estimation of the meridional ekman transport in the ... ·...

Geostrophic  velocity  was  calculated  from  the  density  field  measured  by  the  CTD/uCTD,  ageostrophic  velocity  is  the  difference  between  ADCP  and  geostrophic  velocity.  Assuming  that  in  the  upper  layer,  the  ageostrophic  flow  is  mainly  due  to  the  wind,  Ekman  transport  can  be  calculated  by  integraEng  the  ageostrophic  velocity  verEcally  and  zonally.  

Estimation of the Meridional Ekman Transport in the Tropical Atlantic

by Yao Fu, Johannes Karstensen, Peter Brandt Helmholtz Centre for Ocean Research Kiel Contact: yfu@geomar.de

Reference  

Chereskin,  T.,  and  D.  Roemmich  (1991),  A  Comparison  of  Measured  and  Wind-­‐derived  Ekman  Transport  at  11°N  in  the  AtlanEc  Ocean,  Journal  of  Physical  Oceanography,  21,  869  -­‐  878.  Chereskin,  T.,  W.  Wilson,  L.  Beal  (2002),  The  Ekman  temperature  and  salt  fluxes  at  8°30ʹ′N  in  the  Arabian  Sea  during  the  1995  southwest  monsoon,  Deep  Sea  Research,  49,  1211  -­‐  1230.  Wijffels,  S.,  E.  Firing,  and  H.  Bryden  (1994),  Direct  observaEons  of  the  Ekman  balance  at  10°N  in  the  Pacific,  Journal  of  Physical  Oceanography,  24,  1666  -­‐  1679.  Stewart,  R.  H.  (2008),  IntroducEon  To  Physical  Oceanography.  

IntroducEon  

Steady  wind  over  a  steady  homogeneous  ocean,  the  wind  stress  forcing  is  balanced  by  the  effect  of  Earth’s  rotaEon.  

The  verEcal  integraEon  result  in  a  transport  perpendicular  to  the  wind  blowing  direcEon.  

To  the  right  in  the  northern  hemisphere.      To  the  lee  in  the  southern  hemisphere.  

Ekman  transport  can  be  calculated  separately  in  two  approaches:  

Indirect:  using  wind  stress      

Direct:  using  ageostrophic  velocity      

In  the  tropical  AtlanEc  Ocean,  Ekman  transport  is  an  important  upper  layer  component  of  the  meridional  overturning  circulaEon.  Wind  induced  Ekman  divergence  drives  upwelling  and  poleward  surface  flow,  forming  the  upper  limb  of  the  subtropical  cell.  Generally,  in  the  absence  of  direct  current  measurement,  Ekman  transport  and  the  associated  heat  and  salt  flux  is  derived  from  wind  stress,  SST  and  SSS  based  on  the  satellite  observaEon.  In  this  study,  Ekman  transport  is  calculated  using  direct  velocity  observaEon  along  two  hydrographic  secEons  and  compared  with  the  indirect  approach.    

60oW 50oW 40oW 30oW 20oW 10oW 0o 10oE 20oS

15oS

10oS

5oS

0o

5oN

10oN

15oN

20oN

Longitude

Latit

ude

CTD stations

Data:  Hydrographic  SecEons  along  14.5  N  and  11  S  14.5N  :  Meteor  96  in  May  2013  64  CTD  staEons,  336  underway  CTD  profiles,  Vessel  mounted  ADCP,  Ship-­‐board  wind  measurement.  

11  S  :  Meteor  98  in  July  2013  47  CTD  staEons  near  the  coasts  243  underway  CTD  profiles,  Vessel  mounted  ADCP,  Ship-­‐board  wind  measurement.  

Underway  CTD  (uCTD)  data  are  applied  to  calculate  the  Ekman  transport  at  both  laEtudes.    Satellite  based  wind,  SST,  SSS  products  are  used  as  a  comparison.  

Direct  Method:  Ageostrophic  Velocity  

Mixed  Layer  Depth  and  Top  of  the  Thermocline  

−5 −4 −3 −2 −1 0 1 2 3

0

50

100

150

200

Velocity [cm/s]

Dept

h [d

bar]

Section−averaged velocity profiles

VadcpVgeoVageo

−3 −2 −1 0 1 2 3 4 5

0

50

100

150

200

Velocity [cm/s]

Depth

[dba

r]

Section−averaged velocity profiles

VadcpVgeoVageo14.5  N   11  S  

35 W 30 W 25 W 20 W 15 W 10 W 5 W 0 5 E 10 E

0

50

100

150Longitude

Dep

th [m

]

60W 55 W 50 W 45 W 40 W 35 W 30 W 25 W 20 W

0

50

100

150Longitude

Dep

th [m

]

MLD  depth     TTC  depth    mean  MLD  =  32  m   Mean  TTC  =  57  m  range  =  [17,      57]     range  =  [17,      90]    

mean  MLD  =  47  m  range  =  [15,      95]     range  =  [15,      95]    

Mean  TTC  =  45  m  14.5  N   11  S  

MLD  is  defined  by  a  density  threshold  of  0.03  kg/m3.  TTC  is  defined  by  a  density  gradient  threshold  of  0.01  kg/m4.  

