1. Introduction
Ocean bottom pressure (OBP) measures the total water and air masses above the sea floor, and thus its variations are intrinsically linked to atmospheric, oceanic, and hydrological processes. The OBP anomaly pB is particularly important for the study of transient ocean circulations in deep oceans where conventional observations have been scarce. Changes in pB and ocean circulations are intimately linked through geostrophy, which is the leading dynamical balance in the vast ocean interior away from frictional boundaries. The utility of pB in oceanographic studies has long been recognized. For instance, Eble and Gonzalez (1991) measured pB in the northeastern Pacific Ocean by using high-quality bottom pressure recorders as part of their long-term monitoring program. In addition to some in situ observations (Eble and Gonzalez 1991; Fujimoto 2003; Matsumoto et al. 2006; Peralta-Ferriz and Morison 2010), pB could be estimated indirectly using measurements of sea level and hydrographic data (Ponte 1999; Qu and Song 2009). However, the amplitude of pB variations on seasonal to decadal time scales is typically about one order of magnitude smaller than that due to sea level changes. Therefore, such indirect estimates of pB using sea level and hydrography data are prone to errors in sea level measurements. A new era of pB observations began in 2002 when two satellites were launched through the Gravity Recovery and Climate Experiment (GRACE; Save 2020; Save et al. 2016) to measure changes in Earth’s gravity field.
The GRACE program has revolutionized observations of global pB variations. Direct gravimetric measurements of pB from GRACE have been used in numerous studies in oceanography and other fields in Earth science. For instance, pB data have been used to separate the mass and steric components of sea level rise (SLR) and to examine and quantify contributions from water-mass warming to SLR (Llovel et al. 2014). In addition, GRACE data have been used in various studies of ocean dynamical processes and their variability (e.g., Bentel et al. 2015; Boening et al. 2012; Chambers and Willis 2008, 2009; Johnson and Chambers 2013; Landerer et al. 2015; Piecuch and Ponte 2014; Quinn and Ponte 2012; Song and Zlotnicki 2008), including in the Arctic Ocean where conventional observations have remained hard to obtain (e.g., Armitage et al. 2016; Fournier et al. 2020; Peralta-Ferriz et al. 2014).
In this study, we examine the seasonal variability of pB in the North Pacific Ocean with emphasis on the role of topography in shaping pB responses, both in magnitude and spatial patterns, to wind stress forcing. Previous studies have demonstrated that seasonal and interannual changes of pB in the North Pacific Ocean and Atlantic Ocean circulations are primarily wind-driven (Bingham and Hughes 2006; Piecuch 2015; Ponte 1999; Ponte et al. 2007; Qiu 2002; Song and Qu 2011; Yang 2015; Zhao and Johns 2014). The role of topography in shaping the responses of pB to seasonal wind stress forcing, however, has not been emphasized in previous studies. Gill and Niiler (1973) described a dynamical framework that governs ocean’s responses to seasonal forcing by wind stress. Their study indicates that the barotropic mode dominates the oceanic responses except in equatorial regions. This is indeed consistent with a previous study by Piecuch et al. (2015) who analyzed seasonal pB variability using the GRACE and Estimating the Circulation and Climate of the Ocean, version 4 (ECCOv4, hereinafter), datasets. For barotropic responses in an ocean with varying bathymetry, the ocean’s adjustments to wind stress curl forcing are facilitated through the topographic Sverdrup relation (Holland 1967) and propagated through topographic waves (Gill 1982). Therefore, it is natural to expect that the topography plays a leading role in deciding how the ocean adjusts to seasonal wind stress forcing.
The seasonal pB variability is examined in this paper by using GRACE gravimetric data and ocean-state estimate product ECCOv4 (release 3; Forget et al. 2015; Fukumori et al. 2017). Topographic effects are investigated by using a two-layer wind-driven model (Yang 2015; Yang and Chen 2021). The remainder of this paper is arranged as follows: GRACE, ECCOv4 data, and the linear two-layer model are introduced in section 2, followed by analyses of seasonal pB patterns, discussion of dynamic processes, and mechanisms in section 3. The topographic effects will also be discussed in section 3. Further discussion and conclusions will be given in section 4.
2. Data
The 2002–20 monthly GRACE/GRACE-FO pB data gridded to 0.25° resolution (Save 2020; Save et al. 2016) are used in this study. As explained by Johnson and Chambers (2013), there is a linear trend of about 2 cm decade−1 in equivalent sea level rise in the global-averaged pB. This linear trend from 2002 to 2020 is removed and the monthly pB climatology is computed using the detrended data. Johnson and Chambers (2013) also showed that there is a seasonal cycle of the global-mean pB that is linked to the global hydrological cycle. We compute the basin-averaged pB for the whole North Pacific Ocean from 0° to 60°N,
The monthly climatology of the ocean bottom pressure anomaly in equivalent water thickness [i.e., pB/(ρ0g)] from (a) GRACE gravimetric observations, (b) the ECCOv4 ocean-state estimate, (c) the two-layer-model control run, and (d) the two-layer-model flat-bottom run.
