Professional Documents
Culture Documents
YOHEI ONUKI
Research Institute for Applied Mechanics, Kyushu University, Kasuga, Fukuoka, Japan
TOSHIYUKI HIBIYA
Department of Earth and Planetary Science, Graduate School of Science,
The University of Tokyo, Tokyo, Japan
ABSTRACT
Recent numerical and observational studies have reported that resonant wave–wave interaction
may be a crucial process for the energy loss of internal tides and the associated vertical water mixing in
the midlatitude deep ocean. Special attention has been directed to the remarkable latitudinal de-
pendence of the resonant interaction intensity; semidiurnal internal tides promptly lose their energy to
near-inertial motions through parametric subharmonic instability equatorward of the critical latitudes
298N/S, where half the tidal frequency coincides with the local inertial frequency. This feature contra-
dicts the classical theoretical prediction that resonant wave–wave interaction does not play a major role
in the tidal energy loss in the open ocean. By reformulating the kinetic equation for long internal waves
and developing its calculation method, we estimate the energy decay rates of the low-vertical-mode
semidiurnal internal tides interacting with the ‘‘ubiquitous’’ oceanic internal wave field. The result
shows rapid energy decay of the internal tides, typically within O(10) days for the lowest-mode com-
ponent, near their critical latitudes. This decay time is severalfold shorter than those in the classical
studies and, additionally, varies by a factor of 2 depending on the local depth and density structure. We
suggest from this study that the numerical integration of the kinetic equation is a more effective ap-
proach than recognized to determine the decay parameter of wave energy, which is indispensable for
the global ocean models.
DOI: 10.1175/JPO-D-17-0278.1
Ó 2018 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright
Policy (www.ametsoc.org/PUBSReuseLicenses).
Unauthenticated | Downloaded 10/09/23 09:50 PM UTC
2690 JOURNAL OF PHYSICAL OCEANOGRAPHY VOLUME 48
affects internal tides only equatorward of the latitude equation, which describes the energy transfer rate in a
where half the tidal frequency coincides with the local continuous wave spectrum in general (Hasselmann 1966;
inertial frequency. For the principal lunar semidiurnal Zakharov et al. 1992).
constituent (M 2 tide), this critical latitude is 28.88. Using the kinetic equation, Olbers and Pomphrey
Observations suggest that PSI acts most effectively (1981) and Olbers (1983) estimated the decay rates1 of
near the critical latitude, with an abrupt switch off at internal waves and, surprisingly, concluded that the
higher latitudes and gradual relaxation toward lower resonant interaction for internal tides ‘‘plays no role’’
latitudes (Hibiya and Nagasawa 2004; Kunze et al. in the energy balance of the oceanic internal wave
2006; Hibiya et al. 2007; Qiu et al. 2012). By using a spectrum. Eden and Olbers (2014, hereafter EO14)
numerical model, MacKinnon and Winters (2005) reassessed the decay rates of internal tides by im-
reported that a progressive internal wave in a merid- proving the integration algorithm of the kinetic equa-
ional plane loses a large part of its energy when tion. Nonetheless, their new calculation (Fig. 1c in
passing through the critical latitude. However, a EO14) still did not show the prominent signals of PSI
subsequent observational assessment (Alford et al. near 298N/S. In the end, the discrepancy between the
2007; MacKinnon et al. 2013a) did not show such a observational and theoretical results has not been
catastrophic decay of internal tides. Although the resolved.
energy dissipation rate was largest near 298N, its value To conquer this problem, we have reformulated the
was rather modest compared to that expected. De- kinetic equation for long internal waves and, using
spite the efforts mentioned above, a consensus on the this, reexamined the energy decay of internal tides.
energy loss of internal tides through wave–wave in- As a result, the modest peaks of the decay rates of
teractions is yet to be reached. internal tides are successfully reproduced at their
Previous theoretical studies of PSI have focused critical latitudes. Our estimate is completely different
mainly on the enhancement of disturbance waves from the previous calculation and is compatible with
from an infinitesimal state superimposed on parent the recent observational evidence. In this article,
waves, investigating the ‘‘growth rate’’ of disturbance after a brief description of the formulation and calcu-
wave amplitudes (Sonmor and Klaassen 1997; Young lation methods, the results and their physical impli-
et al. 2008; Dauxois et al. 2018). However, when cations are presented.
