a numerical study on the winter mixed layer on the shelf ... · snapshot of mixed layer depth 20...

A numerical study on the winter mixed layer on the shelf-slope region south of Japan

Hiroshi Kuroda*Takashi Setou* Yuichi Hirota*

Manabu Shimizu* Kazuhiro Aoki*

*: National Research Institute of Fisheries Science, Fisheries Research Agency, Japan

Email: kurocan@affrc.go.jp

Motivation of this study

Distribution of sardine eggs in February 1991(Watanabe et al., 1997)

+January to MarchSpawning season of sardine

+Main spawning groundsouthwest of Japan

Inshore side of the Kuroshio

+Oceanographic conditionDeveloped winter mixed layer

Kuroshio Axis

Missing Link

The aim of this study isto examine properties and dynamics of mixed layer variation

on the basis of high-resolution model

Climatological seasonal cycle of mixed layer depth ( )de Boyer Montégut et al. (2004, JGR)

(m)

oo 22 ×

m

Data Analysis of monthly temperature profileson the region southwest of Japan

R/V Kotaka-maru

Analysis DataStation : Stn.A in Tosa BayPeriod ：1991~2004Frequency ：monthly after 1995Instrument : CTD&STD&XBTVertical resolution ：1m

Definition of Mixed Layer Depth

The same criterion is used inde Boyer Montégut et al. (2004)

Stn.Awater depth: 650m C2.0 o=ΔT

Tosa Bay

Tem

pera

ture

(d

egC

)M

L D

epth

(m

)

ML TemperatureMonthly ML temperature & depthat Stn.A in Tosa Bay

●：monthly meanvertical bar：±1 S.D.

◎：monthly climatologyde Boyer Montégut et al. (2004, JGR)

Time (month)Time (month)

Net Heat Flux at the sea surface

Hea

t flu

x（W

/m2 ）

estimated via bulk formulation (COARE)using JRA25+GHRSST ML Depth

Time (month)

Tem

pera

ture

(d

egC

)M

L D

epth

(m

)

ML TemperatureMonthly ML temperature & depthat Stn.A in Tosa Bay

●：monthly meanvertical bar：±1 S.D.

◎：climatologyde Boyer Montégut et al. (2004, JGR)

Time (month)Time (month)

Net Heat Flux at the sea surface

Hea

t flu

x（W

/m2 ）

estimated via bulk formulation (COARE)using JRA25+GHRSST ML Depth

Time (month)

Within the ML, what can balance with the sea surface cooling

from Jan. to Mar.?

Tem

pera

ture

(d

egC

)M

L D

epth

(m

)

ML TemperatureMonthly ML temperature & depthat Stn.A in Tosa Bay

●：monthly meanvertical bar：±1 S.D.

◎：climatologyde Boyer Montégut et al. (2004, JGR)

Time (month)Time (month)

Net Heat Flux at the sea surface

Hea

t flu

x（W

/m2 ）

estimated via bulk formulation (COARE)using JRA25+GHRSST ML Depth

Time (month)

Can intra-monthly variations affect the monthly variation of

ML from Jan. to Mar.?

Forcings1. Climatological monthly

mean flux at the sea surface 2. Lateral boundary forcings

based on daily mean outputfrom each parent model

one-way nesting system

ROMSRegionalOceanModelingSystem

1/2 degreeBasin-scale(O(103) km)

1/10 degree Mesoscale

(O(102) km)

1/50 degreeSubmesoscale(O(101) km)

Dynamical downscaling system

Spinup Schedule of three models

1/2-degree model

1/10-degree model

9th

1 Jan.

1/50-degree model

15th

1 Jan.20th

1 Jan.

Initial valuewoa2001 Jan.

