Projections of ocean acidification over the next three centuries using a simple global climate carbon-cycle model

Introduction Conclusions References

ning the cumbersome earth system models, we can use a reduced-form model to quickly emulate the CMIP5 models for projection studies under arbitrary emission pathways and for uncertainty analyses of the marine carbonate system.In this study we highlight the capability of Hector v1.1, a reduced-form model, to project changes in the upper ocean carbonate system over the next three centuries.Hector is run under historical emissions and a high emissions scenario (Representative Concentration Pathway 8.5), comparing its output to observations and CMIP5 models that contain ocean biogeochemical cycles.Ocean acidification changes are already taking place, with significant changes projected to occur over the next 300 years.We project a low latitude (> 55 • ) surface ocean pH decrease from preindustrial conditions by 0.4 units to 7.77 at 2100, and an additional 0.27 units to 7.50 at 2300.Aragonite saturations decrease by 1.85 units to 2.21 at 2100 and an additional 0.80 units to 1.42 at 2300.Under a high emissions scenario, for every 1 • C of future warming we find a 0.107 unit pH decrease and a 0.438 unit decrease in aragonite saturations.Hector reproduces the global historical trends, and future projections with equivalent rates of change over time compared to observations and CMIP5 models.Hector is a robust tool that can be used for quick ocean acidification projections, accurately emulating large scale climate models under multiple emission pathways.

Introduction
Human induced activities have led to increasing anthropogenic emissions of greenhouse gases to the atmosphere.In the first decade of the 21st century CO 2 emissions from anthropogenic sources and land use changes accounted for ∼ 9 Pg C yr −1 , with future emission projections of up to 28 Pg C yr −1 by 2100 under a high emissions scenario (RCP 8.5) (Riahi et al., 2011).In response to this increasing atmospheric burden of CO 2 , the oceans are experiencing both physical and biogeochemical changes: surface and deep warming, changes in calcium carbonate saturations, and a decline in pH (Doney, 2010).The oceans represent a major carbon sink and have absorbed 25-30 % of the total anthropogenic carbon emissions since 1750 (Le Quéré et al., 2013;Sabine et al., 2011).There is some concern due to both the physical and chemical changes that the oceans will be less efficient in the uptake of anthropogenic CO 2 as the climate continues to change (Sarmiento and Le Quéré, 1996;Matear and Hirst, 1999;Joos et al., 1999;Le Quéré et al., 2007, 2010).
In particular, the ocean chemistry is quickly changing in response to the continued addition of CO 2 to the atmosphere.Mean surface ocean pH has decreased by 0.1 units relative to the preindustrial (Caldeira et al., 2003).If emission trends continue, ocean acidification will occur at rates and extents that have not been observed over the last few million years (Feely et al., 2004(Feely et al., , 2009;;Kump et al., 2009;Caldeira et al., 2003).
Ocean acidification occurs when CO 2 dissolves in seawater, forming H 2 CO 3 , dissoci- 3 , and 0.5 % in the form of CO 2 (aq), these percentages will change as the oceans take up more carbon.
With declining CO 2− 3 , the stability of biogenic carbonate (CaCO 3 ) is also reduced which is the primary mineral used by marine organisms to build their shells and skeletons.Numerous experiments and observations indicate that a reduction in the surface ocean pH and carbonate saturations may have significant effects on calcifying marine organisms, e.g., phytoplankton and coral reefs, from changes in community structure to suppressed calcification rates (Feely et al., 2004;Riebesell et al., 2000;Fabry et al., 2008;Kleypas et al., 2005;Raven et al., 2005;Dutkiewicz et al., 2015).For example, some coral reefs are believed to already be eroding for parts of the year due to changes in ocean acidification (Yates and Halley, 2006;Albright et al., 2013).Modeling studies project changes in ocean acidification for both the surface and deep oceans.By the end of the century all existing coral reefs will be surrounded by ocean chemistry conditions that are well outside of the preindustrial values and even today's saturations (Ricke et al., 2013).Global surface pH is projected to drop by up to 0.4 units under IS92a scenario and even deep waters will experience pH changes of up to 0.3 units by 2100 under RCP8.5 (Gehlen et al., 2014;Orr et al., 2005).Ocean warming and acidification are consistent across all of the CMIP5 models, however the intensities are strongly dependent upon the scenario (Bopp et al., 2013).These studies indicate that the oceans are already experiencing and will continue to experience significant changes in ocean acidification.
However, these model projections of ocean acidification are primarily from earth system models (ESMs) run under prescribed emission pathways (e.g., RCPs); limiting the analyses that can be conducted to only those scenarios.Here we present Hector, a reduced-form climate model that can emulate the median climate of CMIP5.
Reduced-form models are powerful tools due to their computational efficiency, inexpensiveness to run, and ability to run multiple simulations under arbitrary future climate change emission pathways, allowing us to conduct parameter sensitivity studies Introduction