considering the real ocean system, the growth of
the disturbance waves ceases long after the onset of 2. Formulation and calculation method
PSI, whereas turbulent mixing associated with the
decay of internal tides continues. In terms of accurate In most of the previous studies, including EO14,
ocean circulation modeling, what we need to clarify kinetic equations were designed for interactions
is the global distribution of internal tide energy among vertically propagating waves in a uniformly
transferred to dissipation scales. Actually, Oka and stratified ocean with infinite depth [reviews can be
Niwa (2013) reproduced the Pacific deep circula- found in Müller et al. (1986) and Polzin and Lvov
tion assuming that energy for turbulent mixing is (2011)]. However, in the real ocean, the physical
supplied from internal tides propagating with a con- properties of low-vertical-mode internal waves de-
stant decay time of 30 days (Niwa and Hibiya 2011). pend on the variation of density stratification and
Furthermore, recent studies by Ansong et al. (2015) vertical boundary conditions. To extend the applica-
and Buijsman et al. (2016) showed that proper wave bility, we reformulate the kinetic equation for long
drag, which is highly inhomogeneous in space, is in- inertia–gravity waves, as described in appendix A, as
dispensable to improve the accuracy of the global follows:
internal tide model. Incorporated into the global
internal tide model, the ‘‘decay rate’’ of internal tides 1
provides useful information about the subgrid-scale Polzin and Lvov (2011) considered asymptotically the growth
rate of near-inertial energy fed by a tidal component through PSI,
energy dissipation of internal tides in the world’s specifically (53) in their paper, and found that it becomes infinite
oceans. if the kinetic equation is applied to a situation with a mono-
Unlike the growth rate of disturbances, which can chromatic internal tide. The decay rate of an internal tide, on the
be estimated by linear stability analysis, the decay rate other hand, takes a finite value regardless of the spectrum shape
of an internal tide interacting with a broadly distrib- of the tidal component as far as the overall spectrum is sufficiently
smooth. A more detailed consideration for the growth rate of PSI
uted, continuous spectrum is difficult to assess within in the framework of statistical theory is given by the authors in
the conventional analytical or experimental frame- another paper, Onuki and Hibiya (2018, manuscript submitted to
works. A more suitable theoretical model is the kinetic J. Fluid Mech.).
FIG. 1. Examples of resonant triads. The red curves indicate the loci of k2 5 (k2 , l2 ) that satisfy
(a) k1 2 k2 2 k3 5 0, v1 2 v2 2 v3 5 0 and (b) k1 2 k2 1 k3 5 0, v1 2 v2 1 v3 5 0 for given k1 5 (k1 , 0) and v1 . To
~ 1 5 3:0 m, D
create these figures, f 5 6:16 3 1025 s21 , v1 5 1:41 3 1024 s21 , D ~ 2 5 0:75 m, and D
~ 3 5 0:33 m were
chosen. Note that different scales are used in the two panels.
ð
›n1
5 på jV123 j2 (n2 n3 2 n2 n1 2 n1 n3 )d(k1 2 k2 2 k3 )d(v1 2 v2 2 v3 ) dk2,3
›t j2,3
ð
2 2på jV213 j2 (n1 n3 2 n1 n2 2 n2 n3 )d(k2 2 k1 2 k3 )d(v2 2 v1 2 v3 ) dk2,3 , (1)
j2,3
where n(k, j) represents the wave action density of the time scale much longer than each wave period. Delta
vertical water column per unit area, a function of hori- functions in the integrand specify the condition of non-
zontal wavevector k and vertical mode number j. In the linear resonance and hence restrict the triad interaction on
following, we consider only the baroclinic components some loci in wavevector space (Fig. 1). In principle, (1) can
j $ 1. Frequency v is determined by the dispersion re- be integrated to predict the time evolution of the energy
lation of inertia–gravity waves: spectrum from an arbitrary initial state. Full numerical
integration of (1) is, however, highly difficult because of
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
v(k, j) 5 f 2 1 gD ~ k2 , (2) the complexity and singularity of the kernel function.
j
We now further examine the property of (1). Let us
where f is the Coriolis parameter, D ~ j is the equiva- classify the interaction terms in (1) depending on whether
lent depth of the jth mode, and k 5 jkj is the horizon- n1 is involved or not and then rewrite them as ›n1 /›t 5
tal wavenumber. The coefficient V(k1 , k2 , k3 , j1 , j2 , j3 ) G[n2 , n3 ] 2 n[n2 , n3 ]n1 , where the first term represents
represents the coupling intensity among each triad as the energy gain of the n1 component supplied through the
defined by (A7). In an idealized situation where the resonant interactions with n2 and n3 , and the second term
density stratification is weak and uniform, the co- corresponds to the energy loss, which is proportional to
efficient V123 becomes proportional to the delta-like n1 itself. The coefficient n, a linear functional of the wave
function with respect to j1 , j2 , j3 ; specifically, it takes a action spectrum n(k, j), can then be interpreted as the
nonzero value only if j1 6 j2 6 j3 5 0 is satisfied. How- decay rate of wave energy density.
