1. Introduction
Changes of freshwater flux can lead to changes in ocean salinity, currents, and temperature (e.g., Carton 1991; Reason 1992; Schmitz 1996; Murtugudde and Busalacchi 1998), which may in turn further feed back to the atmosphere (e.g., Zhang et al. 1999; Wu et al. 2010). Modeling studies have shown that a suppression of the freshwater flux in the coupled ocean–atmosphere system can lead to warming over the global oceans with more significant warming in high latitudes because of the changes of ocean diffusive heat transport (Williams et al. 2006, 2007).
Observations indicate a trend of freshening and salinification, respectively, in the high-latitude and subtropical oceans of both hemispheres, signifying an enhanced hydrological cycle over recent several decades (Curry et al. 2003). An enhanced global hydrological cycle has been argued to be a consequence of global warming (e.g., Bosilovich et al. 2005). A series of coupled model studies have attempted to understand responses of the Atlantic meridional overturning circulation (AMOC) to high-latitude freshening and its subsequent global climatic impacts through both oceanic and atmospheric teleconnections (e.g., Dong and Sutton 2002; Zhang and Delworth 2005; Timmermann et al. 2005; Wu et al. 2008).
Recent observational analysis further demonstrates a complicated spatial pattern of trends in freshwater changes (e.g., Yu 2007). One distinct feature is a significant freshwater loss over the global western boundary currents and their extensions (Fig. 1). Over the Kuroshio–Oyashio Extension (KOE), the maximum increase of freshwater flux reaches 1.2 mm day−1 over the past 30 yr, corresponding to a total increase of 0.2 Sv (1 Sv ≡ 106 m3 s−1) freshwater loss. It was hypothesized that the increase of the freshwater loss, mainly the evaporation, is associated with an intensification of warm water transport by the western boundary currents (e.g., Yu 2007). The changes of freshwater flux over the KOE region may induce both barotropic and baroclinic adjustment of the oceanic circulation and subsequent changes of the atmospheric circulation. This will be the major focus of this paper.
After separating from the east coast of Japan, the Kuroshio Current carries nearly 100-Sv warm water eastward into the interior. As the cold dry air from the Eurasia continent comes in contact with the warm water, a large amount of heat and moisture are extracted from the surface, resulting in strong convection (both in the atmosphere and ocean) and rainfall over the KOE region. Therefore, the freshwater flux in the KOE region displays the largest variations over the entire extratropical North Pacific. The vigorous air–sea exchanges of heat and freshwater in the KOE region may convey oceanic variability into the atmosphere. However, studies so far have mainly focused on impacts of SST and heat flux–transport changes in the KOE region on the atmospheric circulation by using either AGCMs (e.g., Peng et al. 1997) or coupled ocean–atmosphere models (e.g., Yulaeva et al. 2001; Liu and Wu 2004; Wu et al. 2005; Kwon et al. 2010). This line of research is motivated by asserting that the air–sea coupling over the KOE region plays an important role in the North Pacific decadal climate variability (e.g., Latif and Barnett 1994; Schneider and Pierce 2002; Pierce et al. 2001; Wu and Liu 2003). The atmospheric responses are determined by not only interactions among midlatitude synoptic eddies, stationary waves, and the jet stream (see a review by Kushnir et al. 2002) but also tropical changes induced by forcing over the KOE region (e.g., Wu et al. 2007; Kwon et al. 2010). Recent modeling studies also highlight roles of salinity changes in North Pacific decadal variability (Zhong and Liu 2009). Yet, processes and mechanisms of coupled ocean–atmosphere response to salinity changes in the midlatitude North Pacific remain poorly understood.
Ocean-modeling studies with idealized freshwater forcing over the KOE region reveal basin-scale changes in temperature, salinity, and circulation over the North Pacific (Huang and Mehta 2005; Huang et al. 2005). Fedorov et al. (2004) suggested that freshening in the northern extratropics can deepen the equatorial thermocline because of a reduction of the subtropical–tropical cell. These OGCM studies highlight the impacts of salinity changes over the extratropical North Pacific on both regional- and basin-scale ocean circulations but are unable to assess the subsequent imprints on the atmospheric circulation. Here, we attempt to assess the coupled ocean–atmosphere responses to recent freshwater flux changes in the KOE region using a fully coupled ocean–atmosphere models with a series of partial-coupling (PC) experiments (Wu et al. 2003). The model explicitly demonstrates that the salinfication in the KOE region can induce a coupled adjustment of ocean and atmosphere circulation involving both the extratropical North and tropical Pacific, with the tropics playing a critical role in regulating the midlatitude atmospheric responses.
The paper is arranged as follows. Section 2 briefly describes the coupled model and experimental setup. Section 3 studies the midlatitude coupled ocean–atmosphere responses to the anomalous freshwater flux forcing. The tropical Pacific responses are discussed in section 4. Section 5 briefly discusses sensitivities to the amplitudes of the freshwater forcing. A summary and some further discussions are given in section 6.
2. Model and experiment design
We use the Fast Ocean–Atmosphere Model (FOAM) version 1.5, a fully coupled model developed at the University of Wisconsin. This is an improved version of the original FOAM (version 1.0), which was described in detail in Jacob (1997). The atmospheric model is a parallel version of the National Center for Atmospheric Research (NCAR) Community Climate Model version 2 (CCM2) but with the atmospheric physics replaced with those of CCM3. The ocean model was developed following the Geophysical Fluid Dynamics Laboratory (GFDL) Modular Ocean Model (MOM). The FOAM version used here has an atmospheric resolution of R15 but with 18 vertical levels and an oceanic resolution of 1.4° × 2.8° with 32 vertical levels. Without flux adjustment, the fully coupled model has been integrated for over 2000 yr without apparent climate shift. FOAM reasonably captures features of the observed climatology (Jacob 1997). Over the North Pacific, the mean total horizontal mass transport in the model is about 40 Sv for the subtropical gyre and 10 Sv for the subpolar gyre, similar to most of the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4) models but weaker than the observed owing to a coarse OGCM resolution. FOAM also reasonably captures climate variability modes including ENSO (Liu et al. 2000) and Pacific decadal climate variability (Wu et al. 2003) etc. It should be noted that in spite of a coarse-resolution (R15) AGCM component, FOAM has been used to explore atmospheric response to midlatitude oceanic anomaly (Liu and Wu 2004) and displays responses similar to other high-resolution AGCMs (e.g., Peng et al. 1997). This is partly attributed to its 18-level vertical resolution, which is normally designed for T42 AGCM model.
