical dilution at the surface due to rainfall (Ishii et al. 2004; Feely et al. 2006). The surface water bodies act as biological pumps as oceans exhibit huge reservoirs for the atmospheric CO2 by the uptake of DIC during photosynthesis of the oceanic biota (Cassar et al. 2004; Blain et al. 2007). The ratio between the carbons emitted to the atmosphere to that of sediment buried showed a relationship of 0.08, indicating the eutrophic lake’s effectiveness as a carbon sink (Yang et al. 2008). Further high pCO2 has increased and calcified the productivity of Emiliania Huxleyi, indicating their response and adaptation to high pCO2 waters. The standard mechanism of pCO2 is generally observed at a depth of 3.5 m below the sea surface, and the role of this interface, the top ocean layer on pCO2, is not considered in climate models. This phenomenon is very significant in ocean regions represented by high rainfall and low wind. The pCO2 at the ocean-atmospheric interface is also altered by the variation in sea skin temperature (Ashton et al. 2016) apart from the turbulence during a rain event (Ho et al. 2004; Zappa et al. 2009). The data on CO2 variations along with the interface during short, intense rain events are scarce.

The DIC and the pCO2 values of the Kuwait Seawater and the Bay water were derived from the results of the previous studies in the region (Bhandary et al. 2018; Ali and Chidambaram 2020). Though there is no significant variation in DIC concentration, the pCO2 of the seawater is relatively higher than the samples representing Kuwait Bay (Fig. 6). This is because the surface dwellers, especially plankton and organism with carbonate shells, remove atmospheric CO2. Subsequently, these dead species, along with their faecal material, settle into the ocean floor. This process facilitates the reduction of pCO2 at the ocean surface. It enhances the pCO2 uptake capacity from the atmosphere and the newly derived DIC from rainwater and surface water sources. The rate of movement of water in the Bay is relatively lesser than the open sea (Pokavanich and Alosairi 2014). This variation in the residence time, current and tidal movement moderates the physical conditions of the Bay and open sea environment leading to the variation in pCO2 values. Further, a sizeable quantum of the organic domestic sewage is discharged into the Bay (Sabarathinam et al. 2019), influencing the pCO2 values of the Bay and the open sea. The precipitation of CaCO3 resulted in the production of CO2 counteracted by the stored and produced organic carbon in the surface of water bodies (Lerman and Mackenzie 2005), and the relationship is explained as follows.

$${\text{H}}_{{2}} {\text{O}} + {\text{CO}}_{{2}} + {\text{CaCO}}_{{3}} \rightleftharpoons {\text{Ca}}^{{{2} + }} + {\text{2HCO}}_{{3}}^{ - } .$$
(5)

Fertilization reduces the pCO2 in these regions and enhances the CO2 in the atmosphere (Ternon et al. 2000). The return of CO2 into the atmosphere through biotic respiration also plays a key role in CO2 dynamics (Cole et al. 2007; Zondervan 2007). In the CaCO3 predominant system, dissolved by acidic rainwater (H2CO3), there is an expected release of CO2 back to the atmosphere.

$${\text{CaCO}}_{{3}} + {\text{H}}_{{2}} {\text{CO}}_{{3}} \rightleftharpoons {\text{Ca}}\left( {{\text{OH}}} \right)_{{2}} + {\text{2CO}}_{{2}} .$$
(6)

CO2 is released into the atmosphere during continental weathering through carbonate precipitation on the ocean floor. In addition, the enhanced CO2 uptake by the photosynthetic process also plays a key role in oceanic CO2, as the planktonic species growth is prominent in Kuwait's open sea environment (Devlin et al. 2019).

The spatial extent of the fresh rainwater on the surface layer governs the pCO2 at the air–water interface. The dilution effect observed along the sea surface reduces the pCO2 along the boundary, and this effect is sustained for a significant period when there is low wind and heavy rainfall (Henocq et al. 2010). Still, the freshwater due to rain can be spread to a larger area through surface currents. In Kuwait, the earlier studies (Pokavanich and Alosairi 2014) indicate that the residence time and the relative water movement is less than other regions globally due to the density influenced by salinity, especially in the Bay (Sabarathinam et al. 2019).