ever, in the real ocean, interaction is always allowed for Now, we separate the action density into those of one
any combination of j1 , j2 , j3 , but their coupling intensity tidal component and other background waves. By as-
varies from place to place depending on the stratifica- suming that the background energy spectrum is isotropic
tion structure and ocean depth. in the horizontal directions, representing the wave-
The kinetic equation describes the statistical evolu- vector k in the polar coordinates (k, u), redefining
tion of the action spectrum (or energy spectrum) on a n(k, j) [ 2pkn(k, j), and performing the integration in (1)
with respect to u2 , u3 , we may write the decay rate of an open ocean far from major wave generation sites. Munk
internal tide as follows: (1981) defined this spectrum in frequency and vertical
mode space as
ð
v3 jV123 j2
n(k1 , j1 ) 5 å (k n 1 k2 n3 ) dk2
j2,3 v2 ,v1 2gD ~ kS 3 2 E(v, j) 5 Ey B(v)H(j) , with v . f and j $ 1, (6)
3 3
ð
v3 jV213 j2 where Ey is the total energy density per unit area, and
1å (2k3 n2 1 k2 n3 ) dk2 , (3)
j2,3 v2 .v1 gD ~ kS B(v) and H(j) are the normalized frequency and vertical
3 3
mode spectra:
where 21/2 21
f2 (j2 1 j2+ )
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi B(v) 5 Cv2p 1 2 2 , H(j) 5 ,
S5
1
2(k1 k2 )2 1 2(k2 k3 )2 1 2(k3 k1 )2 2 k41 2 k42 2 k43 v å(j2 1 j2+ )21
4 j
(4) (7)
is the area of a triangle consisting of k1 , k2 , k3 . The res- where C is the normalization constant. Since hydrostatic
onant conditions yield k3 as a function of the integration approximation eliminates
Ð ‘ the upper limit of the disper-
variable k2 : sion relation, we take f B(v) dv 5 1. Energy density in
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi the ocean has been historically scaled with the local
u buoyancy frequency as E } N (Polzin and Lvov 2011), so
u[v 2 v (k )]2 2 f 2
k3 5 F(k2 ) [ t 1 2 2
. (5) we define Ey as
gD ~
3
ð0
The first term in (3) originates from the sum inter- Ey 5 E0 N(Z) dZ , (8)
2D
action (v1 5 v2 1 v3 , k1 5 k2 1 k3 ), which involves
PSI and takes nonzero values only when v1 . 2f is where E0 5 5:7 3 102 kg m21 s21 is a constant, and Z
satisfied, whereas the second term comes from the represents the vertical coordinate. Defining represen-
difference interaction (v1 5 v2 2 v3 , k1 5 k2 2 k3 ) that tative density rc 5 1:03 3 103 kg m23 and changing the
occurs at all latitudes. In our calculation, both types independent variables in (6), the action spectrum in
of interaction are taken into account so that the decay horizontal wavenumber and vertical mode space is
rates of M2 internal tides take finite values even over derived as
28.88N/S. qffiffiffiffiffiffiffiffi
The intervals of integration are determined from the Ey C g D ~
j
triangle inequality jk1 2 k3 j , k2 , jk1 1 k3 j. Compared n(k, j) 5 v(k, j)2p21 H(j) . (9)
rc
with the original expression (1), the expression (3) is
more simplified, and hence direct numerical computa- In the original expression (Munk 1981), the variable
tion of it is possible without time integration so long parameters in (7) were chosen as p 5 2, j+ 5 3. How-
as we have the background action spectrum n, struc- ever, in this study, we adopt p 5 2, j+ 5 10 as the
ture function Fj (z), and equivalent depth D ~ j for each ‘‘standard’’ values because the recent observa-
vertical mode. tional review in Polzin and Lvov (2011) shows that
The equivalent depth and vertical structure func- j+ typically varies in the range of 3–15 and Olbers
tions are obtained by solving the eigenvalue problem and Eden (2013) also utilized these values for their
for the linear operator D ^ defined as (A3). The geo- calculation.
graphical data for ocean depth in ETOPO1 (Amante Our target is to estimate the value of decay rate n in
and Eakins 2009) and the climatology of temperature (3) for the lowest five modes of M2 internal tides at
and salinity data in the World Ocean Atlas (Locarnini latitudes where they can exist as progressive waves.
et al. 2013; Zweng et al. 2013) are used for the calculation. The spherical domain, except for shallow areas with
Neutral density is successively calculated from the depths less than 500 m, inland seas, and equatorial
surface downward and then interpolated in the vertical areas, is partitioned into 18 3 18 grid cells. Each water
direction to produce sufficiently smooth and fine density column is then divided into 1024 layers to solve the
profiles. eigenvalue problem. Interactions between internal
In this study, we choose the so-called Garrett–Munk tides and background waves of up to 127 vertical modes
spectrum as the background wave field, which has been are examined, with each interval of wavenumber in-
believed to be the ‘‘ubiquitous’’ energy spectrum in the tegration discretized into 1024 grids for the calculation.
FIG. 2. Numerically estimated decay rates n defined in (3) for the lowest five modes of the principal lunar
semidiurnal constituent (M2 tide) interacting with the background wave spectrum. Numbers below the color bar
indicate the e-folding times of the wave energy.
Appendix B presents a method to determine the in- of the energy of high modes is subject to immediate
tervals of wavenumber integration. attenuation, resulting in local energy dissipation near
the generation sites, as anticipated by St. Laurent and
3. Results Garrett (2002).
The decay rates in the proximity of the critical
a. Spherical distribution of decay rates
latitudes2 in the North Pacific vary longitudinally by a
In Fig. 2, the decay rates of the lowest five modes factor of 2 (Fig. 3a). This reflects the variation of
of M2 internal tides are depicted on a global map. Note waveforms that are determined by the depth and density
that a logarithmic color scale is utilized. The typical decay stratification structure. Figure 3b depicts isopycnal sur-
time of the lowest mode component is O(100) days, the faces and the bottom topography for the same region as
same as that in a classical study (Olbers 1983), except near in Fig. 3a. The decay rate of each mode reaches its larg-
the critical latitudes 298N/S, where we find the most rapid est value in shallow areas, especially over the Hawaiian
decay within roughly 20 days for the lowest mode, which is Ridge located in the central Pacific. Another distinct
severalfold shorter than the previous estimates. As the
mode number increases, decay rates increase almost
monotonically. For the fifth mode, the most rapid de- 2
In our 18 resolution calculation, 28.58 is closest to the critical
cay occurs within a few days, implying that a large part latitude 28.88.
