upperlowerlayercouplinginloopcurrenteddies ekman… · ekman forms ekman separation franklin forms...

1.    Dynamics  of  the  Loop  Current  Experiment    

Upper-­‐lower  layer  coupling  in  Loop  Current  Eddies  Ekman  and  Franklin  Kathleen  A.  Donohue1,  Robert  Leben2,  Peter  Hamilton3  and  D.  Randolph  Wa=s1  1Graduate  School  of  Oceanography,  University  of  Rhode  Island,  Narraganse=,  kdonohue@gso.uri.edu  2University  of  Colorado,  Boulder  3SAIC,  Raleigh,  North  Carolina  

B1331  

1.    Dynamics  of  the  Loop  Current  Experiment    

Goals:  Increase  dynamical  understanding  of  Loop  Current,  eddy-­‐shedding   mechanisms,   and   genesis   of   lower-­‐layer  flows.  

Elements:    Moored  arrays  of  current  and  temperature  and  bo=om-­‐mounted   pressure   equipped   inverted   echo  sounders   (PIES)   together   with   remote-­‐sensing   and  numerical-­‐modeling  approaches.    

Deployment:  April  2009  RotaYon:  July  &  November    2010  Recovery:    November  2011  

Here  we  show  results  from  the  first  15  months:  May  2009  to  July  2010.    

Temperature   and   velocity   measurements   from   the  nine   tall   moorings   resolve   the   full   water   column  structure.  

Meanders   (~   7   day   period)   along   the   Loop   Current  periphery  precede  Loop  Current  Eddy  formaYon.  

Upper-­‐layer   Loop   Current   flows   are   not   visually  coherent  with  lower-­‐layer  flows.  

2.    CirculaCon  dominated  by  interacCon  between  the  Loop  Current  and  Eddies  Ekman  and  Franklin  

3.    Upper  and  Deep  StaCsCcs         Principal   axes   of   the   standard   deviaYon   ellipses  and  the  mean  vectors  are  not  aligned  with  those  of  the  upper  layer  (~200  m  depth).  

In   lower   layer,   mean   circulaYon   shows   an   west-­‐east  pa=ern  of  anYcyclone-­‐cyclone  pair.  

Deep   eddy   kineYc   energy   is   high   along   the  northeastern  periphery  of  the  mean  Loop  Current  posiYon.      

4.    Loop  Current  Eddies  Ekman  and  Franklin    

!"#$%"&''(

&'")*"+!,

--./ -0./ -1./

&2.3

&1.3

&0.3

,,"#$%"&''(

&'")*"+!,

,-"#$%"&''(

&'")*"+!,

&2"#$%"&''(

&'")*"+!,

,"4$5"&''(

&'")*"+!,

-"4$5"&''(

&'")*"+!,

,2"4$5"&''(

&'")*"+!,

&&"4$5"&''(

&'")*"+!,

&("4$5"&''(

&'")*"+!,

2"678"&''(

&'")*"+!,

!"#$%&#!'('

!'#)*#+!(

,,-. ,"-. ,/-.

!0-1

!/-1

!"-1

2#3$4#!'('

!'#)*#+!(

((#3$4#!'('

!'#)*#+!(

(,#3$4#!'('

!'#)*#+!(

!0#3$4#!'('

!'#)*#+!(

(#561#!'('

!'#)*#+!(

,#561#!'('

!'#)*#+!(

(0#561#!'('

!'#)*#+!(

!!#561#!'('

!'#)*#+!(

!7#561#!'('

!'#)*#+!(

Array   placed     where   historical   analysis  indicated   eddy   separaYon   was   most   likely   to  occur   and   designed   to   encompass   the   Loop  Current  from  east  to  west.  

Eddy  Ekman  

Aber    the  iniYal  detachment  meanders  develop  (~300  km  wavelength)  along  the  Loop  Current.  

IniYal  detachment  (June-­‐July)  occurs  when  a  trough  forms  in  the  southeast  edge   of   the   Loop   Current.     The   trough   expands   and   ulYmately   causes  detachment.  

SeparaYon  occurs  when  a  frontal  cyclone  along  the  northern  Loop  Current  strengthens  and  moves  southward.  

A   northward   shib   (meander   crest)   of   the   Loop   Current,   lower-­‐layer  thickness  decreases,  and  deep  anYcyclonic  vorYcity  develops  to  conserve  potenYal  vorYcity.  With  favorable  verYcal  offset,    a  lower-­‐layer  anYcyclone  leads  a  meander  crest,  upper-­‐layer  and  lower-­‐layer  eddies  grow.  

Joint  development  also  occurs  when  a  southward  shib   (meander   trough)  increases   lower-­‐layer   thickness   and   generates   a   deep   cyclone   that   leads  the  upper  trough.      

