Effects of upwelling increase on ocean acidification in the California ...

10 downloads 33665 Views 571KB Size Report
Nov 15, 2013 - the California and Canary Current systems, Geophys. Res. Lett., .... California, Los Angeles-Swiss Federal Institute of Tech- nology version of ...
GEOPHYSICAL RESEARCH LETTERS, VOL. 41, 90–95, doi:10.1002/2013GL058726, 2014

Effects of upwelling increase on ocean acidification in the California and Canary Current systems Zouhair Lachkar1 Received 15 November 2013; revised 5 December 2013; accepted 6 December 2013; published 7 January 2014.

[1] Upwelling-favorable winds have increased in several coastal upwelling systems and may further increase in the future. The present study investigates the effects of upwelling intensification on ocean acidification in the California and Canary Current systems (CSs). Model simulations show that the volume of water undersaturated with respect to aragonite almost triples in the California CS under a doubling of wind stress. In contrast, the same wind perturbation results in the disappearance of undersaturation in the Canary CS. These contrasting responses arise from the differences in the relative contributions of circulation and biological processes to aragonite undersaturation in the two systems and the sensitivity of these processes to upwelling intensification. When combined with rising atmospheric CO2 and increased stratification, upwelling intensification accentuates acidification in the California CS and dampens it in the Canary CS. These findings highlight the challenge to predict the future evolution of ocean acidification in regions subject to concurrent disturbances. Citation: Lachkar,

[3] The highly productive marine ecosystems in Eastern Boundary Upwelling Systems (EBUS) such as the California Current system (California CS), the Canary Current system (Canary CS), and the Humboldt and Benguela Current systems are examples of coastal ecosystems that naturally experience low . This is because of the upwelling of deep CO2– 3 -depleted water to the surface in addition to intense organic matter respiration in the subsurface [e.g., Gruber et al., 2012] which releases CO2 and consumes CO2– 3 . Recent observations reveal that aragonite undersaturated waters (arag < 1) are reaching ocean surface in EBUS such as the California CS during the spring/summer upwelling season [Feely et al., 2008]. Furthermore, recent model-based studies predict that the intensity of ocean acidification in EBUS will considerably increase in the future in response to rising atmospheric CO2 concentrations [Gruber et al., 2012; Hauri et al., 2013]. These results were, however, obtained under the assumption of constant oceanic and atmospheric circulations. Yet EBUS are concurrently subject to additional climate change-related perturbations such as increased stratification [e.g., Rykaczewski and Dunne, 2010] and upwelling-favorable wind increase, which might in turn alter the level of acidification these systems will experience in the future. [4] Recent observations suggest an intensification of upwelling-favorable winds over the last decades in several EBUS [e.g., Bakun, 1990; McGregor et al., 2007; Leduc et al., 2010]. This upwelling strengthening has been linked to a global warming-induced increase in the land-sea thermal gradient [Bakun, 1990] and is therefore predicted to increase further in the future under a warmer climate [e.g., Diffenbaugh et al., 2004]. The statistical analyses of different wind and sea surface temperature data products by Narayan et al. [2010] suggest that the observed increases in upwelling-favorable winds are driven more by climate change forcing than natural variability. However, the magnitude of these upwelling changes and their future long-term evolution remain a matter of debate [e.g., Barton et al., 2013; Echevin et al., 2012]. [5] Regardless of the respective contributions of natural variability and long-term climate change to the observed upwelling intensification and the magnitude of the future wind changes, the impacts of such an upwelling increase on ocean acidification in these ecosystems remain unexplored and hence largely unknown. Here I examine how upwelling-favorable wind increase may affect the volume of undersaturated water with respect to aragonite (arag < 1) in the nearshore area and consider the potential impacts of these changes on future habitat size of calcifying species. To this end, I use a comparative approach based on eddyresolving model simulations of two of the four major EBUS, namely, the California CS and the Canary CS, and

Z. (2014), Effects of upwelling increase on ocean acidification in the California and Canary Current systems, Geophys. Res. Lett., 41, 90–95, doi:10.1002/2013GL058726.

