• Keine Ergebnisse gefunden

1)  Variability  in  the  tropical  Atlan7c:   2)  Reduced  gravity  simula7ons  of  the  tropical  Atlan7c  

N/A
N/A
Protected

Academic year: 2022

Aktie "1)  Variability  in  the  tropical  Atlan7c:   2)  Reduced  gravity  simula7ons  of  the  tropical  Atlan7c  "

Copied!
1
0
0

Wird geladen.... (Jetzt Volltext ansehen)

Volltext

(1)

54oW 36oW 18oW 0o 18oE 24oS

12oS 0o 12oN 24oN 36oN

Model domain

Damping of sponge layer [s1 ]

0.5 1 1.5 x 102 −5

Intraseasonal variability in the tropical Atlantic:

Observations vs. reduced gravity simulations

Robert  Kopte,  Peter  Brandt,  Richard  J.  Greatbatch,  Mar7n  Claus  

1)  Variability  in  the  tropical  Atlan7c:   2)  Reduced  gravity  simula7ons  of  the  tropical  Atlan7c  

3)  Comparison  of  reduced  gravity  simula7ons  with  AVISO  SLA:  

4b)  Equatorial  wave  analysis   4a)  Basin  mode  of  the  1

st

 baroclinic  mode  

5)  Summary  and  outlook  

0 50 100 150 200 250 300 350 400

0 1 2 3 4

Spatially averaged amplitude of SLA [cm]

Period [days]

0 50 100 150 200 250 300 350 4000

1 2 3 4 5

Depthaveraged amplitude of zonal vel. at 23°W [cms1 ]

SLA [42W−15E, 10S−10N]

Zonal velocity [23W]

Fig.  1:  Periodogram  of   SLA  in  the  tropical  

Atlan7c  and  zonal  

velocity  at  23°W,    0°N.  

120d  

182d  

365d  

▶  Spectral  peaks  at  annual  and   semi-­‐annual,  and  120-­‐day  

periods  associated  with  4th,   2nd  and  1st  baroclinic  modes  

▶  Peaks  correspond  to  resonant   basin  modes,  composed  of  

equatorial  Kelvin  and  Rossby   waves,  as  well  as  coastally  

trapped  waves[1]  

Fig.  2:  Model  domain  with   sponge  layers  at  northern/

southern  boundaries,  and   area/sec7ons  of  interest.  

▶  To  study  the  intra-­‐seasonal  variability  in  par7cular,  reduced   gravity  model  (RGM)  simula7ons  are  used:  

 

     

▶  The  model  is  run  separately  for  the  first  five  baroclinic  modes   (c1=2.47  m/s,  c2=1.32  m/s,  c3=0.94  m/s,  c4=0.74  m/s,  c5=0.57  m/s),  forced  with  

interannually  varying  wind  stress  from  NCEP  (1990-­‐2014)  

▶  To  allow  for  comparison  of  model  and  observa7ons,  the   model  output  is  ficed  to  AVISO  sea  level  anomaly  (SLA)[2]  

ut f v = gn0 x+sx

0 , vt+f u = gn0 y+sy

0 , t+H(ux+vy) = 0, gn0 = c2n H

0.25°  x  0.25°  

Aviso

40°W 20°W 0° 1997

2000 2003 2006 2009 2012

RGM M1−5

40°W 20°W 0°

−20[cm]

−15

−10

−5 0 5 10 15 20

Aviso J

F M A M J J A S O N D

RGM M1−5

[cm]−8

−6

−4

−2 0 2 4 6 8

Fig.  3:  Hovmöller  plot  of   SLA  along  equatorial  wave   guide  (green  line  in  Fig.  

2).  Le#:  AVISO,  right:  

Reconstruc7on  from  RGM   modes  1-­‐5.  

Fig.  4:  Climatologic  

(1995-­‐  2014)  seasonal   cycle  of  SLA  along  

equatorial  wave  guide.  

Fig.  5:  Climatologic   (1995-­‐2014)  seasonal   cycle  of  band-­‐passed   (25-­‐130d)  SLA  along  

equatorial  wave  guide.    

▶  Dominance  of  the  annual  and  semi-­‐annual  cycle,  well  reproduced   by  the  RGM  (Fig.  3  and  Fig.4)  

▶  In  AVISO,  presence  of  con7nuous  and  recurrent  eastward  

propaga7ons[3],  with  the  intra-­‐seasonal  climatology  represen7ng  

~25%  of  the  seasonal  cycle  amplitude  (Fig.  5)  

RGM M1 120−day 10°S

5°S Eq.

5°N 10°N

50°W 40°W 30°W 20°W 10°W 0° 10°E RGM M1−5

120−day 10°S

5°S Eq.

5°N 10°N [cm]

0 1 2 3 4

AVISO 120−day 10°S

5°S Eq.

5°N 10°N

AVISO 120−day

10°S 5°S Eq.

5°N 10°N

RGM M1−5 120−day

10°S 5°S Eq.

5°N 10°N

Nov/Mar/Jul Dec/Apr/Aug Jan/May/Sep Feb/Jun/Oct

RGM M1 120−day

50°W 40°W 30°W 20°W 10°W 0° 10°E 10°S 5°S Eq.

5°N 10°N

Fig.  6:  Maps  of  amplitude  (le#)  and  phase  (right)  of  120-­‐

day  harmonics  ficed  to  SLA  data.  Top:  Aviso,  middle:  

RGM  modes  1-­‐5,  bo0om:  RGM  mode  1  only.  