Citation: Journal of Physical Oceanography 53, 3; 10.1175/JPO-D-22-0140.1
Next, we analyzed the pB field from ECCOv4 that covers the period of 1992–2015. We follow the same procedure used for analyses of GRACE data to get the spatial anomaly of ECCOv4. The seasonal anomaly of pB/(gρ0) from ECCOv4 is shown in Fig. 1b. The overall spatiotemporal pattern of the monthly anomaly is very similar to that from GRACE (Fig. 1a). The correlation coefficient of seasonal pB variability between GRACE and ECCOv4, shown in Fig. 2a, is overall high especially in the interior ocean. This is consistent with a similar calculation by Piecuch et al. (2015) who used the monthly pB climatology derived from a shorter period of GRACE data (2003–11, as compared with 2002–20 in the present study). Furthermore, Fig. 2a shows that the correlation between ECCOv4 and GRACE remains high on continental margins and in semiclosed marginal seas such as the Okhotsk Sea and Japan Sea, and so on. These regions were excluded in the calculation by Piecuch et al. (2015). The overall good agreement between GRACE and ECCOv4, especially in the vast open ocean, indicates the good quality of ECCOv4 in simulating pB variability but also reflects the fact that GRACE pB data are assimilated in ECCOv4.
Correlation coefficient of seasonal pB variability between (a) GRACE and ECCOv4, (b) the two-layer model and GRACE, and (c) the two-layer model and ECCOv4.
Citation: Journal of Physical Oceanography 53, 3; 10.1175/JPO-D-22-0140.1
The model domain extends from 30°S to 65°N meridionally and from 100°E to 70°W zonally. Lateral boundaries are closed with no normal flow conditions. In addition to the control run that is described above, one additional experiment was conducted by setting the water depth to be constant at 5000 m over the whole model domain. The contrast with the control run is used to illuminate key topographic effects on pB variability.
3. Results
a. Seasonal pB variations in the North Pacific Ocean
The model with realistic topography, that is, the control run, is forced by the monthly wind stress climatology for 50 years until it reaches an equilibrium state. The seasonal variability of pB/(gρ0) from the 50th year is shown in Fig. 1c. Despite its simplicity, the model is able to simulate seasonal pB variability that compares reasonably well to that from GRACE’s gravimetric observations and ECCOv4 ocean-state estimate (Figs. 1a,b). To further quantify the model performance and the consistency between ECCOv4 and GRACE, we compute the correlation coefficient between seasonal pB anomaly from the two-layer model with that from GRACE (Fig. 2b) and ECCOv4 (Fig. 2c). Overall, the correlation is high among GRACE, ECCOv4, and the two-layer model. It is interesting to note that the model–ECCO correlation is noticeably higher than either the model–GRACE or GRACE–ECCO correlation. Low and even negative correlation are found along the eastern boundary and in the northwestern tropical area around 150°E in both Fig. 2a (GRACE–ECCO) and Fig. 2b (model–GRACE). In contrast, the model–ECCO correlation is high over the basin even along boundaries and in marginal seas such as the Japan Sea and Okhotsk Sea (Fig. 2c). It remains to be understood why both ECCOv4 and the two-layer model have lower correlations with GRACE data in those common areas. Given the fact that the model is forced only by wind stress, the overall good agreement between model and data indicates that the seasonal variability of pB in the North Pacific Ocean is predominantly wind-driven. It also demonstrates that this model resolves well key processes and mechanisms that govern the seasonal pB variations and therefore is an appropriate tool for our study.
The seasonal pB variability is characterized by high pressure in the subtropical basin between 10° and 30°N and low pressure in the subpolar basin and equatorial regions in winter (January pB anomaly shown in Figs. 1a–c). This pattern is reversed in summer months such as July–September. Spring and autumn are transitional seasons between these two patterns. Figure 1 shows that the amplitude of the pB anomaly is greater in winter than in summer (e.g., January vs July). In addition, the subpolar low pB in winter is more expansive spatially than the high pB in summer while such area changes are less obvious in the subtropical basin. These seasonal differences in pB amplitude and spatial patterns are directly related to the seasonal anomalies of wind stress curl, which is stronger and more expansive in winter than in summer (Fig. 3). This two-center pattern with opposite pB anomalies in the subtropical and subpolar basins is most noticeable in the winter and summer. In the spring and autumn, however, pB anomalies with the same signs can extend across subpolar and subtropical basins. In November, for instance, a high pB is dominant between 20° and 40°N while low pressure anomalies are seen in the Bering Sea and tropical region.
The seasonal anomaly of surface wind stress curl (OAFlux; 1992–2017 monthly climatology).
Citation: Journal of Physical Oceanography 53, 3; 10.1175/JPO-D-22-0140.1
The seasonal pB variability from GRACE, ECCOv4, and the two-layer model is highly consistent not only in magnitudes but also in spatial patterns, including those on major topographic features like continental slope, ridges, and trenches, indicating that the simple model is adequate for our process studies of topographic effects. Variations in pB reflect changes in the total water mass above the seafloor and, therefore, provide a direct measure for the convergence and divergence of the oceanic mass transports. The ocean responds to changes in wind stress through the Ekman layer. The Ekman pumping and suction, which result from a convergence and divergence of the Ekman layer transport, are the most direct and effective way for redistributing water mass and changing pB. Therefore, it is expected that seasonal pB variations are linked directly to the curl of surface wind stress, which is proportional to the Ekman pumping and suction rate (Ponte 1999; Song and Qu 2011; Qin et al. 2022).
