Role of surface CO2 sources and wind transport on atmospheric CO2 variability over the Bay of Bengal during Southwest Monsoon - using mixing ratio, carbon isotope ratio and model simulated wind field

doi:10.21203/rs.3.rs-1861595/v1

PDF

Research Article

Role of surface CO₂ sources and wind transport on atmospheric CO₂ variability over the Bay of Bengal during Southwest Monsoon - using mixing ratio, carbon isotope ratio and model simulated wind field

https://doi.org/10.21203/rs.3.rs-1861595/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted 02 Aug, 2022

You are reading this latest preprint version

The present study examines the role of surface-CO₂ sources and wind transport on atmospheric CO₂ observations over the Bay of Bengal. The ship-based observations of mixing ratio and stable carbon isotope ratio (δ¹³C) of air CO₂ was carried out at coastal and open ocean location during 17 July 2009 until 18 August 2009, covering the peak months of the Southwest Monsoon season. The average mixing ratio was higher for the coastal sites (407 ± 17 µmol mol^− 1) than the open ocean sites (394 ± 9 µmol mol^− 1), whereas the average δ¹³C value was lower (more depleted in ¹³C) (-8.75 ± 0.53‰) for the coastal sites than the open ocean sites (-8.48 ± 0.04‰). The observations were further used to identify the isotopic ratio of the source CO₂ using the Keeling and the Miller-Tans plot. The presence of two different sources were confirmed from the observations. The coastal observations were affected by continental sources of CO₂ (-19.95 ± 1.7‰), whereas for open ocean sites, oceanic sources (-10.31 ± 0.15‰) played a major role. Although there was a transport of continental air to open ocean, as observed from the WRF model simulated wind, the land-ocean contrast was maintained. It was evident from the Carbon Tracker simulation during the study period. Despite continental air transport, the open ocean observation was influenced by oceanic sources of CO₂. To explore the reason, the predominant wind pattern for the sampling sites were determined from the WRF model simulated winds. A k-means clustering technique was used on the WRF model simulated hourly wind to understand the different atmospheric transport patterns during different days of sampling in July and August. Most of the sampling sites were resided within the clusters whose mean wind direction indicates south-westerly/southerly wind. Thus, irrespective of continental transport, the sampling sites were influenced by oceanic wind during July and August. The influence of oceanic wind allowed the identification of oceanic source CO₂ over the open ocean sites despite the transport from the continent.

Carbon dioxide

mixing ratio

stable carbon isotope

Weather Research and Forecast Model (WRF)

Carbon Tracker

Bay of Bengal

Atmosphere-ocean interaction

The increase in atmospheric CO₂ concentration since the industrial revolution in the 19th century is mainly due to the CO₂ emitted from anthropogenic activity such as fossil fuel combustion, biomass burning, deforestation, land-use changes, etc. (IPCC, 2021). Half of the total anthropogenically emitted CO₂ since 1850 remains in the atmosphere and the other half is taken up by the ocean and land (Takahashi et al. 1997; IPCC, 2013). During 1870–2017, approximately 25% of anthropogenic carbon was captured by the ocean (Le Quéré et al. 2018). Moreover, the retention time of this CO₂ in the ocean is of the order of centuries to millennia (Feely et al. 2001). Thus, the ocean plays an important role in understanding the carbon dynamics and future growth trajectories of atmospheric CO₂ (Le Quéré et al. 2009, 2016).

The Indian Ocean (44°S − 30°N) plays a crucial role in the global carbon cycle. It acts as a net annual sink of CO₂ (Sarma et al. 2013) and accounts for ~ 20% of the global CO₂ uptake (Takahashi et al. 1997). The oceanic fluxes vary largely in the Northern Indian Ocean (Valsala et al. 2012). The north-eastern Indian Ocean (i.e., Bay of Bengal) acts as a net mild sink, whereas the north-western Indian Ocean (Arabian Sea) acts as a net source of atmospheric CO₂ (Valsala and Maksyutov, 2010; Valsala and Maksyutov, 2013; Sarma et al. 2012). The northern Indian Ocean is also uniquely enclosed in the north and east by the Indian Subcontinent. Thus, the transport of continental CO₂ flux also makes this region interesting to study.

In contrast to other tropical oceans, the North-eastern Indian Ocean i.e., the Bay of Bengal (hear after as ‘BoB’) is a small net annual sink of atmospheric CO₂ (Valsa and Maksyutov, 2010; Sarma et al. 2013). The BoB receives significant amount of freshwater from both precipitation and rivers (Sengupta et al. 2006). Strong stratification in the vertical structure of the ocean (George et al. 1994) creates lowers the partial pressure of CO₂ in the surface water (pCO_2sea) in comparison to the atmosphere, thus making BoB a mild net sink of CO₂ (Kumar et al. 1996; Sarma, 2003; Takahashi et al. 2009; Sarma et al. 2012). Although the BoB acts as a net annual sink of CO₂, there exists a considerable temporal and spatial variation of air-sea CO₂ fluxes over BoB (Ye et al. 2018). Based on an observation made at 5°N and 90°E over the central BoB, it was identified as a sink for CO₂ during September to March and a source for CO₂ during April to August (Ye et al. 2018). The pCO₂ in the surface seawater (pCO_2sea) is higher in April–June and lower in December–February, mainly due to higher sea surface temperatures (SSTs) associated with strong solar radiation and weak winds in summer and vice versa in winter (Ye et al. 2018). The western continental shelf of BoB, that is the coastal BoB, can either act as a source or sink depending on the peninsular river discharge (Sarma et al. 2012). From June until September, Southwest monsoon months (mentioned hereafter as ‘SWM’), the pCO_2sea was reported to be lower than the pCO₂ in the atmosphere (pCO_2air), whereas for the discharging peninsular rivers, was an order of magnitude higher than that of pCO_2air. Thus, the river discharge into BoB and the distribution of discharged water along the coast by the East India Coastal Current determine whether the coastal BoB will act as a source or sink during SWM (Sarma et al. 2012).

