Carbon Dioxide – Continental shelves as a Variable Though Increasing Sink for Atmospheric Carbon Dioxide

Australia: The Land Where Time Began

According to Laruelle et al. it has been speculated that in shelf waters the partial pressure of atmospheric CO₂ may lag the rising atmospheric CO₂ concentration. In this paper Laruelle et al. show that this is the case across many shelf regions, which implies that there is a tendency for atmospheric carbon dioxide to be taken up. This analysis is based on long term trends in the atmosphere-ocean gradient of pCO2 (ΔpCO₂) by the use of a global surface ocean database pCO₂ that spans a period of up to 35 years. It was found, by the use of winter data only, that ΔpCO₂ increased in 653 of the 825 0.5^o cells for which a trend could be calculated, with a significant increase being shown that was greater than +0.05 μatm/yr (p<0.05). It is suggested by deseasonalised annual data that the results are similar, though noisier. The idea that shelves might have switched from a source to a sink of CO₂ over the last century would be supported by this being a global trend.

It has been found that the atmosphere’s partial pressure of CO₂ (pCO_2,air) has been increasing at about 1.8 ppm by volume (ppmv) per year over recent decades, which has resulted from human activities, such as burning fossil fuel, deforestation and the production of cement (Takahashi et al., 2009; IPCC). The pCO₂levels in surface water tended to have followed, more or less, those of the atmosphere, in the open ocean (Landschützer, Gruber & Bakker, 2016) in particular, though there is substantial variability, regional and decadal, have been observed (Fay & McKinley, 2013; Feely, 2006). This tracking trend is shown best by the data that were collected at regular intervals at a few ocean time series stations, which now cover more than 30 years (Bates et al., 2014). The close atmospheric tracking of surface water pCO₂ is a consequence of the water residence time of the global surface ocean, which is relatively long, with a time scale that is more than 1 year (Craig, 1963), which is longer than the time scale of about 10 months (Craig, 1963) of the air-sea exchange of CO₂. It is not clear, however, if the pCO₂ of surface water on continental shelves, defined in this paper as shallow regions where the depths are between 20 and 200 m that do not include the very near shore areas, also track the atmospheric pCO₂ increase.

The current understanding of the long-term trend in the pCO₂ of shelves is very limited, as it relies largely on observation from a few time series with records that are shorter than those of the open ocean. Also, the pCO₂ in shelf regions is characterised by high temporal and spatial variability, which makes trend analysis more demanding (8-12). A complementary approach to assessing whether continental shelves show a change in air-sea pCO₂ gradient (ΔpCO₂) = pCO_2,air – pCO₂) over time, is offered by the recent development of a community-driven global ocean pCO₂ data product SOCAT (Surface Ocean CO₂ Atlas (Bakker et al., 2016) ). It allows the reconstructing of the evolution in ΔpCO₂ for 125 regions across the global shelves over a time span of at least a decade, though the data coverage remains sparse within SOCAT. The first aim of the study was to identify if the recent trends that had been observed in ΔpCO₂ support a strengthening or a weakening of the global uptake of CO₂ by shelf regions. Then it was investigated whether there is an emergence of important regional differences from the analysis and if they could discern any global pattern when all the observational evidence was combined. Following this, they reviewed the current knowledge with regard to CO₂ dynamics and then proposed novel observational evidence of rates of change in the air-sea CO₂ gradient from the analysis of the SOCAT database.

It is suggested by syntheses in recent decades that, on a global scale, atmospheric CO₂ is currently being absorbed at a rate of about 0.2 Pg C (Pg = petagram – 10¹⁵ of C) annually (Cai, Dai & Wang, 2006; Borges, Delille & Frankignoulle, 2005; Cai, 2011; Chen et al/. 2013; Laruelle, Dürr, Slomp & Borges, 2010; Laruelle, Lauerwald, Pfeil & Regnier, 2014; Wanninkhof et al., 2013). It is also suggested by the data that mid- to high-latitude shelves are generally a sink for CO₂, while the warm tropical shelves are a moderate source of CO₂ (Cai, 2011; Laruelle, Dürr & Slomp & Borges, 2010; Laruelle et al., 2014), in spite of a high degree of local variability. Therefore, a consensus has emerged with regard to the current strength of the global shelf CO₂ sink and its large-scale variability. This spatial trend in all oceanic basins is clearly supported by continuous high-resolution pCO2 maps for seas of continental shelves that are derived from interpolation of experimental data (Laruelle et al., 2017). Much less is known, however, regarding decadal trends and associated variability in shelf sources of CO₂ and sinks around the globe. Mixed evidence for the magnitude of decadal trends is provided by the limited pCO₂ time series that has been obtained from coastal sites. For the coastal stations Mundia and Iceland Sea it was reported (Bates et al., 2014) small long-term increase rates in pCO₂ (1.3 μatm per year), i.e., a rate that is lower than that of the atmosphere, while showing that the stations Irminger and CARIACO have rates as high as +2.4 and +2.9 μatm per year, respectively. At the SEATS station in the South China Sea a shorter time series over the 1999-2003 period reveals an increase in pCO₂ that is even faster with a rate of +4.2 μatm per year. Such trends from a small number of locations do not allow any conclusion to be drawn in regards to the overall change in shelf air-sea pCO₂ gradient over time, though they are illustrative.