How does pB variability, which reflects changes in the total water mass above seafloor, relate to the Ekman transport? A convergence or divergence of the Ekman transport, for instance, would initially result in an increase or decrease, respectively, of the total water mass above seafloor, that is, a positive pB anomaly. The convergence or divergence also forces a geostrophic flow through Ekman pumping or suction, respectively. On time scales much longer than the inertial period, the geostrophic flow would be established, and its transport divergence or convergence would respectively balance the Ekman transport convergence or divergence exactly (Pedlosky 1996). It is the imbalance between these two transport divergences during the geostrophic adjustment process to a seasonally varying forcing that gives rise to a positive pB anomaly when the wind stress curl is negative and to a negative pB anomaly when the wind stress curl is positive.
The seasonal anomalies of wind stress curl from OAFlux, shown in Fig. 3, reveal that the dominant pattern is an alternation between positive and negative anomalies between the subtropical and subpolar basins. In January, for instance, the curl anomaly is mostly positive to the north of 30°N and negative over the subtropical basin. A positive band of anomalous wind stress curl is also seen in the equatorial region. The seasonally strengthened wind stress curl in subpolar basin results a low pB anomalies north of 30°N in January (Fig. 1). The divergence in the subpolar basin in January is accompanied by a convergence of Ekman transport, that is, negative wind stress curl anomaly, and anomalously high pB in the subtropical basin. Similarly, the seasonally high wind stress curl in January along the equator also results in a low pB in the tropical region. In the summer, such as in July, this seasonal pattern of wind stress curl is reversed and so is pB due to the same Ekman transport mechanism. The seasonal anomalies of pB in the spring and autumn reflect the transitional patterns of wind stress curl between the summer and winter patterns (Figs. 1 and 3).
Overall, the spatiotemporal patterns of the seasonal pB anomalies are consistent with that of wind stress curl, indicating the leading role of Ekman transport in redistributing water mass on the seasonal time scales. However, considerable differences are obvious between spatial patterns of pB and wind stress curl distributions. For instance, the pattern of pB anomaly is slanted along the southeast–northwest direction in the subtropical basin while that of wind stress curl is slanted along the northeast–southwest direction. Such discrepancies are understandable because the ocean’s responses to atmospheric forcing are not stationary but propagate as waves (Gill 1982). Topographically affected pB anomalies and their propagation will be discussed further below.
b. Barotropic and baroclinic components of seasonal pB variability
Seasonal variability of (a) the barotropic and (b) the baroclinic components of the ocean bottom pressure in the two-layer model (shown in Fig. 1c).
Citation: Journal of Physical Oceanography 53, 3; 10.1175/JPO-D-22-0140.1
The discussion above is only valid for short-term variability like the seasonal pB changes. The importance of the baroclinic pB increases when the time scale of the forcing increases. In fact, the barotropic and baroclinic components of pB tend to cancel each other on a very long time scale as explained by Anderson and Gill (1975) for the spinup of circulations in a stratified ocean in response to a steady atmospheric forcing (which is often called the Anderson–Gill spinup in the literature). On time scales longer than the basin-crossing time of the slowest baroclinic Rossby waves (a few decades for the size of Pacific Ocean), geostrophic velocities, and thus pB anomalies, vanish beneath the main thermocline. This is achieved when the baroclinic and barotropic components of geostrophic velocity or pB exactly cancel each other. Figure 5 shows the evolution of pB (Fig. 5a), barotropic pB (Fig. 5b), and baroclinic pB (Fig. 5c) to the annual-mean OAFlux wind stress. At the end of the first year (Fig. 5, left panels), pB anomaly is dominated by the barotropic component. In the fifth year (Fig. 5, center panels), the magnitude of baroclinic pB is comparable to that of barotropic pB, indicating that the importance of baroclinic processes increases over a longer time scale. At the end of the 50th year (Fig. 5, right panels), the barotropic and baroclinic components nearly cancel each other and the total pB anomaly is weak.
The spinup of (a) pB anomaly and (b) its barotropic and (c) baroclinic components.
Citation: Journal of Physical Oceanography 53, 3; 10.1175/JPO-D-22-0140.1
c. Topographic effects on seasonal pB variability
The seasonal pB variability is predominantly barotropic, therefore topography must play a fundamental role in all aspects of pB variations just as it would be in a one-layer homogeneous model. Topographic effects, however, have been largely ignored in previous studies of seasonal pB variability. In an interior ocean away from side boundaries and the equatorial waveguide, the ocean’s response to a wind stress forcing is through propagations of Rossby and inertial gravity waves (Veronis and Stommel 1956). On the seasonal time scale, inertial gravity waves, however, are absent because their minimum frequencies are limited at f. Therefore, Rossby waves are the primary oceanic responses to the seasonal wind stress forcing. Rossby waves travel along PV isolines, which are the geostrophic contours or f/H (here f is the Coriolis parameter and H is water depth) isolines for barotropic waves. Boundary waves will be formed when Rossby waves reach side boundaries. As shown in Fig. 6, geostrophic contours are directly affected by bathymetry especially on continental slopes, midocean ridges and in marginal seas such as the Japan Sea or the South China Sea. This is in stark contrast to an idealized ocean with a flat bottom where Rossby waves travel along the zonally uniform f/H contours.
