Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1225-6692(Print)
ISSN : 2287-4518(Online)
Journal of the Korean earth science society Vol.38 No.7 pp.469-480
DOI : https://doi.org/10.5467/JKESS.2017.38.7.469

Biophysical Effects Simulated by an Ocean General Circulation Model Coupled with a Biogeochemical Model in the Tropical Pacific

Hyo-Jin Park1,2, Byung-Kwon Moon2*, Jieun Wie2, Ki-Young Kim3, Johan Lee4, Young-Hwa Byun4
1Gunsan Dongsan Middle School, Gunsan 54020, Korea
2Division of Science Education & Institute of Fusion Science, Chonbuk National University, Jeonju, 54896, Korea
34D Solution CO., LTD., Seoul 08511, Korea
4National Institute of Meteorological Sciences, Seogwipo 63568, Korea
Corresponding author: moonbk@jbnu.ac.kr+82-63-270-2824+82-63-270-2802
20171010 20171204 20171214

Abstract

Controversy has surrounded the potential impacts of phytoplankton on the tropical climate, since climate models
produce diverse behaviors in terms of the equatorial mean state and El Niño-Southern Oscillation (ENSO) amplitude. We explored biophysical impacts on the tropical ocean temperature using an ocean general circulation model coupled to a biogeochemistry model in which chlorophyll can modify solar attenuation and in turn feed back to ocean physics. Compared with a control model run excluding biophysical processes, our model with biogeochemistry showed that subsurface chlorophyll concentrations led to an increase in sea surface temperature (particularly in the western Pacific) via horizontal accumulation of heat contents. In the central Pacific, however, a mild cold anomaly appeared, accompanying the strengthened westward currents. The magnitude and skewness of ENSO were also modulated by biophysical feedbacks resulting from the chlorophyll affecting El Niño and La Niña in an asymmetric way. That is, El Niño conditions were intensified by the higher contribution of the second baroclinic mode to sea surface temperature anomalies, whereas La Niña conditions were slightly weakened by the absorption of shortwave radiation by phytoplankton. In our model experiments, the intensification of El Niño was more dominant than the dampening of La Niña, resulting in the amplification of ENSO and higher skewness.


초록


    Introduction

    Recent advances in climate modeling have increased our understanding of biophysical feedbacks in the tropical climate. For example, oceanic phytoplankton can modulate the mean state of the tropical Pacific in addition to the El Niño-Southern Oscillation (ENSO) (Park et al., 2014; Zhang et al., 2015 and references therein). Chlorophyll contained in phytoplankton absorbs shortwave radiation for photosynthesis, resulting in surface warming by direct heating (Lewis et al., 1983; Sathyendranath et al., 1991; Strutton and Chavez, 2004). Hence, warmer upper-layer temperatures and deepening thermoclines are expected to occur in the presence of phytoplankton as simulated in some studies (Murtugudde et al., 2002; Marzeion et al., 2005; Wetzel et al., 2006; Patara et al., 2012). However, other studies have simulated sea surface temperature (SST) cooling in the equatorial Pacific after the inclusion of chlorophyll effect (Nakamoto et al., 2001; Manizza et al., 2005; Loptien et al., 2009; Park et al., 2014), mainly due to indirectly induced equatorial upwelling. These contradictory biological impacts indicate that the processes determining chlorophyll’s influence on the ocean’s mean state are somewhat complicated, such that current knowledge may still be insufficient to accurately describe biophysical interactions in models.

    Likewise, the influences of chlorophyll on ENSO variability are also controversial. Given the changed mean state in the equatorial tropics due to phytoplankton, Timmermann and Jin (2002) suggested that La Niña tends to be weakened by the so-called biothermostat process, likely resulting in higher ENSO skewness. Recently, Zhang et al. (2015) reported negative feedback associated with biological heating, which acts to reduce ENSO amplitude. However, another line of studies showed an intensified ENSO due to biological feedbacks (Lengaigne et al., 2007; Loptien et al., 2009; Park et al., 2014). Additionally, the mechanism for ENSO modulation induced by phytoplankton differs among such studies, which may be due to the large spread of modeling strategies as well as the inadequate implementation of biophysical feedbacks in the models.