1. Introduction [2] Rising atmospheric concentrations of carbon dioxide are causing large-scale acidification of open and coastal oceans at unprecedented rates in human history [Caldeira and Wickett, 2003]. The ocean uptake of CO2 draws down the concentration of the carbonate ion (CO2– 3 ) required by marine calcifying organisms that synthesize calcium carbonate (CaCO3 ) minerals such as calcite and the less stable form aragonite to form shells and skeletons [Orr et al., 2005]. The declining CO2– 3 concentrations lower the saturation state of calcium carbonate minerals () defined as the product of the concentrations of CO23 and Ca2+ in seawater relative to the their solubility product (see supporting information for further details on  and its drivers). Waters with  > 1 favor the formation and maintenance of CaCO3 shells, whereas waters with  < 1 are corrosive and favor the dissolution of CaCO3 , thereby putting calcifying species under stress [Fabry et al., 2008]. Additional supporting information may be found in the online version of this article. 1 Environmental Physics, Institute of Biogeochemistry and Pollutant Dynamics, ETH Zurich, Zurich, Switzerland. Corresponding author: Z. Lachkar, Environmental Physics, Institute of Biogeochemistry and Pollutant Dynamics, ETH Zurich, Universitätstrasse 16, CH-8092 Zurich, Switzerland. ([email protected]) ©2013. American Geophysical Union. All Rights Reserved. 0094-8276/14/10.1002/2013GL058726

90

LACHKAR: ACIDIFICATION IN UPWELLING SYSTEMS

Figure 1. Simulated annual mean surface pH in (a) the California CS and (b) the Canary CS. Alongshore averaged DIC (solid line), Alkalinity (dashed line), and arag (color shaded) in (c) the California CS between 34.4ı N and 42ı N and (d) the Canary CS between 18ı N and 22ı N. The red segments in Figures 1a and 1b indicate the northern and southern boundaries of the analysis area in each EBUS. tions for each upwelling system: a control simulation where the wind was left unperturbed and two sensitivity simulations where the wind stress was increased by a factor of 1.5 and 2.0, respectively. This corresponds to increases in wind speed amounting to around 20% and 40%, respectively. These perturbations are of idealized nature. Thus, the results of these experiments should not be viewed as predictions but rather as an exploration of the sensitivity of ocean acidification to changes in upwelling intensity and the key processes that might govern these responses. To compare the impacts of upwelling intensification and those driven by rising atmospheric CO2 concentrations, as well as the combined effects of these two perturbations, two additional simulations were performed under a doubling of the preindustrial CO2 concentration (2  CO2  540 ppm) with, respectively, unperturbed and doubled wind stress. For the double-preindustrial CO2 simulations, the Global Ocean Data Analysis Project-based DIC concentration was increased with the change in DIC between 1995 and 2050, computed from a NCAR global Earth system model simulation that was forced with the IPCC SRES A2 greenhouse gas emissions scenario [Frölicher et al., 2009]. Finally, the robustness of the results was tested through a sensitivity analysis to potential future warming and associated increase in stratification and change in nutrient conditions. Due to computational cost, these simulations were made at 15 km resolution. This coarser resolution version of the model reproduces well the contrasts between the two systems in terms of their sensitivity to upwelling intensification and hence was used to explore the model sensitivity to additional perturbations at an affordable cost (see details in the supporting information).

quantify their sensitivity to idealized wind perturbations both under modern and future increased temperatures and CO2 conditions.

2. Methods [6] The circulation model is based on the University of California, Los Angeles-Swiss Federal Institute of Technology version of the Regional Oceanic Modeling System documented in detail in Shchepetkin and McWilliams [2005]. I use the same model setups for the California and Canary Current systems as the ones described and evaluated by Lachkar and Gruber [2013] (a summary of the model evaluation is presented in the supporting information). Both model grids have a horizontal resolution of 5 km and 32 vertical levels with surface refinement. The ecological-biogeochemical model is a nitrogen-based nutrient-phytoplankton-zooplankton-detritus model [Gruber et al., 2006] with a single phytoplankton functional group. The ecological model is coupled with a carbon cycle module with three state variables, i.e., dissolved inorganic carbon (DIC), alkalinity (Alk), and mineral CaCO3 . The precipitation of mineral CaCO3 is linked to the formation of organic matter via a constant ratio of 0.07 [Hauri et al., 2013]. The sinking materials arriving at the seafloor are collected in a sediment layer, where they are subject to aerobic remineralization and dissolution. Sedimented organic matter is remineralized at a slower rate than in the water column, whereas CaCO3 dissolves slightly faster than in the water column. [7] To explore the effects of the intensification of upwelling-favorable winds on coastal ocean acidification, three experiments were conducted under modern CO2 condi91