As shown in Fig. 1, the annual mean evaporation minus precipitation (EmP) displays a positive trend in the KOE region over recent three decades. To examine climatic response to freshwater flux changes in the KOE region, we impose a freshwater flux forcing in the coupled model, which has the same pattern as the EmP trend in the region (20°–45°N, 130°–180°E) (Fig. 1). The maximum EmP trend in the KOE region is about 1.2mm (day × 30 yr)−1 from 1979 to 2008. To increase the signal-to-noise ratio, the magnitudes of the imposed forcing are amplified by a factor of 2.5 in the experiment. By doing this, it is also assumed that the trend will persist in the next several decades. Nevertheless, the amplification of the forcing magnitudes here is still artificial and mainly for sensitivity studies. The sensitivity to the forcing magnitudes will be also examined. To avoid potential numerical instability, the imposed forcing is gradually reduced to zero within 5° inward from both zonal and meridional boundary, containing two grids and four grids in the longitude and latitude, respectively. It should be noted that in the coupled model, the anomalous EmP forcing is only applied to the oceanic component as a virtual salt flux, with the ocean and the atmosphere remaining fully coupled both locally and elsewhere. The induced oceanic changes by the EmP forcing may feed back to the atmosphere, inducing further changes of the EmP flux. Although the EmP trend varies with the season, the forcing imposed here is fixed to the annual mean trend without seasonal variation. Our experiments demonstrate that the responses under these two different forcings are broadly similar (not shown); therefore, we will only focus on the annual mean forcing.
We conduct a 6-member forced ensemble experiment, with each experiment starting from a different state of the long control simulation and integrated for 30 yr. The difference between the ensemble mean of these forced experiments and the control simulation is taken as the response. This group of ensemble experiments is named as 2.5-EmP experiment. To assess the roles of air–sea coupling and extratropical–tropical teleconnections in coupled ocean–atmosphere responses to the EmP forcing, a series of partial-coupling experiments (Wu et al. 2003) are also conducted, which will be discussed in the corresponding section.
3. Coupled ocean–atmosphere response over the North Pacific
a. Oceanic response
1) Equilibrium response
We first examine the equilibrium responses by focusing on the last 10 yr of the 30-yr model integration. With the positive EmP forcing in the KOE region, the direct response is a significant increase of sea surface salinity (SSS) in the vicinity of the forcing region, with a maximum reaching 0.6 psu (Fig. 2a). The positive salinity anomaly also spreads eastward due to the advection of the mean flow. The salinity response also displays a distinct seasonality in the KOE region with the maximum and minimum occurring, respectively, in fall and spring, while in the midlatitude eastern North Pacific it remains nearly constant in all seasons (Fig. 2b). This can be easily interpreted in terms of the seasonal variation of the mixed layer depth in the KOE region, which deepens in winter and shoals in summer owing to different vertical mixing effects. The salinity anomaly can penetrate to 800-m depth in the KOE region, but it is trapped in the upper 300 m in the east (not shown), indicating a strong west–east asymmetry of vertical mixing in the midlatitude. Associated with an increase of surface salinity, the sea surface height (SSH) decreases accordingly (Fig. 2c), with a maximum decrease of 8 cm collocated with the center of the maximum increase of salinity (Fig. 2a). This originates from the fact that an increase of salinity leads to an increase of density, a sinking of surface water, and thus a decrease of the SSH and a shoaling (deepening) of the upper (lower) ocean pycnocline (Fig. 2d).
The change of salinity significantly modulates the ocean circulation. In the upper ocean, the anomalous flow is characterized by a cyclonic gyre straddling the mean subtropical and subpolar gyre (Fig. 3a), as a result of adjustment to the lowering of the SSH. The cyclonic inter-gyre–gyre leads to a southward shift of the Kuroshio axis by about 2 degrees (Fig. 3c). The response of the oceanic circulation is largely baroclinic (Fig. 3b), as shown in the changes of the isopycnal depths (Fig. 2d), which is consistent with previous studies (Liu 1999). The anomalous flow in the lower layer is characterized by an anticyclonic gyre, with a maximum transport of 1 Sv (not shown). The response of the ocean circulation also includes a barotropic component (Fig. 4a). It should be noted that the barotropic transport here cannot be explained by the classical Goldsbrough–Stommel theory (Goldsbrough 1933; Stommel 1984). This is because in our model, the freshwater forcing is imposed as a virtual salt flux without the input of potential vorticity. The barotropic component of the ocean circulation response is predominantly due to changes of wind stress in response to the EmP forcing. To demonstrate this, we calculate the Sverdrup transport based on the wind stress anomalies (Fig. 4b). The differences between the Sverdrup transport and the simulated barotropic transport are less significant, except in the western boundary regions as expected.