(a) The model domain and its bathymetry, and (b) the distribution of geostrophic contours (f/H, where f is the Coriolis parameter and H is the water depth). The thin black lines are f/H contours ranging from −5 × 10−8 to 5 × 10−8 m−1 s−1 with interval of 2.9412 × 10−9 m−1 s−1. The background colors represent the values of PV ranging from −5 × 10−8 to 5 × 10−8 m−1 s−1 with interval of 5.2632 × 10−9 m−1 s−1. Selected areas along four latitudes, as marked by boxes, are used for investigation of whether pB variability in the Japan Sea (JS0), East China Sea (ECS0), Philippine Sea (PhS0), and South China Sea (SCS0) are connected to variability in the open ocean along the same latitudes and how topography affects such connections (shown in Fig. 11, below).
Citation: Journal of Physical Oceanography 53, 3; 10.1175/JPO-D-22-0140.1
To examine topographic impacts on seasonal pB variability, we conducted an additional model simulation by setting the water depth H = 5000 m uniformly in the model. The coastline, the forcing field and all physical parameters in the model are kept the same as in the control run. The pB from this flat-bottom run shows a dipole pattern between the subtropical and subpolar basins (Fig. 1d), like that from the control run (Fig. 1c). However, significant differences arise in both the amplitude and spatial patterns of pB when compared with that from the control run. The seasonal pB anomalies in the flat-bottom run are aligned more zonally and intensified along the western boundary. The ocean’s barotropic spinup time scale, defined as the time that long Rossby waves take to travel across the basin from the eastern to western boundary (Anderson and Gill 1975), is on the order of days, much less than the seasonal time scale. Therefore, the seasonal pB anomalies, which are predominantly barotropic, can be considered as being in equilibrium with the wind stress forcing. Consequently, it can be considered as being governed by the steady dynamics of Sverdrup balance in the interior and the Munk frictional boundary layer along the western boundary (Pedlosky 1996). Indeed, pB distributions in the flat-bottom run (Fig. 1d) resemble the Sverdrup transport streamfunctions that are expected from a classic wind-driven model.
The STD of the seasonal pB variability from (a) the two-layer-model control run, (b) the two-layer-model flat-bottom run with a constant depth H = 5000 m, (c) satellite gravimetric observations from GRACE, and (d) the ECCOv4 data-assimilated ocean-state estimate.
Citation: Journal of Physical Oceanography 53, 3; 10.1175/JPO-D-22-0140.1
Figure 8 shows the magnitude of the ratio between topographic and planetary beta, that is, |βT/β| where β = df/dy, using model bathymetry. As expected, |βT| is large on steep topographic features such as continental slopes or midocean ridges. In the vast interior region, |βT| is still about 2–5 times as large as β. This ratio is similar to the ratio of the maximum STD of pB between two model runs (Figs. 7a,b). Therefore, the amplified pB variations in the flat-bottom run are mainly due to the fact that the planetary beta is much smaller than the zonally averaged topographic beta and therefore the oceanic responses are larger as explained by Eq. (8).
The ratio between the magnitude of topographic and planetary beta. Note that the magnitude of topographic beta is considerably greater than the planetary beta.
Citation: Journal of Physical Oceanography 53, 3; 10.1175/JPO-D-22-0140.1
Equation (8) is derived from the Sverdrup balance. Here we compare the Sverdrup transport computed directly from the wind stress curl from OAFlux and the model transports, using transports across 45°N as an example. We first calculated the planetary Sverdrup transport streamfunction by integrating the meridional transport from the eastern boundary where transport streamfunction is set to be zero. The integrated Sverdrup transport anomalies from the eastern boundary to 155°E near the coast of Kuril Island off the Okhotsk Sea are shown by the black lines in Fig. 9. The Sverdrup transport anomaly is about 58 Sv (1 Sv ≡ 106 m3 s−1) in January (solid black line) and −20.5 Sv in July (dashed black line). We then computed the meridional transports across 45°N from model runs. The transport anomalies in the flat-bottom run (cyan lines) are about 50 and −19.5 Sv in January and July, respectively, which are close to the Sverdrup transport. In the control run with realistic bathymetry, the transport anomalies are much smaller, about 21 Sv in January and −1.5 Sv in July (red lines). These comparisons indicate that the meridional transport anomalies on the seasonal time scales are determined by the topographic Sverdrup balance instead of the planetary one.
The seasonal anomalies of the meridional transport at 45°N that is integrated from the eastern boundary, for January (solid lines) and July (dashed lines). The lines represent the planetary Sverdrup transports (black lines), the transport anomalies from the flat-bottom run (cyan lines) and transport anomalies in the control run with realistic topography (red lines).