FIG. 3. (a) Estimated decay rates of the lowest five modes of the M2 internal tide along
28.58N in the Pacific. The red curve indicates the rate of the lowest mode. (b) The colored
contours are isopycnal surfaces of neutral density, successively calculated downward from the
surface, taking into account the compressibility of seawater by using data from the World
Ocean Atlas. The contour increment is 0.25 kg m23. The bottom topography in ETOPO1 is
also shown. These data are utilized to solve the eigenvalue problem for the vertical modes.
feature is the eastward increase of the decay rates of the b. Vertical distribution of interaction intensity
lowest mode, which seems to be associated with vari-
Kinetic equation (1) is defined in horizontal wavevector
ation of the density stratification structure. At this
and vertical mode space. In this expression, the coupling
latitudinal band, the subtropical gyre causes the pyc-
coefficient V123 involves the vertical coordinate Z in
nocline to become gradually shallower and sharper
the form of vertically integrated
Ð 0 triple products of vertical
eastward. In the eastern area, the shallow pycnocline
structure functions, R123 5 2D F1 (Z)F2 (Z)F3 (Z) dZ. To
confines the vertical structure of the lowest-mode
investigate the vertical dependence of the interaction
waves near the surface, reducing the equivalent depth
among internal waves, we now heuristically interchange
and hence restricting the horizontal wavelength of in-
the order of summation and integration so as to rewrite
ternal tides. Because shorter waves tend to cause stronger
the decay rate of internal tides, n, as
nonlinear interactions, the decay rate of the lowest mode
is sensitive to pycnocline depth. Although other higher ð0
modes also show longitudinal variations, a clear tendency n(k, j) 5 n0 (Z; k, j) dZ . (10)
2D
as found for the lowest mode is no longer recognized. The
effect of the variation of stratification and topography is The integrand of this expression, n0 , is interpreted as
further investigated in the following. interaction intensity, specifically written as
"ð ð #
v3 jV~123 j2 v3 jV~213 j2
n (Z; k1 , j1 ) 5 åF1 (Z)F2 (Z)F3 (Z)R123
0
(k3 n2 1 k2 n3 ) dk2 1 (2k3 n2 1 k2 n3 ) dk2 ,
j2,3 ~
v2 ,v1 2gD3 k3 S
~
v2 .v1 gD3 k3 S
(11)
FIG. 4. Vertical cross sections of interaction intensity n0 defined in (11) are depicted along (left) 179.58W and
(right) 28.58N for the lowest five modes of M2 internal tides. In hatched areas, n0 takes negative values. Black curves
show the isopycnal surfaces with an interval of 0.5 kg m23.
FIG. 5. (a) Zonal averages of the interaction intensity n0 along 28.58N in the Pacific. (b) Cumulative contribution
of n0 integrated from the surface to each depth, normalized by the value integrated over the total depth, as defined in
(12). The gray dotted lines indicate nc 5 0:8 in the horizontal axis and 1000-m depth in the vertical axis. In both
panels, those for the lowest five modes of M2 internal tides are plotted.
energy decay time of 15 days or longer, is the optimal candidate for the energy loss of low-mode internal tides
value to tune their global tide model with the historical and has been investigated theoretically and numerically
current meter records in the world’s oceans. Kelly et al. (Kerry et al. 2014; Dunphy and Lamb 2014; Dunphy
(2013) calculated the ‘‘residence time’’ of the global et al. 2017; Wagner et al. 2017). Nevertheless, resources
internal tide energy in terms of a simple expression of are not enough to see the relative importance between
(total internal tide energy)/(internal tide generation the wave–wave and wave–vortex interactions in various
rate) as 7–21 days. Zhao et al. (2016) also estimated the regions. We expect that the recent fundamental efforts
residence time of internal tide energy to be 1–1.5 days, mentioned above will be combined with high-resolution
much shorter than the previous ones. Caution is needed eddy-resolving global ocean models (Qiu et al. 2018)
in that these globally averaged values involve energy and the upcoming Surface Water and Ocean Topogra-
dissipation in the coastal regions. Far from lateral phy (SWOT) mission (Fu and Ubelmann 2014) to reveal
boundaries, the typical decay time of internal tides may this undetermined process in the near future.
become substantially longer than the average. In reality, The shortest decay time of the lowest-mode internal
signals of internal tides detected by the satellite altim- tides, 20 days in our estimate, suggests that the role of
eters propagate O(1000) km in the open ocean, surviving PSI is still limited even near 28.88N/S. Nonetheless, the
several tens of days (Zhao et al. 2016). modest peaks of the energy dissipation rates at 298N
Based on a semianalytical model, Kelly et al. (2013) on the path of internal tides propagating northward
estimated the scattering coefficients of mode-1 internal from Hawaii, reported from the Internal Waves Across
tides over small bottom topography. Their result, de- the Pacific (IWAP) experiment (Alford et al. 2007;
fined as the energy loss of an internal tide per unit MacKinnon et al. 2013a,b), could be reasonably ex-
length, is translated to the decay rate per unit time by plained in terms of the present result. Moreover, our
multiplying the group velocity of inertia–gravity wave at global estimate suggests that the high energy dissipation
each location, yielding the typical decay time of internal rates near the generation sites of internal tides between
tides as O(10) days. Mathur et al. (2014) also discussed 208 and 308N reported by a series of observations by
the scattering of the internal tides due to finite-size to- Hibiya and Nagasawa (2004) and Hibiya et al. (2006,
pography with more realistic stratification to find that 2007) are ascribed to the energy loss of high-mode in-
the scattering coefficient is dominantly determined by ternal tides due to PSI, which is a more efficient energy
the largest topography on the path of internal tides. In pathway than that for far-propagating low modes.
