Dynamics of IO annual and interannual circulation Jay McCreary Alpine Summer School: Monsoon Systems Valsavarenche Val d Aosta, Italy June 8 13, 2009
References 1) (MKM93) McCreary, J.P., P.K. Kundu, and R. Molinari, 1993: A numerical investigation of dynamics, thermodynamics and mixedlayer processes in the Indian Ocean. Prog. Oceanogr., 31, 181 244. 2) (SM01) Schott, F., and J. P. McCreary, 2001: The monsoon circulation of the Indian Ocean. Prog. Oceanogr., 51, 1 123. 3) (SXM09) Schott, F.A., S.-P. Xie, and J.P. McCreary, 2009: Indian Ocean circulation and climate variability. Rev. Geophys., 47, RG1002, doi:10.1029/2007rg000245.
Wind forcing Seasonally reversing monsoon winds Reversing cross-equatorial winds Equatorial zonal winds In the SIO, the trades are relatively steady throughout the year. Through Ekman pumping, they raise the thermocline in a band from 5 10ºS, forming a ridge.
Annual sea-level response Courtesy of Jerome Vialard Courtesy of Jerome Vialard
Issues Does the Indian Ocean impact climate variability, either locally or remotely? What dynamic and thermodynamic processes impact biological activity in the Indian Ocean? A hierarchy of models has been used to study IO processes. The MKM model is a variable-temperature, 2½-layer model with an embedded mixed layer.
Introduction 1)Dynamical building blocks 2)Annual cycle 3)Interannual variability
1) Dynamical building blocks a) Interior ocean dynamics b) Coastal ocean dynamics c) Equatorial ocean dynamics
1½-layer model If a particular phenomenon is surface trapped, it is often useful to study it with a model that focuses on the surface flow. Such a model is the 1½-layer model. Its equations are where the pressure is The model allows water to transfer into and out of the layer by means of an across-interface velocity, w 1.
1a) Interior ocean dynamics
Response to a switched on zonal wind When β = 0 (f-plane), easterly winds force northward Ekman drift (northern hemisphere), which spins up two counterrotating, geostrophic gyres, a process known as Ekman pumping. When β 0, Rossby waves extend the response west of the wind region, and adjust the response to a state of Sverdrup balance.
Response to a switched on zonal wind The initial response is the same as on the f-plane. Subsequently, the radiation of Rossby waves adjusts the circulation to Sverdrup balance. β-plane f-plane
1b) Coastal ocean dynamics
Forcing by a band of alongshore wind τ y All the solutions discussed in this part of my talk are forced by a band of alongshore winds of the form, Since this wind field is x-independent, it has no curl. Therefore, the response is entirely driven at the coast by onshore/offshore Ekman drift. The time dependence is either switched-on or periodic Y(y)
Response to switched-on τ y In a 2-dimensional model (x, h), alongshore winds drive upwelling and coastal currents only locally, in the region of the wind. The offshore decay scale is the Rossby radius of deformation. 1½-layer model f-plane f-plane In a 3-d model (x, y, h) with β = 0, in addition to local upwelling by w e, coastal Kelvin waves extend the response north of the forcing region. The pycnocline tilts in the latitude band of the wind, creating a pressure force to balance the wind stress.
Response to switched-on τ y β-plane When β 0, Rossby waves carry the coastal response offshore, leaving behind a state of rest in which p y balances τ y everywhere. A fundamental question about eastern-boundary currents, then, is: Why do they exist at all?
1c) Equatorial ocean dynamics
Equatorial dynamics In response to forcing by a patch of easterly wind, Kelvin and Rossby waves radiate from the forcing region, reflect from basin boundaries, and eventually adjust the system to a state of Sverdrup balance.
Spin-up of an inviscid, baroclinic mode d (1 month) In response to forcing by a patch of easterly wind, Kelvin and Rossby waves radiate from the forcing region, reflect from basin boundaries, and eventually adjust the system to a state of Sverdrup balance. d (6 months) Rossby wave Kelvin wave If the wind oscillates at the annual (or semiannual) period, these adjustments continue indefinitely. Coastal Kelvin waves continuously radiate from the equator around the perimeter of the basin, followed by the propagation of Rossby waves into the basin interior. d (1 year) d (5 years) Sverdrup flow Equatorial jet Reflected Rossby-wave packet
2) Annual cycle a) Model hierarchy b) Surface circulation c) Heat budget
2a) Model hierarchy
1½-layer model If a particular phenomenon is surface trapped, it is often useful to study it with a model that focuses on the surface flow. Such a model is the 1½-layer model. Its equations are where the pressure is The model allows water to transfer into and out of the layer by means of an across-interface velocity, w 1.
2½-layer model If a phenomenon involves two layers of circulation in the upper ocean (e.g., a surface coastal current and its undercurrent), then a 2½-layer model may be useful. Its equations can be summarized as where i = 1,2 is a layer index, and the pressure gradients in each layer are Note that when water entrains into layer 1 (w 1 > 0), layer 2 loses the same amount of water, so that mass is conserved.