2) Transient response
The SST response exhibits substantial year-to-year variability, while the SSS response remains relatively steady (Figs. 6a,c). A close examination reveals that the initial anomalous warming is attributed to both mixing and surface heat flux and is countered by the advection (Fig. 6a). The warming cannot be sustained by mixing and heat flux because of the overwhelming effect of the cold advection, which ultimately leads to the emergence of the cold anomalies (Fig. 6a). Consistent with the analysis of the equilibrium responses, the local cooling is largely associated with anomalous meridional advection (Fig. 6b). Salinity changes are largely attributed to vertical mixing in the KOE region, while the mean zonal advection plays a major damping role by advecting the salt anomalies to the east (Figs. 6c,d).
b. Atmospheric response
The EmP forcing in the KOE region also exerts discernible changes in the air–sea fluxes and atmospheric circulation, with the strongest signals in winter (Fig. 7). Over the North Pacific, the heat flux responses are characterized by downward (upward) anomalies over the anomalous cooling (warming) region with a magnitude of 6–10 W m−2, indicating a forcing of the ocean on the atmosphere (Fig. 7b). The precipitation responses over the North Pacific are denoted by anomalous dry conditions in the west and wet conditions in the east with a magnitude of about 0.2–0.3 mm day−1 (Fig. 7a). This suggests a positive freshwater feedback operating locally in the KOE region: a positive (negative) EmP forcing leads to a cooling (warming) and thus a decrease (an increase) of precipitation, which further amplifies the EmP forcing. This local freshwater feedback is consistent with a recent study on the western tropical Pacific by Wu et al. (2010).
Over the North Pacific, the atmospheric circulation displays an equivalent batrotropic trough in both the low and upper troposphere (Fig. 8). The magnitude of the geopotential height anomalies is about 9 and 14 m at 850 and 250 mb, respectively. The dynamics controlling the atmospheric response to midlatitude SST anomaly has been suggested to be associated with interaction among synoptic eddies, stationary waves, and jet stream etc. (e.g., Kushnir et al. 2002). A consensus is that the atmosphere internal variability may play an important role in shaping the pattern of forced response (Palmer 1999). Therefore, we use a method suggested by Deser et al. (2004), which projects the atmosphere response onto the leading mode of atmospheric internal variability, to partition the atmospheric response into direct and indirect parts. In our model, the leading mode of the Northern Hemisphere atmospheric internal variability in winter resembles the northern annular mode (NAM; Thompson and Wallace 2000), with two distinctive centers of the same polarity located over the North Atlantic and the North Pacific, respectively. The projection of the atmospheric responses onto the leading mode (indirect response) and the residual (direct response, the total atmospheric response minus the indirect response) are demonstrated in the middle and right panel of Fig. 8. It can be seen that the atmospheric responses are dominated by direct responses, while the indirect responses are very weak. This appears to contradict previous studies (e.g., Deser et al. 2004; Li et al. 2009), which support a dominant role of the internal mode in dictating the midlatitude atmospheric response. The inconsistency here is largely because of the fact that the atmospheric responses in the North Pacific and the North Atlantic have opposite polarity while the leading internal mode has the same polarity (Fig. 8, middle panel). Note that the EmP forcing in the KOE region induces substantial rainfall anomalies in the western tropical Pacific (Fig. 7a). In the following section, we will demonstrate that the opposite responses in the North Pacific and the North Atlantic are predominantly mediated by the tropical Pacific.
c. Coupling effect
To further demonstrate the effects of air–sea coupling over the North Pacific and the tropical Pacific on both the ocean and atmospheric responses, we conducted two PC experiments (Wu et al. 2003). The first PC experiment, named PC_TP, is configured as the 2.5-EmP experiment but with ocean and atmosphere decoupled in the tropical Pacific (15°N–15°S). In the tropical Pacific, the atmospheric model sees a prescribed annual cycle of SST that is obtained from the control simulation, and the ocean model is forced by the full atmosphere–ocean flux calculated by the atmospheric model. The second PC experiment, named PC_NTP, is similar to the PC_TP but with ocean and atmosphere further decoupled in the North Pacific. Each PC experiment also has 6-member ensemble runs with each run integrated for 30 yr.
In the absence of tropical air–sea coupling, the SST responses over the North Pacific are dominated by basin-scale warm anomalies, with cooling virtually absent in the KOE region (Fig. 9a versus Fig. 3a). The significant reduction of the cooling can be also seen in the subsurface (Fig. 9b versus Fig. 5a). Perhaps the most striking change is the surface wind over the North Pacific, which displays an anomalous anticyclone as a sharp contrast to the anomalous cyclone in the presence of air–sea coupling in the tropical Pacific (Fig. 9a versus Fig. 7b). The anticyclonic wind anomalies can force an anomalous anticyclonic gyre (Fig. 9c). In the KOE region, the warm advection by the anomalous northward wind-driven flow should counter the cold advection by the anomalous southward salinity-driven flow, leading to a substantial reduction of the local cooling.
The modulation of the atmospheric responses by the tropics is reflected in changes of both 850- and 250-mb geopotential height. Now the atmospheric responses become more NAM-like, with an equivalent barotropic high over both the North Pacific and the North Atlantic and an equivalent barotropic low capping the Arctic (Fig. 10, left panel). A further partition of the atmospheric circulation anomalies reveals a dominant role of the internal mode in governing the responses (Fig. 10, middle and right panel). This warm-ridge response has been identified in different AGCM studies (e.g., Peng et al. 1997) and coupled modeling studies (e.g., Liu and Wu 2004). To more clearly demonstrate the mediation of the tropical Pacific, the differences of the atmospheric geopotential height anomalies between the fully coupled forced run and the PC_TP experiment are calculated. The pattern displays a Rossby wave train resembling the Pacific–North American (PNA) teleconnection pattern (Horel and Wallace 1981), which is apparently associated with the positive rainfall anomalies in the western tropical Pacific (Fig. 7a).
The local coupled feedbacks on the North Pacific are further assessed in a comparison between PC_NTP and PC_TP experiment. In the absence of local air–sea coupling (PC_NTP), the extratropical North Pacific displays similar responses as those in the presence of the air–sea coupling (PC_TP) (not shown), suggesting an overwhelming effect of the imposed freshwater forcing in driving the changes in the ocean circulation and subsurface temperature. However, the local air–sea coupling also exerts discernible effects. In general, the coupling tends to slacken the salinity-driven cyclonic gyre and reduce the subsurface cooling in the KOE region owing to an emergence of the anomalous anticyclonic wind stress in response to the freshwater forcing. To this end, the local air–sea coupling plays a negative feedback to the freshwater-induced oceanic responses.