LACHKAR: ACIDIFICATION IN UPWELLING SYSTEMS

3. Results [8] Simulated corrosive surface waters with pH below 8 stretch several hundreds of kilometers along the Californian Coast from Point Conception ( 34ı N) in the south to the California-Oregon border ( 42ı N, Figures 1a and 1b). In contrast, equally low pH waters (pH < 8) occupy a much smaller area in the Canary CS essentially around Cape Blanc (20ı N–22ı N) and sparsely along the Mauritanian coast. In the offshore region, surface pH ranges between 8.02 and 8.06 in most of the California CS, while it generally exceeds 8.06 in the Canary CS. These differences in surface distributions of pH are associated with substantial contrasts in the vertical distributions of arag , DIC, and Alk (Figures 1c and 1d). In the California CS, the aragonite saturation horizon (arag = 1) is found in the model at depths ranging between 200 and 250 m in the offshore region and at around 100 m in the nearshore area. Recent observations show that waters with arag < 1 can be found at shallower depths (0–50 m) in several nearshore locations during the upwelling season [Feely et al., 2008]. In the Canary CS, undersaturated (arag < 1) waters are found exclusively in the region between 18ı N and 22ı N around Cape Blanc at relatively shallow depths (75–150 m) and occupy a substantially smaller volume confined to few tenths of kilometers off the continental shelf. The distribution of supersaturated water (arag > 1) similarly exhibits large contrasts between the two systems. 2– [9] The generally lower carbonate CO2– 3 ([CO3 ]  Alk - DIC) concentrations and more widespread undersaturation (arag < 1) in the California CS relative to the Canary CS are driven by the substantially lower upper ocean Alk in the Pacific relative to the Atlantic. Indeed, Alk varies roughly between 2200 and 2280 mmol m–3 in the top 300 m of California CS domain, whereas it ranges in the southern Canary CS between 2320 mmol m–3 at 300 m in the nearshore area and over 2440 mmol m–3 between 50 and 100 m in the offshore area. In contrast, the concentration of DIC is generally higher in the Canary CS than in the California CS and hence does not contribute to the Canary system’s higher carbonate ion concentration. The differences in Alk and DIC between the Atlantic and Pacific essentially reflect the differences in stratification and water age between the two basins.

Figure 2. Volume fractions of seawater in different ranges of arag in the top 200 m, 50 km nearshore of (a) the central California CS and (b) the southern Canary CS. Figures 2a and 2b depict water volume fractions corresponding to seasonal and annual means as simulated under modern and increased wind forcing. to only 40% under double wind stress forcing. These annual mean changes are associated with seasonal changes with different magnitudes. In the California CS, the largest aragonite undersaturation expansion occurs during the maximum upwelling season in spring and summer with the volume of arag < 1 water increasing in the two seasons from 15% and 13% under modern conditions to 37% and 39% under double wind stress forcing, respectively. In the southern Canary CS, the decline in volumes of undersaturation (arag < 1) and strong supersaturation (arag > 2) is more pronounced in the fall and winter seasons.

3.1. Changes in Ocean Acidification and Calcifiers Habitat Size [10] In the California CS, the fraction of undersaturated water (arag < 1) in the top 200 m, 50 km nearshore, expands from around 10% on annual mean in the control simulation to around 20% under 50% stronger wind stress and 27% in the double wind stress scenario (Figure 2). This expansion of low-arag water in the upper ocean develops along with a decrease of the fraction of the strongly supersaturated water (arag > 2) which drops from around 5% of the total volume in the control simulation to around 0.5% in the double wind stress case. In contrast, similar upwelling intensification leads in the southern Canary CS to a reduction of the volume fraction of undersaturated water which shrinks from around 10% under modern winds to less than 1% under double wind stress forcing. Additionally, the volume fraction of strongly supersaturated water (arag > 2) gets also reduced from around 45% in the control simulation

3.2. Drivers of CaCO3 Undersaturation in EBUS: Advection Versus Biology [11] Two processes contribute to CaCO3 undersaturation in coastal upwelling systems: (i) large scale advection of DIC-enriched, carbonate-depleted waters from deeper layers and (ii) biological production and subsequent remineralization of organic matter which locally increases DIC concentration and draws down carbonate levels. Additionally, in EBUS with wide continental shelves such as the Canary 92

LACHKAR: ACIDIFICATION IN UPWELLING SYSTEMS Ω arag a

b

2 1.9

100

100

1.8

Depth (m)

Depth (m)

1.7 200

1.6

200

1.5 1.4 300

300

1.3 1.2 1.1

400

400

500

300

150

1 500

0

300 150 Distance (km)