Projection of total CO2 against the ratio of (HCO3 −/ Ca2+ + Mg2+)

Total CO2 in rainwater samples were projected against the ratio of (HCO3−/ Ca2+ + Mg2+). Ca2+ and Mg2+ are the common neutralizing ions present in rainwater due to the carbonate dust in the atmosphere. The concentration of HCO3− varies dissolution of the dust depending on the reaction time with the rainwater and the pH. The pH is also governed by the neutralizing ions present in the rainwater, similar to the study in groundwater containing dominant ions of Ca2+ and Mg2+ (Li et al. 2013, 2019). Hence, the HCO3−/( Ca2+ + Mg2+) ratio will provide a relative variation trend of Ca2+ + Mg2+and the influence of carbonate dust in rainwater. The average values of the samples in the study area were compared to global values and that of the GCC region. The average total CO2 value of 5 mg/L, with an average ratio of 1.70 of HCO3− and (Ca2+ + Mg2+), was observed in the Indian rainwater samples (Fig. 7). The Indian southwest monsoon showers reflected alkaline nature on long-term observation. The linear relation between HCO3− and (Ca2+ + Mg2+) in Indian rainwater samples indicated carbonate dust as a crustal source, which tends to neutralize the H+ ions present in rainwater (Rastogi and Sarin 2005).

Fig. 7
figure 7

Influence of the carbonate dust on the total CO2 in rainwater, a comparison of the Kuwait rainwater ratio to global values(calculated as average from all areas) and to those representing different land use patterns such as Baharin (Naik et al. 2017), Brazil (Facchini Cerqueira et al. 2014), Iran (Naimabadi et al. 2018), King of Saudi Arabia (KSA) (Ahmed et al. 1990), (Alabdula’aly and Khan 2000), Spain (Moreda-Piñeiro et al. 2014), Yemen (Saleh et al. 2017), North and Central USA, Romania, India, Mix, Mining, Forest (Liu et al. 2010)

Full size image

On the contrary acidic pH was observed in the southern coastal region of India during the southwest monsoon season due to the minor influence of Na, and Cl ions due to sea spray, in the coastal rain showers (Chidambaram et al. 2014). The HCO3−/(Ca2+ + Mg2+) ratio of samples from USA, Spain and Brazil were within a range of 0.2 to 0.1 values (Fig. 6). The acidity of precipitation in the United States was reported to be changed due to NH3 and alkaline soil dust in the atmosphere, which tend to neutralize the anthropogenically derived acids (Munger 1982). Similar observations on the variation of acidity of rainfall in Brazil showed that sugarcane burning, soil dust and industrial emissions are the main sources that control the acidity of rainwater (Lara et al. 2001). The studies on precipitation in Tibet indicated that alkalinity prevailed in the samples though pCO2 decreases with an increase in elevation, attributed to the presence of CaCO3 rich dust in the atmosphere (Zhang et al. 2003).

The sample of Iran had the least ratio of 0.01, with the lowest total CO2 concentration of 0.42 mg/L. The average values of total CO2 in the samples of Kuwait ranged from 20 to 30 mg/L and the ratio of HCO3−/(Ca2+ + Mg2+) was higher in the residential regions than the industrial regions. The studies related to CO2 emissions in the urban region of Japan observed higher values in areas using reclaimed water (Shimizu et al. 2013). It leads to the fact that the average sample values from GCC (Kingdom of Saudi Arabia, Bahrain and Yemen) have total CO2 values ranging from > 15 mg/L to less than 90 mg/L and the ratio of HCO3−/(Ca2+ + Mg2+) fall predominantly between 1 and 2. Thus, a distinct signature of the samples representing the Arabian Peninsula region was observed. Further, the average values of samples of the present study were well below the global mean value of total CO2 and HCO3−/(Ca2+ + Mg2+) in the rainwater samples. The sample representing a mixed land use pattern had the higher value of both the variables (Fig. 6).

The coastal rainwater is also affected by the concentration of Na+ and Cl− due to the sea spray effect (Chidambaram et al. 2014). The increase of these ions also alters the total ionic concentration of rainwater. The CO2 content is influenced predominantly by carbonate dust than the sea spray process. The impact of sea spray is observed only in the regions with a predominant sea breeze. Hence, it is inferred that Ca2+ and Mg2+ ions play a significant role in neutralization than Na+ and Cl−. The rainwater compositions reported in the peninsular region show a characteristic zone of representation in the plot indicating the ratio range between 1 and 2 and the total CO2 between 15 and 90 mg/L.