The ocean–atmosphere adjustment in response to the KOE salinification also demonstrates the air–sea coupled feedbacks (Fig. 11). In the initial five years, the midlatitude North Pacific is dominated by warm anomalies (Fig. 11a1). This initial adjustment is largely controlled by an intensification of vertical mixing due to the salinification in the KOE region. The surface heat flux tends to enhance the warming in the KOE region (Fig. 11a2). It is noted that negative wind stress curl anomalies also develop over the North Pacific, although the statistical significance is weak (Fig. 11a4). During the second five years, temperature anomalies at both surface and subsurface display a more pronounced west–east seesaw pattern because of an advection of anomalous cyclonic gyre forced by the salt flux (Figs. 11b1,3). Both surface heat flux and wind stress curl anomalies remain similar to those in the first five years, in spite of some notable differences (Figs. 11b2,4). The negative wind stress curl anomalies should sustain the subsurface warming by accelerating the subtropical gyre to counter anomalous cold advection of the salinity-driven cyclonic gyre circulation (Figs. 11c3, d3). Associated with the recurrence of warming in the KOE region, the negative wind stress anomalies are also further substantiated (Figs. 11c4, d4), which should reinforce the warming in the KOE region. In addition, the positive downward heat flux in the KOE region should also help sustaining the anomalous anticyclonic wind (Liu and Wu 2004). Therefore, the ocean–atmosphere adjustment forced by the salt flux in the KOE region involves three phases: fast local vertical mixing adjustment, slow salinity-driven, and subsequent wind-driven circulation adjustment.
In summary, the North Pacific ocean–atmosphere responses to the salinity forcing in the KOE region are regulated by both extratropical and tropical air–sea coupling. While the extratropical air–sea coupling tends to suppress the salinity-forced oceanic response, the tropical air–sea coupling tends to sustain the oceanic response. In the next section, we will investigate processes leading to tropical responses.
4. Tropical Pacific response
Several pathways by which the North Pacific air–sea interaction can affect the tropical Pacific have been proposed (see a review by Liu and Alexander 2006). These pathways include the shallow subtropical–tropical cell (STC; McCreary and Lu 1994; Liu 1994), oceanic planetary–Kelvin wave (e.g., Lysne et al. 1997), surface ocean–atmosphere coupled process (e.g., Vimont et al. 2001; Wu et al. 2007), and the atmospheric bridge (e.g., Barnett et al. 1999). Recently, Wu et al. (2007) proposed a relay teleconnection that unifies the surface ocean–atmosphere coupled process controlled by wind–evaporation–SST (WES; Xie and Philander 1994) feedback and the STC. How does the salinification in the KOE region affect the tropical Pacific?
With the ocean–atmosphere fully coupled in both the tropical and North Pacific, tropical response is characterized by a La Niña–like pattern, with warm anomalies in the west and cold anomalies in the east (Fig. 12a). An elimination of air–sea coupling in both the tropical and North Pacific eliminate the warm anomalies in the western tropical Pacific but significantly intensify subsurface warming and cooling in the west and east, respectively (Fig. 12b). Therefore, it can be readily concluded that the surface warming in the western tropical Pacific is attributed to the air–sea coupling in both the tropical and the North Pacific, while the subsurface warming in the western tropical Pacific originates from the extratropical–tropical oceanic teleconnection.
The warm anomalies in the western tropical Pacific are largely associated with the surface-coupled WES mechanism, which conveys the extratropical SST anomalies to the tropics. To demonstrate this process, we examine the seasonal development of SST, surface wind, and latent heat flux (Fig. 13). In early winter [October–December (OND)], the imposed positive salt flux forcing in the KOE region induces warming in the central and eastern subtropical North Pacific (Fig. 13a) but is primarily limited to north of 20°N. Associated with the warm anomalies, anomalous cyclonic winds develop in the east, which decelerate the northeast trades and thus reduce the oceanic latent heat loss (Fig. 13b). Note that both wind and downward heat flux anomalies have extended to the lower latitude, which should favor an equatorward extension of the warm anomalies in the following months (Fig. 13c). At the same time, westerly wind anomalies develop in the western tropical Pacific and reduce the latent heat loss (Fig. 13d). This can further substantiate equatorward progression of the warm anomalies (Fig. 13e). The warming in the western tropical Pacific peaks in February–April (FMA) with a maximum of 0.2°C that is about 30% of the warming amplitude in the midlatitude. However, the warming in the western tropical Pacific cannot be sustained owing to an enhancement of the latent heat loss as a result of air–sea temperature contrast increase (Figs. 13f,g). In addition, the warming in the western tropical Pacific also intensifies the zonal SST gradient, triggering the easterly anomalies in the east (Fig. 13f). This will enhance the equatorial upwelling, initiate a development of cold anomalies in the east (Fig. 13g), and a further intensification of the easterly anomalies (Fig. 13h).
The seasonal evolution of SST and surface wind anomalies in the equator can be more clearly seen in Fig. 14. In general, the warming in the west prevails in the first half year, while the cooling in the east prevails from April to October. A westward propagation for both the warm and cold anomalies can be also identified, signifying the effect of the Bejerknes feedback. Overall, the zonal SST gradient is intensified, consistent with earlier studies by Fedorov et al. (2004, 2007). In their studies, they conducted some idealized OGCM studies to show that freshening in the extratropics can suppress the equatorial zonal SST gradient and attributed that to changes of the upper-ocean meridional overturning circulation. Here, we demonstrate that salinity changes over the extratropical ocean can affect the equatorial zonal SST gradient through the surface-coupled WES mechanism. This surface-coupled process should work together with the oceanic teleconnection to maintain the equatorial thermocline and zonal SST gradient anomalies.
