www.ecco-group.org/abstracts/woce_lee.txt

Interannual Variation of Mixed-layer Heat Balance:
contrasting different climate phenemena and the roles of linear
versus nonlinear effects

Tong Lee, Ichiro Fukumori, Benyang Tang, and Dimitris Menemenlis

Variation of mixed-layer heat balance are studied using an ECCO
assimilation product (http://www.ecco-group.org).  The analysis
focuses on the similarity and difference associated with El
Nino-Southern Oscillation (ENSO), Indian-Ocean Dipole (OID), and
Pacific Decadal Oscillation (PDO), and the relative importance of
linear versus nonlinear advective tendencies. In the eastern and central
equatorial Pacific, the warming and cooling are caused by the combined  
effect of oceanic advection and diffusion.  Surface heat flux opposes   
the change. The opposite is true for the western equatorial Pacific.    
The relative contribution of different components of oceanic advection  
is spatially dependent.  Zonal advection is dominant in the central
equatorial Pacific where zonal temperature gradient associated with
the warm-pool edge is large.  Vertical advection is more important in
the eastern equatorial Pacific where upwelling and vertical
temperature gradient are large.  Diffusion has a smaller magnitude
than advection, but is comparable to the temporal change of
mixed-layer temperature (MLT).  Advection of anomalous temperature
gradient by mean flow and advection of mean temperature gradient by
anomalous flow have comparable magnitude, both being in-phase with the
change of MLT. Nonlinear tendency (advection of anomalous temperature 
gradient by anomalous flow) is sizable and counteracts the two linear 
tendencies above.  Different from the tropical Pacific, nonlinear effect
in mid-latitude Pacific is very small; the advection of mean temperature
gradient by anomalous flow has the largest contribution to the total
advection of heat.  The role of total oceanic advection in MLT balance is
similar for ENSO, IOD, and PDO. However, that of surface heat flux is
different.

1	Interannual Variation of Mixed-layer Heat Balance:
2	contrasting different climate phenemena and the roles of linear
3	versus nonlinear effects
4
5	Tong Lee, Ichiro Fukumori, Benyang Tang, and Dimitris Menemenlis
6
7	Variation of mixed-layer heat balance are studied using an ECCO
8	assimilation product (http://www.ecco-group.org). The analysis
9	focuses on the similarity and difference associated with El
10	Nino-Southern Oscillation (ENSO), Indian-Ocean Dipole (OID), and
11	Pacific Decadal Oscillation (PDO), and the relative importance of
12	linear versus nonlinear advective tendencies. In the eastern and central
13	equatorial Pacific, the warming and cooling are caused by the combined
14	effect of oceanic advection and diffusion. Surface heat flux opposes
15	the change. The opposite is true for the western equatorial Pacific.
16	The relative contribution of different components of oceanic advection
17	is spatially dependent. Zonal advection is dominant in the central
18	equatorial Pacific where zonal temperature gradient associated with
19	the warm-pool edge is large. Vertical advection is more important in
20	the eastern equatorial Pacific where upwelling and vertical
21	temperature gradient are large. Diffusion has a smaller magnitude
22	than advection, but is comparable to the temporal change of
23	mixed-layer temperature (MLT). Advection of anomalous temperature
24	gradient by mean flow and advection of mean temperature gradient by
25	anomalous flow have comparable magnitude, both being in-phase with the
26	change of MLT. Nonlinear tendency (advection of anomalous temperature
27	gradient by anomalous flow) is sizable and counteracts the two linear
28	tendencies above. Different from the tropical Pacific, nonlinear effect
29	in mid-latitude Pacific is very small; the advection of mean temperature
30	gradient by anomalous flow has the largest contribution to the total
31	advection of heat. The role of total oceanic advection in MLT balance is
32	similar for ENSO, IOD, and PDO. However, that of surface heat flux is
33	different.
34