    It is thus necessary to further investigate the role of biophysical processes in the tropical Pacific. This study examined the influence of phytoplankton on the tropical climate and ENSO variability using an ocean general circulation model (OGCM) coupled with a state-of-the-art biogeochemical model (described below). Moreover, we analyzed the ENSO modulation with respect to the vertical baroclinic mode (Cane, 1984), offering a new perspective for understanding the influence of ocean biology on tropical variability.

    Model experiments and data

    We used the Geophysical Fluid Dynamics Laboratory (GFDL) Modular Ocean Model Version 4 (MOM4p1), which is based on the hydrostatic primitive equation systems using an explicit free surface (Griffies et al., 2009; https://www.gfdl.noaa.gov/mom-ocean-model). This model implements a tripolar grid with a uniform longitudinal resolution of 1° and non-uniform latitudinal resolution of 1/3° near equator, gradually increasing to 1° at 30°N and 30°S. There are 50 vertical levels with 10 m spacing in the upper ocean (<225 m), increasing to about 360 m in the deepest ocean. The model also includes a global dynamic/ thermodynamic sea ice model, Sea Ice Simulator (SIS), which is linked to ocean physics via a flux coupler module. The ocean biogeochemistry component coupled to MOM4p1 is Tracers of Ocean Phytoplankton with Allometric Zooplankton (TOPAZ) version 2.0 (Dunne et al., 2013), which considers thirty tracers to represent the ocean biogeochemical and ecological cycles.

    To investigate the biological impacts on ocean physics, we performed two different experiments using the MOM4p1: a control run without biogeochemical processes (NoBGC), and an experimental run using the with interactive TOPAZ module (BGC). In the former, TOPAZ was turned off and therefore no chlorophyll existed in the modeled ocean. In the latter, the simulated chlorophyll concentration could affect the ocean by modifying solar light absorption as described in (Manizza et al., 2005). Therefore, differences between the BGC and NoBGC runs enabled us to isolate changes due to biophysical feedbacks. The two experiments were forced using inter-annually varying atmospheric data from the Coordinated Ocean-ice Reference Experiments (COREII) during 1950–2008 (Large and Yeager, 2009) and the simulation results are analyzed. The OGCM initial state was spun up for 20 years with climatological CORE-II forcings fixed at year 1949.

    The observational data set for monthly chlorophyll concentrations are derived from the Sea-viewing Wide Field-of-view Sensor (SeaWiFS) measurement from October 1997-December 2007 (McClain et al., 1998). The original chlorophyll data with a 9-km latitude/ longitude grid is interpolated onto the regular 1°×1° grid by using a bilinear interpolation method. We also used monthly mean data of the Extended Reconstructed Sea Surface Temperature (ERSST) with 2 latitudelongitude grid during 1950-2008 (Smith et al., 2008).

    Changes in mean states

    Fig. 1 compares annual mean surface chlorophyll concentration from 1998-2008, as simulated by the BGC experiment, with that observed by the SeaWiFS satellite. The model reasonably reproduces the observed chlorophyll patterns with low concentrations at subtropical gyres and high concentrations in the tropical oceans, which are related to inactive subtropical vertical mixing and strong equatorial upwelling, respectively (Chavez et al., 1999). These features indicate that the upwelling of nutrient-enriched waters provide a favorable environment for phytoplankton growth, which is also shown by the vertical distribution of chlorophyll in the “cold tongue” region of the equatorial Pacific (Fig. 1d). Compared with the observational data, however, the simulated chlorophyll values are lower in northern coastal regions and higher in the North Pacific, eastern tropical Pacific, and Southern Ocean (Fig. 1c). These biases are partly due to unrealistic nutrient supply (Vichi and Masina, 2009) or inadequate mixing (Schneider et al., 2008).