Summary  and  Outlook  The  meridional  Ekman  transport  along  two  transatlanEc  secEons  is  esEmated  using  direct  and  indirect  method,  respecEvely.  The  underway  CTD  data  provide  consistent  results  compared  with  the  regular  CTD  data  in  esEmaEng  the  Ekman  transport.  At  both  laEtudes,  the  Ekman  transport  extended  beyond  the  mixed  layer.  In  the  direct  method,  the  Ekman  flux  is  sensiEve  to  the  choice  of  integral  depth,  the  top  of  the  thermocline  appears  to  be  a  reasonable  choice  for  the  integraEon  of  the  ageostrophic  velocity.  Though  in  these  two  cases,  the  Ekman  fluxes  using  the  SST  and  SSS  are  not  significantly  different  from  using  a  layer  of  temperature  and  salinity.  

In  the  next  step,  the  observed  Ekman  transport  and  fluxes  will  be  compared  with  GECCO/GECCO2  assimilaEon  products.    Eventually  the  study  will  be  extended  to  the  full  water  depth  and  different  components  related  to  the  meridional  overturning  circulaEon  will  be  esEmated  using  observaEonal  data,  and  the  variability  of  the  MOC  at  14.5  N  will  be  analyzed  with  model  data.  

Ekman  Heat  and  Salinity  Flux  

ME  :  Ekman    volume  transport  [Sv]  He      :  Ekman  heat  flux  [PW]  SE        :  Ekman  salinity  transport  [109  kg/s]  

θΕ        :  Transport  weighted  temperature  [C]  sE      :  Transport  weighted  salinity  [psu]  

!3

θΕ sE ME HE SE θΕ sE ME HE SE

TTC 25,777 36,203 5,731 0,606 0,213 25,671 36,694 -11,959 -1,259 -0,449

Const 25,618 36,171 6,342 0,666 0,235 26,051 36,836 -11,499 -1,228 -0,434

Surface 25,907 36,207 5,731 0,609 0,213 25,672 36,668 -11,959 -1,259 -0,449

Sat 25,893 36,266 5,731 0,608 0,213 25,768 36,390 -11,959 -1,263 0,446

Surface 25,731 36,160 6,264 0,661 0,232 25,419 36,694 -12,971 -1,352 -0,488

Sat 25,571 36,158 6,264 0,657 0,232 25,278 36,370 -12,996 -1,347 -0,484

14,5 N

Direct

11 S

Indirect

Ekman  Theory  

At  14.5  N,  The  calculaEon  based  on  the  uCTD  data  shows  good  consistency  with  the  calculaEon  based  on  the  CTD  data.  The  top  of  the  thermocline  appears  to  be  a  bener  choice  for  integraEng  the  ageostrophic  velocity  than  either  the  MLD  or  a  constant  depth.  

60 W 55 W 50 W 45 W 40 W 35 W 30 W 25 W 20 W−1

0

1

2

3

4

5

6

7

8

Longitude

ship windsatellite windageo 50 mageo MLDageo TTCageo uCTD TTC

35 W 30 W 25 W 20 W 15 W 10 W 5 W 0 5 E 10 E−14

−12

−10

−8

−6

−4

−2

0

2

Longitude

ship windageo 70mageo MLDageo TTC

14.5  N   11  S  

CumulaEve  Transport  [Sv]  

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011−25

−20

−15

−10

−5

0

Time [yr]

Tran

spor

t [Sv

]

Time series of Ekman Transport at 11S2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011−5

0

5

10

15

20

25

Time

Tran

spor

t

Ekman Transport from NCEP CFSR wind

Long-­‐term  Variability  of  the  Northward  Ekman  Transport    

14.5  N  

11  S  

The  wind  stress  data  from  NCEP  CFSr.  The  seasonal  cycle  dominates  the  variability.      At  14.5  N,  the  mean  Ekman  Transport  is  7.9  Sv  with  3.5  Sv  STD.    At  11  S,  the  mean  is  -­‐10.4  Sv  with  3.3  Sv  STD.