The atmospheric dynamics of BoB are equally interesting and emphasize the need for monitoring CO₂ in the atmospheric marine boundary layer. The BoB experiences south-westerly wind in SWM and north-easterly wind in winter (December to February). During SWM, the BoB experiences the most intense deep convection and precipitation in the tropical oceans (Webster 1995). The deep convection varies at a 30–90 day time scale, and it can carry the near-surface trace gases like CO₂ to the mid and upper troposphere (Li et al. 2010). During SWM, there exists a noticeable land-ocean difference in the mixing ratio of CO₂ and CH₄ over the Indian subcontinent and the adjacent BoB (Kumar et al. 2013; Guha et al. 2018), primarily driven by wind and the contrast in surface CO₂ fluxes between land and ocean. This contrast is usually preserved in the vertical due to strong convection (Li et al. 2010) and thus the CO₂ mixing ratio measured near the ocean surface is detectable by model and satellite-based observations (Kumar et al. 2013).

Although, previous studies over BoB have extensively investigated the biogeochemical cycle of CO₂ in the surface water as well as with increasing depth of water (Sarma et al. 2012, 2013), there is an acute scarcity in understanding the atmospheric CO₂ variability in the marine boundary layer over BoB and thus the understanding is limited both spatially and temporally, especially during SWM. The Continental Tropical Convergence Zone (CTCZ) expedition of 2009 (https://incois.gov.in/portal/datainfo/ctczdata.jsp) gave an opportunity to understand the atmospheric CO₂ variability over BoB during July 17th until August 18th. Although a month-long observation campaign is not sufficient to identify intra-seasonal variability, the aim of this study is to understand how the surface fluxes and/or the transport from continent affect the atmospheric CO₂ variability over BoB. We studied combined measurements of mixing ratio and stable carbon isotope ratio (δ¹³C) of atmospheric CO₂ at coastal and open ocean locations to identify the predominant sources of CO₂ over BoB. We further analysed the WRF model simulated winds to discern the contribution of BoB sources and sinks compared to the influence of continental fluxes at the sampling sites at the open ocean locations. One of the common techniques to address this problem is Lagrangian back trajectory analysis (Draxler et al. 2010; Liu et al. 2019). The Lagrangian back trajectories of air parcels are often used to determine the origins of air masses using model simulated wind fields (Gheusi and Stein 2002; Guha et al. 2017). However, in this study we adopted spatio-temporal clustering (in particular, k-means clustering) of the simulated wind fields to analyse dominant wind patterns and their persistence over a larger BoB study region. Clustering analyses have been used in previous studies to identify clusters in meteorological observations (e.g., Kaufmann and Whiteman 1999, Clifton et al. 2012). Unlike back trajectory modelling, k-means clustering is rather simple and does not involve additional atmospheric models other than model simulated winds. One of the other qualitative goals of this study was to find regions of homogenous wind patterns over a period. This provides a general understanding of the atmospheric dynamics required to deploy such future experiments.

2.1 Air sample collection, extraction, and analysis of CO₂:

During the CTCZ national cruise expedition of 2009 two research vessels, the ORV Sagar Kanya (SK261) and the Sagar Nidhi were deployed to make observations over the BoB during SWM. The air sampling was conducted from July 17th until August 18th, 2009, along the cruise track shown in Fig. 1. Air samples were collected from 12 different sites over BoB (Fig. 1). The geographic location of the sampling sites was determined using high-resolution GPS on board throughout the cruise. Sampling was done mostly in the afternoon time (between 12:00–15:00 Indian Standard Time), as at that time the boundary layer remain at higher height and air is mostly well mixed within the boundary layer. Out of these sites, 4 sites were located along the western coast of BoB, termed as “coastal sites” and the rest 8 sites were in the open ocean, termed as “open ocean sites”. The open ocean sites were selected in such a way that they were at least ~ 350 km away from the coast. Air samples were collected from the ship deck at ~ 10 m above the sea surface. To avoid contamination of sampling with the ship engine exhaust, the sampling place was decided based on the prevailing wind direction. The location chosen was to be approximately 80 m upwind of the ship exhaust funnels. Air samples were collected in a 1 litre Pyrex glass flask fitted with a high-vacuum Viton O-ring stopcock (Brandon Scientific, Australia) (Guha and Ghosh, 2010) for preventing air exchange between the flask and outside (Sturm et al. 2004). A detail of the air sampling protocol can be found in Guha and Ghosh 2010; 2013; 2015. Within a week after the cruise period, CO₂ was extracted from the glass flask using a cryogenic extraction and separation protocol (Guha and Ghosh, 2013). The mixing ratio of the extracted CO₂ was determined volumetrically using Baratron manometer, whereas the carbon isotopic ratio, δ¹³C (i.e., ¹³C/¹²C ratio in the sample with respect to a standard, as δ = (^13C/12Csample / ^13C/12Cstandard – 1) of the was measured using dual inlet peripheral of isotope ratio mass spectrometer, IRMS MAT253 (Thermo-Fisher, Bremen, Germany) (Guha and Ghosh, 2013). The experimental set up was calibrated using primary standard NBS19 and a cross-calibration of our measurement with JRAS reference air mixtures provided by MPI-BGC (Wendeberg et al. 2013) was conducted to determine the system offset. The precision for mixing ratio and δ¹³C measurements was ± 2 µmol mol^− 1 and 0.02‰, determined based on repeat measurements of CO₂ extracted from JRAS standard (Guha and Ghosh, 2013). The observational data is given in Table 1.