    Since phytoplankton containing chlorophyll absorb solar radiation with shallower attenuation depth, higher phytoplankton concentrations generally lead to a warmer ocean surface (Lewis et al., 1990). This can be seen in Fig. 2a, which shows biologically-induced heating in the tropical Pacific SST (T~0.28°C throughout the region covered by Fig. 2a) as modeled by the BGC experiment. Interestingly, while the BGC equatorial surface warming is most prominent in the western Pacific, eastern Pacific, and coastal regions in comparison to NoBGC, a slight cooling appears in the central Pacific (~150°W) despite larger chlorophyll concentrations occurring there as compared to the western Pacific (Fig. 1b). This central Pacific cooling can be attributed to enhanced zonal cold advection by anomalous westward currents along the equator, as shown in Fig. 2b. The thermocline warms and deepens (Fig. 2c) as the equatorial divergence is diminished (Fig. 2d), which indicates a reduction in the equatorial upwelling due to biological heating. This weaker upwelling would result in surface warming in the equatorial Pacific as previous studies have shown (Murtugudde et al., 2002; Wetzel et al., 2006; Lengaigne et al., 2007).

    However, it is worth noting that the SST response to biological heating simulated in this study (Fig. 2a) is somewhat different in two aspects from that of previous studies. First, while our model shows a large increase in SST over the western Pacific, most studies show pronounced SST warming confined to the cold tongue region (e.g., Fig. 5 in Murtugudde et al., 2002) where the highest chlorophyll concentrations exist, suggesting a relevant role of phytoplankton in surface radiative heating. Second, the slight decrease in SST in the central Pacific that occurs along with the deepening of thermocline in this study (as discussed above) has not been simulated in others. These discrepancies imply that a different mechanism is likely acting on the simulation of the tropical mean state in our model.

    The central Pacific cooling (shown in Fig. 2a) along with the presence of chlorophyll implies an indirect dynamical cooling effect that overwhelms the tendency for heating due to absorption of shortwave radiation by phytoplankton (Nakamoto et al., 2001; Manizza et al., 2005; Loptien et al., 2009; Park et al., 2014). In previous studies, this surface cooling has been attributed to the enhanced equatorial upwelling due to net poleward volume transport in the mixed layer, which is related to the shoaling of the thermocline (e.g., Loptien et al., 2009). However, our model experiments simulated a deepening of the thermocline (Fig. 2c) under biological influences, indicating that a different mechanism is likely to be important. While some modeling studies also produced a deepening of the mixed layer depth (Murtugudde et al., 2002; Marzeion et al., 2005) when considering ocean biology, this was associated with surface warming, opposite to the pattern shown in Fig. 2a. These contradictory results might arise from the different modeling strategies, reflecting our insufficient knowledge on phytoplankton-light feedback processes.

    Fig. 2b shows a strengthened westward current over the central Pacific where the negative SST difference occurs, indicating that negative temperature advection is crucial. To check the role of temperature advection on such a SST cooling in the central Pacific (Fig. 2a), we analyzed the relationship between the difference in monthly surface zonal current and difference in surface temperature over the central Pacific (Fig. 3a). Note that a large number of differences in both monthly current and SST are negative, which is consistent with Figs. 2a and 2b. There is a remarkable correspondence between the zonal current and SST differences. This implies that the stronger westward surface current modeled by the BGC experiment relative to NoBGC contributes to cooling the central Pacific through the enhancement of negative temperature advection.

    Furthermore, Fig. 3b clearly shows that a deeper thermocline leads to more warming at depth. To further explore this pattern, we investigated changes in zonal and meridional circulations (Fig. 4). Zonal currents showed considerable changes in the BGC model run: the westward equatorial upper oceanic current was enhanced in the central Pacific and the Equatorial Undercurrent (EUC) was relocated to deeper ocean areas (Fig. 4a). This downward shift of the EUC is consistent with the thermocline deepening shown in Fig. 2c. Unlike the aforementioned studies that focused on an enhanced upwelling related to surface cooling, our BGC simulation did not show a distinct increase of the equatorial upwelling in comparison to NoBGC (Fig. 4b). Rather, a weak anomalous downwelling appears to be associated with the changes in the meridional equatorial (5°S to 5°N) current in the upper (<100 m) ocean. Thus, the surface cooling in the tropical Pacific simulated when biological impacts were included cannot be driven by changes in the vertical advection of cold water. Again, we argue that the westward equatorial surface current gets stronger in BGC (Fig. 3a) and that this brings colder water from the eastern Pacific region. Concurrently, the eastward zonal current above the EUC peak (~100 m depth) is also weakened in BGC run, which leads to the accumulation of heat in the western Pacific warm pool region. This probably contributes to the increase in SST over the western Pacific (Fig. 2a).