Variable-temperature, 2½-layer model If a phenomenon involves upwelling and downwelling by w 1, it is useful to allow temperature (density) to vary within each layer. Equations of motion of are where the terms ensure that heat and momentum are conserved when w 1 causes water parcels to transfer between layers.
Variable-temperature, 2½-layer model Because T i varies horizontally, the pressure gradient depends on z [i.e., p z = gρ ( p) z = g ρ], within each layer. So, the equations use the depth-averaged pressure gradients in each layer, where the density terms are given by
MKM 2½-layer model (with mixed layer) (a) (b) mixed layer T 1t + v 1 T 1 = κ h 2 T 1 w e θ(w e )(T 1 T 2 )/h 1 fossil seasonal layer thermocline w k w e T 1
2b) Surface circulation
January March
2c) Heat budget
Surface heat flux (July) The annual-mean surface heat flux into the IO is dominated by July contributions. So, the heat flux into the ocean is caused by oceanic upwelling. Advection then spreads cool SSTs away from the upwelling region, causing heating over a larger area.
Surface heat flux (annual mean) There is a net annual-mean heat flux into the Indian Ocean, that vanishes when cooling due to upwelling is dropped from the model. In this model, then, the annual-mean heating happens entirely because of upwelling. How model dependent is this result? Perhaps in this model it is overemphasized because heating in the 5 10 S band is too strong.
Cross-equatorial Ekman/Sverdrup flow The IO winds circulate clockwise (anticlockwise) about the equator during the summer (winter). The annual-mean winds have the summer pattern.
Cross-equatorial Ekman/Sverdrup flow EQ Ekman Transport Wind (boreal Winter) EQ Ekman Transport Wind (boreal Summer, annual mean) Ekman transport appears to be involved off the equator. But, what dynamics are involved near the equator?
Cross-equatorial Ekman/Sverdrup flow Consider forcing by τ x that is antisymmetric about the equator τ x = X x) Y ( y) = τ X ( x) y / ( 0 L The Sverdrup transport is V but V can be rewritten V 1 x = τ β τ 0X = βl y y τ 0 X / y = βl x 1 τ = β y x τ = f Thus, for this special wind the Sverdrup and Ekman transports are equal. It follows that the concept of Ekman flow can be extended to the equator, since τ x tends to zero as f does.
Consider the equations for a 1½-layer model, Then, Cross-equatorial Ekman/Sverdrup flow fv h t + g' h fu + g' h y + ( hu) x x = τ = τ x y + ( hv) / h, / h, y = 0. 2 x h + ( β / f ) g ' hh = ( τ / f ) = t ( x y w e For a τ x that is antisymmetric about the equator, w e τ 0 y = = 0, L y βy and so h never changes in response to this wind! So, no geostrophic currents are ever generated, and the total flow field is entirely Ekman drift.
3) Interannual variability a) Sea level variability b) ENSO warming c) IOD timing
3a) Sea-level variability
Sea-level variability (4 8 years) Courtesy of Irina Sakova
Sea-level variability (3 + 1.5 years) Courtesy of Irina Sakova
Sea-level variability (1.5 years) Courtesy of Irina Sakova
3b) ENSO warming
ENSO warming Figure 12: Correlations between Nino3 SST for Nov(0)-Jan(1) with SST in the eastern equatorial Pacific (160W 120W, 5S 5N; black), the tropical IO (40 100E, 20S 20N; red), the Southwest IO (50 70E, 15 5S; green), and the eastern equatorial IO (90 110E, 10S Eq.; blue). What processes cause the warming? What causes the delay? What are the impacts of these SST anomalies on the atmosphere?
ENSO warming January June August November The thermocline ridge is shallow throughout the year. One can expect that ENSO-related IO winds generate large SST anomalies there. Where else might large SST anomalies be expected?
ENSO warming Figure 13: Spatial structures of the first combined EOF during MAM after El Nino for a) upper ocean heat content (HC in ºC), b) SST (ºC), and c) wind stress (dyn/cm 2 ). The shading in c) indicates regions of wind stress magnitude larger than 0.1 dyn/cm2 (from Huang and Kinter, 2002). The change in heat content in the western IO suggests that the thermocline has deepened there and, hence, the importance of ocean dynamics. SST is also warm in the region. The structure of the wind anomalies suggests a linkage to SST.
ENSO warming r(z20,sst) There is a close connection between thermocline depth and SST in the western, tropical IO at interannual time scales. Figure 14: a) Annual-mean depth of the 20ºC isotherm (contours in m) and correlation of its interannual anomalies with local SST (color shades) (from Xie et al., 2002).