Table 1

Mixing ratio and δ¹³C value of air CO₂ measured during CTCZ cruise expedition, 2009
	Sample ID	Date	Time	Latitude (°N)	Longitude (°E)	mixing ratio (µmol mol^− 1)	δ¹³C (VPDB)
Open Ocean Sites	SK1	17-Jul-09	12:30	11.8	85.8	401	-8.51
	SK2	18-Jul-09	12:40	11.0	89.0	403	-8.51
	SK3	23-Jul-09	12:40	18.98	89.1	398	-8.50
	SK4	30-Jul-09	14:55	19.0	90.6	403	-8.52
	SN1	26-Jul-09	14:45	6.2	86.0	381	-8.43
	SN2	10-Aug-09	14:00	8.0	91.5	393	-8.48
	SN3	11-Aug-09	15:00	8.0	93.0	393	-8.47
	SN4	17-Aug-09	14:30	9.0	85.0	383	-8.41
Average						394	-8.48
Standard Deviation						± 9	± 0.04
Coastal Sites	SK5	7-Aug-09	13:00	20.7	88.3	395	-8.32
	SK6	11-Aug-09	15:00	19.9	86.4	390	-8.33
	SK7	16-Aug-09	14:30	17.2	82.8	434	-9.41
	SK8	17-Aug-09	14:05	13.5	80.8	408	-8.95
Average						407	-8.75
Standard Deviation						± 19	± 0.53

2.2 Simulated CO₂ from Carbon Tracker

To identify the spatial variability in atmospheric CO₂, we used simulated CO₂ from the Carbon Tracker, CT-2015 version (Peters et al. 2007). Carbon Tracker is a global CO₂ measurement and modelling system, created by National Oceanographic and Atmospheric Administration/Earth System Research Laboratory (NOAA/ESRL). It generates simulated CO₂ mole fraction using the Carbon Tracker optimized surface fluxes and simulated atmospheric transport. This helps to track the global distribution of carbon uptake and release from land ecosystems and oceans around the world. We used the Carbon Tracker mole fraction (µmol mol^− 1 of CO₂) response of individual fluxes, like fossil fuel, biosphere, and fire, which is created using optimized surface fluxes and simulated atmospheric transport. We took an average of fossil fuel, biosphere, and fire mole fraction component and averaged them. This averaged value was considered in this study as ‘land contribution of CO₂’, as these are the three predominant CO₂ sources over land.

2.3 WRF Model simulation and clustering analysis

To explore the dominant wind patterns over BoB during the period of our study and to analyze wind patterns over the BoB sampling sites, we utilized simulated wind data from the Weather Research and Forecasting model (WRF), version 3.9.1.1 (Skamarock et al. 2008).

The WRF model is a mesoscale numerical weather prediction system with a fully compressible, non-hydrostatic Euler equation with vertical coordinate. In this study, we used two nested domains, where the coarse domain (d01) uses a horizontal grid spacing of 12km and the inner domain (d02) 4km grid spacing. The nested domain configuration is shown in Supplementary material (SI-1). The “one-way” nesting method is used, in which the inner domain is constrained by the outer domain through nudging of the boundary conditions that drive the meteorology (Soriano et al. 2004). The feedback from the inner domain to the coarse domain was not allowed. We only analysed the inner domain (d02) as it covers the area of interest. The physical parameterization used for WRF simulation is given in Table 2. The model simulation is done separately for the month of July (start from: 2009-07-01, 00:00:00 UTC, run through: 2009-08-01, 00:00:00 UTC) and August (start from: 2009-08-01, 00:00:00 UTC, run through: 2009-09-01, 00:00:00 UTC) to capture the dominant wind pattern more precisely.

Table 2

WRF physical parameterizations:
Parameter	Option
Land Surface Model	Unified Noah (Tewari et al. 2004)
Planetary boundary layer	Yonsei University (YSU; Hong et al. 2006) scheme
Surface layer	Revised MM5 (Jimenez et al. 2012)
Cumulus	Kain–Fritsch (KF; Kain, 2004)
Microphysics	WSM-3 (Hong et al. 2004)
Shortwave radiation physics	Dudhia Shortwave Scheme (Dudhia, 1989)
Longwave radiation physics	Rapid Radiative Transfer Model (RRTM) (Mlawer eta al. 1997)
Initial and boundary conditions	NCEP Global Final Analysis (FNL)

We used hourly U, V wind at 10 m above ground from the WRF simulation over the entire study period and perform a spatio-temporal clustering. Cluster analysis can identify groups with similar wind speed and wind direction. A wide array of clustering methods has been proposed thus far to find groups in n-dimensional data. In this study we adopted the ubiquitous k-means clustering technique (Kaufman and Rousseeuw, 2005; Jain et al. 2000). K-means clustering splits the data into k many discrete and non-overlapping buckets or clusters. The k-means clustering chooses a set of k cluster centers and assignment of data points to cluster centers by minimizing a global dissimilarity. We used the squared Euclidean distance as the dissimilarity between the data points in a particular cluster and that cluster’s center. Given the number of clusters, global dissimilarity is then computed as the sum over the cluster dissimilarities. To get closer to the global minima while minimizing the global dissimilarity, the k-means algorithm is usually run multiple times with multiple random initializations of the cluster center to get the least global dissimilarity. We used the R-package clusterR (Mouselimis, 2020) to perform the k-means clustering. We chose the optimal number of clusters based on the ‘distortion_fk’ criterion (Pham et al. 2005) with efficient K-means + + initializations (Arthur and Vassilvitskii, 2007). Once we have determined the optimal number of clusters, we used the standard random-initialization with 50 repetitions to obtain the final set of clusters.