    Given the strong air-sea interactions due to the already warm state in the warm pool, the above increase in SST due to ocean biology can have a significant impact on the global atmosphere. More studies using atmosphere-ocean coupled models with biophysical feedbacks are needed to examine whether western Pacific warming due to biology still occurs, and to understand how phytoplankton interacts with the climate system in such cases.

    Influences on ENSO

    We also investigated the effects of phytoplankton on ENSO variability, since chlorophyll-induced heating can modulate ENSO variability by changing the abovementioned mean state. Fig. 5 shows the distributions of standard deviation for each experiment as well as for observational data. Both the BGC and NoBGC runs accurately reproduced the overall pattern of tropical interannual variability, despite extending these too far into the western Pacific compared to observational data, which is a common problem in climate models (e.g., Yu and Kim, 2010; Moon, 2007); the causes for this are being examined. The standard deviations of the NINO3 (NINO4) regions from BGC, NoBGC, and observational data are 1.29 (1.00), 1.20 (0.79), and 0.94 (0.71) °C, respectively. Note that the inclusion of biological processes changes the amplifying variability by 7.5% (from 1.29 to 1.20 oC) in the NINO3 region and by 26.6% (from 0.79 to 1.00°C) in the NINO4 region. In this study, we focused on the NINO3 SST changes in association with ENSO modulation due to biology-induced heating, because concentration of the equatorial chlorophyll is highest in cold tongue region (Fig. 1d).

    It is evident that ocean models can accurately depict the NINO3 SST variability compared with observational data, regardless of whether ocean biology is included or not (Fig. 6). For example, both BGC and NoBGC capture a pronounced interannual oscillation associated with El Niño and La Niña events as indicated by the vertical colored bars. Note that, although our models were integrated for the period from 1950 to 2008, Fig. 6 shows only the period from 1980 to 2008 for clarity. We have checked the entire simulation data and found that the following results are not significantly changed by this choice.

    Since our model experiments use the same atmospheric forcings without air-sea interactions, it is not surprising that the simulated ENSO variations are very similar to each other and the observational data. However, the models do, to some extent, also produce ENSO changes due to biological heating. In particular, El Niño is amplified in the BGC simulation relative to NoBGC with positive values in the averaged differences of the NINO3 index during two selected El Niño periods (see the numbers 0.19 and 0.27 in Fig. 6). In the two La Niña phases indicated by blue bars in Fig. 6, these differences are 0.10 and 0.12, showing the slight weakening of La Niña. These results also indicate that the amplification of ENSO shown in Fig. 5c results mainly from the strengthening of El Niño. This intensified ENSO due to biological heating is consistent with previous studies (Marzeion et al., 2005; Lengaigne et al., 2007; Loptien et al., 2009; Park et al., 2014), although the causal mechanism is significantly different. For example, Loptien et al. (2009) argued that La Niña events are stronger with shoaling of the thermocline in the eastern Pacific under conditions of biologically induced heating.

    As mentioned above, however, the amplitude of La Niña events was weakened by phytoplankton in this study. Instead, El Niño was amplified along with the biologically-induced deepening of the thermocline (Fig. 2c). Consequently, ENSO skewness increases from 0.55 (NoBGC) to 0.59 (BGC) in association with chlorophyll-induced heating (Fig. 7). This phase dependency of the ENSO modulation on ocean biology implies that the lack of ENSO asymmetry in climate models (e.g., An et al., 2005) is presumably caused by the fact that most climate modes do not implement an interactive marine biogeochemistry model.

