Lecture 15. Subduction and obduction

21
1 Lecture 15. Subduction and obduction 4/17/2006 8:43 PM 1. Introduction A) Iselin's (1939) model. A major conceptual difficulty in getting subsurface layer in motion is that they are not directly in contact with the atmospheric forcing locally. However, by looking at any hydrographic section, one realizes that most isopycnals are in contact with the atmosphere, primarily at high latitudes, i.e., they are somehow ventilated through the outcropping window. Iselin (1939) made a link between the T-S relation found in a vertical section and the wintertime mixed layer at higher latitudes. His schematic picture for this ventilation process is shown in Fig. 1 in Lecture 13. The arrows indicate the speculated motion. In modern terminology the basic idea is that within the subtropical gyre water is pushed downward into the thermocline by Ekman pumping and then downwells along isopycnals as it moves southward induced by the Sverdrup dynamics. The particles' motion after their ejection from the base of the mixed layer is confined within the corresponding isopycnal surface because mixing is relatively weak within the main thermocline. The weakness of mixing in the upper ocean and below the mixed layer has been confirmed by observation, such as the recent tracer release experiments. Fig. 1. Iselin’s conceptual model for water mass formation due to water sinking along isopycnal surfaces (Iselin, 1939). Iselin's model was the first prototype for water mass formation in the oceans; however, it was incomplete in two aspects. First, Iselin neglected the mixed layer that plays a most important role in water mass formation. Second, since mixed layer depth and density change greatly from season to season, it was not clear how to make the link between water mass properties and mixed layer properties at winter as suggested by Iselin. B) How to calculate the water mass formation rate? According to Iselin's model, the Ekman pumping rate might serve as the water mass formation rate. Although, this incorrect concept had dominated for a long time, it turned out that the Ekman pumping rate is not exactly the rate of water mass formation. A better way is to calculate the mass flux across the base of the mixed layer. Mixed layer models have been developed, and they can provide the seasonal cycle of the entrainment/detrainment rate across the base of the mixed layer. Can we use the annually integrated entrainment/detrainment rate as the local water mass formation rate? The answer is NO. Water leaving the mixed layer may not enter the permanent pycnocline; instead, some of it may be re-entrained into the mixed layer

Transcript of Lecture 15. Subduction and obduction

Page 1: Lecture 15. Subduction and obduction

1

Lecture 15. Subduction and obduction 4/17/2006 8:43 PM

1. Introduction A) Iselin's (1939) model. A major conceptual difficulty in getting subsurface layer in motion is that they are not directly in contact with the atmospheric forcing locally. However, by looking at any hydrographic section, one realizes that most isopycnals are in contact with the atmosphere, primarily at high latitudes, i.e., they are somehow ventilated through the outcropping window. Iselin (1939) made a link between the T-S relation found in a vertical section and the wintertime mixed layer at higher latitudes. His schematic picture for this ventilation process is shown in Fig. 1 in Lecture 13. The arrows indicate the speculated motion. In modern terminology the basic idea is that within the subtropical gyre water is pushed downward into the thermocline by Ekman pumping and then downwells along isopycnals as it moves southward induced by the Sverdrup dynamics. The particles' motion after their ejection from the base of the mixed layer is confined within the corresponding isopycnal surface because mixing is relatively weak within the main thermocline. The weakness of mixing in the upper ocean and below the mixed layer has been confirmed by observation, such as the recent tracer release experiments.

Fig. 1. Iselin’s conceptual model for water mass formation due to water sinking along isopycnal surfaces (Iselin, 1939). Iselin's model was the first prototype for water mass formation in the oceans; however, it was incomplete in two aspects. First, Iselin neglected the mixed layer that plays a most important role in water mass formation. Second, since mixed layer depth and density change greatly from season to season, it was not clear how to make the link between water mass properties and mixed layer properties at winter as suggested by Iselin. B) How to calculate the water mass formation rate? According to Iselin's model, the Ekman pumping rate might serve as the water mass formation rate. Although, this incorrect concept had dominated for a long time, it turned out that the Ekman pumping rate is not exactly the rate of water mass formation. A better way is to calculate the mass flux across the base of the mixed layer. Mixed layer models have been developed, and they can provide the seasonal cycle of the entrainment/detrainment rate across the base of the mixed layer. Can we use the annually integrated entrainment/detrainment rate as the local water mass formation rate? The answer is NO. Water leaving the mixed layer may not enter the permanent pycnocline; instead, some of it may be re-entrained into the mixed layer

Page 2: Lecture 15. Subduction and obduction

2

downstream. Similarly, water entrained into the mixed layer may not come from the permanent pycnocline; instead it may be water temporarily detrained upstream. C) Subduction/obduction rate A modified conceptual model is shown in Fig. 2. The upper ocean is divided into four layers, the Ekman layer, the mixed layer, the seasonal pycnocline, and the permanent pycnocline. Water mass exchange between the mixed layer and the seasonal pycnocline is called entrainment/detrainment, while water mass exchange between the seasonal pycnocline and the permanent pycnocline is called subduction/obduction. Accordingly, the annual mean subduction rate is defined as the total amount of water going from the mixed layer, passing through the seasonal pycnocline, to the permanent pycnocline irreversibly in one year. Similarly, the annual mean obduction rate is defined as the total amount of water going from the permanent pycnocline, passing through the seasonal pycnocline, to the mixed layer irreversibly in one year.