The k-means clustering technique was employed on the hourly U, V wind components in July and August to group the data into different atmospheric transport patterns. Clustering was done separately for July and August. For each month, hourly U, and V data were vectorized and stacked to create a dataset with two features or columns. Then k-means clustering was performed on the entire spatio-temporal dataset. Therefore, one specific cluster can not only cover a region but also spans over a period. This means, two clusters may cover same region over different days. This is quite plausible as the wind field over a fixed location can greatly vary over time and thus can belong to different clusters on different days. A total of 14 clusters were obtained for each month. For each cluster, we obtained the summary statistics of the wind directions, speeds and the centroids over the hours and days it was present. This provided us with an overview of the location, timing, and strength of the wind clusters.

3.1 Identifying the land contribution of CO₂ from Carbon Tracker:

The land contribution of CO₂ mole fraction from Carbon Tracker simulation, averaged over the study period, is shown in Fig. 1, as a base map. The land contribution of CO₂ is found to be comparatively higher near to the coastal area than the open ocean. This is due to the proximity of the coastal area to the Indian subcontinent. Although there exists no in-situ fossil fuel, biosphere or fire sources over the open ocean, traces of land contribution of CO₂ are found to be present over open ocean. This prevailing westerly wind has transported the continental CO₂ to the open ocean. This is evident in the WRF model simulated 10 m above ground wind vector, as shown in Fig. 1 with black arrows. Moreover, the oceanic fluxes also play a role in controlling the CO2 variability. Thus, a contrasting amplitude in land and ocean fluxes is expected to affect the observation of CO₂ between coast and the open ocean. This is further explored using observation on mixing ratio and δ¹³C of atmospheric CO₂ at coastal and open ocean sites.

3.2 Identifying the isotopic signature of the source-CO₂

The average mixing ratio values for the coastal and open ocean sites are 407 ± 17 µmol mol^− 1 and 394 ± 9 µmol mol^− 1 respectively. The δ¹³C of CO₂ is found to be more depleted in ¹³C (more negative) for the coastal sites compared to the open ocean sites. The average δ¹³C values of CO₂ for the coastal and open ocean sites are − 8.75 ± 0.53‰ and − 8.48 ± 0.04‰ respectively. The mixing ratio of CO₂ is found to be higher for the coastal sites compared to the open ocean sites. Observations at the coastal area is affected largely by land contribution i.e., the continental sources of CO₂, whereas over open ocean the land contribution is diluted as evident from the Carbon tracker data also. The open ocean sites show a relatively less variability (spread) in both mixing ratio and δ¹³C of CO₂. Along with the transport of continental CO₂, the oceanic flux is also expected to play a major role in composing the open ocean CO₂. To address this, we have identified the sources of CO₂ for both the open ocean and coastal sites, separately. We have used an inverse relationship between both mixing ratios and δ¹³C value of CO₂ to identify the isotopic ratio of CO₂ in source air, popularly known as a Keeling plot (Keeling, 1958). The Keeling plot for both the open ocean (blue circles) and coastal sites (black circles) is shown in Fig. 2a. A Geometric Mean Regression, GMR method is used to fit the data and the intercept of the regression line provides the isotopic composition of source CO₂. Usually, an ordinary least square method is used to fit the data and the method assumes that the independent variable is not associated with any error (Sokal and Rohlf, 1995). In this case, the independent variable that is mixing ratio has an analytical error, so instead of ordinary least square method we have used GMR fit to the data (Pataki et al. 2003). The δ¹³C value of source-CO₂ is found to be and − 20.2 ± 1.8‰ (R² = 0.95, p < 0.05) and − 10.3 ± 0.15‰ (R² = 0.96, p < 0.05) for the coastal and open ocean sites respectively, as shown in Fig. 2a. According to the Miller and Tans, 2003, in keeling plot method the source isotopic ratio is determined by extrapolating the 1/CO₂ mixing ratio and this could introduce large error in source CO₂ estimation. So, we have further used Miller-Tans method to estimate the δ¹³C value of source-CO₂ (Miller and Tans, 2003). The relationship between mixing ratios and mixing ratio x δ¹³C values of CO₂ is plotted in Fig. 2b. The GMR method is used, and the slopes of the regression lines provide the δ¹³C value of source CO₂. The δ¹³C value of source CO₂ is found to be -19.9 ± 1.7‰ (R² = 0.98, p < 0.05) and − 10.3 ± 0.15‰ (R2 = 0.99, p < 0.05) for coastal and open ocean sites, respectively. Miller-Tans method has allowed explicit subtraction of background values for both CO₂ and δ¹³C and thus provided suitable platform where the background is not constant through the time or space, as relevant to this study. Two different sources are found to influence observations at coastal and open ocean sites. For the coastal sites, the source estimated (~-20‰) is representative of continental sources (combination of sources like biomass, fossil fuel, cement etc.) (Guha and Ghosh, 2015). The source identified for open ocean sites (~-10‰) is comparatively enriched than typical anthropogenic source signature and rather indicative of a kinetic air-sea gas exchange that is ~ -2‰ fractionation on top of a background value of ~ -8‰ (Siegenthaler and Munnich, 1981; Wanninkhof 1985; Zhang et al. 1995). This implies an influence of oceanic fluxes (https://www.esrl.noaa.gov/gmd/ccgg/isotopes/c13tellsus.html) on the air-CO₂ variability for open ocean sites. A direct measurement of isotopic ratio of CO₂ flux from ocean surface might confirm it quantitatively. Nevertheless, the presence of two different sources for the coastal and open ocean sites are confirmed from the observation of mixing ratio and δ¹³C of CO₂. Apart from different sources, the wind can also play an essential role on air CO₂ variability over the open ocean sites of BoB either by transporting continental air to open ocean sites or allowing the in-situ oceanic fluxes to influence the observations over open ocean sites. To understand the wind transport for the open ocean sites we have used a cluster analysis of wind.