To examine the role of the oceanic teleconnection, a heat budget analysis is conducted for the upper 30–120 m in the eastern equatorial Pacific (Fig. 15).The cooling is predominantly associated with both anomalous meridional and vertical advection, indicating an acceleration of the meridional overturning circulation. The acceleration of the meridional overturning circulation cannot be attributed to the upper-ocean wind-driven STC because the northeast trades are decelerated as a result of the surface WES process (Fig. 13), which should lead to a slowdown of the STC. Indeed, this is associated with a spinup of thermohaline circulation driven by the imposed salinity flux (Fig. 16). In the Pacific, the meridional overturning circulation is dominated by two shallow antisymmetric cells in the Northern and Southern Hemisphere, without an apparent deep interhemispheric thermohaline component (Fig. 16a). As a result, the poleward heat transports are roughly antisymmetric about the equator, indicating a strong heat divergence in the equator (Fig. 16c). This is very different from the Atlantic where the northward heat transport dominates the tropics. The imposed salinity flux in the extratropical North Pacific drives a deep thermohaline cell, with a maximum transport of about 2 Sv at 700-m depth around 40°N (Fig. 16b). This leads to an increase of the poleward heat transport by about 10% at 8°N (Fig. 16d) and a cooling of the equatorial thermocline. It is interesting to note that in the PC_NTP experiment where air–sea coupling is suppressed in both the tropical and extratropical North Pacific, the poleward heat transport is increased by about 25% at 8°N, much larger than that in the fully coupled forced run (Fig. 16d). This is due to the weakening of the shallow wind-driven STC in response to the deceleration of the northeast trades triggered by the WES process in the fully coupled forced run (Fig. 13).
The heat budget analysis also displays a warming effect of the mean meridional and vertical advection (Fig. 15). This is associated with subduction of the warm anomalies from the midlatitudes to the tropics. The extratropical–tropical subduction bridge has a general warming effect on the equatorial thermocline, countering the cooling effect due to the spinup of the meridional overturning circulation.
In summary, the tropical temperature response to a positive EmP forcing over the KOE region is determined by different dynamic processes including surface WES feedback, subsurface subduction process, and salinity-driven meridional overturning circulation. The WES process can covey the warming anomalies to the western tropical Pacific at seasonal time scales, and enhances the tropical zonal SST gradient effectively, while the oceanic subduction and density-driven overturning circulation play a competing role in the equatorial thermocline. Furthermore, the WES process also slackens the northeast trades, which can also slow down the wind-driven STC and favor a warming of the equatorial thermocline (Wu et al. 2007).
5. Sensitivity to EmP forcing amplitude
To assess the sensitivity of the coupled ocean–atmosphere response to the amplitudes of the imposed EmP forcing, an additional set of ensemble experiments are carried, which have the same EmP forcing pattern as in the 2.5-EmP experiment but with the forcing magnitudes amplified fivefold (named as 5-EmP experiment). The oceanic changes appear to be linear, as shown in both the SST and subsurface temperature anomalies with the amplitudes doubled (Fig. 17a versus Fig. 3a, Fig. 17b versus Fig. 5a).
In contrast to the oceanic response, the atmospheric responses appear to be not sensitive to the forcing amplitudes. Compared with the 2.5-EmP experiment, the magnitudes of the North Pacific geopotential height anomalies remain virtually unchanged (Figs. 17c,d versus Fig. 8, left panel), although the amplitudes of the EmP forcing are doubled. A similar tendency is also found for the surface wind (not shown). One possible interpretation is that extratropical atmospheric response is not only determined by local oceanic changes but also the changes in the western tropical Pacific. The net response is determined by these two competing effects.
6. Summary and discussion
Observations have indicated a trend of freshwater loss in the global western boundary current extension regions over recent several decades. With a fully coupled climate model, we attempt to assess the coupled ocean–atmosphere responses to the freshwater flux changes in the KOE region. The model explicitly demonstrates that the positive EmP forcing in the KOE region can set up a cyclonic gyre straddling the subtropical and subpolar gyre, which sustains southward cold advection in the west, northward warm advection in the interior, and thus a formation of temperature dipole in the midlatitudes. The EmP forcing also induces an equivalent barotropic low in the atmosphere, which is primarily controlled by the western tropical Pacific response. In the absence of the tropical air–sea coupling, the atmospheric response to the positive EmP forcing in the KOE region resembles the NAM, which has a negative feedback to the EmP forced oceanic changes in the North Pacific.
The salinification in the KOE region can strengthen the tropical Pacific zonal SST gradient through several competing processes including surface-coupled WES mechanism, subduction of extratropical warm anomalies, and a spinup of the density-driven meridional overturning circulation.
The WES mechanism intensifies the equatorial zonal SST gradient but at the same time it tends to deepen the equatorial thermocline by reducing the STC and thus the poleward heat transport. The competing effects also occur for the subduction of the extratropical warm/salty anomalies and the density-driven meridional overturning circulation. The PC_NTP experiment clearly demonstrates a dominant role of the density-driven meridional overturning circulation in controlling the changes of poleward heat transport.
The strengthening of both equatorial zonal SST gradient and thermocline also intensifies ENSO. To assess the impacts of the salinifcation in the KOE region on ENSO, we calculate the leading EOF mode of 13-month low-pass-filtered SST of the tropical Pacific obtained from the control simulation and the reconstructed database on the 6-member forced ensemble runs. The reconstructed data are a combination of the last 20 yr of the 6-member forced ensemble runs. In general, the model-simulated ENSO in the control run exhibits a pattern extending to the far west and an magnitude of about 60% of the observed ENSO (Fig. 18a), similar to the earlier version (Liu et al. 2000). The imposed positive EmP forcing in the KOE region leads to an increase of the ENSO amplitude by about 20% (Fig. 18b). The strengthening of ENSO in the 2.5-EmP experiment is consistent with a strengthening of thermocline in the eastern equatorial Pacific.
The salinification over the KOE region may partly contribute to the recent SST trend in the tropical and extratropical North Pacific basin. Observations indicate that from 1979 to 2008 the equatorial Pacific SST gradient increases by about 0.6°C and the extratropical North Pacific SST warms by about 0.7°C (Fig. 19). Our experiments suggest the salinification over the KOE region may account for about 10%–15% of the increase of the tropical Pacific zonal SST gradient and the warming in the extratropical North Pacific. In addition, the salinity-forced cyclonic gyre may potentially affect the gyre circulation change in warm climate. Recent observations indicate a poleward expansion of the Hadley circulation (Fu et al. 2006; Lu et al. 2007), which can potentially induce a poleward expansion or shift of the subtropical gyre. Climate models demonstrate such a poleward expansion of the subtropical gyre in warm climate. Here, our modeling study suggests that the poleward shift of the subtropical gyre in warm climate can be partly offset by the freshwater changes in warm climate.