Fig. 2. Water mass formation and elimination through subduction and obduction process. 2. The Stommel demon A major technical difficulty was that water properties and mixed layer depth vary considerably in a year. By carefully analyzing the processes involved, Stommel (1979) was able to show that a process is at work that selects only the late winter water for actual subduction into the permanent pycnocline, Fig. 3. As will be discussed later, the effective detrainment period is marked by the Lagrangian trajectories of water particles released from the base of the mixed layer. The basic mechanism is that the mixed layer reaches its annual maximum density and depth at late winter, so there is a very thick layer of almost vertically homogenized water. When spring comes,

Page 3: Lecture 15. Subduction and obduction

3

the mixed layer shoals very quickly (as indicated by the sharp turning of the mixed layer depth in the upper and lower panels of Fig. 3) and leaves the homogenized water behind, so that water subducted has properties very close to those of the late winter mixed layer. It is readily seen that if the time evolution of the mixed layer depth really is like a δ function, i.e., 0TΔ → , the subducted water would have the properties of late winter water. Stommel called this process a demon; thus, it seems appropriate for us to call it the "Stommel demon" to honor his contribution.

Fig. 3. The Stommel demon: how the subduction selects the mixed layer properties at late winter. Notice that the subduction process in the oceans is a very complicated process involving the seasonal cycle. In fact, now the challenging problem is to find out the annual mean subduction rate including the seasonal cycle. In one way, this mean can be considered as some kind of weighted average of the instantaneous detrainment rate. Choosing the late winter properties is equivalent to using a δ function as the weight-function. Stommel's suggestion has been used extensively in almost all theoretical models of the ventilated pycnocline. What Stommel suggested gives an elegant solution to this rather intricate problem. This can be thought of as the lowest-order solution. The next step is to find out a weight-function that is better than the δ function. In other words, we would like to know the next-order correction to the subduction rate calculated according to the Stommel formula. As such a correction must include the seasonal cycle, it is not an easy problem. 3. Subduction Let us begin with a layered model without a seasonal cycle. In such a model the ventilation process can be divided into two steps. First, ventilation occurs when water flows downward from the mixed layer into a layer below. Second, water in each density layer follows a southward motion

Page 4: Lecture 15. Subduction and obduction

4

induced by Sverdrup dynamics. Thus, water within a dense layer will move underneath the next lighter layer. This process of the submersion of a denser layer under a lighter one is called subduction. The term subduction has been used in geology to describe the similar process during the movement of tectonic plates. According to this strict classification, as the number of layers increases, the first stage (ventilation) becomes shorter and shorter. It is readily seen that for a continuously stratified ocean, these two stages will merge, and we use the term subduction and reserve the term ventilation for the general case of either subduction or obduction. The seasonal cycle plays one of the most important roles in the upper ocean dynamics, so we must include the seasonal cycle in our subduction model. The most important parameter describing the subduction/ventilation process is the subduction rate. The instantaneous detrainment rate is defined as the volume flux of water per unit horizontal area (Cushman-Roisin, 1987)

( )/mb mbD w v h h t= − + ⋅∇ + ∂ ∂ (1)

where 0

mb e hw w vdz

−= − ∫ and mbv are the vertical and horizontal velocity at the base of the

mixed layer, and h is the mixed layer depth, Fig. 4. The first term on the right-hand side is contribution due to vertical pumping at the base of the mixed layer, which is slightly smaller than the Ekman pumping rate due to the geostrophic flow in the mixed layer. The second term is due to the lateral induction. The third term is due to the mixed layer detrainment.

Fig. 4. The definition of instantaneous detrainment rate. If there were no seasonal cycle, the subduction rate would be the same as the detrainment rate,

( )mb mbS D w v h= = − + ⋅∇ (2) Thus, this equation can be used for calculating the subduction rate, if there is no seasonal cycle; however, the strong seasonal cycle in the oceans make the subduction rate calculation much more complicated. Another parameter commonly used in the description of tracer ventilation is the so-called ventilation rate of an individual isopycnal or water mass, defined as

( )r

water mass VolumeVS

= (3)

where S is the subduction velocity defined above. Physically, the ventilation rate determines the average time (in units of years) it takes to renew the entire water mass through the ventilation process, or the average time water particles remain in a water mass category, for example, Jenkins (1987).