their calculation, scattering coefficients on the northern To discuss the ‘‘near-field tidal mixing’’ caused by
side of Hawaii, where no isolated ridges exist, are con- wave–wave interactions, our theoretical model requires
sistent with those obtained by Kelly et al. (2013). The further improvement. In the formulation of the kinetic
effect of the slowly varying eddy field is also a key equation, we treat independently the eigenmodes by
assuming that the phase of each wave component is (x, Z). Here we use the notation of horizontal vector
statistically uncorrelated, thus excluding coherent wave rotation as (a, b)? [ (2b, a). The spatial domain is
structure. In the real ocean, internal tides near the gener- bounded by the flat bottom boundary and the upper free
ation site are mostly correlated, sometimes creating a surface. Under the condition that the potential vorticity is
beam structure obliquely emanating from the bottom uniform on each isopycnal surface, the equation of mo-
(Nash et al. 2006). Using a numerical model, Nikurashin tion can be described by the canonical equations, in which
and Legg (2011) reproduced local dissipation of internal horizontal velocity potential and vertical thickness are
tides at rough topography, demonstrating bottom- conjugate variables (Lvov and Tabak 2004). Here, for
intensified energy dissipation induced by PSI. The convenience of analysis, the vertical coordinate Z is re-
present statistical theory cannot cope with this spa- garded as the reference level of each isopycnal surface in
tially confined interaction process. Occurrence of the state of rest, taking values in (2D, 0). The density r is
resonant interaction near an internal-tide generation assumed to be a prescribed nonincreasing function of Z.
site was also identified through bispectrum analysis of Letting h be the vertical excursion of each isopycnal
observation data (Sun and Pinkel 2013). A suitable surface, the thickness variable is defined as g [ ›h/›Z. By
theoretical model for near-field resonant interaction further employing the Boussinesq approximation, the
remains to be constructed. time evolution of the horizontal velocity potential
Finally, the present estimate of decay rates should be f(x, Z) and the thickness g(x, Z) can be written as
interpreted as a lower bound because in this calculation
we utilized the most basic spectrum as the background ›f dH ›g dH
52 , 5 , (A1)
field. At major dissipation sites of internal tides, higher ›t dg ›t df
wave energy with a distorted spectrum has been reported
with the Hamiltonian
(e.g., Hibiya et al. 2012). Enhanced background waves
will absorb the internal tide energy more rapidly. To ð0 ð
1 ^ dx dZ ,
detect the optimal energy balance between the internal H5 [(1 1 g)j=f 1 f =? D21 gj2 1 gg Dg]
2 2D
tides and the background spectrum, we must construct a
time-evolving model as EO14 attempted. Such a model (A2)
enables identification of the energy transfer route not ^ is
where = [ (›x , ›y ), D [ ›2x 1 ›2y . The linear operator D
only in physical space but also in mode-and-wavenumber defined such that
space, which is linked to the subsequent energy cascade
that causes actual wave dissipation and resulting water ðZ ð0
^ r(Z) r(Z0 )
mixing. This study suggests that the numerical integration Dg(Z) 5 g(Z0 ) dZ0 1 g(Z0 ) dZ0 ,
rc 2D Z rc
of the kinetic equation is a more effective approach than (A3)
recognized to determine the decay parameter of internal
waves, which is indispensable for internal tide models to where rc is the representative density of seawater. One
reproduce deep mixing in the global ocean. may find that Montgomery potential gDg ^ appears as a
pressure force in the first equation of (A1). The kinetic
Acknowledgments. The authors express their gratitude plus the available potential energy in the whole domain
to Shinichiro Kida and two anonymous reviewers for their coincides with rc H . It is noted that our formulation is
invaluable comments on the original manuscript. Numerical equivalent to that in Lvov and Tabak (2004), except for
calculations were carried out on the Fujitsu PRIMEHPC the vertical boundary conditions. If the vertical domain
FX10 System (Oakleaf-FX) in the Information Technol- is extended to 2‘ , Z , ‘ and the vertical coordinate is
ogy Center, The University of Tokyo. This study was transformed to the density r, the Hamiltonian (A2) re-
supported by JSPS KAKENHI Grants JP16H02226 and duces to the classical one. The linear Hermitian operator
JP18H04918. This paper forms part of the first author’s D^ defined in (A3) is a key ingredient in the present
doctoral dissertation at The University of Tokyo com- formulation, which specifies the eigenbasis of the linear
pleted in 2017. internal wave field, as we shall see below.