(momentum=0)

Initial valuefrom 1/10-degree

model output

Initial valuefrom 1/2-degree

model output

Total: 5-year simulationAnalysis: 2nd year to 5th year

External ForcingsSea surface flux : climatological monthly mean

Lateral Boundary : daily mean from parent model

Kuroshio axis positionmodel

observation1992-2003

50%

o

101

Kuroshio axis distance vs.Surface current in Tosa Bay

Kuroshio axis distance vs.Subsurface temperature

in Tosa Bay

o

501

o

501

model

obs.

model

obs.

Kuroda et al. (ready for submit)

Time (month) Time (month)

(m)

(deg

C)

Monthly mean time series of mixed layer at Stn.A in Tosa Bay

12-hourly time series of mixed layer at Stn.A in Tosa Bayfrom the 2nd to 5th year

ML

temperature

(deg.C)

ML

dep

th (m

)

1 Jan.1 Jan. 1 Jan.

gray area: September to March

1 Jan.

ML depth ML temperature

2nd 3rd 4th 5th

ML depthML temperature

data analysis

simulation

climatology

de Boyer Montégutet al. (2004)

12-hourly time series of mixed layer at Stn.A in Tosa Bayfrom the 2nd to 5th year

ML

temperature

(deg.C)

ML

dep

th (m

)

1 Jan.1 Jan. 1 Jan.

gray area: September to March

1 Jan.

ML depth ML temperature

2nd 3rd 4th 5th

9 10 11 12 1 2 3 4Time (month)

ML

dep

th (m

)

ML

temperature

(deg.C)

Snapshot of Mixed Layer Depth20 January in the 3rd year

9 10 11 12 1 2 3 4Time (month)

ML

dep

th (m

)

ML

temperature

(deg.C)

Stn.A

Kuroshio

Spatial features

Patch-like deeper MLStreak-like shallower MLalong the Kuroshio front

m

Daily snapshots ofSST (upper panels) and ML depth (lower panels)

20 Jan. 21 Jan. 22 Jan. 23 Jan.

m

Co

Correlation coefficient of mixed layer depth for Stn.A from January to March

dotted circlearound Stn.A

small: radius = 10kmlarge：radius = 20km

e-folding scale=10-20km

0.3

0.20.1

20km0

0.2

0.2>

scale

Heat balance of mixed layer temperature at a fixed station (Stn.A)

( ) ( )

( ) ( )( ) Tth

hT

hdzT

h

Th

Thz

qCt

T

h

hzz

Δ∂∂

−∇∇+⋅∇−

∇−∇+⎥⎥⎦

⎤

⎢⎢⎣

⎡

∂∂

=∂

∂

∫−

−==

111

111

HHH

0

zz0zzp

ML

κ

κκρ

u

shortwavepenetration

longwavesensible and

latent heat fluxes

advection entrainment

vertical diffusion

at ML bottom

horizontaldiffusion

Each term is diagnostically estimated from model output by a method of Kim et al. (2006, JPO)

Heat flux through the sea surface

dzTh

Th∫−

=0

ML1

Term estimation of the mixed-layer-temperature equation on the slope of Tosa BayPart 1: Monthly Mean

( ) ( ) ( ) ( )( ) Tth

hT

hdzT

hT

hT

hzq

CtT

hhzz Δ

∂∂

−∇∇+⋅∇−∇−∇+⎥⎥⎦

⎤

⎢⎢⎣

⎡

∂∂

=∂

∂∫−

−==

111111HHH

0

zz0zzp

ML κκκρ

u

2. Heat flux through

the sea surface

4.Advection

5.Horizontal

diffusion

1. Tendency

6.Entrainment

3Verticaldiffusion

Oct.

Nov.

Dec.

Jan.

Feb.

Mar.

( )610C −×so

( )610C −×so

( )610C −×so

( )610C −×so

( )610C −×so

( )610C −×so

MLTemp.