Page 5: Lecture 15. Subduction and obduction

5

In earlier models, the mixed layer thickness was set to zero for simplicity, so the only term contributing to subduction was the vertical pumping, which was the same as the Ekman pumping since the mixed layer thickness was zero. One of the pitfalls in the subduction process is this notion that the subduction velocity seems to be the same as the Ekman velocity. Actually, the mixed layer depth is non-zero and it varies with time and location, so each term on the right-hand side of (1) contributes differently. First, because the mixed layer has a finite thickness, the vertical velocity at the base of the mixed layer is slightly smaller than the Ekman pumping velocity. Second, the depth advection term actually contributes to subduction substantially. In the North Atlantic the wintertime mixed layer deepens northward from 100 meters to about 400 meters within 3000 kilometers, so we estimate that the mixed layer slope is about 0.0001. Given that the meridional velocity in the mixed layer is about 1 cm/sec, we conclude that the contribution from vertical pumping and that from the lateral induction are the same order. According to a calculation by Huang (1990), the contribution from the vertical pumping amounts to 12.1 Sv, and that from the lateral induction is about 12.7 Sv. When the time-dependent term is non-zero, the situation becomes much more complicated. First of all, there is the seasonal pycnocline between the mixed layer and the permanent pycnocline. The seasonal pycnocline plays the role of a buffer, i.e., the mass exchange between the mixed layer and the permanent pycnocline must go through the seasonal pycnocline. Thus, a complete picture must consist of four layers, as shown in Fig. 2. We will call the mass flux from the seasonal pycnocline to the permanent pycnocline subduction, while the mass exchange between the mixed layer and the seasonal pycnocline is called detrainment/entrainment. In general, the subduction rate is different from the detrainment rate because they represent different processes! Motions in the mixed layer have a broad spectrum in space and time; there are two prominent cycles in the mixed layer, i.e. the diurnal cycle and the annual cycle. For simplicity, we will discuss the seasonal cycle only. Since we have assumed that the flow in the permanent pycnocline is time independent, the subduction across the base of the seasonal pycnocline does not vary with time. However, water mass exchange across the base of the mixed layer has a prominent seasonal cycle. In earlier theory of the mixed layer, the seasonal cycle was separated into two phases, detrainment and entrainment. However, advancement of the subduction theory suggests that the seasonal cycle in the subtropical basin can be further divided into three phases (Cushman-Roisin, 1987). Between late winter and early fall, detrainment is activated due to the Ekman pumping and mixed layer shoaling. This period can be further divided into two sub-phases: From late winter to early spring, water entering the seasonal pycnocline from the mixed layer will eventually reach the permanent pycnocline; this process is called the effective detrainment. From early spring to early fall, water entering the seasonal pycnocline will be re-taken by rapid mixed layer deepening during the winter season, resulting in temporary (ineffective) detrainment (Fig. 5). From early fall to late winter, the mixed layer deepens rapidly -- the entrainment phase. It appears that the temporal and spatial inhomogeneity of motions in the mixed layer can create a fairly complicated detrainment/entrainment process, and comprehensive understanding of the intermittent and sporadic nature of detrainment/entrainment and its contribution to subduction is yet to come through observations and theoretical investigations. 4. Subduction rate defined as an integral property Were the mixed layer to overlay a stagnant ocean, the subduction rate would be a purely local property. In the oceans the mixed layer overrides currents in the seasonal/permanent pycnocline. As soon as water particles are left behind the mixed layer, they are carried downstream

Page 6: Lecture 15. Subduction and obduction

6

by the currents. There is no chance that water particles could be overtaken at the same location where they first left the mixed layer. This situation is very similar to a human breathing air. A person confined in a small box may breathe the same air again and again. A jogger can never inhale the same air he exhales. In the oceans, the mixed layer can only overtake the water pumped down from upstream, but not the water pumped down at the same location. Although a person sitting at a station may know the local rate of mixed layer entrainment/detrainment as a function of time, that person cannot be sure how much of this water really reaches the permanent pycnocline. To obtain the correct answer, one has to check at stations downstream because subduction is a non-local process. The subduction rate can be defined in different ways depending on the coordinates used. First, subduction rate can be defined in Lagrangian coordinates (Woods and Barkmann, 1986).

1 TL

L o

hS wdtT T

Δ= +∫ (4)

where T is the time of the average, taken as one year because of the seasonal cycle, and LhΔ means the mixed layer depth change accumulated over a one-year trajectory in Lagrangian coordinates. Thus, this definition includes both the temporal average and the spatial average over a one-year trajectory. A schematic picture in Fig. 5 illustrates this definition for a two-dimensional case. An instrument called a Bobber is released in late winter when effective detrainment starts at a station. This instrument can be checked continuously by acoustic signals. If we were following the instrument, we would see that during the first part of the trajectory effective detrainment takes place, i.e. the mixed layer retreats and leaves stratified water behind. During the second half of the trajectory mixed layer entrainment takes place and re-takes part of the water which entered the seasonal pycnocline (at an earlier time). Thus, the seasonal cycle can be divided into three phases, i.e., the effective detrainment phase during which water that left the mixed layer flows geostrophically into the seasonal pycnocline and enters the permanent pycnocline irreversibly, the ineffective detrainment phase during which water that entered the seasonal pycnocline will be re-taken later (at a downstream location), and the entrainment phase. Calculation of the subduction rate requires accurate information about the kinematic structure of the mixed layer and the velocity field in the pycnocline. Because such detailed information is very difficult to obtain from oceanic climatology, a simplified formula is

,1 ,0 ,1 ,0( ) ( )tr tr m mL

d d h hS

T− − −

= (5)

where ,0trd and ,1trd are the depth of the trajectory at the beginning and end of one year, ,0mh and

,1mh the depth of the mixed layer at the beginning and end of one year, T=1 year is the duration of the motion.

Page 7: Lecture 15. Subduction and obduction

7

Fig. 5. The annual mean subduction rate defined in the Lagrangian coordinates. The annual mean subduction rate can be defined also in Eulerian coordinates. In this case we are standing by a fixed station. In order to calculate the annual subduction rate, we will monitor the local mixed layer detrainment and entrainment. In addition, we have to follow the trajectories of particles released from our station to see whether these particles will eventually enter the permanent pycnocline or whether they will be overtaken by the mixed layer entrainment downstream (Fig.6). Similar to the case in Lagrangian coordinates, the seasonal cycle in the mixed layer can be divided into three phases, effective detrainment, ineffective detrainment, and entrainment (Fig. 6). The annual mean subduction rate is defined as

1 E

S

T

E TS Ddt

T= ∫

indicate the starting and ending times of the effective detrainment. In Fig. 5 we have assumed a simple case for the northern part of the subtropical basin where the mixed layer slope increases northward and the Ekman pumping velocity increases southward.