ENSO warming r(z20,sst) Precipitation 15 The thermocline dome provides a window for coupling ocean dynamics (thermocline depth) to SST and, hence, to atmospheric convection. Xie et al. (2002)
ENSO warming Xie et al. (2002) Figure 17: Correlation with eastern Pacific SST during Oct-Dec (months 10 12) as a function of x and t: Z20, SST, and rainfall averaged from 8 12ºS. Shading denotes where correlation exceeds 0.6 with Z20 in (a) and (b), and with SST in (c). The SWIO delay results from a downwelling Rossby wave generated in the southeastern IO, which deepens the pycnocline in the 5 10ºS ridge after its arrival there. As a result, SST warms there, and this warming increases rainfall.
ENSO warming Figure 15: Partial correlation of 1000 hpa winds (vectors) and wind curl (colors) with a) an IOD index and b) NINO3; only correlations at 99% level are shown (from Yu et al., 2005). ENSO is associated with positive wind curl. It forces a downwelling Rossby wave that propagates into the SWIO to impact the 5 10ºS ridge several months later. IOD ENSO
ENSO warming R Nino NIO Regressions of Nino SST during NDJ(0) on SST, wind and solar SWIO radiation ( precipitation) during May-Jun(1). The There SWIO are also convection precipitation induces a local anomalies cyclonic in the circulation. tropical WNP. It also forces As discussed a cross-equatorial next, they appear response to with be remotely northeasterlies generated throughout by the IO the warming NIO. They via the weaken radiation the of SWM, an causing atmospheric the NIO warm to Kelvin warm for wave. the second time. Du et al. (2009, J Climate)
ENSO warming 850-250 mb temp surface winds precip Xie et al. (2009, J Climate) A warm atmospheric Kelvin wave radiates from the TIO, and because of damping generates northeasterlies and surface divergence on its northern flank, which suppresses convection over the tropical NWP.
ENSO warming p, (u, v) The response is represented simply by the Gill atmospheric model, which models the response of a baroclinic mode of the atmosphere to a prescribed heating. The solution has a local response and radiates damped Kelvin and Rossby waves, which are damped by Newtonian cooling ( κp).
ENSO warming SLP & sfc wind The atmospheric model is dry, linearized about the NCEP mean state for JJA, with 20 vertical levels. It is forced by a prescribed diabatic heating that extends throughout the troposphere, to model deep convection. The figure shows the response when to a symmetric, an additional basin-scale diabatic heating over is imposed the TIO. in As the in the tropical Gill model, WNP a that Kelvin is proportional wave radiates to the into average the tropical wind WP. Northeasterlies convergence in the develop blue area. on its In northern this case, flank, a pronounced due to the anticyclonic background circulation winds and to develops, damping. consistent with the observations. Xie et al. (2009, J. Climate)
ENSO warming Summary Second warming JJA(1) Kelvin wave Rossby wave Ocean-atmosphere interaction in the SWIO is anchored by warming due to an oceanic Rossby wave. It is key to the persistence of the warming throughout the IO during JJA(1). The IO warming excites an atmospheric Kelvin wave, which causes low-level divergence in the tropical NWP, reducing rainfall and generating an anticyclone there.
3c) IOD timing
Figure 18: IOD pattern during Sept-Nov. a) Regression of Z20 (shading in m) and surface wind velocity (m/s) on the first principal component of Z20. b) Correlation of precip (shading) and SST (contours at 0.3, 0.6 and 0.9 and negative dashed) with the first principal component of Z20. From Saji et al. (2006a). The typical IOD develops from Sept Nov, with dipole anomalies in SST, Z20, and precipitation and anomalous equaotorial easterlies. These changes are all consistent with the occurrence of Bjerknes feedback. IOD timing
IOD timing Courtesy of Jerome Vialard Why does the IOD develop in the fall (Sept Nov)? One reason is because the normal annual cycle of SST in the EEIO is coolest (and Z20 is shallowest) at that time. Thus, the monsoon forcing creates a window in which IOD events can develop. Courtesy of Jerome Vialard
Summary
Dynamical building blocks The circulation in the northern IO is remarkably linear, because it is located in the tropics and the winds are so highly variable. The ocean responds differently to 1) coastal winds, 2) wind-stress curl in the interior ocean, and 3) equatorial winds. These forcings all generate baroclinic waves (coastal and equatorial Kelvin waves, Rossby waves) that impact the ocean remotely. Annual cycle Each of these forcing mechanisms contributes significantly to circulations in the northern Indian Ocean at some times and locations. The annual-mean, surface heat flux into the IO is due to upwelling of cool, subsurface water. Southward, Ekman/Sverdrup transport across the equator closes provides the sink to balance the heat input. Interannual variability ENSO warming is dynamically anchored along 5 10ºS by the deepening of the thermocline ridge there. This warming may feedback to impact the atmosphere in the western Pacific. The IOD is phase-locked to occur during Fall by the annual cycle.