Our study here mainly focuses on the effects of salinity changes induced by freshwater flux without including latent heating associated with evaporation. A significant trend has been also detected for the latent heat loss in the KOE region over three recent decades, with a pattern similar to the freshwater flux trend and a maximum increase of 20 W m−2 from 1979 to 2008 (not shown). Assessing effects of this enhanced latent heat loss remains a challenge and beyond the scope of this paper because it needs to synthesize the effects of water vapor and diabatic heating changes in the atmosphere associated with evaporation changes. In general, the latent heating component should substantiate the changes induced by the salinity flux component associated with the EmP trend in the fact that both lead to an increase of density, a sinking of surface water, and thus a decrease of the SSH and a shoaling (deepening) of the upper (lower) ocean pycnocline. Therefore, our current study may underestimate the impacts of such freshwater flux changes on the trends of both ocean and atmospheric circulation changes in warm climate. Future studies should synthesize effects of both freshwater flux and latent heat flux associated with ocean water cycle changes in warm climate.
Acknowledgments
This work is supported by the China National Key Basic Research Program (2007CB411800) and the China National Natural Science Foundation Distinguished Young Investigator Project (40788002). We appreciate comments from three anonymous reviewers, which improved the paper substantially.
REFERENCES
Adler, R. F., and Coauthors, 2003: The version 2 Global Precipitation Climatology Project (GPCP) monthly precipitation analysis (1979–present). J. Hydrometeor., 4 , 1147–1167.
Barnett, T. P., D. W. Pierce, M. Latif, D. Dommenget, and R. Saravanan, 1999: Interdecadal interactions between the tropics and midlatitudes in the Pacific basin. Geophys. Res. Lett., 26 , 615–618.
Bosilovich, M., S. Schubert, and G. Walker, 2005: Global changes of the water cycle intensity. J. Climate, 18 , 1591–1608.
Carton, J. A., 1991: Effect of seasonal surface freshwater flux on sea surface temperature in the tropical Atlantic Ocean. J. Geophys. Res., 96 , 12593–12598.
Curry, R., B. Dickson, and I. Yashayaev, 2003: A change in the freshwater balance of the Atlantic Ocean over the past four decades. Nature, 426 , 826–829.
Deser, C., G. Magnusdottir, R. Saravanan, and A. Phillips, 2004: The effects of North Atlantic SST and sea ice anomalies on the winter circulation in CCM3. Part II: Direct and indirect components of the response. J. Climate, 17 , 877–889.
Dong, B. W., and R. T. Sutton, 2002: Adjustment of the coupled ocean–atmosphere system to a sudden change in the thermohaline circulation. Geophys. Res. Lett., 29 , 1728. doi:10.1029/2002GL015229.
Fedorov, A. V., R. C. Pacanowski, S. G. Philander, and G. Boccaletti, 2004: The effect of salinity on the wind-driven circulation and the thermal structure of the upper ocean. J. Phys. Oceanogr., 34 , 1949–1966.
Fedorov, A. V., M. Barreiro, G. Boccaletti, R. Pacanowski, and S. G. Philander, 2007: The freshening of surface waters in high latitudes: Effects on the thermohaline and wind-driven circulations. J. Phys. Oceanogr., 37 , 896–907.
Fu, Q., C. M. Johanson, J. M. Wallace, and T. Reichler, 2006: Enhanced mid-latitude tropospheric warming in satellite measurements. Science, 312 , 1179–1179.
Goldsbrough, G. R., 1933: Ocean currents produced by evaporation and precipitation. Proc. Roy. Soc. London, 141A , 512–517.
Horel, J. D., and J. M. Wallace, 1981: Planetary-scale atmospheric phenomena associated with the southern oscillation. Mon. Wea. Rev., 109 , 813–829.
Huang, B., and V. M. Mehta, 2005: Response of the Pacific and Atlantic Oceans to interannual variations in net atmospheric freshwater. J. Geophys. Res., 110 , C08008. doi:10.1029/2004JC002830.
Huang, B., V. M. Mehta, and N. Schneider, 2005: Oceanic response to idealized net atmospheric freshwater in the Pacific at the decadal time scale. J. Phys. Oceanogr., 35 , 2467–2486.
Jacob, R., 1997: Low frequency variability in a simulated atmosphere ocean system. Ph.D. thesis, University of Wisconsin—Madison, 155 pp.
Kushnir, Y., W. A. Robinson, I. Blade, N. M. J. Hall, S. Peng, and R. Sutton, 2002: Atmospheric GCM response to extratropical SST anomalies: Synthesis and evaluation. J. Climate, 15 , 2233–2256.
Kwon, Y. O., M. A. Alexander, N. A. Bond, C. Frankignoul, H. Nakamura, B. Qiu, and L. Thompson, 2010: Role of Gulf Stream and Kuroshio–Oyashio systems in large-scale atmosphere–ocean interaction: A review. J. Climate, 23 , 3249–3281.
Latif, M., and T. P. Barnett, 1994: Causes of decadal climate variability over the North Pacific and North America. Science, 266 , 634–637.
Li, C., L. Wu, Q. Wang, L. Qu, and L. Zhang, 2009: An intimate coupling of ocean–atmospheric interaction over the extratropical North Atlantic and Pacific. Climate Dyn., 32 , 753–765.
Liu, Z., 1994: A simple model of mass exchange between the subtropical and tropical ocean. J. Phys. Oceanogr., 24 , 1153–1165.
Liu, Z., 1999: Forced planetary wave response in a thermocline gyre. J. Phys. Oceanogr., 29 , 1036–1055.