Fig. 6. Finding the critical trajectory that defines the end of the effective detrainment period by tracing trajectories released at a fixed station and regular time interval.

Page 8: Lecture 15. Subduction and obduction

8

The two additional terms are defined as

maxM mb mbdS w v hdy

= + (6)

F mb mbdS w v hdy

= + (7)

where h is the annual mean mixed layer depth. Note that both of these definitions treat the subduction in the local sense, so these two subduction rates do not include the average over the trajectory downstream. As a result, the rates calculated from these two equations are smaller than the rates calculated from the above defined ES and LS , both of which include the contribution resulting from the spatial variance of the Ekman pumping velocity and mixed layer depth, Fig. 7. Thus, one should not use a simple annual mean for calculating the so-called annual mean circulation; there is always some nonlinear effect, which must be investigated carefully.

Fig. 7. An example of the seasonal cycle of the detrainment/ entrainment and different annual mean subduction rates. 5. Obduction Upwelling/entrainment prevails in subpolar basins. Similar to what happens during the ineffective detrainment period, water entrained into the mixed layer may not actually come from the permanent pycnocline instead, it may come from the seasonal pycnocline whose water was detrained from the mixed layer previously (Woods, 1985; Cushman-Roisin, 1987). In order to clarify the physical processes involved in entrainment, we introduce the term, obduction. Obduction has been used in geology in describing the process of upward thrusting of a crystal plate over the margin of an adjacent plate. Here obduction is borrowed to describe the process in which water from the permanent pycnocline upwells into the mixed layer and flows over the adjacent layers of water (Fig. 8). Although obduction is basically a continuous process between the permanent pycnocline and the seasonal pycnocline, effective entrainment from the seasonal pycnocline to the mixed layer occurs only during part of the entrainment period. Water from the permanent pycnocline, which has not been exposed to the surface processes, is entrained into the mixed layer through the pycnocline. During the rest of the entrainment period (i.e. the ineffective entrainment period), water that has been exposed to air-sea interactions in the past year enters entering the mixed layer from the seasonal pycnocline.

Page 9: Lecture 15. Subduction and obduction

9

Fig. 8. An example of obduction, upwelling with a uniform upwelling of 18 m/year. The mixed layer depth has a simple sinusoidal cycle. Using a special term, obduction, helps us to clarify the irreversible mass flux from the permanent pycnocline to the mixed layer. For example, although mixed layer entrainment takes place in subtropical basins during a seasonal cycle, in most places water entrained into the mixed layer actually comes from the seasonal cycle, so there is no obduction. In fact, obduction takes place only within the subpolar basins and the subtropical-subpolar boundary regions, as will be shown through processing the climatological data. The obduction rate can be defined in a way similar to the subduction rate. Though obduction is an antonym of subduction, obduction is not simply subduction with an opposite sign. There are two major differences between these two terms. First, the physical processes involved in subduction and obduction are different. Subduction takes place in the subtropical basins, where water geostrophically flows down into the permanent pycnocline. As a result, water subducted to the permanent pycnocline carries late winter mixed layer properties. In comparison, obduction takes place in the subpolar basin, where water from the permanent pycnocline below flows geostrophically upward into the seasonal pycnocline and eventually enters the mixed layer. As water enters the mixed layer, it quickly loses its identity as a result of strong mixing, and it is impossible to keep track of the trajectory of an individual water parcel afterwards. The difference in the physics is reflected in the mathematical formulation of the suitable boundary value problems for the pycnocline structure in the subtropical and subpolar basins. The pycnocline equation is a nonlinear hyperbolic equation (Huang, 1988a, b). In the subtropical basin, density is specified as an upper boundary condition because the upper surface is the upstream boundary. In the subpolar basin, the upper surface density cannot be specified. In fact, the mixed layer density is determined by the dynamics of the permanent pycnocline and is a part of the solution. Second, the times when effective detrainment and effective entrainment take place are different. It is well known that effective detrainment takes place after late winter when the mixed layer reaches its annual maximum depth and density and starts to retreat. Effective entrainment takes place between late fall and early winter when the mixed layer deepens quickly, but before it reaches its annual maximum depth and density. In calculating the obduction rate, it is important to trace back to the origin of the entrained water. To make our presentation clearer, we will assume t=0 at late winter, say March 1. We will

Page 10: Lecture 15. Subduction and obduction

10

begin with a simple case where the mixed layer depth has a simple sinusoidal cycle, and its amplitude is spatially uniform. The upwelling rate is 18 /m year , and it is also uniform along the trajectory. Because the mixed layer is relatively shallow, we will assume that the vertical velocity is approximately the same as the Ekman upwelling rate everywhere along the trajectory. Below the base of the mixed layer mixing is negligible, so the identity of water parcels is preserved. As a result, the particle trajectory can be used to trace the origin of the water before it is entrained into the mixed layer. As soon as water parcels enter the mixed layer, they lose their identity because of the strong vertical mixing within the mixed layer, and it is impossible to continue tracking the paths of water particles. During spring and early summer the mixed layer retreats and leaves the stratified water behind, so this is the period of detrainment, although it is only a temporary detrainment in the present case. Beginning in early fall, the mixed layer deepens and entrainment takes place. The water entering the mixed layer during the first period does not really come from the permanent pycnocline. In fact, these water parcels were detrained into the seasonal pycnocline at an earlier time and at some upstream locations, as indicated by the top five lines in Fig. 8. Such water was contaminated by the mixed layer in the past year, so this is not real effective entrainment. It is only within the second phase of entrainment that water from the permanent pycnocline enters the mixed layer, as indicated by the lower trajectories. For the situation shown in Fig. 8 a simple calculation shows that for the case with a uniform upwelling of 10m/year, the effective entrainment starts at 0.8762ST = and ends at 1.0088ET = . On an ordinary calendar, the effective entrainment starts at January 18 and ends at March 4, and the duration is about 44 days. The annual obduction rate is O(10 m/year) , the same as the Ekman upwelling rate. The contrasts between subduction and obduction are the following: Table 1. Ventilation: Subduction vs. Obduction Subduction Obduction Mass flux From Mixed Layer to