Eddy  Franklin  

Similar   to   Eddy   Ekman,   a   growing   and   southward   propagaYng  meander  along  the  west  Florida  slope  iniYates  the  first  detachment.  

Also   similar   to  Eddy  Ekman,  a   second  detachment  occurs  when  a   frontal  cyclone   along   the   northern   Loop   Current   strengthens   and   moves  northward.  

A  deep  anYcyclone  is  offset  from  an  upper  crest,  a  deep  cyclone  offset  from  upper  trough  and  deep  anYcyclone  offset  from  upper  crest.  

A   strong   deep   cyclone   with   swirl   speed   in   excess   of   20   cm/s   develops  leading  an  upper   trough,   the  pair   jointly   intensify,  move   south  and  Eddy  Franklin  detaches  from  the  Loop  Current.  

We  acknowledge  support  from  BOEMRE  contract  M08PC20043.  See  related  posters:  Rosburg,  K.,    Performance  evaluaYon  of  HYCOM  in  GOM,  B1333  Hamilton,  P.,  Loop  Current  Eddies  Ekman  and  Franklin,  A  historical  perspecYve,  B1322  

•   Preliminary  invesYgaYons  show  strong  interacYon  of  upper-­‐layer  flows  with  deep  currents.  •   Upper  and  lower-­‐layer  responses  during  Ekman  and  Franklin  detachments  are  organized  and  suggesYve  of  baroclinic  instability.  •   Loop  Current  meanders  are  an  important  mechanism  to  transfer  energy  to  the  lower  layer.  •   Four  Loop  Current  Eddies  formed  in  the  Gulf  of  Mexico  during  the  30-­‐month  observaYonal  Yme-­‐frame,  analysis  is  ongoing.  

18 MAY 2010

20 cm s!1

88°W 87°W 86°W

25°N

26°N

27°N

Pressure (dbar)!0.2 0 0.2

25 MAY 2010

20 cm s!1

88°W 87°W 86°W

Upper-­‐layer   and   lower-­‐layer   coupling   revealed   from  mapped   bo=om  pressure  (shaded  color  contours)  and  currents  (gray  vectors)  as  well  as  measured  deep  currents  (black  vectors).      

Mapped   round-­‐trip   travel   acousYc   travel   Yme   (thick   lines)     with   a  contour  interval  of  .002  s  comparable  to  ~9  cm  in  sea  surface  height.  

The   17   cm   contour   (black)   denotes   the   locaYon   of   the   Loop  Current  and  Loop  Current  eddy  fronts  in  the  alYmeter-­‐derived  sea  surface  height.  

!"!#$%!&''(

&')*++

&,)*++

-')*++

(').++ /'

).++

&/!#$%!&''(

(').++ /'

).++

!&!#$0!&''(

(').++ /'

).++

&1!#$0!&''(

(').++ /'

).++

'(!2$3!&''(

(').++ /'

).++

&-!2$3!&''(

(').++ /'

).++

'1!456!&''(

(').++ /'

).++

447+89:;!-' !!' !' -'

Eddy   Franklin   underwent     five   detachment/rea=achment  cycles  unYl  final  separaYon  in  late  September  2010.    

!"!#$%!!&'&

!&()**

!+()**

,&()**

-&(.** "&

(.**

'!!/01!!&'&

-&(.** "&

(.**

!2!/01!!&'&

-&(.** "&

(.**

&-!345!!&'&

-&(.** "&

(.**

!,!345!!&'&

-&(.** "&

(.**

&6!347!!&'&

-&(.** "&

(.**

!'!347!!&'&

-&(.** "&

(.**

889*:;<=!,& !'& '& ,&

A2

90oW 88

oW 86

oW 84

oW 82

oW 80

oW

20oN

22oN

24oN

26oN

28oN

30oN

Short MooringTall MooringPIES

0

200

400

600

depth

[m

]

Ekman

Forms

Ekman

SeparationFranklin Forms

tem

pera

ture

[oC

]

4

12

20

28

!40!20

02040

2000 m

velo

city [cm

s!1]

07/01/09 10/01/09 01/01/10 04/01/10 07/01/10!40!20

02040

2935 m

!1000

100

152 m

!1000

100

352 m

velo

city [cm

s!1]

A2  Tall  Mooring  

24oN

26oN

28oN

89oW 87

oW 85

oW

10 cms!1

mean Loop Current

eddy kinetic energy [cm 2 s!2

]

0 0.004 0.008 0.012 0.016

Deep    Upper    

24oN

26oN

28oN

89oW 87

oW 85

oW

10 cms!1

mean Loop Current

tau [s]1.12 1.122 1.124 1.126 1.128