Distance (km) c

d

100

Depth (m)

100

Depth (m)

0

200

200

300

300

400

400

500

300

150

0

500

Distance (km)

300

150

0

Distance (km)

Figure 3. Alongshore averaged arag in the (a and c) California CS and (b and d) Canary CS as simulated in the control (Figures 3a and 3b) and abiotic (Figures 3c and 3d) simulations. Averaging is done between 34.4ı N and 42ı N in the California CS and between 18ı N and 22ı N in the Canary CS. CS, a substantial fraction of organic matter remineralization occurs on the seafloor in the sediment, drawing down undersaturation in the overlying water column. [12] To separate these effects and assess their relative importance in each EBUS, a series of additional simulations was undertaken. First, to separate the effects of biology and advection on CaCO3 undersaturation, abiotic simulations (where DIC and Alk behave like passive tracers) are contrasted in each EBUS to the standard simulations where biology is taken into account. Second, to evaluate the importance of the wide continental shelf in the case of the Canary CS for CaCO3 undersaturation, an additional Canary CS simulation was made where only the production of DIC through organic matter respiration in the sediment is switched off. To reduce the computational cost, these sensitivity simulations were performed at a coarser resolution of 15 km. [13] In the California CS, the shutdown of the biological pump within the domain results in a weakening of the vertical gradient in arag and a deepening of the saturation horizon by around 50 m (Figure 3). Yet undersaturated waters still fill most of the domain at relatively shallow depths ( 250 m) even in the absence of local production and remineralization. This indicates that the CaCO3 undersaturation in the California CS is largely driven by upwelling and large-scale advection of carbonate-depleted waters, whereas local respiration plays a relatively minor role. In contrast, turning off biological production and respiration in the Canary CS entirely removes undersaturation developing around the continental shelf. This indicates that low arag levels are primarily produced in the southern

Canary CS by local respiration of organic matter. Similar removal of aragonite undersaturation is obtained when only the production of DIC through respiration in the sediment is hindered (Figure S2). This indicates that in the Canary CS, most of water column aragonite undersaturation is driven by organic matter remineralization in the underlying sediment layer. 3.3. Mechanisms of Aragonite Undersaturation Changes [14] In the California CS, changes in aragonite undersaturation are primarily driven by changes in advection and upwelling of high-DIC, low-arag water. A Lagrangian diagnostic of the vertical circulation in central California CS reveals that wind increases lead to a substantial deepening of the upwelling. Indeed, Lagrangian tracking of virtual particles released in the subsurface layer shows that the proportion of deep upwelling waters (i.e., particles whose original depth 1 month earlier is at least 50 m deeper than their actual depth) increases from 1% to 20% of the total water masses in the layer between 100 and 200 m as we double the wind stress (Figure S3). This deepening of upwelling results in an increased advection of DIC-enriched water from depth and hence explains the expansion of carbonatedepleted waters and the shoaling of the arag saturation horizon under increased winds. [15] In the southern Canary CS, changes in aragonite undersaturation are essentially induced by changes in local respiration, in particular on the continental shelf. In the nearshore 100–200 m layer where most of the undersaturation develops in the Canary CS, biological depletion of 93

LACHKAR: ACIDIFICATION IN UPWELLING SYSTEMS

et al., 2011; Lachkar and Gruber, 2011, 2012, 2013]. For instance, the southern Canary CS exhibits a higher productivity than the California CS, associated with a more efficient nutrient utilization by biology. These differences have been attributed to various contrasting features between the two EBUS including shelf topography, eddy activity, and basin-scale forcings [Marchesiello and Estrade, 2009; Gruber et al., 2011; Lachkar and Gruber, 2011]. This study highlights how these differences can also affect the vulnerability of the two systems to ocean acidification in the context of upwelling intensification. [20] Although future ocean acidification in EBUS will be primarily driven by the rise of atmospheric CO2 , local wind changes might accentuate or dampen this large-scale trend in different EBUS. The direction of this change will vary from one EBUS to another depending on local factors such as the shelf topography as well as larger-scale conditions of Alk and DIC in each basin.

carbonate decreases by around 13–15% under double wind stress forcing. Most of this decrease (70%) is due to declining respiration on the continental shelf. The remaining 30% is due to increased Alk production in the water column (Table S2). The decrease in respiration is due to a reduction in the inventory of organic carbon by 15% in the top 200 m water column and 20% in the sediment on the continental shelf. This decrease reflects a drop in the coastal retention of organic matter as indicated by a reduction in the ratio of vertical export to new production (–22%) as well as an increase in the offshore export of organic matter at 50 km from the coast (+27%). Finally, despite the overall decreasing respiration, the fraction of strongly supersaturated water (arag > 2) also decreases under increased winds because of enhanced mixing and dispersion (Table S3).