3.3 k-means clustering of wind to explore the dominant transport to the sampling sites

Total 14 different clusters with different wind pattern are found for both July and August. Out of all sampling days in July, three days (July 17, July 23, and July 30) are plotted in Fig. 3. The spatial distribution of all the 14 clusters obtained at the sampling time are shown in Fig. 3 (A1, A2 and A3) and the mean wind direction for the respective clusters are shown in Fig. 3 (B1, B2 and B3). Similarly for August, three days (August 10, August 11 and August 17) are plotted in Fig. 4. The spatial distribution of all the 14 clusters is shown in Fig. 4 (A1, A2 and A3) and the mean wind direction for the respective clusters are shown in Fig. 4 (B1, B2 and B3). The sampling sites are found to be located near/within the clusters as shown in both Fig. 3 and Fig. 4 confirming the predominant wind for the sampling site at the time of sampling. Although at the beginning of the cruise season (from July 17), the sites were influenced by south-westerly wind, as the days progressed towards August, the sampling sites predominantly experienced southerly wind. While the south-westerly wind carry the continental air to the sites, this continental transport was largely reduced latter in the cruise season due to the presence of southerly wind bringing oceanic air to the sites. This predominance of oceanic wind at open ocean sites largely reduced the continental source influence and eventually allowed capturing oceanic source influence in the atmospheric CO₂ observation.

The observation on mixing ratio and δ¹³C of CO₂ over BoB during SWM allowed identification of possible sources of CO₂ for both coastal and open ocean sites. Based on the source identification from observation, the coastal sites observations were found to be controlled by the continental sources, whereas the observation on open ocean sites by oceanic sources of CO₂. Despite the transport of continental CO₂ over open ocean as shown from Carbon Tracker, the open ocean observation could identify the influence of oceanic source of CO₂. The cluster analysis on WRF simulated wind showed dominance of southerly wind (mostly from ocean) for majority of the sampling sites. The dominance of wind from oceanic regime allowed identification of oceanic source influence in open ocean observation, despite continental transport of CO₂. In an ideal situation, a clustering analysis with forward model simulated winds can provide a detailed design to decide on the sampling sites prior to such a cruise expedition. The k-means clustering of the wind field can help identifying the suitable sites which can minimizes the similarities in observation between different sites while maximizing the oceanic contribution if the target is to identify oceanic sources. Thus, this study also provides guidance for utilizing wind from model simulation to decide on the potential sampling sites prior to cruise expedition. A more intensive observational network over BoB is needed to study the carbon cycle, role of the land-ocean interactions on climate, and the ocean’s ability as a sequester of anthropogenic CO₂ in a warming climate.

Acknowledgments

We thank the ARMEX project for conducting the CTCZ Cruise Expedition. We thank the Chief Scientists Dr. Surya Chandra Rao, IITM Pune and Dr. Vinayachandran, CAOS, Indian Institute of Science for giving us the opportunity to carry out this research. We thank the ORV Sagar Kanya and Sagar Nidhi ship management, officials and crew members for their enormous support. We thank Dr. A. Chakraborty, Prof. D. Sengupta (CAOS). Prof Raghuram Murtugudde (ESSIC, University of Maryland), Dr. Vinu Valsala (IITM, Pune) for valuable suggestions, Mrs. U. Sivanandan and Mr. R. Rangarajan for sample collection onboard Sagar Nidhi.

Ethical Approval:

The paper reflects the authors' own research and analysis in a truthful and complete manner.

Consent to Participate:

I, Tania Guha voluntarily agree to participate in this research study.

Consent to Publish:

I, Tania Guha, give my consent for the publication of identifiable details, within the text (“Material”) to be published in the above Journal.

Funding:

This work was supported by funding from the Ministry of Earth Science, Government of India, and project MoES/ATMOS/PP-IX/09. We declare that the manuscript is original and not currently being considered for publication elsewhere. We have no conflict of interest associated with the manuscript. As a corresponding author, I hereby confirm that the manuscript has been read and approved by all the co-authors for submission.

Competing Interests:

The authors have no relevant financial or non-financial interests to disclose.

Authors Contribution:

The first three authors have contributed to the study conception and design. Material preparation, data collection and analysis were performed by [Tania Guha], and [Subhomoy Ghosh]. The first draft of the manuscript was written by [Tania Guha] and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Availability of data and materials:

Data can be made available on request.

Arthur, D., & Vassilvitskii, S., 2007. k-means++: The advantages of careful seeding. Stanford. Proceedings of the eighteenth annual ACM-SIAM symposium on Discrete algorithms. Society for Industrial and Applied Mathematics Philadelphia, PA, USA., 1027–1035. https://dl.acm.org/doi/10.5555/1283383.1283494.

Clifton, A., Lundquist, J.K., 2012. Data Clustering Reveals Climate Impacts on Local Wind Phenomena. J Appl Meteorol Climatol 51(8), 1547–1557. https://doi.org/10.1175/JAMC-D-11-0227.1.

Draxler, R.R., Ginoux, P., Stein, A.F., 2010. An empirically derived emission algorithm for wind-blown dust. J. Geophys. Res. 115, D16212. https://doi.org/10.1029/2009JD013167.

Dudhia, J., 1989. Numerical study of convection observed during the Winter Monsoon Experiment using a mesoscale two–dimensional model. J. Atmos. Sci. 46, 3077–3107. https://doi.org/10.1175/1520-0469(1989)046<3077:NSOCOD>2.0.CO;2.

Feely, R.A., Sabine, C.L., Takahashi, T., Wanninkhof, R., 2001. Uptake and storage of carbon dioxide in the ocean: The global CO₂ survey. Oceanography 14(4):18–32, https://doi.org/10.5670/oceanog.2001.03.