Correlation analysis

Spearman correlation

Spearman correlation analysis was adopted to study the correlation between major ions and total CO2 in rainwater of residential and industrial regions using SPSS 2.0 software. The output data (Table 2) from the residential region exhibited that pH was strongly associated with EC, Alkalinity, ions such as Na+, Ca2+, F−, NO3−, and SO42− and weakly correlated with Mg2+ and total CO2. Likewise, EC and TDS were correlated to alkalinity, ions such as Na+, K+, Ca2+, Mg2+, and major anions F−, Cl−, NO3− and SO42. Alkalinity was strongly correlated to total CO2 and ions Na+, Ca2+, Mg2+, NO3− and SO42- but weekly associated with F− and Cl−. Alkaline soil dust eroded to the atmosphere due to wind, and a few gases neutralizing the anthropogenic acidic concentrations increases the pH and the dissolution of CO2 (Munger, 1982). Na+ is dependent on K+, Ca2+, Mg2+, F−, Cl−, Br−, NO3− and SO42. A significant observation from the Spearman correlation result was total CO2 is strongly associated with pH, EC, Alkalinity, Na+, K+, Ca2+, Mg2+, NO3−, but weekly correlated to F−, Cl− and SO42−. Also, it was observed that NH4+ was negatively associated with temperature. It means pH, ions, and alkalinity are the major influencing factors for the dissolution of CO2. The cations associated with the bicarbonates are mainly from the soil eroded in dust form that contributes to CO2 (Cerqueira et al. 2014). The industrial region exhibited a similar relation to samples from the residential region, but significant positive correlations (Table 2a, b) was observed between temperature, total CO2 and negative correlation with PO43−.

Table 2 a Spearman correlations of rainwater chemistry and total CO2 of residential region
Full size table

Principal component analysis (PCA)

The PCA was also applied to evaluate the relationships between the analytes present in the rainwater (Wu et al. 2014, 2020; Li et al. 2019; Ren et al. 2021). The factor analysis adopting varimax rotation (Table 3) resulted in six residential and four industrial regions with an eigenvalue greater than 1. The first factor accounted for approximately 43.0% of the variance, reflecting a strong positive association of total CO2 with EC, alkalinity, and major ions. It indicates that the ions and bicarbonates are purely from crustal sources. The crustal source of dust in Kuwait is the primary source of contamination in rainwater (Samayamanthula et al. 2021). The abundances of ions and alkaline nature in precipitation of semi-arid region, India suggested that the ions are mostly crustal and eroded as soil dust (Cerqueira et al. 2014). The second major factor loadings contribute to approximately 12.2% of the variance with a strong association between pH, alkalinity, and total CO2. This factor depicts clearly that as pH increases, alkalinity increases and simultaneously influences the total CO2. The bicarbonates are strongly associated with Mg2+and K+ which neutralizes the acid originated from dust (Rastogi and Sarin 2005; Cerqueira et al. 2014). Only 10% contributes from the third factor with an impact of ammonium and nitrite ions. Temperature influence on the rainwater was observed from the fourth factor with 9.8%. There is no representation of ions or pH, temperature, EC in the fifth factor. Bromide and phosphate ions were associated with the sixth factor with a contribution of 7.3%.

Table 3 Rotated Component Matrixa of rainwater chemistry and total CO2 of residential and industrial Regions in Kuwait
Full size table

In the industrial region (Table 3), the first factor loadings show 52.2% of the variance with an association between EC, TDS, alkalinity, ions and total CO2, temperature, and nitrite. According to the second-factor loadings, temperature influenced nitrite and weakly correlated with Mg2+ with 13.9% variance. The third factor indicates that temperature and total CO2 are negatively associated with PO43− with 11.7% variance. Finally, pH was correlated to Mg2+ with a variance of 10.7% according to the fourth-factor loadings. The report of IPCC (2013) states that the increase of CO2 in the atmosphere results in increased soil and air temperature. On the other hand, the elevated temperature enhances soil respiration leading to the addition of CO2 to the atmospheric stock.