Conclusions References
Tables Figures

Back Close
Full and uncertainty analyses (Senior and Mitchell, 2000;Ricciuto et al., 2008;Irvine et al., 2012).This study builds upon Hartin et al. (2015), which introduced Hector v1.0, an opensource, object-oriented simple climate model with the capabilities of projecting changes in the surface ocean carbonate system over the next three centuries.This work is timely due to the fact that the recent CMIP5 process included numerous ESMs that contain dynamic ocean biogeochemistry.
Other simple models have modeled the complexity of the nonlinear carbonate system through mixed layer Impulse Response Functions (IRF) calculating air-sea fluxes (Joos et al., 1996(Joos et al., , 2001;;Meinshausen et al., 2011) and evaluating the parameters of the carbonate system by back calculating from the ocean uptake of CO 2 (Tanaka et al., 2007;Harman et al., 2011).The IRF method has been widely used across the scientific community, as it is cost-effective to run, provides surface to deep mixing estimates, and can also be used to look at oceanic uptake of conservative tracers.However, the carbonate system is not directly calculated and many effects like temperature effects on CO 2 solubility are typically parameterized.The carbonate system is strongly dependent upon temperature where pCO 2 changes by about 4.2 % per Kelvin (Copin-Montegut, 1988;Takahashi et al., 1993).While these models are able to reproduce changes in the global climate system, details in the carbonate system (HCO 2− 3 , CO 3 , pH, pCO 2 , and alkalinity) are not actively solved for.
The remainder of the paper is organized as follows.Section 2 describes the model components focusing on the ocean carbon cycle and carbonate chemistry.Section 3 presents the model experiments and comparison data used and lastly, Sect. 4 describes the main results and a discussion.

Model description -carbon cycle
The carbon component in Hector contains three carbon reservoirs: a single well-mixed atmosphere, a land component consisting of vegetation, detritus, and soil, and an Introduction

Conclusions References
Tables Figures

Back Close
Full ocean component consisting of four boxes (high and low surface boxes, an intermediate box, and a deep box) (Fig. 1).The change in atmospheric carbon is a function of the anthropogenic emissions (F A ), land-use change emissions (F LC ), and atmosphereocean (F O ) and atmosphere-land (F L ) carbon fluxes.The default model time step is 1 year.
The terrestrial cycle in Hector contains vegetation, detritus, and soil, all linked to each other and the atmosphere by first-order differential equations.Vegetation net primary production is a function of atmospheric CO 2 and temperature.Carbon flows from the vegetation to detritus and to soil and loses fractions of carbon to heterotrophic respiration on the way.An "earth" pool debits carbon emitted as anthropogenic emissions, allowing a continual mass-balance check across the entire carbon cycle.Atmosphereland fluxes at time t are calculated by: where NPP is the net primary production and RH is the heterotrophic respiration summed over user-specified n groups (i.e., latitude bands, political units, or biomes) (Hartin et al., 2015).