George, M.D., Kumar, M.D., Naqvi, S.W.A., Banerjee, S., Narvekar, P.V., de Sousa, S.N., Jayakumar, D.A., 1994. A study of the carbon dioxide system in the northern Indian Ocean during premonsoon. Mar. Chem. 47, 243-254. https://doi.org/10.1016/0304-4203(94)90023-X.

Gheusi, F., Stein, J., 2002. Lagrangian description of air flows using Eulerian passive tracers. Q J R Meteorol Soc 128(579), 337–360.http://doi.org/10.1256/00359000260498914.

Guha, T., Ghosh, P., 2010. Diurnal variation of atmospheric CO₂concentration and δ¹³C in an urban atmosphere during winter—role of the Nocturnal Boundary Layer. J Atmos Chem. 65, 1-12. http://doi.org/10.1007/s10874-010-9178-6.

Guha, T., Ghosh, P., 2013. An experimental set up for carbon isotopic analyses of atmospheric CO₂ and an example of ecosystem response during solar eclipse 2010. J. Earth Syst. 122(3), 623-638. http://doi.org/10.1007/s12040-013-0303-1.

Guha, T., Ghosh, P., 2015. Diurnal and seasonal variation of mixing ratio and d¹³C of air-CO₂ observed at an urban station Bangalore, India. Envir Sci & Pollut Research 22(3), 1877-90. http://doi.org/10.1007/s11356-014-3530-3.

Guha, T., Lin, C.T., Bhattacharya, S.K., Mahajan, A.S., Ou-Yang, C.F., Lan, Y., Hsu, S.C., Liang, M.C., 2017. Isotopic ratios of nitrate in aerosol samples from Mt. Lulin, a high altitude station in Central Taiwan. Atmospheric Environment 154, 53-69. https://doi.org/10.1016/j.atmosenv.2017.01.036.

Guha, T., G., Tiwari, Y.K., Valsala, V., Mahajan, A.S., Lin, X., Ramonet, M., Datye, A., Ravi Kumar, K., 2018. What controls the atmospheric methane seasonal variability over India? Atmospheric Environment. 175, 83-91. https://doi.org/10.1016/j.atmosenv.2017.11.042.

Hong, S.Y., Dudhia, J., Chen, S.H., 2004. A revised approach to ice microphysical processes for the bulk parameterization of clouds and precipitation. Mon. Wea. Rev. 132, 103–120. https://doi.org/10.1175/1520-0493(2004)132<0103:ARATIM>2.0.CO;2.

Hong, S.Y., Noh, Y., Dudhia, J., 2006. A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Wea. Rev. 134, 2318–2341. https://doi.org/10.1175/MWR3199.1.

IPCC, 2021. Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Masson-Delmotte, V., Zhai, P., Pirani, A., Connors, S.L., Péan, C., Berger, S., Caud, N., Chen, Y., Goldfarb, L., Gomis, M.I., Huang, M., Leitzell, K., Lonnoy, E., Matthews, J.B.R., Maycock, T.K., Waterfield, T., Yelekçi, O., Yu, R., Zhou, B. (eds.)]. Cambridge University Press. In Press.

IPCC, 2013: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V., Midgley, P.M. (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535. http://doi:10.1017/CBO9781107415324.

Jain, A.K., 2010. Data clustering: 50 years beyond K-means. Pattern recognition letters 31(8), 651-666. http://doi.org/10.1016/j.patrec.2009.09.011.

Jimenez, P.A., Dudhia, J., Gonzalez–Rouco, J.F., Navarro, J., Juan P.M., Garcia–Bustamante, E., 2012. A revised scheme for the WRF surface layer formulation. Mon. Wea. Rev., 140, 898–918. https://doi.org/10.1175/MWR-D-11-00056.1.

Kain, J.S., 2004. The Kain–Fritsch convective parameterization: An update. J. Appl. Meteor. Climatol. 43, 170–181. https://doi.org/10.1175/1520-0450(2004)043<0170:TKCPAU>2.0.CO;2.

Kaufman, L., Rousseeuw, P.J., 2009. Finding groups in data: an introduction to cluster analysis 344, John Wiley & Sons, ISBN 0-471-73578-7.

Keeling, C.D., 1958. The concentration and isotopic abundance of carbon dioxide in rural areas. Geochimica Cosmochimica Acta. 13, 322-334. https://doi.org/10.1016/0016-7037(58)90033-4Get rights and content.

Kumar, M. D., Naqvi, S.W.A., George, M.D., Jayakumar, D.A., 1996. A sink for atmospheric carbon dioxide in the northeast Indian Ocean. J. Geophys. Res. 101, 18121-18125. https://doi.org/10.1029/96JC01452.

Kumar, K.R., Tiwari, Y.K., Valsala, V., Murtugudde, R., 2013. On understanding the land–ocean CO2 contrast over the Bay of Bengal: A case study during 2009 summer monsoon, Environ Sci Pollut Res 21(7), 5066-75. http://doi.org/10.1007/s11356-013-2386-2.

Le Quéré, C., Raupach, M., Canadell, J., et al. 2009. Trends in the sources and sinks of carbon dioxide. Nature Geoscience 2, 831–836. https://doi.org/10.1038/ngeo689.

Le Quéré, C., Andrew, R.M., Canadell, J.G., Sitch, S., Korsbakken, J.I., Peters, G.P., et al. 2016. Global carbon budget 2016. Earth System Science Data, 8, 605–649. https://doi.org/10.5194/essd-8-605-2016.