Citation: Journal of Physical Oceanography 53, 3; 10.1175/JPO-D-22-0140.1
In addition to large differences in the amplitude of pB variability, the spatial pattern of pB differs significantly between two model runs (Figs. 1 and 7). We postulate that this pattern difference is mainly due to topographic effects on Rossby wave propagation. The largest differences between the control run and flat-bottom run are on shelves, marginal seas as shown in Fig. 10. Seasonal pB variability in the control run is distributed more coherently along the coast and broadly on continental shelves. For instance, pB in the Okhotsk and Japan Seas are connected especially along boundaries, which is consistent with a study by Kida et al. (2016), who showed that seasonal transport between Japan Sea and the western Pacific Ocean is influenced by atmospheric forcing in the Okhotsk Sea. The pB anomaly is also coherent along the coast of East China Sea. In contrast, seasonal pB variations in the flat-bottom run are distributed zonally and changes at the western boundary are directly connected to pB variations in the interior along the same latitudes. The difference is mainly due to topographic effects on Rossby wave’s propagation pathways. Long barotropic Rossby waves propagate along f/H contours in a direction with higher f/H on their right-hand side. Therefore, pB anomalies on the continental margins in the west Pacific propagate equatorward from higher to low latitudes (e.g., Andres et al. 2011; Ma et al. 2010). Indeed, our analyses using Hovmöller diagram (not shown) reveal that signals of pB variations do propagate from the Bering Sea to the Okhotsk Sea, Japan Sea and even to the East China Sea in the model. We conducted an additional experiment (not shown) in which the bottom drag coefficient cD is increased by a factor of 10 only in the Bering Sea to the north of the Aleutian Islands. The amplitude of pB variability is considerably reduced not only in Bering Sea but along the coast in the Okhotsk and Japan Seas. In contrast to the control run, pB variations in those marginal seas in the flat-bottom model run are traced to interior forcing at the same latitudes.
Contrast between (a) the control run and (b) the flat-bottom run in the western Pacific Ocean. The thin black lines are f/H contours with values marked in the panels (m−1 s−1). The f/H contours range from −5 × 10−8 to 5 × 10−8 m−1 s−1 with interval of 1.6949 × 10−9 m−1 s−1. The background shading is the monthly climatology of the ocean bottom pressure anomaly in equivalent water depth.
Citation: Journal of Physical Oceanography 53, 3; 10.1175/JPO-D-22-0140.1
Another interesting feature is that the STD of pB in the Philippine Sea is relatively small in GRACE, ECCOv4 and model control run. The Philippine Sea is segregated from the central and eastern Pacific Ocean by the Izu–Bonin–Mariana Arc and Trenches (Fig. 6a). On the eastern side of the trenches the pB variability is relatively large as shown in Fig. 7. The monthly pB anomalies on two sides of the Izu–Bonin–Mariana Arc and Trenches are somewhat disconnected because of these major topographic features (Fig. 10), which disrupt the connections of geostrophic contours (Fig. 6b). In contrast, the pB variability is large in the Philippine Sea in the flat-bottom run (Figs. 7 and 10). This is expected due to the westward propagation of long Rossby waves that result in a western intensification in both the mean and anomaly fields. Such westward propagation is disrupted and detoured by topographically shaped geostrophic contours in the control run.
Figures 7 and 10 show that such topographic effects are more significant in areas that are on the western sides of major topographic features, such as in the Japan Sea, East China Sea, South China Sea and Philippine Sea. To illustrate this, we compare the averaged pB in selected areas in the Japan Sea (marked by box JS0 in Fig. 6b), East China Sea (box ECS0), Philippine Sea (box PhS0), and South China Sea (box SCS0), and compare each of them with pB variations at selected areas across the basin at the same latitudes (see Fig. 6b for the locations of the selected areas).
Figure 11 shows the four pairs of comparison of pB variability between the control run and the flat-bottom run along these four selected latitudes. The major differences between the control run and the flat-bottom run are clearly illuminated in these comparisons. First, the amplitude of the seasonal variability in the control run is much smaller in all these four seas than that in the flat-bottom run. This is consistent with the STD distribution shown in Fig. 7. Second, the phase of the seasonal variability is distinctly different between these two experiments in the Japan Sea and the East China Sea. The Philippine Sea and the South China Sea are less restricted by topography in their connections, in terms of geostrophic contours, to the broader Pacific Ocean. Therefore, the difference between the control run and the flat-bottom run is smaller but still significant. Third, the pB variability in JS0, ECS0, PhS0, and SCS0 in the flat-bottom run is connected closely to variations across the basin along the same latitudes. This connection is expected through the link by long Rossby waves that propagate westward along each latitude. Such connections are absent for JS0 and ECS0 in the control run and weak for PhS0 and SCS0. The North Pacific Ocean has a very complex bathymetry and thus complex and intertwined geostrophic contours. The averaged f/H value in JS0, for instance, is connected to the broader Pacific Ocean through multipathways. Variability of pB along each of such pathways could potentially affect pB changes in JS0. The main objective of this study is to highlight the importance of topography in shaping the seasonal pB variability in the North Pacific Ocean. Specific contributions from pB propagation along each pathway will be conducted in follow-on studies.