Ocean component
The ocean component is modeled after Lenton (2000), Knox and McElroy (1984) and Sarmiento and Toggweiler (1984) Once carbon enters the system it is circulated between the boxes via advection and water mass exchange, simulating a simple thermohaline circulation.We do not explicitly model diffusion.Simple box-diffusion models and "HILDA" (e.g., Siegenthaler and Joos, 1992) type models, are typically in good agreement with observations but are more computationally demanding than a simple box model (Lenton, 2000).The change in carbon of any box i is given by the fluxes in and out, with F atm→i as the atmospheric carbon flux: More specifically, the change in change in carbon of box i is related to the transport (T i →j ) in Sverdrups (Sv -m 3 s −1 ) between i and j , the volume of i (V i ), and the total carbon in i (including any air-sea fluxes) (C i ); dynamics of ocean uptake of CO 2 is strongly dependent on this downward transport rate of CO 2 laden waters from the surface ocean to depth.There are 4 measurable parameters of the carbonate system in seawater: DIC, TA, pCO 2 and pH, and any pair can be used to describe the entire carbonate system.Within Hector, DIC and TA are used to solve for surface ocean pH and pCO 2 values.These detailed carbonate chemistry equations are based on numeric programs from Zeebe and Wolf-Gladrow, 2001 (A).We have simplified these equations by neglecting the effects of pressure, since we are only concerned with the surface ocean.A best-fit alkalinity (2311.0mol kg −1 for HL and 2435.0 mol kg −1 for LL), is solved for at the end of spinup, that when calculated with an initial DIC input for each surface box results in a pre-industrial net zero flux of carbon over the global ocean.Hector is run to equilibrium in a perturbation-free mode, prior to running the historical period, ensuring that Hector is in steady-state (Hartin et al., 2015;Pietsch and Hasenauer, 2006).
where k is the CO 2 gas-transfer velocity, α is the solubility of CO 2 in seawater (K 0 ), and the ∆pCO 2 is the difference in [CO 2 ] between the atmosphere and ocean.The 585 is a unit conversion factor and Sc is the Schmidt number.The Schmidt number (A1) is calculated from Wanninkhof (1992) based on the temperature of each surface box.The average wind speed (U 10 ) of 6.7 m s −1 is the same over both surface boxes (Table 1).We assume surface waters are fully equilibrated with the overlying atmosphere given our time step of 1 year; the average time for surface waters to come into equilibrium (Broecker and Peng, 1982).pH (total scale), HCO − 3 , and CO 2− 3 are calculated using the H + ion, solved for in a higher order polynomial (A1).Aragonite and calcite are the two primary carbonate minerals within seawater.The degree of saturation in seawater with respect to calcite (Ω Ca ) and aragonite (Ω Ar ) is calculated from the product of the concentrations of calcium [Ca  2+ ] and carbonate ions 3 ], divided by the solubility (K sp ).The calcium concentration is based on equations from Riley and Tongudai (1967) at a constant salinity of 35.If Ω = 1, the solution is at equilibrium, and if Ω > 1 (Ω < 1) the solution is supersaturated (undersaturated) with respect to the mineral.