Thermocline From Thermocline to Mixed layer

Water mass Formation Elimination Time Spring Winter Atmospheric forcing Heating Cooling Mixed layer Shoaling Deepening Trajectory tracking Downstream Upstream 6. Obduction rate defined as an integral quantity The obduction rate can be defined slightly differently, depending on the coordinates used. First, paralleling Woods (1985), the obduction rate can be defined in the Lagrangian sense. Accordingly, the annual mean obduction rate is

01 LL trT

hO w dtT T−

Δ= +∫ (8)

where T is the time of the average, taken as one year because of the seasonal cycle, the subscript tr indicates the critical trajectory in Fig. 8 (which marks the end of obduction), and LhΔ indicates the mixed layer depth change accumulated over one year's Lagrangian trajectory. Note that both terms include the temporal average over the past year and the spatial average over the one-year trajectory.

Page 11: Lecture 15. Subduction and obduction

11

Second, paralleling Cushman-Roisin (1987), an instant entrainment rate in Eulerian coordinates can be defined as

/mb mbE w u h h t= + ⋅∇ +∂ ∂ (9) where the subscript mb indicates the base of the mixed layer. Because the instantaneous entrainment rate fluctuates greatly during one seasonal cycle, it is not convenient to use it as an index for the water mass conversion rate; however, it is more meaningful to use the annual mean obduction rate defined as

1 E

S

T

E TO Edt

T= ∫ (10)

where ST and ET are the times when the effective entrainment started and ended, E is the instantaneous entrainment rate defined in (9). Again, a pitfall in calculating the annual mean obduction rate is using a simple Eulerian mean over the whole year

01E T

O EdtT −

= ∫ (11)

Substituting (9) would lead to a wrong estimation of E mb mbO w u h= + ⋅∇ . In general, this substitution tends to underestimate the annual mean obduction rate. The major differences between these two definitions of obduction rate follow. In Eulerian coordinates, water parcels entrained into the mixed layer at one station are monitored, and trajectories of parcels are traced back upstream for one year to determine whether the water comes from the permanent pycnocline or not. Accordingly, the beginning and the end of the effective entrainment are determined, and the annual obduction rate can be calculated by (10). In Lagrangian coordinates, water parcels entrained into the mixed layer at a station are monitored to determine the critical trajectory that marks the end of obduction. Given this trajectory, one can trace back along the trajectory for one year and calculate the obduction rate by (8). Calculating the annual mean subduction rate according to these two definitions requires accurate information on the spatial and temporal evolution of the mixed layer, and such detailed information is almost impossible to obtain from any climatic data. Thus, we propose a simple way of calculating the annual mean Lagrangian obduction rate by

, 1 ,0 , 1 ,0( ) ( )tr tr m mL

d d h hO

T− −− − −

= (12)

where d and h are the depth of the trajectory and the mixed layer base, subscript and 0 and -1 indicates the fact that these quantities are calculated at the beginning of the year and the previous location by upstream-tracing of the critical trajectory for one year. Although obduction rates calculated according to these definitions are slightly different, their difference is quite small compared with the errors existing in present-day climatology data. Therefore, our definition (12) can serve as a convenient and practical tool for calculating the subduction rate from climatological data with errors comparable to those of other processes. 7. Water mass formation rate balance Water mass formation and elimination in the oceans take place in the oceans through subduction and obduction processes discussed in this lecture. With the mixed layer and the Stommel demon these processes can be schematically viewed as in Fig. 9.

Page 12: Lecture 15. Subduction and obduction

12

Fig. 9. Define the subduction and obduction rate in the Lagrangian coordinate. In a steady state, the water mass volume between two density surfaces must be in balance, Fig. 10:

0Sub Ob DM LFM M M M− + + = (13) For a basin with open boundaries, the contribution of the last term should be balance by the mass flux through the mixed layer, i.e.,

LF MLM M=∫ ∫ ∫ ∫ (14)

The contribution due to the diapycnal mixing integrated over all the density range also vanishes because this term can transform water mass between different density categories. Thus, for a basin with open boundaries the total amount of subduction should equal to the total amount of obduction plus the mixed layer inflow from the side boundary

Sub Ob MLM M M= +∫ ∫ ∫ ∫ ∫ ∫ (15)

For a closed basin, there is no mixed layer inflow through the side boundaries; thus, the total amount of subduction and obduction should be exactly balanced:

Sub ObM M=∫ ∫ ∫ ∫ (16)

Equation (15) or (16) is one of the most important constraints for the water mass formation rates in a open (closed) basin, and we will show that this constraint is satisfied in our diagnostic calculation for the Pacific.