Some data-driven analyses have also been attempted in order to decipher the rate of increase of pCO₂ in continental shelf settings. It was suggested by data from 2 large shelf seas that were semi-enclosed, the North Sea and the Baltic Sea, and from the Bering Sea that a rapid increase in pCO₂ may be exhibited (Thomas et al., 2007; Tseng et al., 2007) towards atmospheric values, thereby lowering the air-sea pCO₂ gradient over time. Contrasting with this, another study carried out in the North Sea (Salt et al., 2013) and reports from the warm Caribbean Sea (Park & Wanninkhof, 2012) which were mostly from areas that are deeper than the shelf depths, as defined for this study, the coast of Japan (Ishii et al., 2011), West Antarctica Peninsula (Hauri et al., 2015), and the Scotia Shelf (Shadwick et al., 2010), showed that the increase in the pCO₂ of the surface of the ocean lags well behind that of the atmosphere, which makes the areas either an increased sink (Pacific coast of Japan, the Coast of West Antarctic Peninsula, and Puerto Rico) or a source that is decreased (Scotian Shelf) for atmospheric CO₂. It is suggested by a recent study, however, that the margin of Japan as a whole tracks roughly the increase in atmospheric CO₂. These regional analyses, overall, highlight that trends in CO₂ sources and sinks appear to be highly variable, both within the same shelf and across different shelf systems.

Attempts have also been made to investigate the change in shelf air-water CO₂ exchange by the use of models. Mackenzie et al. used a box model to suggest that shelves may have turned from being a CO₂ source in the preindustrial time to a sink at the present and that the rate of CO₂ uptake would increase over time (Andersson et al., 2004). A conceptual model was provided by Bauer et al. (Bauer et al., 2013) & Cai (14), suggesting an increasing global shelf sink of CO₂ with time as a result of the pCO₂ atmospheric increase. An eddy-resolving global model was used recently to simulate the flux of anthropogenic CO₂ into the coastal ocean (Bourgeois et al., 2016). According to Laruelle et al. the latter can be viewed as an open ocean model that has been extended to the coast that is lacking a few, though important processes in the nearshore environments. In particular, the global model is lacking in detailed sediment interactions, the handling of river fluxes, and processes of shallow calcification, which were, however, captured in the box model that was spatially and temporally crude (Bauer et al., 2013; Mackenzie & Lerman, 2011). Nonetheless, according to Laruelle et al. it is consistently shown by both approaches that the CO₂ uptake by shelf water increases with increasing atmospheric CO₂ concentrations. No consensus emerges, however, as to whether past and future decadal changes in shelf ΔpCO₂ and, therefore, absorption of CO₂ per unit area, will increase at a rate that is faster or slower when compared with the global open ocean.

There are 2 main mechanisms that have been proposed to explain the evolution of the CO₂ sink on the continental shelf. The first mechanism relies on the efficiency of the physical pump and more specifically, on different timescales of the air-water and on exchanges of CO2 in shelf water-open water (Bauer et al., 2013; Cai, 2011). The pCO₂ increase in waters on the shelves may be slower than the atmospheric pCO₂ increase, even if it is assumed that no change in biology and physics occurs over time, in situations where the CO₂ exchange rate across the shelf is faster than with the atmosphere (Bauer et al., 2013; Cai, 2011). The accumulation of anthropogenic CO₂ in shallow waters in these margins would be limited and would help maintain a significant gradient of air-water CO₂ which would favour an efficient uptake of anthropogenic CO₂. Contrasting with this, where the cross-shelf export is not able to keep up with the air-sea flux of anthropogenic CO₂ that is increasing, CO₂ may accumulate and the increase in CO₂ would follow the atmosphere as a result of this bottleneck in offshore transport (Bourgeois et al., 2016).

The stimulation of the biological pump is relied on by the second mechanism. Anthropogenic inputs seriously influence many continental shelves and have a higher biological production today than what they had in preindustrial times (Mackenzie, Lerman & DeCarlo, 2011; Walsh, 1988). Therefore, according to Laruelle et al. on the shelves net ecosystem metabolism (NEM) could have shifted from net heterotrophy to net autotrophy and the change could have been sufficiently large to reverse the air-sea CO₂ flux from a source in preindustrial times to a sink under conditions of the present. On the shelves, net ecosystem calcification (NEC) also plays a significant role in the air-water CO₂ exchange, though the contribution of the carbonate pump to changes in air-water fluxes over the historical period are likely to be smaller than the biological pump (Andersson & Mackenzie, 2001).

The aim of this paper was to present the first analysis that was observation based of decadal trends in global shelf pCO₂. The results presented in the regional and global analysis of Laruelle et al. are derived primarily from winter data when the photosynthetic activity is generally at its weakest, and when there are the most intensive exchanges of coastal ocean waters with the open ocean, and as a consequence, the strongest impact on the accumulation of CO₂ in the global ocean (Laruelle & Lauerwald, 2014; Walsh, 1988). Trends result from this that tend to be clearer. In order to check on these wintertime analyses Laruelle et al. also used results from an analysis using deseasonalised data for all seasons, confirming that their choice for wintertime only does not result in artefacts. This does not, however, suggest that winter contributes more than other seasons to overall annual trend.

Sources & Further reading