Model experiments and data sources
The Hector code is open-source and available at https://github.com/JGCRI/hector.The repository includes all model code needed to compile, as well as, all input files and R scripts to process the model output.For this study we run Hector v1.1, Git Commit #, with updated ocean temperature to better match the CMIP5 mean.Hector is run under prescribed emissions from 1850-2300 for all four Representative Concentration Pathways (RCP 2.6, RCP 4.5, RCP 6, RCP 8.5) (Moss et al., 2010;Fujino et al., 2006;van Vuuren et al., 2007;Clarke et al., 2007;Wise et al., 2009;Riahi et al., Introduction Conclusions References Tables Figures
Comparison data is obtained from a series of observational surface data and a suite of CMIP5 models.Surface ocean observations of DIC, pCO 2 , pH, Ω Ar , and Ω Ca are from ocean time-series stations in both the high and latitude oceans; Hawaii Ocean Time Series (HOT), Bermuda Atlantic Time Series (BATS), the European Station for Time Series in the Ocean at the Canary Islands (ESTOC), the Irminger Sea, and the Iceland Sea (Table 3) (Bates, 2007;Fujieki et al., 2013;Dore et al., 2009;Santana-Casiano et al., 2007;Olafsson, 2007a, b;Gonzalez-Davila, 2009).The time-series data are annually averaged over the upper 100 m of the water column.The carbonate parameters not found in Table 3 are computed from temperature, salinity, and carbonate parameter pairs using the CO2SYS software (Lewis and Wallace, 1998).Lastly, a longer record  of pH and Ω Ar from Flinder's Reef in the western Coral Sea is used in the comparison (Pelejero et al., 2005).We use rates of change (∆) over a 20 year period  to quantify how well Hector does at simulating the observed changes in the ocean carbonate parameters.
We also compare Hector to a suite of 15 CMIP5 Earth System Models (Table 4) (Taylor et al., 2012).The CMIP5 output is available from the Program for Climate Model Diagnostics and Intercomparison (http://cmip-pcmdi.llnl.gov/cmip5/).The CMIP5 data are converted to annual global, high latitude and low latitude averages over the upper 100 m of water depth, with one standard deviation of the annual averages and CMIP5 model range calculated using the RCMIP5 package (https://github.com/JGCRI/RCMIP5).All CMIP5 comparisons used in this study are from model runs with prescribed atmospheric concentrations.We acknowledge that this is not a perfect comparison between emissions forced Hector and concentration forced CMIP5.However, very few CMIP5 models were run under prescribed emissions.We use a combination of root mean square error (RMSE) and rates of change (∆) as our metrics to characterize how well Hector compares to the CMIP5 median.Introduction