Liu, Z., and L. Wu, 2004: Atmospheric response to North Pacific SST: The role of ocean–atmosphere coupling. J. Climate, 17 , 1859–1882.
Liu, Z., and M. A. Alexander, 2006: Atmospheric bridge, oceanic tunnel, and global climatic teleconnections. Rev. Geophys., 45 , RG2005. doi:10.1029/2005RG000172.
Liu, Z., J. Kutzbach, and L. Wu, 2000: Modeling climatic shift of El Niño variability in the Holocene. Geophys. Res. Lett., 27 , 2265–2268.
Lu, J., G. A. Vecchi, and T. Reichler, 2007: Expansion of the Hadley cell under global warming. Geophys. Res. Lett., 34 , L06805. doi:10.1029/2006GL028443.
Lysne, J., P. Chang, and B. Giese, 1997: Impact of the extratropical Pacific on equatorial variability. Geophys. Res. Lett., 24 , 2589–2592.
McCreary, J. P., and P. Lu, 1994: On the interaction between the subtropical and the equatorial oceans: The subtropical cell. J. Phys. Oceanogr., 24 , 466–497.
Murtugudde, R., and A. J. Busalacchi, 1998: Salinity effects in a tropical ocean model. J. Geophys. Res., 103 , 3283–3300.
Palmer, T. N., 1999: A nonlinear dynamical perspective on climate prediction. J. Climate, 12 , 575–591.
Peng, S., W. A. Robinson, and M. P. Hoerling, 1997: The modeled atmospheric response to midlatitude SST anomalies and its dependence on background circulation states. J. Climate, 10 , 971–987.
Pierce, D., T. Barnett, N. Schneider, R. Saravanan, D. Dommenget, and M. Latif, 2001: The role of ocean dynamics in producing decadal climate variability in the North Pacific. Climate Dyn., 18 , 51–70.
Rayner, N., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108 , 4407. doi:10.1029/2002JD002670.
Reason, C., 1992: On the effect of ENSO precipitation anomalies in a global ocean GCM. Climate Dyn., 8 , 39–47.
Schmitz, W. J., 1996: On the World Ocean circulation. Vol. 1: Some global features/North Atlantic circulation. Woods Hole Oceanographic Institute Tech. Rep. WHOI-96-O3, 140 pp.
Schneider, N., and D. W. Pierce, 2002: Anatomy of North Pacific decadal variability. J. Climate, 15 , 586–605.
Stommel, H. M., 1984: The delicate interplay between wind stress and buoyancy input in ocean circulation: The Goldsbrough variation. Tellus, 46A , 111–119.
Thompson, D. W. J., and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13 , 1000–1016.
Timmermann, A., S. An, U. Krebs, and H. Goosse, 2005: ENSO suppression due to a weakening of the North Atlantic thermohaline circulation. J. Climate, 18 , 3122–3139.
Vimont, D., J. M. Wallace, and D. S. Battisti, 2001: Footprinting: A seasonal connection between the midlatitudes and tropics. Geophys. Res. Lett., 28 , 3923–3927.
Williams, P. D., E. Guilyardi, R. T. Sutton, J. M. Gregory, and G. Madec, 2006: On the climate response of the low-latitude Pacific Ocean to changes in the global freshwater cycle. Climate Dyn., 27 , 593–611.
Williams, P. D., E. Guilyardi, R. T. Sutton, J. M. Gregory, and G. Madec, 2007: A new feedback on climate change from the hydrological cycle. Geophys. Res. Lett., 34 , L08706. doi:10.1029/2007GL029275.
Wu, L., and Z. Liu, 2003: Decadal variability in the North Pacific: The eastern North Pacific mode. J. Climate, 16 , 3111–3131.
Wu, L., Z. Liu, R. Gallimore, R. Jacob, D. Lee, and Y. Zhong, 2003: Pacific decadal variability: The tropical Pacific mode and the North Pacific mode. J. Climate, 16 , 1101–1120.
Wu, L., D. E. Lee, and Z. Liu, 2005: The 1976/77 North Pacific climate regime shift: The role of subtropical ocean adjustment and coupled ocean–atmosphere feedbacks. J. Climate, 18 , 5125–5140.
Wu, L., Z. Liu, C. Li, and Y. Sun, 2007: Extratropical control of recent tropical Pacific decadal climate variability: A relay teleconnection. Climate Dyn., 28 , 99–112.
Wu, L., C. Li, C. Yang, and S.-P. Xie, 2008: Global teleconnections in response to a shutdown of the Atlantic meridional overturning circulation. J. Climate, 21 , 3002–3019.
Wu, L., Y. Sun, J. Zhang, L. Zhang, and S. Minobe, 2010: Coupled ocean–atmosphere response to idealized freshwater forcing over the western tropical Pacific. J. Climate, 23 , 1945–1954.
Xie, S.-P., and S. G. H. Philander, 1994: A coupled ocean–atmosphere model of relevance to the ITCZ in the eastern Pacific. Tellus, 46A , 340–350.
Yu, L., 2007: Global variations in oceanic evaporation (1958–2005): The role of the changing wind speed. J. Climate, 20 , 5376–5390.
Yu, L., and R. A. Weller, 2007: Objectively analyzed air–sea heat flux for the global ice-free oceans (1981–2005). Bull. Amer. Meteor. Soc., 88 , 527–539.
Yulaeva, E., N. Schneider, D. Pierce, and T. Barnett, 2001: Modeling of North Pacific climate variability forced by oceanic heat flux anomalies. J. Climate, 14 , 4027–4046.
Zhang, R., and T. Delworth, 2005: Simulated tropical response to a substantial weakening of the Atlantic thermohaline circulation. J. Climate, 18 , 1853–1860.
Zhang, R., L. M. Rothstein, and A. J. Busalacchi, 1999: Interannual and decadal variability of the subsurface thermal structure in the Pacific Ocean: 1961–90. Climate Dyn., 15 , 703–717.
Zhong, Y., and Z. Liu, 2009: On the mechanism of Pacific multidecadal climate variability in CCSM3: The role of the subpolar North Pacific Ocean. J. Phys. Oceanogr., 39 , 2052–2076.