ML Depth

Term estimation of the mixed-layer-temperature equation on the slope of Tosa BayPart 2: 12-hour mean

( ) ( ) Tth

hdzT

hT

hzq

CtT

hz Δ

∂∂

−⋅∇−∇+⎥⎥⎦

⎤

⎢⎢⎣

⎡

∂∂

≈∂

∂∫−

=

1111 0

0zzp

ML uκρHeat flux through

the sea surfaceAdvectionTendency Entrainment

( )610C −×so

12-hour time series from January to March in the 3rd year

Jan. Feb. Mar.

The sign of tendency isreversed.

minus TendencySurface heat fluxAdvectionEntrainment

( )

),,(

1 0

wvu

dzTh h

=

⋅∇− ∫−

u

u

HL uuuu ′+′+=

HL TTTT ′+′+=

( )( )

HHLHH

HLLLL

HL

HLHL

'

'

TTT

TTT

TTT

TTTT

′′⋅∇+′⋅∇+′⋅∇+

′′⋅∇+′⋅∇+′⋅∇+

′⋅∇+′⋅∇+⋅∇=

′+′+′+′+⋅∇=⋅∇

uuu

uuu

uuu

uuuu

( )

( ) ( ) ( )

( ) ( ) ( )

( ) ( ) ( )∫∫∫

∫∫∫

∫∫∫

∫

−−−

−−−

−−−

−

′′⋅∇−′′⋅∇−′⋅∇−

′′⋅∇−′′⋅∇−′⋅∇−

′⋅∇−′⋅∇−⋅∇−=

⋅∇−

0

HH

0

LH

0

H

0

HL

0

LL

0

L

0

H

0

L

0

0

111

111

111

1

hhh

hhh

hhh

h

dzTh

dzTh

dzTh

dzTh

dzTh

dzTh

dzTh

dzTh

dzTh

dzTh

uuu

uuu

uuu

u

Decompositionof monthly mean advection

Original term

AverageLow Frequency(>30-day period)

High Frequency(Intra-monthly)

If h is constant, Orange Terms=0

( )

( ) ( ) ( )

( ) ( ) ( )

( ) ( ) ( )∫∫∫

∫∫∫

∫∫∫

∫

−−−

−−−

−−−

−

′′⋅∇−′′⋅∇−′⋅∇−

′′⋅∇−′′⋅∇−′⋅∇−

′⋅∇−′⋅∇−⋅∇−=

⋅∇−

0

HH

0

LH

0

H

0

HL

0

LL

0

L

0

H

0

L

0

0

111

111

111

1

hhh

hhh

hhh

h

dzTh

dzTh

dzTh

dzTh

dzTh

dzTh

dzTh

dzTh

dzTh

dzTh

uuu

uuu

uuu

u ( )( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )HHLHH

HLLLL

HL

,,,

,,,

,,,

,

TfTfTf

TfTfTf

TfTfTf

Tf

′′+′′+′+

′′+′′+′+

′+′+=

uuu

uuu

uuu

u

Contribution of decomposed components (February)

( )610C −×so

4-ye

ar m

ean

deco

mpo

sed

valu

eIn

Feb

ruar

y

( )Tf ,u

( )H,Tf ′u

( )L,Tf ′u

( )Tf ,Hu′

( )Hu Tf ′′ ,H

( )LH ,Tf ′′u

( )Tf ,Lu′

( )Hu Tf ′′ ,L

( )LL ,Tf ′′u

3rd-order upstream

2rd-order centered

Intra-monthly eddy heat advection

ConclusionsSeasonal cycle of ML depth and temperature exhibits two regimes

1. September to NovemberMonthly mean ML depth increases & ML temperature decreases

2. January to MarchMonthly mean ML depth and temperature become constant

For the monthly mean heat balance in the ML Heat flux through the sea surface + entrainment ~

advection

There is dominant intra-monthly submesoscale variability associated with the Kuroshio or its frontal disturbances

eddy heat advection

Intra-monthly variation contributes significantly to the monthly/seasonal variation of heat balance within the ML

Thank you for your attention