Conclusions References
Tables Figures

Back Close
Full  .Hector captures the trend in DIC concentrations for both the high and low latitude surface ocean with an average RMSE of 4.6 µmol kg −1 when compared to CMIP5 models over the historical period (Fig. 2).Low latitude DIC is slightly higher than the CMIP5 range, but rates of change are similar between 1850 and 2100, 1.27 yr −1 for Hector and 1.24 yr −1 for CMIP5 (Table 5).To obtain a steady state, Hector is initialized with carbon values slightly higher than the average CMIP5 values.Hector accurately tracks the pCO 2 in both the high and low latitude surface ocean with similar rates of change from 1850-2300 (Fig. 3).There is a low bias in Hector compared to CMIP5 models after 2100, highlighted by the higher RMSE; 1.4 µatm between 2006-2100 increasing to 6.0 µatm between 2006-2300.This is due to the low bias in projected atmospheric [CO 2 ] over the same time period (Hartin et al., 2015).
The oceanic uptake of CO 2 since the preindustrial has caused the marine carbonate system to shift to lower pH and lower [CO 2− 3 ].Hector accurately captures the decline in pH compared to CMIP5 and observations from BATS, HOT, ESTOC, Irminger Sea, Iceland Sea, and Flinders Reef (Fig. 4).Since the pre-industrial, surface ocean pH decreased by 0.08 units, corresponding to a 24 % increase in  , 2004;Sabine et al., 2004;Caldeira et al., 2003;Orr et al., 2005) that estimate an average global decrease in pH of 0.1 or a 30 % increase in H + .The Flinder's Reef pH record provides a natural baseline to compare future trends in ocean acidification.While we don't expect to match exactly, as this reef site is heavily influenced by coastal dynamics, internal variability, and upwelling, rates of change from the preindustrial (1750) to 1988 are the same (0.0002 yr −1 ) for both Hector and Flinder's Reef (Table 5).Over the limited observational record from both the Pacific and Atlantic Oceans (1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006), Hector accurately simulates the change in pH (−0.0015 yr  5) and findings from intermediate complexity models (Montenegro et al., 2007).At approximately 2050, atmospheric [CO 2 ] is double the pre-industrial concentrations, corresponding to a 0.20 pH decrease to 7.96.Shortly after this doubling, pH values are well outside the lowest observed natural variability found in Flinder's Reef.Aragonite and calcite are forms of biogenic calcium carbonate.Formaninferia and coccolithophorids are composed of calcite the less soluble form of biogenic calcium carbonate, while corals and pteropods are composed of aragonite.Hector accurately simulates the decline in saturations (Ω Ar and Ω Ca ) from 1850-2300 under RCP8.5 to CMIP5 and observations (Fig. 5).Since the preindustrial, surface low (high) latitude Ω Ar decreased by 0.4 (0.3) units, with similar rates for CMIP5.Rates of change over a 14 year period for Hector (−0.007 yr to aragonite and calcite over a 14 year period (1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005) (Feely et al., 2012), while the average decrease in Hector is between 19 and 25 %.Saturations Of both Ar and Ca decrease rapidly over the next 100 years in both the high and low latitude.Hector accurately captures the decline in saturations with low RMSE values for both Ω Ar (0.027) and Ω Ca (0.012).Under RCP8.5 Hector projects that low latitude Ω Ar will decrease by 1.85 units to 2.21 by 2100 and by 2.6 units to 1.42 by 2300.For low latitude Ω Ca , Hector projects low latitude Ω Ca decrease by 2.88 units to 3.34 by 2100 and by over 4.09 units to 2.31 by 2300.A lowering of Ω Ar from approximately 4 to 3 is predicated to lead to significant reductions in calcification rates in tropical reefs (Kleypas et al., 1999;Silverman et al., 2009).In agreement with Roy et al. (2015) and Ricke et al. (2013) by the end of the 21st century (2072 under RCP8.5)Hector projects that the low latitude oceans Ω Ar will drop below 3, well outside of the preindustrial values of Ω Ar > 3.5 and the Ω Ca high latitude will drop below 2. While at the end of the 21st Century, the oceans are not undersaturated (Ω < 1), the threshold for biogenic carbonate precipitation is species dependent and can be significantly higher than 1 when combined with other factors.For example, some coral reef communities need to develop in waters that have Ω Ar > 3.3 (Pelejero et al., 2010;Hoegh-Guldberg et al., 2007;Kleypas et al., 1999).The lowest observed Ω Ar found in individual coral reef ecosystems was Ω Ar = 2.85 (Shamberger et al., 2011).
Figure 7  into one global value, also making it easier to compare to Bopp et al. (2013).This is an area of future research to better emulate the high latitude surface ocean temperature.Lastly, Fig. 8 highlights pH and Ω Ar projections under all four RCPs from 1850 to 2300.Over the last 20 years both pH and Ω Ar have declined sharply and will continue to rapidly decline under RCP 4.5, 6.0 and 8.5.

Conclusions
We developed a simple, open-source, object oriented carbon cycle climate model, Hector, that reliably reproduces the median of the CMIP5 climate variables (Hartin et al., 2015).The ocean component presented in this study, directly calculates the upper ocean carbonate system (pCO 2 , DIC, pH, Ω Ar , Ω Ca ).Under all four RCPs, pH, Ω Ar , and Ω Ca decrease significantly outside of their preindustrial values.In the near future the open ocean and coral reef communities are likely to experience pH and carbonate saturation levels that are unprecedented in the potentially the last 2 million years (Hönisch et al., 2009).Even at a best case scenario, RCP 2.6 (Fig. 8), pH will drop to 7.73 by 2100 and to 7.43 by 2300.This may result in drastic changes to marine ecosystems in particular the CaCO − 3 secreting organism.For example, the rate of coral reef building decreases, calcification rates of planktonic cocolithophores and foraminifera decreases, changes in trophic level interactions and ecosystems, have all been proposed to be potential consequences of ocean acidification (Cooley and Doney, 2009;Silverman et al., 2009;Fabry et al., 2008;Riebesell et al., 2000).Organic carbon, CaCO 3 sediment interactions, and changes in ocean circulation are not currently simulated within Hector.We assume negligible CaCO 3 interactions on hundred year time scales; however, this is a necessary component under interglacial and glacial cycles.We neglect any climate feedbacks on the carbon cycle resulting from changes in ocean circulation.CMIP5 models show up to a 60 % decrease in the Atlantic meridional overturning circulation by 2100 (Cheng et al., 2013).While this may have a significant impact on the uptake and transport of carbon, in Hector v1.1, we