Annual mean EmP trend during 1979–2008 (×30). Unit is mm day−1 yr−1. The precipitation data are from the Global Precipitation Climatology Project (GPCP) product (Adler et al. 2003), and the evaporation data are from the global ocean–atmosphere flux data product developed recently by the Woods Hole Oceanographic Institute (WHOI–OAFlux, Yu and Weller 2007).
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Annual mean (a) SSS, (c) SSH changes in the fully coupled 2.5-EmP run. Contour interval is 0.2 psu and 0.02 m for SSS and SSH, respectively. (b) Seasonality of SSS response averaged over a midlatitude belt (35°–45°N). (d) Density–longitude diagram of annual mean isopycnal depth averaged over the same latitudinal belt. Unit is m. Gray shading indicates >90% confidence level.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Ocean current and temperature response in the fully coupled 2.5-EmP run. (a) Annual mean SST anomaly (C) and currents in the upper 500 m (vectors). Gray shading indicates >90% confidence level. (b) Annual mean subsurface zonal velocity averaged between 140° and 180°E. (c) Zonal velocity averaged over (140°E–180°) in the upper 200 m in different run. Unit for currents is m s−1.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Changes of ocean transport in the fully coupled 2.5-EmP run: (a) barotropic transport and (b) Sverdrup transport calculated based on wind stress changes. Unit is Sv.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
(a) Annual mean subsurface temperature averaged over (35°–45°N). Heat budget over (b) the KOE region and (c) the eastern subtropical gyre over the North Pacific (see 2 boxes in Fig. 3a) in the fully coupled 2.5-EmP run. Ta, HFLX, Mix denote temperature anomaly, surface net heat flux forcing [positive down, scaled by 1/(ρcph)], and mixing, respectively. The rest of the terms describe advection, with capital (U, V, W, T) and lower case (u, υ, w, t) letters denoting mean and anomaly, respectively. Units for temperature and heat budget term are °C, and 5 × 10−8 °C s−1, respectively.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Time-integrated (a) heat budget and (c) salinity budget in the KOE region in the fully coupled 2.5-EmP run. (b),(d) The different terms of advection are depicted. The capital and lower case letters denote mean and anomaly. Units for time-integrated heat and salinity budget term are °C and psu, respectively.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Atmospheric response in the fully coupled 2.5-EmP run. (a) Winter precipitation anomalies. (b) Winter turbulent heat flux (contours, positive downward) and wind (vectors, m s−1) anomaly. Units for precipitation and turbulent heat flux are 0.1 mm day−1 and 2 W m−2, respectively. Shaded areas in (a) and (b) as well as vectors displayed in (b) exceed the 90% significance level based on a Student’s t test.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Projection of the total (left) JFM geopotential height anomalies at (top) 850 (upper) and (bottom) 250 mb into the Northern Hemisphere leading EOF mode in (middle) the control run (indirect response) and (right) the residual (direct response) in the fully coupled 2.5-EMP run. Contour interval is 2 m at 850 mb and 4 m at 250 mb. Area shaded exceeds 90% significance level based on a Student’s t test.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Annual mean responses in PC_TP run. (a) SST (contours with interval at 0.1 °C) and wind (vectors, unit m s−1) anomaly, (b) subsurface temperature anomaly (°C) averaged between 35° and 45°N, (c) anomalous currents in the upper 500 m between PC_TP run and fully coupled 2.5-EmP run (m s−1).
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
As in Fig. 8, but for the PC_TP experiment.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Evolution of air–sea coupled response in PC_TP run in the first 2 decades. (top to bottom) SST, heat flux (HF, positive downward), heat content (HC) and wind stress curl (WSC), respectively. Units are °C, W m−2, 108 J m−2, and 10−8 N m−3 for SST, HF, HC, and WSC, respectively. Area shaded exceeds the 90% significance level based on a Student’s t test.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Annual mean response of tropical temperature averaged over (5°S–5°N) in (a) 2.5 EmP and (b) PC_NTP. Contour interval is 0.02°C.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
(left) Seasonal mean SST and latent heat flux (contours, positive downward) and (right) surface wind (vectors) response [(a),(b) October–December (OND), (c),(d) December–February (DJF), (e),(f) FMA, and (g),(h) May–July (MJJ)] in the 2.5-EmP experiment. Units for SST and heat flux are °C, W m−2, respectively. Area shaded and vectors displayed exceed the 90% significance level based on a Student’s t test.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Seasonal evolution of tropical SST (contours) and surface wind (vectors, m s−1) averaged over (5°N, 5°S) in the fully coupled 2.5-EmP run. Shaded area exceeds 90% significance level.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Heat budget terms in the upper 120 m of the eastern tropical Pacific (see box in Fig. 12a). The meaning of each term denoted in the y axis is the same as that in Figs. 5b,c.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Transport of (a) mean and (b) anomalous meridional overturning circulation (Sv) over the Pacific sector in the 2.5-EmP experiment. (c) Mean and (d) anomalous poleward heat transport (PW) over the Pacific sector in the 2.5-EmP and PC_NTP experiment.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
(a) SST and upper 500-m ocean currents (same as Fig. 3a), (b) longitude–depth profile of temperature in the midlatitude (same as Fig. 5a), and geopotential height anomalies at (c) 850 and (d) 250 mb in the fully coupled 5-EmP run (as in Fig. 8 left panel). The units for temperature, geopotential height, and ocean current are °C, m, and m s−1, respectively.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Leading EOF mode of 13-month low-pass-filtered SST in the tropical Pacific: (a) control run and (b) the difference between the forced ensemble runs (2.5 EmP) and the control run. Unit is °C.
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1
Observed SST trend in the tropical and extratropical Pacific from 1979 to 2008. Unit is °C (30 yr)−1. The SST is from U.K. Hadley Centre Global Sea Ice and Sea Surface Temperature dataset (HadISST; Rayner et al. 2003).
Citation: Journal of Climate 24, 5; 10.1175/2010JCLI3835.1