AVISO Frequency [days1 ]

Wavenumber [10−3 km−1]

−1.5 −1 −0.5 0 0.5 1 1.5

0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045

0.05 −10 −20 20 10

Wavelength [degrees longitude]

RGM

Period [days]

Wavenumber [10−3 km−1]

−1.5 −1 −0.5 0 0.5 1 1.5

365 100 50 30

Log

10 of spectral density [cm2/cpd/km−1]

6 7 8 9 10

−10 −20 20 10

Wavelength [degrees longitude]

Fig.  7:  Mean  (5°S-­‐5°N,  see  horiz.  black  lines  in  Fig.  2)  wave-­‐

number-­‐frequency  diagram.  Le#:  Aviso,  right:  RGM  (modes  1-­‐5).  

The  white  lines  represent  the  theore7cal  dispersion  curves  of   the  Kelvin  wave  and  the  first  three  meridional  Rossby  waves.  

▶  Basin-­‐wide  structure  of  SLA  

associated  with  the  1st  baroclinic   mode  with  variability  detectable   as  far  as  10°S  

▶  Consistent  pacern  in  the  RGM,  

although  with  considerable  lower   amplitude  

▶  Mode  1  in  the  RGM  simula7ons   explains  most  of  the  structure   seen  in  the  “full”  model  

▶  Averaged  over  the  equatorial  belt  (5°S-­‐5°N),  zonal   wavenumber-­‐frequency  diagrams  of  SLA  exhibit  

spectral  peaks  near  the  theore7cal  dispersion  curves   of  the  first  baroclinic  mode  equatorial  Kelvin  and  

Rossby  waves  

▶  In  the  RGM,  total  energy  is  lower,  however  there  is   considerable  energy  in  the  mixed  Rossby-­‐gravity  

wave  range  

RGM M1−5

EAST −−4N−−> WEST −−EQ−−> EAST −−4N−−>WEST

Month

0E 20W 40W40W 20W 0E 0E 20W 40W

Jan Apr Jul Oct

AVISO

Month

Jan Apr Jul Oct

[cm]−4

−3

−2

−1 0 1 2 3 4

Fig.  8:  Climatologic  (1995-­‐2014),  band-­‐passed   (25-­‐130d)  SLA  along  4°N-­‐Equ.-­‐4°N  (see  Fig.  2).  

Top:  Aviso,  bo0om:  RGM  modes  1-­‐5.  

▶  Only  ~50%  of  the  intra-­‐seasonal  signal   amplitude  is  reproduced  by  the  RGM,  

however  the  phase-­‐lock  of  the  propaga7ons   appears  to  be  consistent  

▶  Intra-­‐seasonal  SLA  variability  in  the  tropical  

Atlan7c  is  essen7ally  wind-­‐driven,  as  it  can  be   reproduced  by  reduced  gravity  simula7ons,  

although  with  weaker  amplitudes  

▶  Possible  reasons  for  discrepancies  to  be  tested:  

▶  Bad  choice  /  spa7al  variability  of  phase   speeds,  which  leads  to  the  missing  of     resonance  to  a  periodic  forcing  

▶  Uncaptured  (i.e.  non-­‐linear)  effects  of  the   North  Equatorial  Counter  Current  (NECC)  

on  westwards  propaga7ng  Rossby  waves  at  

~4°N  (Fig.  8)    

References  

[1]  Cane  MA,  Moore  DW  (1981):  A  Note  on  Low-­‐Frequency  Equatorial  Basin  Modes.  J  Phys  Oceanogr  11:1578-­‐1584  doi:10.1175/1520-­‐0485  

[2]  The  al7meter  products  were  produced  by  Ssalto/Duacs  and  distributed  by  Aviso,  with  support  from  Cnes    (hcp://www.aviso.al7metry.fr/duacs/)  

[3]  Polo  I,  Lazar  A,  Rodriguez-­‐Fonseca  B,  Arnault  S  (2008):  Oceanic  Kelvin  waves  and  tropical  Atlan7c  intraseasonal  variability:  1.  Kelvin  wave  characteriza7on.  J  Geophys  Res-­‐Oceans  113:18  doi:10.1029/2007jc004495  

Aviso

40J°W 20°W 0° F

M A M J J A S O N D

RGM M1−5

−40 −20 0 [cm]−2

−1.5

−1

−0.5 0 0.5 1 1.5 2

Referenzen

ÄHNLICHE DOKUMENTE

Abstract. An existing one-dimensional chemical model has been extended with dynamics and isotopic chemistry to simulate chemical production, vertical ascent, and diffusion

More detailed information on the stratifi- cation during the experimentwas given by Peters (1978). 16 day average vertical profiles of temperature T, salinity S,

The wavelet method hence is useful to verificate that the regions of highest predictability are associated with i) the states during the most prominent El Niño and La Niña events,

 Low oxygen zones are created just below the mixed-layer, in the euphotic zone of high productive anticyclonic modewater eddies (oxygen at 42 and 170m, salinity, meridional

Helmholtz Centre for Ocean Research Kiel Contact: yfu@geomar.de. Reference

Adding all natu- ral and anthropogenic factors, we estimate a total modelled warming of 0.5 K decade −1 around the TTL (Table 3), which is less than the observed 0.9 K decade −1

After the first three profiles the ship’s acoustic instruments (H YDROSWEEP , P ARASOUND , ADCPs) were switched off as suggested by the manufacturer due to possible

The associated spatial correlation pattern between the decadal cell strength index (blue curve in Figure 1) and the observed SSTs closely resembles that shown in Figure 4a in