Comparisons of pB variability between the control run (left panel of each pair) and the flat-bottom run (right panel of each pair) along latitudes of (a) the Japan Sea, (b) the East China Sea, (c) the Philippine Sea, and (d) the South China Sea. Colored lines are seasonal variabilities of averaged pB anomalies in each box that is shown in Fig. 6b, and the legends in the figure represent the averaged pB of each of those boxes.
Citation: Journal of Physical Oceanography 53, 3; 10.1175/JPO-D-22-0140.1
4. Summary
In this study we examine the seasonal variability of the ocean bottom pressure pB in the North Pacific Ocean and how this variability affected by topography. Satellite gravimetric observations from GRACE and a data-assimilated ocean-state estimate (ECCOv4) are used to characterize and quantify seasonal pB variability, which can be described as alternations of high and low anomalies among the subpolar, subtropical, and equatorial regions. A two-layer wind-driven ocean model is used to examine forcing mechanisms and topographic effects. Two experiments were conducted, the control run using the realistic topography and the flat-bottom run assuming a constant ocean depth of 5000 m. The seasonal pB variability from the control run compares remarkably well to that from GRACE observations and ECCOv4 ocean-state estimate, lending good confidence that the model is appropriate for the purpose of our study. The flat-bottom run with the same surface forcing, however, simulates a pB variability that is distinctly different from the observed one, confirming the hypothesis for this study that the topography plays a leading role in shaping responses of pB to the atmospheric forcing.
The model is forced by the wind stress alone. The good agreement of pB variability between the control run, GRACE and ECCOv4 indicates that the wind stress is the leading forcing mechanism for the ocean bottom pressure changes on the seasonal time scale. Our analyses indicate that the seasonal pB variability is dominated by the barotropic component. Some relatively small but not negligible contributions from the baroclinic processes are seen in the equatorial region and along the western continental margins. The dominance of the barotropic mode implies that topography does play an important role in both the amplitude and the pattern of pB variability. A scaling analysis of the barotropic Sverdrup relation shows that the amplitude of pB variability is inversely proportional to the value of beta–a measure for the gradient of the background PV. It is shown that the topographic beta is several times larger than the planetary beta in the vast interior of the North Pacific Ocean. Therefore, the amplitude of pB variability would be much smaller when realistic topography is used in the model. This explains in terms of dynamics why the pB variability in the flat-bottom run is several times greater than that from GRACE, ECCOv4, and the control run.
Our analyses indicate that the STD of pB variability is typically small in areas on the western side of major topographic features such as ridges and trenches. We use 4 selected areas for this study, the Japan Sea (JS0), the East China Sea (ECS0), the Philippine Sea (PhS0), and the South China Sea (SCS0). It is found that the amplitudes of the variability in these four areas are not always well connected to that in the open ocean along the same latitudes. This contrasts with that from the flat-bottom run where the zonal propagation by long planetary Rossby waves links pB variability in the western basin to the interior changes along the same latitudes. In addition, it is found that the phases of the seasonal pB variability in JS0, ECS0, PhS0, and SCS0 are distinctly different between the control and flat-bottom runs. The main objective of this study is to examine the forcing mechanism and the basinwide topographic impacts on seasonal pB variability. The North Pacific Ocean has a very complex bathymetry and thus complex and intertwined propagation pathways for oceanic variability. More detailed and focused studies are needed for pB variability in particular marginal seas. Some of such regionally focused studies are still ongoing and will be reported in future papers.
Acknowledgments.
Author Chen is supported by Qingdao Postdoctoral Application Grant Program (862105040003) and Laoshan Laboratory. Author Yang was supported for this study by the WHOI-OUC Collaborative Initiative and W. Van Alan Clark Chair for Excellence in Oceanography from WHOI. We are very grateful to two anonymous reviewers whose constructive comments have helped us to improve the paper.
Data availability statement.
The GRACE data used in this study can be found online (http://www2.csr.utexas.edu/grace/RL06_mascons.html). The ECCOv4 data used in this study are the version-4 release3 of ECCO data from NASA Jet Propulsion Laboratory (https://ecco.jpl.nasa.gov/drive/files/Version4/Release3/).
REFERENCES
Anderson, D. L. T., and A. E. Gill, 1975: Spin-up of a stratified ocean, with applications to upwelling. Deep-Sea Res. Oceanogr. Abstr., 22, 583–596, https://doi.org/10.1016/0011-7471(75)90046-7.
Andres, M., Y. Kwon, and J. Yang, 2011: Observations of the Kuroshio’s barotropic and baroclinic responses to basin‐wide wind forcing. J. Geophys. Res., 116, C04011, https://doi.org/10.1029/2010JC006863.
Armitage, T. W. K., S. Bacon, A. L. Ridout, S. F. Thomas, Y. Aksenov, and D. J. Wingham, 2016: Arctic sea surface height variability and change from satellite radar altimetry and GRACE, 2003–2014. J. Geophys. Res. Oceans, 121, 4303–4322, https://doi.org/10.1002/2015JC011579.