Le Quéré, C., Andrew, R.M., Friedlingstein, P., Sitch, S., Hauck, J., Pongratz, J., et al. 2018. Global carbon budget 2018. Earth Syst. Sci. Data Discuss. 2018, 1–3. http://doi: 10.5194/essd-2018-120.

Li, K.F., Tian, B.,Waliser, D.E., Yuk Yung, L., 2010. Tropical midtropospheric CO₂ variability driven by the Madden–Julian Oscillation. Proc Natl Acad Sci 107, 19,171–19,175. http://doi.org/10.1073/pnas.1008222107.

Liu, K., Tianning, W., Wang, X., Song, Z., Zong, C., Wei, N., Li, D., 2019. Environ Sci Technol 53 (18), 10612-10619. http://doi.org/10.1021/acs.est.9b03427.

Mlawer, E.J., Taubman, S.J., Brown, P.D., Iacono, M.J., Clough, S.A.,1997. Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated–k model for the longwave. J. Geophys. Res. 102, 16663–16682. http://doi.org/10.1029/97JD00237.

Miller, J.B., Tans, P.P., 2003. Calculating isotopic fractionation from atmospheric measurement at various scales. Tellus 55B, 207–214. https://doi.org/10.3402/tellusb.v55i2.16697.

Mouselimis, L., 2020. ClusterR: Gaussian Mixture Models, K-Means, Mini-Batch-Kmeans, K-Medoids and Affinity Propagation Clustering. R package version 1.2.2. https://CRAN.R-project.org/package=ClusterR.

Pataki, D.E., Ehleringer, J.R., Flanagan, L.B., Yakir, D., Bowling, D.R., Still, C.J., Buchmann, N., Kaplan, J.O., Berry, J.A., 2003. The application and interpretation of Keeling plots in terrestrial carbon cycle research. Global Biogeochemical Cycles 17 (1), 1022. https://doi.org/10.1029/2001GB001850.

Peters, W., Jacobson, A.R., Sweeney, C., Arlyn, E.A., Conway, T.J., Masarie, K., Miller, J.B., Bruhwiler, L.M.P., Petron, G., Hirsch, A.I., Douglas, E.J.W., van der Werf G.R., Randerson, J.T., Wennberg, P.O., Krol, M.C., Tans, P.P., 2007. An atmospheric perspective on North American carbon dioxide exchange: CarbonTracker. P Natl Acad Sci 104, 18925–18930. https://doi.org/10.1073/pnas.0708986104.

Pham, D.T., Dimov, S.S., Nguyen, C.D., 2005. Selection of K in K-means clustering. Proceedings of the Institution of Mechanical Engineers, Part C. Journal of Mechanical Engineering Science 219(1), 103-119. http://doi.org/10.1243/095440605X8298.

Sarma, V.V.S.S., 2003. Monthly variability in the surface pCO₂ and net air–sea CO₂ flux in the Arabian sea. J. Geophys. Res. 108, 3255. http://doi.org/10.1029/2001JC001062.

Sarma, V.V.S.S., Krishna, M.S., Rao, V.D., Viswanadham, R., Kumar, N.A., Kumari, T.R., Gawade, L., Ghatkar, S., Tari, A., 2012. Sources and sinks of CO₂ in the west coast of Bay of 10 Bengal, Tellus B, 64, 10961. http://doi.org/10.3402/tellusb.v64i0.10961.

Sarma, V.V.S.S., Lenton, A., Law, R., Metzl, N., Patra, P.K., Domey, S., Lima, I.D., Dlugokencky, E., Ramonet, M., Valsala, V., 2013. Sea-air CO₂ fluxes in the Indian Ocean between 1990 and 2009. Biogeosciences Discussions 10 (7), 10759-10810. http://doi.org/10.5194/bgd-10-10759-2013.

Sariano, C., Jorba, O., Baldasano, J.M., 2004. One-Way Nesting Versus Two-Way Nesting: Does It Really Make a Difference? In book: Air Pollution Modeling and Its Application XV. http://doi.org/ 10.1007/0-306-47813-7_18.

Sengupta, D., Bharath Raj, G.N., Shenoi, S.S.C., 2006. Surface freshwater from Bay of Bengal runoff and Indonesian Throughflow in the tropical Indian Ocean. Geophys. Res. Lett., 33, L22609. http://doi.org/10.1029/2006GL027573.

Skamarock, W.C., Klemp, J.B., Dudhia, J., Gill, D.O., Barker, D.M., Duda, M.G., Huang, X.-Y., Wang, W., Powers, J.G., 2008. A Description of the Advanced Research WRF Version 3. NCAR Tech. Note NCAR/TN-475+STR, 113. http://dx.doi.org/10.5065/D68S4MVH.

Siegenthaler, U., Miinnich K.O., 1981. ¹³C/¹²C fractionation during CO₂ transfer from air to sea. In Scope 16: Bolin B brium constants for the dissociation of carbonic acid in sea water media. Deep Sea Res 34, 1733-1743. https://scope.dge.carnegiescience.edu/SCOPE_16/SCOPE_16_1.5.05_Siegenthaler_249-257.pdf.

Sokal, R., Rohlf, J., 1995. Biometry: The Principles and Practice of Statistics in Biological Research. 3rd Edition, W.H. Freeman and Co., New York.

Sturm, P., Leuenberger, M., Sirignano, C., Neubert, R.E.M., Meijer, H.A.J., Langenfelds, R., Brand, W.A., Tohjima, Y., 2004. Permeation of atmospheric gases through polymer O-rings used in flasks for air sampling. J Geophys Res 109, D04309. http://doi.org/10.1029/2003jd004073.

Takahashi, T., Feely R.A., Weiss R.F., Wanninkhofs R.H., Chipman D.W., Sutherland S.C., Takahashi T.T., 1997. Global air-sea flux of CO₂: An estimate based on measurements of sea–air pCO₂ difference. Proc. Natl. Acad. Sci. 94, 8292-8299. https://doi.org/10.1073/pnas.94.16.8292.