    To examine the mechanism behind the changes in ENSO with the presence of chlorophyll, we analyzed Hovmöller plots of the ocean evolution associated with the 1997/98 El Nino (Fig. 8). In early 1997, westerly wind anomalies in the western and central equatorial Pacific led to downwelling thermocline anomalies and ultimately the rapid development of El Niño (Fig. 8a and 8c). Note that the eastwardtraveling sea level height in BGC is larger than NoBGC (arrow in Fig. 8b), indicating the increase in amplitude of downwelling Kelvin waves due to phytoplankton heating. According to previous studies (Dewitte, 2000; Yeh et al., 2001; Moon et al., 2004), oceanic thermal structure and the corresponding changes in vertical baroclinic modes are important factors modulating the ENSO amplitude. Specifically, the enhanced higher-order baroclinic modes lead to a larger amplitude of downwelling Kelvin waves forced by westerly wind anomalies and thus stronger El Niño events. Fig. 9 shows the first three eigenfunctions An(Z) corresponding to the stability structure in the equatorial Pacific. As the vertical mode number n increases, the number of nodes crossing the zero line also increase so that higher modes are associated with the smaller scale motions. Moon et al. (2004) suggested that the wind stress acting on each baroclinic mode depends on the surface values of vertical baroclinic structure, i.e., (0). Note that the surface value of the vertical structure function ((0)) for the second mode significantly increases in the central Pacific due to biological heating. It is worth to note that this enhancement of the second baroclinic mode can have larger responses in sea surface height to wind forcing (Fig. 8b), and consequently more strongly intensify the El Niño signal.

    In contrast to the significant increase in El Niño SST anomalies, the effects of phytoplankton lead to the weakening of La Niña as explained above. Fig. 8b illustrates the overall warming in the eastern equatorial Pacific and confirms the damping effect on La Niña variability. This result is in agreement with Timmermann and Jin (2002), emphasizing the role of direct heating from phytoplankton on La Niña phases. That is, enhanced easterly winds cause stronger equatorial upwelling associated with La Niña conditions. This leads to an increase in chlorophyll concentrations in the equatorial Pacific, since stronger upwelling brings the abundant nutrients supply from deep waters as shown in Fig. 8c. Note that our model well captures the observed chlorophyll pattern (Fig. 8e). Thus, more shortwave radiation is absorbed by phytoplankton at the ocean surface, which results in surface warming and in turn weakens La Niña conditions (Fig. 8c). Without the enhanced downwelling Kelvin waves due to altered baroclinic structure, the ENSO amplitude would be reduced due to this shortwave heating (see Fig. 3 in Timmermann and Jin, 2002).

    Summary and discussion

    We investigated the influence of phytoplanktoninduced changes in the tropical oceans using the MOM4p1 ocean model coupled with a state-of-the-art biogeochemistry model (TOPAZ). Our results show that ocean biology can impact the mean state in the tropical Pacific and ENSO variability, due to the modification of light absorption in the upper ocean due to phytoplankton. This effect leads to SST warming and thermocline deepening in the equatorial western Pacific, which are associated with the accumulation of heat due to stronger westward currents. These intensified currents also contribute to a slight SST cooling in the central Pacific through the enhancement of negative SST advection from the eastern Pacific cold tongue region. Unlike previous studies that focus on equatorial upwelling as a reason for SST cooling due to biophysical feedbacks, our model experiments do not show such a change. Instead, in our results equatorial downwelling occurs in the central Pacific in association with the meridional convergent surface currents. This result shows that the biologically-induced changes in the tropical Pacific appear to be quite variable across different models. Thus, the intercomparison projects, such as the Climate Model Intercomparision Project (CMIP) and the Atmospheric Model Intercomparision Project (AMIP), are needed to investigate the role of biogeochemistry in the equatorial Pacific through a set of coordinated simulations, diagnostics and evaluations.

    These biological feedbacks also affect ENSO amplitude and skewness. The ENSO amplification is associated with a stronger El Niño signal, as the eastward-propagating downwelling Kelvin waves become more intensified due to the phytoplanktoninduced changes in the vertical structure. Indeed, the second baroclinic mode is more favored in the equatorial Pacific when the biophysical feedbacks are included, resulting in stronger El Niño events. Although the higher contribution of the second baroclinic mode could cause an increase in La Niña conditions, in our results the La Niña SST anomalies were weakened through the direct heating effect. This effect is associated with the additional shortwave heating in the chlorophyll-rich upper ocean, which is produced by upwelling of nutrient-rich deep waters during a La Niña period.