We now write the eigenvalues and corresponding eigen-
functions of D^ as D ~0 . D~ 1 . . . . and F0 (Z), F1 (Z), . . .,
APPENDIX A
which are generally called equivalent depths and ver-
tical structure functions. Without loss of generality, we
Derivation of the Kinetic Equation
can assume that the eigenfunctions are a complete or-
Let us consider a rotating stratified ocean with the thogonal set. Let us introduce a complex variable
hydrostatic approximation in the isopycnal coordinate a(k, j), which is a linear combination of f and g expanded
in vertical eigenmodes and on a horizontal Fourier basis; Now we assume that the nonlinearity is sufficiently
that is, weak: H 2 H 3 . An additional assumption that the
phase of each component is statistically independent
pffiffiffiffi ð 0 ð
v allows us to define the action density n such that
a5 pffiffiffi g(x, Z)Fj (Z)e2ikx dx dZ hay (k, j)a(k0 , j0 )i 5 (2p)2 n(k, j)d(k 2 k0 )djj0 , where the
2p 2k 2D
ð0 ð angle brackets indicate an ensemble average. The wave
ik
1 pffiffiffiffiffiffi f(x, Z)Fj (Z)e2ikx dx dZ . (A4) energy density in a vertical water column per unit area is
‘ Ð
2p 2v 2D approximately Ey 5 rc åj50 v(k, j)n(k, j) dk. Following
the conventional procedure (e.g., Zakharov et al. 1992),
Here we utilize the dispersion relation of the long
kinetic equation (1) is readily derived.
inertia–gravity wave v(k, j), defined in (2). The canon-
ical equations (A1) are consequently rewritten as
APPENDIX B
›a dH
i 5 y, (A5)
›t da
Interval of Wavenumber Integration
where the Hamiltonian H 5 H 2 1 H 3 is expressed in The interval of wavenumber integration in (3) is
terms of a as specified by the inequality
ð
H 2 5 å vay a dk and (A6a) jk1 2 k3 j , k2 , jk1 1 k3 j , (B1)
j
of grids in the integration, 1024 as related in the text, is ——, N. Furuichi, and R. Robertson, 2012: Assessment of fine-
reduced to half. scale parameterizations of turbulent dissipation rates near
mixing hotspots in the deep ocean. Geophys. Res. Lett., 39,
L24601, https://doi.org/10.1029/2012GL054068.
REFERENCES Kelly, S. M., N. L. Jones, J. D. Nash, and A. F. Waterhouse, 2013:
The geography of semidiurnal mode-1 internal-tide energy
Alford, M. H., J. A. MacKinnon, Z. Zhao, R. Pinkel, J. Klymak, and loss. Geophys. Res. Lett., 40, 4689–4693, https://doi.org/10.1002/
T. Peacock, 2007: Internal waves across the Pacific. Geophys. grl.50872.
Res. Lett., 34, L24601, https://doi.org/10.1029/2007GL031566. Kerry, C. G., B. S. Powell, and G. S. Carter, 2014: The impact of
Amante, C., and B. W. Eakins, 2009: ETOPO1 1 Arc-Minute Global subtidal circulation on internal-tide-induced mixing in the
Relief Model: Procedures, data sources and analysis. NOAA Philippine Sea. J. Phys. Oceanogr., 44, 3209–3224, https://doi.org/
Tech. Memo. NESDIS NGDC-24, 25 pp., https://www.ngdc. 10.1175/JPO-D-13-0249.1.
noaa.gov/mgg/global/relief/ETOPO1/docs/ETOPO1.pdf. Kunze, E., E. Firing, J. M. Hummon, T. K. Chereskin, and A. M.
Ansong, J. K., B. K. Arbic, M. C. Buijsman, J. G. Richman, J. F. Thurnherr, 2006: Global abyssal mixing inferred from lowered
Shriver, and A. J. Wallcraft, 2015: Indirect evidence for sub- ADCP shear and CTD strain profiles. J. Phys. Oceanogr., 36,
stantial damping of low-mode internal tides in the open ocean. 1553–1576, https://doi.org/10.1175/JPO2926.1.
J. Geophys. Res. Oceans, 120, 6057–6071, https://doi.org/ Locarnini, R. A., and Coauthors, 2013: Temperature. Vol. 1, World
10.1002/2015JC010998. Ocean Atlas 2013, NOAA Atlas NESDIS 73, 40 pp., https://
Arbic, B. K., A. J. Wallcraft, and E. J. Metzger, 2010: Concurrent data.nodc.noaa.gov/woa/WOA13/DOC/woa13_vol1.pdf.
simulation of the eddying general circulation and tides in a Lvov, Y. V., and E. G. Tabak, 2004: A Hamiltonian formulation for
global ocean model. Ocean Modell., 32, 175–187, https://doi.org/ long internal waves. Physica D, 195, 106–122, https://doi.org/
10.1016/j.ocemod.2010.01.007. 10.1016/j.physd.2004.03.010.
Buijsman, M. C., and Coauthors, 2016: Impact of parameterized ——, K. L. Polzin, and N. Yokoyama, 2012: Resonant and near-
internal wave drag on the semidiurnal energy balance in a resonant internal wave interactions. J. Phys. Oceanogr., 42,
global ocean circulation model. J. Phys. Oceanogr., 46, 1399– 669–691, https://doi.org/10.1175/2011JPO4129.1.