Conclusions References
Tables Figures

Back Close
Full hold ocean circulation constant with time and accurately simulate global variables out to 2100 with a slight bias after 2100.We also note that other factors such as eutrophication, river discharge, and upwelling will likely increase the probability that coastal regions will experience the effects of ocean acidification sooner than the projected open ocean values in Hector (Ekstrom et al., 2015).
This study is timely because the CMIP5 archive, includes a large suite of ESMs that contained dynamic biogeochemistry, allowing us to study future projections of the marine carbon cycle.Rather than running the earth system models, we can use Hector to quickly emulate the CMIP5 median for projection studies under different emission pathways and uncertainty analyses of the marine carbonate system.Due to Hector's simplistic nature and fast execution times, Hector has the potential to be a critical tool to the decision-making, scientific, and integrated assessment communities, allowing for further understanding of future changes to the marine carbonate system.

Appendix: Carbonate Chemistry
Modified from Zeebe and Wolf-Gladrow (2001 This equation results in a higher order polynomial equation for H + , in which the roots (1 positive, 4 negative) are solved for.Once H + is solved for, pH, pCO 2 , HCO − 3 , and CO 2− 3 can be determined.

Conclusions References
Tables Figures

Back Close
Full 1. K H and K 0 are similar equations calculating Henry's constant or the solubility of CO 2 , however they return different units (mol kg −1 atm −1 and mol L −1 atm −1 ) (see Weiss, 1974 for equations and coefficients).K H is used to solve pCO 2 while K 0 is used to solve air-sea fluxes of CO 2 .
2. The Schmidt number is taken from Wanninkhof (1992) for coefficients of CO 2 in seawater.

Conclusions References
Tables Figures

Back Close
Full Discussion Paper | Discussion Paper | Discussion Paper | 8.2, 88 % of the DIC is in the form of HCO − 3 , 11 % in the form of CO 2− Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | consisting of four boxes; two surface boxes (high and low latitude), an intermediate and deep box, simulated a simple thermocline circulation.The cold high latitude surface box makes up 15 % of the ocean, representing 19274 Discussion Paper | Discussion Paper | Discussion Paper | the subpolar gyres (> 55 • ), while the warm low latitude surface box makes up 85 % of the ocean.The temperatures of the surface boxes are linearly related to the global atmospheric temperature change, and are initialized at 2 • C in the high latitude and 22 • C for the low latitude box.This temperature gradient sets up a flux of carbon into the cold high latitude box and a flux out of the warm low latitude box.The ocean-atmosphere flux of carbon is the sum of all the surface fluxes (F i = 2).