4. Discussion and Conclusions [16] Additional concurrent changes in the physical (e.g., warming) and chemical (e.g., nutrients and carbon) boundary conditions may further influence the future levels of ocean acidification and aragonite saturation state in these systems. [17] A series of sensitivity simulations performed at 15 km horizontal resolution indicates that the effects of upwelling intensification on acidification are a quite robust feature which is replicated under a wide range of physical and chemical conditions (Figure S4). The effects of increased stratification are small in the Canary CS and appear uncertain in the California CS. Deep-penetrating temperature anomalies seem to counteract the effects of upwelling intensification in this system, while shallow warming has little impact on undersaturation. Under increased stratification, doubling the wind stress results in the California CS in an expansion of the volume fraction of aragonite undersaturated water to occupy between 30% and 50% of the top 200 m while it leads to a vanishing of undersaturation in the Canary CS. Finally, these simulations reveal that substantially larger changes in ocean acidification are likely to result from the future atmospheric CO2 rise (Figure S4). [18] In a set of additional high-resolution (5km) sensitivity simulations, it is found that the fraction of aragonite undersaturated water (top 200 m) increases in the California CS from 10% under modern conditions to around 75% under a doubling of the preindustrial CO2 concentration ( 540 ppm), which is far more severe than the increase associated with upwelling intensification. In contrast, the volume of undersaturation rises in the southern Canary CS from 10% under current CO2 conditions to around 20% under doubling of preindustrial CO2 concentration, which is comparable in magnitude to the effect of doubling wind stress. In the California CS, the combination of increasing atmospheric CO2 concentrations and increasing upwelling-favorable winds accentuate aragonite undersaturation with the fraction of undersaturated water increasing up to 83% under 2 CO2 , 2 conditions. In contrast, in the Canary CS the two perturbations offset each other resulting in almost no change in the volume of undersaturation (+0.5%). [19] Previous comparative analyses of the central California CS and southern Canary CS have revealed substantial differences between these two systems [e.g., Carr and Kearns, 2003; Marchesiello and Estrade, 2009; Gruber

[21] Acknowledgments. I would like to thank the Editor and the two anonymous reviewers for their valuable comments and suggestions. I am deeply thankful to N. Gruber for his encouragements and helpful comments on the manuscript. Support for this research has come from the Swiss Federal Institute of Technology Zurich (ETH Zurich). Computations were performed at the central computing cluster of ETH Zurich, Brutus. I thank D. Loher and M. Münnich for their help with ROMS and G. Turi for her helpful comments on the manuscript. I am grateful to B. Blanke and N. Grima for making their ARIANE code available and to T. Frölicher, F. Joos, and M. Steinacher for providing me with the future carbon boundary conditions. [22] The Editor thanks two anonymous reviewers for their assistance in evaluating this paper.