    Note that the above-mentioned two processes tend to exert a counteracting influence on ENSO: an amplifying effect associated with enhancement of the second vertical mode and negative feedback associated with direct heating; this tends to reduce ENSO variability. In our experiment, the amplifying effect seems to overwhelm the direct heating, leading to ENSO amplification as well as an increase in skewness. It therefore follows that inconsistent modeling results due to biophysical feedbacks might result from model biases associated with the vertical baroclinic mode. The definitive mechanisms, however, still remain to be determined with more complex climate models. Further studies using biological models and air-sea coupled modes are required to shed more light on the influences of biophysical feedbacks in the Earth’s oceans and climate system.

    Acknowledgment

    This research was supported by the project “Research and Development for KMA Weather, Climate, and Earth System Services” (NIMS-2016- 3100) of the National Institute of Meteorological Sciences/Korea Meteorological Administration.

    Figure

    JKESS-38-469_F1.gif

    Annual mean surface chlorophyll concentration (CHL, mg m−3 ) from 1998-2007: (a) SeaWiFS satellite observations, (b) BGC simulation results (averaged over the upper 25 m), (c) difference between the two (BGC data minus observation data), and (d) vertical distribution of chlorophyll along the equator in the Pacific (BGC data).

    JKESS-38-469_F2.gif

    Annual mean (1950-2008) differences in the tropical Pacific between two model runs: BGC (accounting for biogeochemical processes including chlorophyll) and NoBGC (control), showing BGC minus NoBGC for (a) SST ( °C), (b) surface zonal current (cm/s), (c) vertical temperature profile along the equator ( °C), and (d) surface meridional current (cm/s). The contour lines in (b) and (d) represent the averaged surface zonal and meridional currents from NoBGC, respectively. The two solid lines in (c) represent the mean depth of 20 °C isotherms (i.e., thermocline depth) from the BGC (green) and NoBGC (blue) model runs.

    JKESS-38-469_F3.gif

    Scatter plots and regression lines comparing (a) monthly surface temperature differences versus surface zonal current differences (BGC minus NoBGC) in the equatorial Pacific (160 o W to 140 o W) and (b) temperature differences at 200 m depth versus thermocline depth differences (180 o to 140 o W).

    JKESS-38-469_F4.gif

    Vertical distribution of annual mean differences (BGC minus NoBGC) of (a) zonal current along the equator (color shading, cm/s) and (b) meridional current averaged from 140 o W to 150 o W (color shading, cm/s).

    JKESS-38-469_F5.gif

    Standard deviations ( °C) of the SST anomalies for (a) BGC model results, (b) NoBGC model results, (c) their difference (BGC-NoBGC), and (d) observational data (ERSST). Note the different color scale for observational data (d) and model data (a, b). Green and blue boxes in (c) represent the NINO4 (5 o N to 5 o S, 160 o E to 150 o W) and NINO3 (5 o N to 5 o S, 150 o W to 90 o W) regions, respectively.

    JKESS-38-469_F6.gif

    Time series of the NINO3 SST index ( °C) from 1980-2008 for observational, BGC, and NoBGC data, along with the difference between BGC and NoBGC. Here we present the last 29-year result from the entire simulation (1950-2001) for clear comparison. Each of two El Niño (1982/83 and 1997/98) and La Niña (1988/89 and 1999/ 2000) events are highlighted in pink and blue, respectively. The numbers inside the highlighted areas represent the mean differences of the NINO3 index ( °C) for the corresponding El Niño and La Niña periods.

    JKESS-38-469_F7.gif

    Histograms of NINO3 SST anomalies from the (a) BGC and (b) NoBGC experiments. Solid lines indicate the fitted Gaussian curves. Bin size is 0.15 °C.

    JKESS-38-469_F8.gif

    Longitude-time Hovmöller diagram of (a) SST (°C, shading) and zonal wind stress (N/m2 , contour) anomalies for BGC, (b) SST ( °C) and sea level height (m) difference between BGC and NoBGC, (c) Chlorophyll (mg m−3 , shading) and thermocline anomalies (m, contour) for BGC, and (d) short wave heating (W/m2 ) for BGC run in the upper (0-50 m) equatorial Pacific from 1997-2001. (e) Observed chlorophyll (SeaWiFS) anomaly along the equator (October 1997 -December 2001). Arrow in (b) indicates a relatively larger SST anomaly for BGC than NoBGC in association with the wind-forced downwelling Kelvin wave.