1419, https://doi.org/10.1175/JPO-D-15-0074.1. MacKinnon, J. A., and K. B. Winters, 2005: Subtropical catastrophe:
Dauxois, T., S. Joubaud, P. Odier, and A. Venaille, 2018: Instabilities Significant loss of low-mode tidal energy at 28.98. Geophys. Res.
of internal gravity wave beams. Annu. Rev. Fluid Mech., 50, Lett., 32, L15605, https://doi.org/10.1029/2005GL023376.
131–156, https://doi.org/10.1146/annurev-fluid-122316-044539. ——, M. H. Alford, R. Pinkel, J. Klymak, and Z. Zhao, 2013a: The
de Lavergne, C., G. Madec, J. L. Sommer, A. J. G. Nurser, and latitudinal dependence of shear and mixing in the Pacific
A. C. N. Garabato, 2016: The impact of a variable mixing ef- transiting the critical latitude for PSI. J. Phys. Oceanogr., 43,
ficiency on the abyssal overturning. J. Phys. Oceanogr., 46, 3–16, https://doi.org/10.1175/JPO-D-11-0107.1.
663–681, https://doi.org/10.1175/JPO-D-14-0259.1. ——, ——, O. Sun, R. Pinkel, Z. Zhao, and J. Klymak, 2013b:
Dunphy, M., and K. G. Lamb, 2014: Focusing and vertical mode Parametric subharmonic instability of the internal tide at
scattering of the first mode internal tide by mesoscale eddy 298N. J. Phys. Oceanogr., 43, 17–28, https://doi.org/10.1175/
interaction. J. Geophys. Res. Oceans, 119, 523–536, https://doi.org/ JPO-D-11-0108.1.
10.1002/2013JC009293. ——, and Coauthors, 2017: Climate process team on internal wave-
——, A. L. Ponte, P. Klein, and S. L. Gentil, 2017: Low-mode in- driven ocean mixing. Bull. Amer. Meteor. Soc., 98, 2429–2454,
ternal tide propagation in a turbulent eddy field. J. Phys. Oce- https://doi.org/10.1175/BAMS-D-16-0030.1.
anogr., 47, 649–665, https://doi.org/10.1175/JPO-D-16-0099.1. Mathur, M., G. S. Carter, and T. Peacock, 2014: Topographic
Eden, C., and D. Olbers, 2014: An energy compartment model for scattering of the low-mode internal tide in the deep ocean.
propagation, nonlinear interaction, and dissipation of internal J. Geophys. Res. Oceans, 119, 2165–2182, https://doi.org/
gravity waves. J. Phys. Oceanogr., 44, 2093–2106, https://doi.org/ 10.1002/2013JC009152.
10.1175/JPO-D-13-0224.1. Melet, A., S. Legg, and R. Hallberg, 2016: Climatic impacts of
Fu, L.-L., and C. Ubelmann, 2014: On the transition from profile parameterized local and remote tidal mixing. J. Climate, 29,
altimeter to swath altimeter for observing global ocean surface 3473–3500, https://doi.org/10.1175/JCLI-D-15-0153.1.
topography. J. Atmos. Oceanic Technol., 31, 560–568, https:// Müller, M., J. Y. Cherniawsky, M. G. G. Foreman, and J.-S. von
doi.org/10.1175/JTECH-D-13-00109.1. Storch, 2012: Global M2 internal tide and its seasonal vari-
Hasselmann, K., 1966: Feynman diagrams and interaction rules of ability from high resolution ocean circulation and tide mod-
wave-wave scattering processes. Rev. Geophys., 4, 1–32, https:// eling. Geophys. Res. Lett., 39, L19607, https://doi.org/10.1029/
doi.org/10.1029/RG004i001p00001. 2012GL053320.
Hibiya, T., and M. Nagasawa, 2004: Latitudinal dependence of Müller, P., G. Holloway, F. Henyey, and N. Pomphrey, 1986: Non-
diapycnal diffusivity in the thermocline estimated using a linear interactions among internal gravity waves. Rev. Geo-
finescale parameterization. Geophys. Res. Lett., 31, L01301, phys., 24, 493–536, https://doi.org/10.1029/RG024i003p00493.
https://doi.org/10.1029/2003GL017998. Munk, W., 1981: Internal waves and small-scale processes. Evolu-
——, ——, and Y. Niwa, 2006: Global mapping of diapycnal dif- tion of Physical Oceanography, B. A. Warren and C. Wunsch,
fusivity in the deep ocean based on the results of expendable Eds., MIT Press, 264–291.
current profiler (XCP) surveys. Geophys. Res. Lett., 33, Nash, J. D., E. Kunze, C. M. Lee, and T. B. Sanford, 2006: Structure of
L03611, https://doi.org/10.1029/2005GL025218. the baroclinic tide generated at Kaena Ridge, Hawaii. J. Phys.
——, ——, and ——, 2007: Latitudinal dependence of diapycnal Oceanogr., 36, 1123–1135, https://doi.org/10.1175/JPO2883.1.
diffusivity in the thermocline observed using a microstructure Nikurashin, M., and S. Legg, 2011: A mechanism for local dissipation
profiler. Geophys. Res. Lett., 34, L24602, https://doi.org/10.1029/ of internal tides generated at rough topography. J. Phys. Oce-
2007GL032323. anogr., 41, 378–395, https://doi.org/10.1175/2010JPO4522.1.