References Bakun, A. (1990), Global climate change and intensification of coastal ocean upwelling, Science, 247, 198–201, doi:10.1126/ science.247.4939.198. Barton, E. D., D. B. Field, and C. Roy (2013), Canary Current upwelling: More or less?, Prog. Oceanogr., 116, 167–178, doi:10.1016/ j.pocean.2013.07.007. Caldeira, K., and M. E. Wickett (2003), Anthropogenic carbon and ocean pH, Nature, 425, 365–365. Carr, M., and E. J. Kearns (2003), Production regimes in four Eastern Boundary Current systems, Deep Sea Res., Part II, 50, 3199–3221, doi:10.1016/j.dsr2.2003.07.015. Diffenbaugh, N. S., M. A. Snyder, and L. C. Sloan (2004), Could CO2 induced land-cover feedbacks alter near-shore upwelling regimes?, Proc. Natl. Acad. Sci. U.S.A., 101, 27–32. Echevin, V., K. Goubanova, A. Belmadani, and B. Dewitte (2012), Sensitivity of the Humboldt Current system to global warming: A downscaling experiment of the IPSL-CM4 model, Clim. Dyn., 38, 761–774, doi:10.1007/s00382-011-1085-2. Fabry, V. J., B. A. Seibel, R. A. Feely, and J. C. Orr (2008), Impacts of ocean acidification on marine fauna and ecosystem processes, ICES J. Mar. Sci., 65(3), 414–432, doi:10.1093/icesjms/fsn048. Feely, R. A., C. L. Sabine, J. M. Hernandez-Ayon, D. Ianson, and B. Hales (2008), Evidence for upwelling of corrosive “acidified” water onto the continental shelf, Science, 320, 1490–1492, doi:10.1126/science.1155676. Frölicher, T. L., F. Joos, G.-K. Plattner, M. Steinacher, and S. C. Doney (2009), Natural variability and anthropogenic trends in oceanic oxygen in a coupled carbon cycle-climate model ensemble, Global Biogeochem. Cycles, 23, GB1003, doi:10.1029/2008GB003316. Gruber, N., H. Frenzel, S. C. Doney, P. Marchesiello, J. C. McWilliams, J. R. Moisan, J. Oram, G.-K. Plattner, and K. D. Stolzenbach (2006), Eddy-resolving simulation of plankton ecosystem dynamics in the California Current System, Deep Sea Res., Part I, 53, 1483–1516, doi:10.1016/j.dsr.2006.06.005. Gruber, N., Z. Lachkar, H. Frenzel, P. Marchesiello, M. Munnich, J. C. McWilliams, T. Nagai, and G.-K. Plattner (2011), Mesoscale eddyinduced reduction of biological production in coastal upwelling systems, Nat. Geosci., 4(11), 787–792. Gruber, N., C. Hauri, Z. Lachkar, D. Loher, T. L. Frolicher, and G.-K. Plattner (2012), Rapid progression of ocean acidification in the California Current System, Science, 337, 220–223, doi:10.1126/science.1216773.

94

LACHKAR: ACIDIFICATION IN UPWELLING SYSTEMS Marchesiello, P., and P. Estrade (2009), Eddy activity and mixing in upwelling systems: A comparative study of Northwest Africa and California regions, Int. J. Earth Sci., 98, 299–308, doi:10.1007/s00531-0070235-6. McGregor, H. V., M. Dima, H. W. Fischer, and S. Mulitza (2007), Rapid 20th-century increase in coastal upwelling off Northwest Africa, Science, 315, 637–639, doi:10.1126/science.1134839. Narayan, N., A. Paul, S. Mulitza, and M. Schulz (2010), Trends in coastal upwelling intensity during the late 20th century, Ocean Sci., 6, 815–823, doi:10.5194/os-6-815-2010. Orr, J. C., et al. (2005), Anthropogenic ocean acidification over the twentyfirst century and its impact on calcifying organisms, Nature, 437(7059), 681–686, doi:10.1038/nature04095. Rykaczewski, R. R., and J. P. Dunne (2010), Enhanced nutrient supply to the California Current ecosystem with global warming and increased stratification in an earth system model, Geophys. Res. Lett., 37, L21606, doi:10.1029/2010GL045019. Shchepetkin, A. F., and J. C. McWilliams (2005), The regional oceanic modeling system (ROMS): A split-explicit, free-surface, topographyfollowing-coordinate oceanic model, Ocean Modell., 9, 347–404.

Hauri, C., N. Gruber, M. Vogt, S. C. Doney, R. A. Feely, Z. Lachkar, A. Leinweber, A. M. P. McDonnell, M. Munnich, and G.-K. Plattner (2013), Spatiotemporal variability and long-term trends of ocean acidification in the California Current System, Biogeosciences, 10, 193–216, doi:10.5194/bg-10-193-2013. Lachkar, Z., and N. Gruber (2011), What controls biological productivity in coastal upwelling systems? Insights from a comparative modeling study, Biogeosciences, 8(10), 2961–2976. Lachkar, Z., and N. Gruber (2012), A comparative study of biological production in eastern boundary upwelling systems using an artificial neural network, Biogeosciences, 9, 293–308, doi:10.5194/bg-9293-2012. Lachkar, Z., and N. Gruber (2013), Response of biological production and air-sea CO2 fluxes to upwelling intensification in the California and Canary Current Systems, J. Mar. Syst., 109-110, 149–160, doi:10.1016/j.jmarsys.2012.04.003. Leduc, G., C. T. Herbert, T. Blanz, P. Martinez, and R. Schneider (2010), Contrasting evolution of sea surface temperature in the Benguela upwelling system under natural and anthropogenic climate forcings, Geophys. Res. Lett., 37, L20705, doi:10.1029/2010GL044353.

95