Page 13: Lecture 15. Subduction and obduction

13

Fig. 10. Water mass balance, where Sub indicates subduction, Ob for obduction, DM for diapycnal mixing flux, and LF for lateral flux through the side boundary. 8. Examples of subduction/obduction This is a natural place to examine the ventilation rate at the place where the Kuroshio (or the Gulf Stream) separates from the western boundary. A prominent maximum in the late winter mixed layer depth exists off the coast of Japan, where there is strong cooling by cold, dry polar air from the continent. For simplicity, we assume that the along-current-path distribution of the late winter mixed layer depth can be represented by the profile shown in Fig. 11a. Also assume the seasonal cycle of the mixed layer depth, f(t), has the typical saw-tooth profile shown in Fig. 11b (the thin line). The mixed layer pattern in this case can be expressed as

2( / )min min( , ) 100 (1 0.001 ) 100 ( )x x

mh x t h x e h f t− Δ⎡ ⎤= + × + + −⎣ ⎦ (17)

where min 40h m= m is the mixed layer depth minimum, and xΔ is the width of the Gaussian profile.

Page 14: Lecture 15. Subduction and obduction

14

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1−2000

−1000

0

1000

2000

3000

4000S

ubdu

ctio

n ra

te (

m/y

r)

Effective

Detrainment

Ineffective

Detrainment Ineffective

Entrainment

Effective

EntrainmentT1

Tede

Tees

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1−200

−150

−100

−50

0

Mix

ed la

yer

dept

h (m

)

Time (yr)

−2.5 −2 −1.5 −1 −0.5 0 0.5 1 1.5 2 2.5−200

−150

−100

−50

0

50

100

150

200

SL

SE

OE

OL

S−

X(1000km)

Rat

e (m

/yr)

Fig. 11. An example for subduction/obduction at the Kuroshio Extension. a) Mixed layer depth in late winter (solid curve) and in summer (dashed line). Trajectories of water particles are depicted by horizontal arrows. b) Instantaneous ventilation rate at station x=0 as a function of time; unit in m/year. The thin solid line indicated the seasonal cycle of the local mixed layer depth. c) Annual mean subduction (positive) and obduction (negative) rates: heavy solid lines indicate the Lagrangian rates, long-dashed line the Eulerian rates, and the short-dashed line is the rate calculated using the local winter mixed layer slope as defined in Eq. (18) Where the Kuroshio separates from the coast, water travels relatively quickly. When combined with the strong mixed layer depth gradient, it makes a large contribution to the ventilation rate. In comparison, the contribution due to Ekman pumping is negligible because this is close to the inter-gyre boundary. To illustrate our basic idea, we will neglect the vertical pumping and assume water parcels travel 1000 km in one year. Our focus is on the station at the center of the winter mixed layer trough, x=0. The instantaneous Eulerian subduction/obduction rate calculated by (1) is shown in Fig. 11b (the thick line). Note that the time axis t=0 corresponds to March 1.

Page 15: Lecture 15. Subduction and obduction

15

Effective detrainment begins at t=0 when the mixed layer starts to shoal rapidly. The end of the effective detrainment, 0.37ed

eT = , is calculated by tracing water particles released from this station at a time interval of 0.001tΔ = . The ineffective detrainment period is between ed

eT and 1 0.336T = , because the water subducted during this period is entrained into the downstream mixed layer at a later time. Entrainment starts at 1T . The beginning of the effective entrainment is calculated by tracing water particles upstream for one year. It occurs in this case at 0.716ee

sT = . Note that although detrainment takes place for one-third of a year, the effective detrainment is rather short, about 13 days. Similarly, entrainment takes place during two-thirds of a year, but only the last one-half of this period, about 103 days, is effective entrainment. At x=0 as shown in Fig. 11b, effective entrainment takes place in winter. Near the end of winter, the entrainment mode slows down and eventually switches to the detrainment mode. The transition between these two modes is so quick, water that entered the mixed layer remains close to the location where it leaves the permanent pycnocline and this same water is left behind and geostrophically flows into the permanent pycnocline thereafter. Using this example, we have shown that obduction and subduction can take place at the same location, although water masses involved come through effective detrainment and entrainment during different phases of the seasonal cycle. The spatial distribution of the annual mean subduction/obduction rates is shown in Fig. 11c, in which the subduction rate is defined as positive and the obduction rate is defined as negative. The solid lines denote the rates calculated in Lagrangian coordinates, whereas the long-dashed lines denote those calculated in Eulerian coordinates. Here, simplified versions of (12), i.e.,

,0 ,1( ) /L m mS h h T= − and ,0 , 1( ) /L m mO h h T−= − , have been used in computing the Lagrangian rates. Note that the subduction/obduction rates calculated using these two formulas agree well both in magnitude and in spatial distribution. The slight difference is because the two definitions use different weighting on the annual means. Four dynamically distinct regions appear in Fig. 11c. In the upstream direction, there is a region of pure obduction because the mixed layer depth increases monotonically downstream. In the middle of the mixed layer depth trough, there is an ambiductive region where both subduction and obduction take place and compensate for each other. There is a narrow band downstream where only subduction occurs. Farther downstream, an insulated region exists where there is neither effective subduction nor effective obduction. For comparison, the subduction/obduction rate calculated using just the annual mean velocity and winter mixed layer depth

intw erS u h= − ⋅∇ (18) is shown in Fig. 11c by the short-dashed line. Such a formula is similar to the one used by Marshall et al. (1993) in their data analysis on subduction in the North Atlantic and the one used in the analytical study of the subtropical North Pacific by Huang and Russell (1994). Although this definition is much easier to apply, it inflates the subduction/obduction rate near the mixed layer depth fronts. In addition, such a definition eliminates possible overlapping and exclusion of subduction and obduction because it neglects the seasonal cycle and the along-trajectory changes. Note that calculating the annual mean subduction/obduction rate in Eulerian coordinates requires detailed information of the seasonal cycle. On the other hand, the simplified formula (12) in Lagrangian coordinates requires the annual mean velocity and winter mixed layer properties only. Such a calculation can provide much more accurate information about the ventilation rates, although caution must be taken to set the rate to zero whenever it is negative.