The residential and industrial regions of PCA reflected that the dissolution of ions occurs from the same source in both regions. Additionally, the industrial region was governed by wind speed and direction. Hence, it is clear that pH influences the dissolution of CO2. It should also be noted that the biogenic emission and the fossil-derived carbon contributes to the organic carbon stock in the atmosphere (Tsigaridis and Kanakidou 2003; Spracklen et al. 2008; De Gouw and Jimenez 2009). Predominantly the primary organic carbon is converted to secondary organic aerosols in the atmosphere (Hallquist et al. 2009). Organic carbon removal from the atmosphere through precipitation also plays a significant role in carbon flux transfer (Iavorivska et al. 2016). The marine region had lesser organic carbon values than that of the continental precipitation, and it varied with the distance from the coast. The volatile organic gases in air and water are expected to vary by chemical reactions (Ervens et al. 2011). The chemistry of rainwater is chiefly governed by natural than anthropogenic sources (Liu et al. 2010). It is also mainly due to the prevalence of CaCO3 dust in that atmosphere, either calcite or dolomite dust. But the predominant extreme acidic rain events in the industrial regions are mainly in the downwind direction.

Conclusions

The study has derived the following conclusions:

  1. 1.

    Rainwater representing both residential and industrial regions acts as a sink for atmospheric CO2.

  2. 2.

    Total CO2 was correlated with pH, alkalinity and major ions. The mean pH of rainwater in the study was observed to be in alkaline nature due to neutralizing ions such as Ca2+ and Mg2+; while only Shuaiba, industrial region reflected acidity. The temporal observation from 2005 to 2019 reflected an increasing trend of pH and alkalinity, which influenced the dissolution of CO2.

  3. 3.

    Statistical correlation and PCA analysis inferred that the main contribution of CO2 is from eroded crustal sources along with soil dust. Atmospheric dust from the inland contributed higher amounts of DIC and values greater than 800 µmoles/L were identified in the study region located in western parts and south of Kuwait Bay. Wind speed, its direction, temperature and sand storms also influenced the dissolution of CO2 in rainwater.

  4. 4.

    DIC values in rainwater samples tend to be higher in the initial spells and reduced in the subsequent events. The total DIC flux derived from the rainwater during 2018 and 2019 were estimated to be 1,85,652 µmole/L and 39,362 µmoles/L.

  5. 5.

    A definite trend with a wide range of pCO2 was observed in 2019 compared to 2018 and 2005 values. The pCO2 of rainwater in Kuwait on comparison with global equilibrium values of pCO2 of CO2–H2O, CaCO3–CO2–H2O were found to be lower except for one sample.

  6. 6.

    The spatial distribution of pH, DIC, pCO2 and total CO2 inferred that Shuaiba, an industrial region had acidic pH (5.69), which led to the highest pCO2 (388.73 ppmV), while Ishbiliya, the residential region was noticed with alkaline pH containing high DIC and total CO2 (1147.92µmole/L and 50.41 mg/L respectively).

  7. 7.

    DIC and pCO2 values of the groundwater are generally higher than the sea and Bay waters. DIC in groundwater is mainly governed by lithology and dissolution temperature. The precipitation of carbonates releases CO2 to the atmosphere.

  8. 8.

    Sea skin temperature and the CO2 values along the sea-air interface play a key role in the carbon dynamics of the oceanic atmosphere. The CO2 in the sea is governed by anthropogenic stress, residence time, amount of rainfall and wind speed.

  9. 9.

    The rainwater samples of GCC region had a definite ratio of HCO3−/(Ca2+ + Mg2+) and a wide range of dissolved CO2 depending on local conditions. The interplay of terrestrial CO2 and oceanic CO2 contributes to the atmospheric sink, which is later reflected in the total CO2 of rainwater.

Overall, the study infers that the natural processes have a predominant control of CO2 in the rainwater of Kuwait. The study has been carried out with the available data set for inorganic carbon values in the samples collected from 2018 November to 2019 December. The sample’s organic carbon concentration and the stable carbon isotopes will also yield information on the sources and the atmospheric processes. Samples during a pandemic would be more significant to note if there has been a substantial variation in the dissolved CO2 values in the rainwaters. The rainwater study in arid regions is scarce. The data on the pCO2 variation in rainwater of other arid regions for and space would help assess the impact of aeolian dust and gasses. The sampling on each rain event and samples representing the same event in different time intervals can be focused on in future studies.