Takahashi, T., Sutherland, S.C., Wanninkhof, R., Sweeney, C., Feely, R.A., Chipman, D.W., Hales, B., Friederich, G., Chavez, F., Sabine, C., Watson, A., Bakker, D.C.E., Schuster, U., Metzl, N., Yoshikawa-Inoue, H., Ishii, M., Midorikawa, T., Nojiri, Y., Kortzinger, A., Steinhoff, T., Hoppema, M., Olafsson, J., Arnarson, T. S., Tilbrook, B., Johannessen, T., Olsen, A., Bellerby, R., Wong, C.S., Delille, B., Bates, N.R., and de Baar, H.J.W., 2009. Climatological mean and decadal change in surface ocean pCO₂ , and net sea-air CO₂ flux over the global oceans. Deep-Sea Research Part. II, 56, 554–577, http://doi.org/10.1016/J.Dsr2.2008.12.009.

Tewari, M., Chen, F., Wang, W., Dudhia, J., LeMone, M.A., Mitchell, K., Ek, M., Gayno, G., Wegiel, J., Cuenca, R.H., 2004. Implementation and verification of the unified NOAH land surface model in the WRF model. 20th conference on weather analysis and forecasting/16th conference on numerical weather prediction, 11–15. http://n2t.net/ark:/85065/d7fb523p.

Valsala, V., Maksyutov, S., 2010. Simulation and assimilation of global ocean pCO₂ and air–sea CO₂ fluxes using ship observations of surface ocean pCO₂ in a simplified biogeochemical offline model. Tellus B, 62, 821–840. https://doi.org/10.1111/j.1600-0889.2010.00495.x.

Valsala, V., Maksyutov, S., Murtugudde, R., 2012. A window for carbon uptake in the southern subtropical Indian Ocean. Geophys. Res. Lett. 39, L17605, http://doi.org/10.1029/2012GL0528570.

Valsala, V., Maksyutov, S., 2013. Inter-annual variability of the air–sea CO₂ flux in the north Indian Ocean. Ocean Dynam. 63, 165–178. http://moeseprints.incois.gov.in/id/eprint/374.

Wanninkhof, R., 1985. Kinetic fractionation of the carbon isotopes ¹³C and ¹²C during transfer of CO₂ from air to seawater. Tellus B 37B(3). https://doi.org/10.1111/j.1600-0889.1985.tb00061.x.

Webster, P.J., 1995. The annual cycle and the predictability of the tropical coupled ocean atmosphere system. Meteorol Atmos Phys 56, 33–55. https://doi.org/10.1007/BF01022520.

Wendeberg, M., Richter, J.M., Rothe, M., Brand, W.A., 2013. Jena Reference Air Set (JRAS): a multi-point scale anchor for isotope measurements of CO₂ in air. Atmos. Meas. Tech., 6, 817–822. http://doi.org/10.5194/amt-6-817-2013.

Ye, H., Sheng, J., Tang, D., Morozov, E., Kalhoro, M.A., Wang, S., Xu, H., 2018. Examining the impact of tropical cyclones on air-sea CO₂ exchanges in the Bay of Bengal based on satellite data and in situ observations. Journal of Geophysical Research: Oceans, 123.https://doi.org/10.1029/2018JC014533.

Zhang, J., Quay, P.D., Wilbur, D.O., 1995. Carbon isotopic fractionation during gas-water exchange and dissolution of CO₂. Geochimica Cosmochimica Acta. 59(1), 107-114. https://doi.org/10.1016/0016-7037(95)91550-D.

Supplementary Material is not available with this version.

PDF

Version 1

posted 02 Aug, 2022

You are reading this latest preprint version

Role of surface CO₂ sources and wind transport on atmospheric CO₂ variability over the Bay of Bengal during Southwest Monsoon - using mixing ratio, carbon isotope ratio and model simulated wind field

Status:

Version 1

Abstract

Figures

1. Introduction

2. Methods

2.1 Air sample collection, extraction, and analysis of CO₂:

2.2 Simulated CO₂ from Carbon Tracker

2.3 WRF Model simulation and clustering analysis

3. Results And Discussion

3.1 Identifying the land contribution of CO₂ from Carbon Tracker:

3.2 Identifying the isotopic signature of the source-CO₂

3.3 k-means clustering of wind to explore the dominant transport to the sampling sites

4. Conclusions

Declarations

References

Supplementary Information

Status:

Version 1

Role of surface CO2 sources and wind transport on atmospheric CO2 variability over the Bay of Bengal during Southwest Monsoon - using mixing ratio, carbon isotope ratio and model simulated wind field

Status:

Version 1

Abstract

Figures

1. Introduction

2. Methods

2.1 Air sample collection, extraction, and analysis of CO2:

2.2 Simulated CO2 from Carbon Tracker

2.3 WRF Model simulation and clustering analysis

3. Results And Discussion

3.1 Identifying the land contribution of CO2 from Carbon Tracker:

3.2 Identifying the isotopic signature of the source-CO2

3.3 k-means clustering of wind to explore the dominant transport to the sampling sites

4. Conclusions

Declarations

References

Supplementary Information

Status:

Version 1

Role of surface CO₂ sources and wind transport on atmospheric CO₂ variability over the Bay of Bengal during Southwest Monsoon - using mixing ratio, carbon isotope ratio and model simulated wind field

2.1 Air sample collection, extraction, and analysis of CO₂:

2.2 Simulated CO₂ from Carbon Tracker

3.1 Identifying the land contribution of CO₂ from Carbon Tracker:

3.2 Identifying the isotopic signature of the source-CO₂