Page 16: Lecture 15. Subduction and obduction

16

9. Ventilation of the North Atlantic and North Pacific Ventilation in the North Atlantic and North Pacific is examined by analyzing the Levitus (1982) climatological data and the Hellerman & Rosenstein (1983) wind stress data (Fig. 12a). The calculation is based on the geostrophic velocity, using the 2000-m level as the reference, and the one-year trajectories are shown in Fig. 12b. Note that during one-year period, water particles can travel over a great distance on the order to 1000 km; thus, using the local gradient of the mixed layer depth, as in Eq. (6), is inaccurate for the subduction/obduction calculation.

Fig. 12. a) Ekman pumping rate in the North Atlantic, in m/yr; b) one-year trajectories of water particles released from the base of the Mixed layer in March, with the along isopycnal geostrophic velocity calculated using the 2000-m depth as the reference level. (Qiu and Huang, 1995). Ventilation in the North Atlantic and North Pacific can be classified into four types of different regions: the subductive regions, the obductive regions, the ambiductive regions where both subduction (in spring) and obduction (in winter) take place, and the insulated regions where neither subduction nor obduction happens (Figure 13). Although the total subduction rates in these two oceans are comparable, the obduction rates are considerably different (Table 2). In the North Atlantic, obduction is strong (23.45 Sv), which is consistent with the notion of the fast thermohaline circulation and the short renewal time of the subpolar water masses in the Atlantic basin. Obduction is weak in the North Pacific (7.83Sv), consistent with the sluggish thermohaline circulation and the slower renewal process of the subpolar water masses there.

Page 17: Lecture 15. Subduction and obduction

17

Fig. 13 Annual mean ventilation rate for the North Atlantic. Subduction rate and its two components are shown in (a) the vertical pumping rate, (b) the lateral induction term, and (c) the subduction rate. Stippled regions in (c) indicate zero subduction rate. Obduction rate and its two components are shown in (d) the vertical pumping term, (e) the lateral induction term, and (f) the obduction rate. Stippled regions in (f) indicate zero obduction rate. The dotted line with crosses in (d) indicates the southern limit of the obduction zone. (Qiu and Huang, 1995). The most interesting features of these maps are the ambiductive regions in the oceans, where the local water mass conversion rate can reach 80 m/year . These local conversion rate maxima are very closely related to the heat flux maxima from the ocean to the atmosphere, as indicated by the dashed lines in Figs. 14.

Page 18: Lecture 15. Subduction and obduction

18

Fig. 14. Local mass conversion rate in the North Atlantic, defined within the ambiductive region, with contour interval of 40m/yr. The shaded areas are insulated regions where neither subduction nor obduction takes place. The dashed lines indicate the annual heat loss from the ocean to the atmosphere, in units of 2/w m , adapted from Hsiung (1985). (Qiu and Huang, 1995).

Fig. 15. Subduction/obduction rates per 0.2 θσ interval as a function of density for the North Atlantic (upper) and the North Pacific (lower). (Qiu and Huang, 1995).

Page 19: Lecture 15. Subduction and obduction

19

The water mass formation (destruction) rate computed as the sum of subduction (obduction) rate integrated over the corresponding outcropping area is plotted in Fig. 15. The peaks of the subduction rate correspond to the subtropical mode water in the North Atlantic and North Pacific. A second peak in the North Atlantic indicates the subpolar mode water, which has no corresponding part in the North Pacific because shallow marginal seas, such as Sea of Okhotsk, were not included in the calculation. Table 2. Ventilation rates for the North Atlantic and the North Pacific, in Sv: N. Atlantic N. Pacific Sum A) Subduction Ekman pumping -22.2 -30.8 -53.0 Vertical pumping 17.5 25.1 42.6 Lateral Induction 9.5 10.1 19.6 Total Subduction 27.0+3.1* 35.2 65.3 B) Obduction Ekman pumping 2.8 3.6 6.4 Vertical pumping -0.62 3.1 2.5 Lateral induction 24.1 4.7 28.8 Total Obduction 23.5 7.8 31.3 C)Local Conversion 4.0 3.5 7.5 * 3.1 Sv is the sum of the localized subduction south of Iceland (1.9 Sv) and in the Labrador Sea (1.2 Sv). One of the most interesting features from Table 2 is that Obduction rate is 10 times larger than the total Ekman upwelling rate. This indicates the important dynamical role of the sloping mixed layer depth. It is clear that using the term “Ekman upwelling” to describing water mass destruction in the ocean is not acceptable. Appendix A. On the Eulerian subduction/obduction rate Subduction rate in Eulerian coordinates can be defined in two ways: A) Subduction going through the upper surface of the permanent pycnocline at a given location (our equation (4.5)). This is the definition used by Marshall et al. (1993), Huang (1990), Huang and Russell (1994). Notice that there is nothing wrong about this definition. In fact, this definition gives more accurate information of how much water enters the permanent pycnocline at each location. In addition, the subduction rate calculated in this way is exactly the amount of subduction rate into a given density layer assuming that density does not vary in the permanent pycnocline. The shortcoming of this definition is that water subducted into permanent pycnocline at the given station does not necessary come from the local mixed layer detrainment. In fact, most of the water entering permanent pycnocline at the given station actually comes from upstream mixed layer detrainment. Therefore, such a subduction rate has no direct link with the local mixed layer detrainment. B) Subduction rate equal to the effective detrainment rate at a local station, this is the definition we have been used. This definition is defined as how much water is actually leaving the mixed layer and eventually enters the permanent pycnocline somewhere downstream. The advantage of this definition is the direct link with the local mixed layer dynamics. The disadvantage is that subduction rate defined in this way is not really a local property; instead, it is an integrated average