Niwa, Y., and T. Hibiya, 2011: Estimation of baroclinic tide energy circulation model. J. Geophys. Res., 117, C10024, https://doi.org/
available for deep ocean mixing based on three-dimensional 10.1029/2012JC008170.
global numerical simulations. J. Oceanogr., 67, 493–502, https:// ——, J. G. Richman, and B. K. Arbic, 2014: How stationary are the
doi.org/10.1007/s10872-011-0052-1. internal tides in a high-resolution global ocean circulation
——, and ——, 2014: Generation of baroclinic tide energy in a model? J. Geophys. Res. Oceans, 119, 2769–2787, https://doi.org/
global three-dimensional numerical model with different 10.1002/2013JC009423.
spatial grid resolutions. Ocean Modell., 80, 59–73, https://doi.org/ Simmons, H. L., R. W. Hallberg, and B. K. Arbic, 2004: Internal
10.1016/j.ocemod.2014.05.003. wave generation in a global baroclinic tide model. Deep-Sea
Oka, A., and Y. Niwa, 2013: Pacific deep circulation and ventilation Res. II, 51, 3043–3068, https://doi.org/10.1016/j.dsr2.2004.09.015.
controlled by tidal mixing away from the sea bottom. Nat. Sonmor, L. J., and G. P. Klaassen, 1997: Toward a unified theory of
Commun., 4, 2419, https://doi.org/10.1038/ncomms3419. gravity wave stability. J. Atmos. Sci., 54, 2655–2680, https://doi.org/
Olbers, D. J., 1983: Models of the oceanic internal wave field. Rev. 10.1175/1520-0469(1997)054,2655:TAUTOG.2.0.CO;2.
Geophys., 21, 1567–1606, https://doi.org/10.1029/RG021i007p01567. St. Laurent, L., and C. Garrett, 2002: The role of internal tides in
——, and N. Pomphrey, 1981: Disqualifying two candidates for the mixing the deep ocean. J. Phys. Oceanogr., 32, 2882–2899, https://
energy balance of oceanic internal waves. J. Phys. Oceanogr., doi.org/10.1175/1520-0485(2002)032,2882:TROITI.2.0.CO;2.
11, 1423–1425, https://doi.org/10.1175/1520-0485(1981)011,1423: Sun, O. M., and R. Pinkel, 2013: Subharmonic energy transfer from
DTCFTE.2.0.CO;2. the semidiurnal internal tide to near-diurnal motions over
——, and C. Eden, 2013: A global model for the diapycnal diffu- Kaena Ridge, Hawaii. J. Phys. Oceanogr., 43, 766–789, https://
sivity induced by internal gravity waves. J. Phys. Oceanogr., doi.org/10.1175/JPO-D-12-0141.1.
43, 1759–1779, https://doi.org/10.1175/JPO-D-12-0207.1. Wagner, G. L., G. Ferrando, and W. R. Young, 2017: An asymp-
Polzin, K. L., and Y. V. Lvov, 2011: Toward regional character- totic model for the propagation of oceanic internal tides
izations of the oceanic internal wavefield. Rev. Geophys., 49, through quasi-geostrophic flow. J. Fluid Mech., 828, 779–811,
RG4003, https://doi.org/10.1029/2010RG000329. https://doi.org/10.1017/jfm.2017.509.
Qiu, B., S. Chen, and G. S. Carter, 2012: Time-varying parametric Young, W. R., Y.-K. Tsang, and N. J. Balmforth, 2008: Near-in-
subharmonic instability from repeat CTD surveys in the ertial parametric subharmonic instability. J. Fluid Mech., 607,
northwestern Pacific Ocean. J. Geophys. Res., 117, C09012, 25–49, https://doi.org/10.1017/S0022112008001742.
https://doi.org/10.1029/2012JC007882. Zakharov, V. E., V. S. L’vov, and G. Falkovich, 1992: Kolmogorov
——, ——, P. Klein, J. Wang, H. Torres, L.-L. Fu, and Spectra of Turbulence I: Wave Turbulence. Springer, 264 pp.
D. Menemenlis, 2018: Seasonality in transition scale from Zhao, Z., M. H. Alford, J. B. Girton, L. Rainville, and
balanced to unbalanced motions in the World Ocean. H. L. Simmons, 2016: Global observations of open-ocean
J. Phys. Oceanogr., 48, 591–605, https://doi.org/10.1175/ mode-1 M 2 internal tides. J. Phys. Oceanogr., 46, 1657–
JPO-D-17-0169.1. 1684, https://doi.org/10.1175/JPO-D-15-0105.1.
Shriver, J. F., B. K. Arbic, J. G. Richman, R. D. Ray, E. J. Metzger, Zweng, M., and Coauthors, 2013: Salinity. Vol. 2, World
A. J. Wallcraft, and P. G. Timko, 2012: An evaluation of the Ocean Atlas 2013: NOAA Atlas NESDIS 74, 39 pp., https://
barotropic and internal tides in a high-resolution global ocean data.nodc.noaa.gov/woa/WOA13/DOC/woa13_vol2.pdf.