)
Volume transports are tuned to yield an approximate flow of 100 Pg C from the surface high latitude to the deep ocean at steady state, simulating deep water formation.Discussion Paper | Discussion Paper | Discussion Paper | The alkalinity values are within the range of open ocean observations, 2250.0-2450.0mol kg −1of solution(Key et al., 2004)  and are held constant with time in Hector.We assume negligible carbonate precipitation/dissolution or alkalinity runoff from the land surface over our period of interest (100-300 years).From this, Hector actively solves for pCO 2 , pH (total scale), and HCO − 3 , CO 2− 3 , and aragonite (Ω Ar ) and calcite saturations (Ω Ca ) in the surface ocean boxes.pCO 2 is calculated from the concentration of [CO * 2 ] and the solubility of CO 2 in seawater, based on salinity, temperature, and pressure.[CO * 2 ] is calculated from DIC and the first and second dissociation constants of carbonic acid fromMehrbach et al. (1973), refit byLueker et al. (2000) (A1).pCO 2 is needed to calculate atmosphere ocean carbon fluxes(Takahashi et al., 2009): product of k and α results in Tr, the sea-air gas transfer coefficient, where Tr (gC m −2 Figures Back Close Full Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | month Discussion Paper | Discussion Paper | Discussion Paper | [H + ] concentrations and an 8 % decrease in [CO 2− 3 ].This is in close agreement with numerous studies (Discussion Paper | Discussion Paper | Discussion Paper | et al.
. More observations in the North Pacific show surface changes of pH of up to 0.06 units between 1991 and 2006 with an average rate of −0.0017 yr −1 (Byrne et al., 2010).Rates of change in high latitude pH over the same time period are −0.0018yr −1 for Hector and CMIP5.Under RCP 8.5, Hector projects a decrease of over 0.40 units to 7.77 from 1850 to 2100 and by over 0.6 units to 7.5 by 2300 in low latitude ocean pH, similar to CMIP5 (Table −1 ) agree well with CMIP5 (−0.006 yr −1 ) and HOT (−0.010 yr −1 ).Repeat oceanographic surveys in the Pacific Ocean observed an average 0.34 % yr −1 decrease in the saturation state of surface seawater with respect Discussion Paper | Discussion Paper | Discussion Paper | highlights the relationship between surface temperature change and surface carbonate chemistry changes across the 4 RCPs.Under RCP 8.5, for every one degree of surface warming surface in Hector (CMIP5), pH declines by 0.107 (0.122) units (change relative to 1990-1999 plotted over 2006-2100).This is similar to Bopp et al., who calculated a global change of 0.125 units • C −1 across the CMIP5 models.Under RCP 8.5, for every one degree of surface warming surface in Hector (CMIP5), aragonite saturation declines by 0.438 (0.432) units.For calcite saturations (not shown), for every one degree of surface warming in Hector (CMIP5), calcite saturations decrease by 0.681 (0.673) units.Our high latitude ocean box warms faster than the rest of the world's oceans, therefore, we chose to combine both the high and low latitude oceans Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |

Figure 1 .Figure 3 .Figure 5 .Figure 6 .
Figure1.Representation of the steady state ocean carbon cycle in Hector.For details on the terrestrial component seeHartin et al. (2015).The atmosphere consists of one well-mixed box, connected to the surface ocean via air-sea fluxes of carbon.The ocean consists of four boxes, with advection (represented by solid arrows) and water mass exchange (represented by dashed arrows) simulating thermohaline circulation (see Table2for description of parameters).The carbon exchange from the high latitude to deep box is initially tuned to ∼ 100 Pg C yr −1 .The inorganic carbon cycle is solved for in the high and low latitude surface boxes.At steady state, there is a flux of carbon from the atmosphere to the high latitude surface box, while the lowlatitude surface ocean releases carbon to the atmosphere.The lower left number represents the initial carbon pools of each box in Pg C yr −1 and the lower right hand numbers are the depth of each box in meters.
Hartin et al. (2015)conducted a thorough analysis of Hector v1.0 demonstrating how it can reproduce the historical trends and future projections of atmospheric [CO 2 ], radiative forcing, and global temperature under the RCPs.For this discussion we focus on the upper ocean high and low latitude inorganic carbon chemistry under RCP 8.5, comparing to a suite of earth system models included in the CMIP5 archive and observations.Hector's primary carbonate parameter outputs are summarized in Table 5. Figures 2-6 compare Hector to observations and CMIP5 median, one standard deviation and model spread.DIC and pCO 2 , functions of the inorganic carbon species in seawater, are directly related to rising temperatures and atmospheric [CO 2 ]

Table 2 .
Initial model conditions prior to the spinup phase.Carbon values change slightly after spinning up to a steady state.

Table 3 .
Time-series information and carbonate parameters from each location.