Page 20: Lecture 15. Subduction and obduction

20

over the downstream neighborhood. Furthermore, water subducted at one station is the water detrained during the effective detrainment period. Although the effective detrainment is relatively short, about one to three months, water properties, such as temperature, salinity, and density may very during this period. As a result, water subducted at one station does not belong to a single density category, i.e., subduction rate's distribution in density coordinates calculate by this definition is not very accurate! According to this argument, we may want to be a little careful, when we discuss the distribution of subduction/obduction rate in the density coordinate, such as our last figure. 1) Subduction/Obduction can happen at the same location in our definition. 2) They may or may not correspond to slightly different density. 3) If we look at what happen at the upper surface of the permanent pycnocline at a given station, the mass flux can have only one sign, either subduction or obduction, so Marshall et al. (1993) is right in this sense! 4) If we are only interested in what happen to the permanent pycnocline, subduction's distribution should be calculated in definition A). However, that Marshall et al. (1993) did not calculated the basin sum of subduction rate nor the sum in term of density. In summary, our last figure is not a very accurate account for the mass balance of the permanent pycnocline; however, it is not entirely wrong. First, we have been standing at the sea level, and we have been analyzing the mass flux leaving and coming into the mixed layer, which we defined as subduction/obduction for a station. This figure is what you would see if you are meteorologist because it is really the mass flux leaving the mixed layer. In this sense subduction/obduction can still happen at the same station, and at roughly the same density interval, so this is the way to explain the figure. In addition, subduction and obduction may still happen at the same density interval (AT DIFFERENT LOCATIONS!). There are so many different ways of doing the calculation, they should not conflict with each other; instead they should compromise and enhance each other! Of course, there is always the problem that the subduction/obduction period is not extremely short, so the density does vary slightly, make our calculation a little inaccurate, otherwise there is really nothing wrong. It is a matter of definition! Appendix B. Potential vorticity in the ventilated thermocline For simplicity, let us assume a two-dimensional case.

Fig. B1. A sketch for the potential vorticity formed during subduction. Using the density conservation, we have

Page 21: Lecture 15. Subduction and obduction

21

0

mf uQw u h

ρρ

⋅∇=

+ ⋅∇

Thus, potential vorticity in the ventilated thermocline is linearly proportional to the meridional density gradient of the mixed layer density, and inversely proportional to the sum of the vertical velocity at the base of the mixed layer plus the horizontal increment of the mixed layer depth. Therefore, low potential vorticity water is formed when: i) There is low meridional gradient of mixed layer density. ii) Strong Ekman pumping. References: Cushman-Roisin, B., 1987: Subduction. In: Dynamics of the oceanic surface mixed layer.

Proceedings Hawaiian Winter Workshop, 181-196. Hsiung, J., 1985. Estimates of global oceanic meridional heat transport. J. Phys. Oceanogr., 15,

1405-1413. Huang, R. X., 1988a. On boundary value problems of the ideal-fluid thermocline. J. Phys.

Oceanogr., 18, 619--641. Huang, R. X., 1988b. Ideal-fluid thermocline with weakly convective adjustment in a subpolar

basin. J. Phys. Oceanogr., 18, 642--651. Huang, R. X., 1990. On the three-dimensional structure of the wind-driven circulation in the North

Atlantic. Dyn. Atmos. Oceans, 15, 117--159. Huang, R. X., and S. Russell, 1994. Ventilation of the subtropical North Pacific. J. Phys. Oceanog.,

24(12), 2589--2605. Jenkins, W. J., 1987: 3H and 3He in the beta triangle: observations of gyre ventilation and oxygen

uilization rates. J. Phys. Oceanogr., 17, 763-783. Marshall, J. C., A. J. G. Nurser, and R. G. Willaims, 1993. Inferring the subduction rate and period

over the North Atlantic, J. of Phys. Oceanogr., 23, 1315-1329. Qiu, B. and R. X. Huang, 1995: Ventilation of the North Atlantic and Pacific: subduction and

obduction. J. Phys. Oceanogr., 25, 2374-2390. Woods, J. D., 1985: The physics of pycnocline ventilation. Coupled Ocean-Atmosphere Models. J.

C. Nihoul, Ed., Elsevier Sci. Pub., 543-590. Woods, J. D. and W. Barkmann, 1986: A Lagrangian mixed layer model of Atlantic 18oC water

formation, Nature, 319, 574-576.