    JKESS-38-469_F9.gif

    Vertical structure functions ((z), see Moon et al. (2004) for details) for BGC (solid lines) and NoBGC (dashed lines), calculated from vertical profiles of the Brunt- Väisälä frequency for the equatorial Pacific (160 o E to 160 o W). Red, green, and blue lines indicate the first, second, and third baroclinic modes, respectively. Note that the surface amplitude of the vertical structure function ((0)) for BGC is larger than NoBGC (particularly the second mode), indicating that the second mode of BGC is significantly enhanced compared to NoBGC.

    Table

    Reference

    1. AnS.I. HamY.G. KugJ.S. JinF.F. KangI.S. (2005) El Niño-La Niña asymmetry in the coupled model intercomparison project simulations. , J. Clim., Vol.18 ; pp.2617-2627
    2. CaneM.A. (1984) Modeling sea-level during El NiA o. , J. Phys. Oceanogr., Vol.14 ; pp.1864-1874
    3. ChavezF.P. StruttonP.G. FriederichC.E. FeelyR.A. FeldmanG.C. FoleyD.C. McPhadenM.J. (1999) Biological and chemical response of the equatorial pacific ocean to the 1997-98 El NiA o. , Science, Vol.286 ; pp.2126-2131
    4. DewitteB. (2000) Sensitivity of an intermediate oceanatmosphere coupled model of the tropical pacific to its oceanic vertical structure. , J. Clim., Vol.13 ; pp.2363-2388
    5. DunneJ.P. JohnJ.G. ShevliakovaE. StoufferR.J. KrastingJ.P. MalyshevS.L. MillyP.C. SentmanL.T. AdcroftA.J. CookeW. DunneK.A. GriffiesS.M. HallbergR.W. HarrisonM.J. LevyH. WittenbergA.T. PhillipsP.J. ZadehN. (2013) Gfdl ?(tm)s esm2 global coupled climate-carbon earth system models. Part II: Carbon system formulation and baseline simulation characteristics. , J. Clim., Vol.26 ; pp.2247-2267
    6. GriffiesS.M. SchmidtM. HerzfeldM. (2009) Elements of mom4p1. Gfdl ocean group technical report, Vol.6 ; pp.444
    7. LargeW.G. YeagerS.G. (2009) The global climatology of an interannually varying air-sea flux data set. , Clim. Dyn., Vol.33 ; pp.341-364
    8. LengaigneM. MenkesC. AumontO. GorguesT. BoppL. AndreJ.M. MadecG. (2007) Influence of the oceanic biology on the tropical pacific climate in a coupled general circulation model. , Clim. Dyn., Vol.28 ; pp.503-516
    9. LewisM.R. CarrM.E. FeldmanG.C. EsaiasW. McclainC. (1990) Influence of penetrating solarradiation on the heat-budget of the equatorial pacificocean. , Nature, Vol.347 ; pp.543-545
    10. LewisM.R. CullenJ.J. PlattT. (1983) Phytoplankton and thermal structure in the upper oceanconsequences of nonuniformity in chlorophyll profile. , J. Geophys. Res. Oceans, Vol.88 ; pp.2565-2570
    11. LoptienU. EdenC. TimmermannA. DietzeH. (2009) Effects of biologically induced differential heating in an eddy-permitting coupled ocean-ecosystem model. , J. Geophys. Res. Oceans, Vol.114 ; pp.C06011
    12. ManizzaM. Le QuereC. WatsonA.J. BuitenhuisE.T. (2005) Bio-optical feedbacks among phytoplankton,upper ocean physics and sea-ice in a global model. , Geophys. Res. Lett., Vol.32 ; pp.L05603
    13. MarzeionB. TimmermannA. MurtuguddeR. JinF.F. (2005) Biophysical feedbacks in the tropical pacific. , J. Clim., Vol.18 ; pp.58-70
    14. McClainC.R. CleaveM.L. FeldmanG.C. GreggW.W. HookerS.B. KuringN. (1998) Science quality seawifs data for global biosphere research. , Sea Technol., Vol.39 ; pp.10-16