Bentel, K., F. W. Landerer, and C. Boening, 2015: Monitoring Atlantic overturning circulation and transport variability with GRACE-type ocean bottom pressure observations – A sensitivity study. Ocean Sci., 11, 953–963, https://doi.org/10.5194/os-11-953-2015.
Bingham, R. J., and C. W. Hughes, 2006: Observing seasonal bottom pressure variability in the North Pacific with GRACE. Geophys. Res. Lett., 33, L08607, https://doi.org/10.1029/2005GL025489.
Boening, C., J. K. Willis, F. W. Landerer, R. S. Nerem, and J. Fasullo, 2012: The 2011 La Niña: So strong, the oceans fell. Geophys. Res. Lett., 39, L19602, https://doi.org/10.1029/2012GL053055.
Chambers, D. P., and J. K. Willis, 2008: Analysis of large-scale ocean bottom pressure variability in the North Pacific. J. Geophys. Res., 113, C11003, https://doi.org/10.1029/2008JC004930.
Chambers, D. P., and J. K. Willis, 2009: Low-frequency exchange of mass between ocean basins. J. Geophys. Res., 114, C11008, https://doi.org/10.1029/2009JC005518.
Cushman-Roisin, B., and J.-M. Beckers, 2011: Introduction to Geophysical Fluid Dynamics: Physical and Numerical Aspects. Academic Press, 828 pp.
Eble, M. C., and F. I. Gonzalez, 1991: Deep-ocean bottom pressure measurements in the northeast Pacific. J. Atmos. Oceanic Technol., 8, 221–233, https://doi.org/10.1175/1520-0426(1991)008<0221:DOBPMI>2.0.CO;2.
Forget, G., J.-M. Campin, P. Heimbach, C. N. Hill, R. M. Ponte, and C. Wunsch, 2015: ECCO version 4: An integrated framework for non-linear inverse modeling and global ocean state estimation. Geosci. Model Dev., 8, 3071–3104, https://doi.org/10.5194/gmd-8-3071-2015.
Fournier, S., T. Lee, X. Wang, T. W. K. Armitage, O. Wang, I. Fukumori, and R. Kwok, 2020: Sea surface salinity as a proxy for Arctic Ocean freshwater changes. J. Geophys. Res. Oceans, 125, https://doi.org/10.1029/2020JC016110.
Fujimoto, H., 2003: Ocean bottom pressure variations in the southeastern Pacific following the 1997–98 El Niño event. Geophys. Res. Lett., 30, 1456, https://doi.org/10.1029/2002GL016677.
Fukumori, I., O. Wang, I. Fenty, G. Forget, P. Heimbach, and R. M. Ponte, 2017: ECCO version 4 release 3. ECCO Tech. Rep., 10 pp., https://hdl.handle.net/1721.1/110380.
GEBCO Compilation Group, 2020: The GEBCO_2020 Grid - A continuous terrain model of the global oceans and land. British Oceanographic Data Centre, accessed 12 June 2020, https://doi.org/10.5285/a29c5465-b138-234d-e053-6c86abc040b9.
Gill, A. E., 1982: Atmosphere–Ocean Dynamics. Academic Press, 662 pp.
Gill, A. E., and P. P. Niiler, 1973: The theory of the seasonal variability in the ocean. Deep-Sea Res. Oceanogr. Abstr., 20, 141–177, https://doi.org/10.1016/0011-7471(73)90049-1.
Holland, W. R., 1967: On the wind-driven circulation in an ocean with bottom topography. Tellus, 19A, 582–600, https://doi.org/10.3402/tellusa.v19i4.9825.
Johnson, G. C., and D. P. Chambers, 2013: Ocean bottom pressure seasonal cycles and decadal trends from GRACE Release-05: Ocean circulation implications. J. Geophys. Res. Oceans, 118, 4228–4240, https://doi.org/10.1002/jgrc.20307.
Kida, S., B. Qiu, J. Yang, and X. Lin, 2016: The annual cycle of the Japan Sea throughflow. J. Phys. Oceanogr., 46, 23–39, https://doi.org/10.1175/JPO-D-15-0075.1.
Landerer, F. W., D. N. Wiese, K. Bentel, C. Boening, and M. M. Watkins, 2015: North Atlantic meridional overturning circulation variations from GRACE ocean bottom pressure anomalies. Geophys. Res. Lett., 42, 8114–8121, https://doi.org/10.1002/2015GL065730.
Llovel, W., J. K. Willis, F. W. Landerer, and I. Fukumori, 2014: Deep-ocean contribution to sea level and energy budget not detectable over the past decade. Nat. Climate Change, 4, 1031–1035, https://doi.org/10.1038/nclimate2387.
Ma, C., D. Wu, X. Lin, J. Yang, and X. Ju, 2010: An open-ocean forcing in the East China and Yellow Seas. J. Geophys. Res., 115, C12056, https://doi.org/10.1029/2010JC006179.
Matsumoto, K., T. Sato, H. Fujimoto, Y. Tamura, M. Nishino, R. Hino, T. Higashi, and T. Kanazawa, 2006: Ocean bottom pressure observation off Sanriku and comparison with ocean tide models, altimetry, and barotropic signals from ocean models. Geophys. Res. Lett., 33, L16602, https://doi.org/10.1029/2006GL026706.