    15. MoonB.K. (2007) Characteristics of the simulated ENSO in a CGCM. , Journal of Korean Earth Science Society, Vol.28 ; pp.343-356[in Korean with English abstract].
    16. MoonB.K. YehS.W. DewitteB. JhunJ.G. KangI.S. KirtmanB.P. (2004) Vertical structure variability in the equatorial pacific before and after the pacific climate shift of the 1970s. , Geophys. Res. Lett., Vol.31 ; pp.L03203
    17. MurtuguddeR. BeauchampJ. McClainC.R. LewisM. BusalacchiA.J. (2002) Effects of penetrative radiation on the upper tropical ocean circulation. , J. Clim., Vol.15 ; pp.470-486
    18. NakamotoS. KumarS.P. OberhuberJ.M. IshizakaJ. MuneyamaK. FrouinR. (2001) Response of the equatorial pacific to chlorophyll pigment in a mixed layer isopycnal ocean general circulation model. , Geophys. Res. Lett., Vol.28 ; pp.2021-2024
    19. ParkJ.Y. KugJ.S. SeoH. BaderJ. (2014) Impact of bio-physical feedbacks on the tropical climate in coupled and uncoupled gcms. , Clim. Dyn., Vol.43 ; pp.1811-1827
    20. PataraL. VichiM. MasinaS. FogliP.G. ManziniE. (2012) Global response to solar radiation absorbed by phytoplankton in a coupled climate model. , Clim. Dyn., Vol.39 ; pp.1951-1968
    21. SathyendranathS. GouveiaA.D. ShetyeS.R. RavindranP. PlattT. (1991) Biological-control of surface-temperature in the arabian sea. , Nature, Vol.349 ; pp.54-56
    22. SchneiderB. BoppL. GehlenM. SegschneiderJ. FrolicherT.L. CaduleP. FriedlingsteinP. DoneyS.C. BehrenfeldM.J. JoosF. (2008) Climateinduced interannual variability of marine primary and export production in three global coupled climate carbon cycle models. , Biogeosciences, Vol.5 ; pp.597-614
    23. SmithT.M. ReynoldsR.W. PetersonT.C. LawrimoreJ. (2008) Improvements to NOAA ?(tm)s historical merged land-ocean surface temperature analysis (1880-2006). , J. Clim., Vol.21 ; pp.2283-2296
    24. StruttonP.G. ChavezF.P. (2004) Biological heating in the equatorial pacific: Observed variability and potential for real-time calculation. , J. Clim., Vol.17 ; pp.1097-1109
    25. TimmermannA. JinF.F. (2002) Phytoplankton influences on tropical climate. , Geophys. Res. Lett., Vol.29 ; pp.2014
    26. VichiM. MasinaS. (2009) Skill assessment of the pelagos global ocean biogeochemistry model over the period 1980-2000. , Biogeosciences, Vol.6 ; pp.2333-2353
    27. WetzelP. Maier-ReimerE. BotzetM. JungclausJ. KeenlysideN. LatifM. (2006) Effects of ocean biology on the penetrative radiation in a coupled climate model. , J. Clim., Vol.19 ; pp.3973-3987
    28. YehS.W. DewitteB. JhunJ.G. KangI.S. (2001) The characteristic oscillation induced by coupled processes between oceanic vertical modes and atmospheric modes in the tropical pacific. , Geophys. Res. Lett., Vol.28 ; pp.2847-2850
    29. YuJ.Y. KimS.T. (2010) Identification of centralpacific and eastern-pacific types of enso in cmip3 models. , Geophys. Res. Lett., Vol.37 ; pp.L15705
    30. ZhangR.H. GaoC. KangX.B. ZhiH. WangZ.G. FengL.C. (2015) Enso modulations due to interannual variability of freshwater forcing and ocean biology-induced heating in the tropical pacific. , Sci. Rep., Vol.5