Pedlosky, J., 1996: Ocean Circulation Theory. Springer-Verlag, 453 pp.
Peralta-Ferriz, C., and J. Morison, 2010: Understanding the annual cycle of the Arctic Ocean bottom pressure. Geophys. Res. Lett., 37, L10603, https://doi.org/10.1029/2010GL042827.
Peralta-Ferriz, C., J. H. Morison, J. M. Wallace, J. A. Bonin, and J. Zhang, 2014: Arctic Ocean circulation patterns revealed by GRACE. J. Climate, 27, 1445–1468, https://doi.org/10.1175/JCLI-D-13-00013.1.
Piecuch, C. G., 2015: Bottom-pressure signature of annual baroclinic Rossby waves in the northeast tropical Pacific Ocean. J. Geophys. Res. Oceans, 120, 2449–2459, https://doi.org/10.1002/2014JC010667.
Piecuch, C. G., and R. M. Ponte, 2014: Annual cycle in southern tropical Indian Ocean bottom pressure. J. Phys. Oceanogr., 44, 1605–1613, https://doi.org/10.1175/JPO-D-13-0277.1.
Piecuch, C. G., I. Fukumori, R. M. Ponte, and O. Wang, 2015: Vertical structure of ocean pressure variations with application to satellite-gravimetric observations. J. Atmos. Oceanic Technol., 32, 603–613, https://doi.org/10.1175/JTECH-D-14-00156.1.
Ponte, R. M., 1999: A preliminary model study of the large-scale seasonal cycle in bottom pressure over the global ocean. J. Geophys. Res., 104, 1289–1300, https://doi.org/10.1029/1998JC900028.
Ponte, R. M., K. J. Quinn, C. Wunsch, and P. Heimbach, 2007: A comparison of model and GRACE estimates of the large-scale seasonal cycle in ocean bottom pressure. Geophys. Res. Lett., 34, L09603, https://doi.org/10.1029/2007GL029599.
Qin, J., X. Cheng, C. Yang, N. Ou, and X. Xiong, 2022: Mechanism of interannual variability of ocean bottom pressure in the South Pacific. Climate Dyn., 59, 2103–2116, https://doi.org/10.1007/s00382-022-06198-0.
Qiu, B., 2002: Large-scale variability in the midlatitude subtropical and subpolar North Pacific Ocean: Observations and causes. J. Phys. Oceanogr., 32, 353–375, https://doi.org/10.1175/1520-0485(2002)032<0353:LSVITM>2.0.CO;2.
Qu, T., and Y. T. Song, 2009: Mindoro Strait and Sibutu Passage transports estimated from satellite data. Geophys. Res. Lett., 36, L09601, https://doi.org/10.1029/2009GL037314.
Quinn, K. J., and R. M. Ponte, 2012: High frequency barotropic ocean variability observed by GRACE and satellite altimetry. Geophys. Res. Lett., 39, L07603, https://doi.org/10.1029/2012GL051301.
Save, H., 2020: CSR GRACE and GRACE-FO RL06 Mascon Solutions v02. The University of Texas at Austin, accessed 1 July 2020, http://www2.csr.utexas.edu/grace/RL06_mascons.html.
Save, H., S. Bettadpur, and B. D. Tapley, 2016: High-resolution CSR GRACE RL05 mascons. J. Geophys. Res. Solid Earth, 121, 7547–7569, https://doi.org/10.1002/2016JB013007.
Song, Y. T., and V. Zlotnicki, 2008: Subpolar ocean bottom pressure oscillation and its links to the tropical ENSO. Int. J. Remote Sens., 29, 6091–6107, https://doi.org/10.1080/01431160802175538.
Song, Y. T., and T. Qu, 2011: Multiple satellite missions confirming the theory of seasonal oceanic variability in the northern North Pacific. Mar. Geod., 34, 477–490, https://doi.org/10.1080/01490419.2011.590110.
Veronis, G., and H. Stommel, 1956: The action of variable wind stresses on a stratified ocean. J. Mar. Res., 15, 45–75.
Yang, J., 2015: Local and remote wind stress forcing of the seasonal variability of the Atlantic Meridional Overturning Circulation (AMOC) transport at 26.5°N. J. Geophys. Res. Oceans, 120, 2488–2503, https://doi.org/10.1002/2014JC010317.
Yang, J., and K. Chen, 2021: The role of wind stress in driving the along‐shelf flow in the northwest Atlantic Ocean. J. Geophys. Res. Oceans, 126, e2020JC016757, https://doi.org/10.1029/2020JC016757.
Yu, L., and X. Jin, 2012: Buoy perspective of a high‐resolution global ocean vector wind analysis constructed from passive radiometers and active scatterometers (1987–present). J. Geophys. Res., 117, C11013, https://doi.org/10.1029/2012JC008069.
Zhao, J., and W. Johns, 2014: Wind-driven seasonal cycle of the Atlantic meridional overturning circulation. J. Phys. Oceanogr., 44, 1541–1562, https://doi.org/10.1175/JPO-D-13-0144.1.
