Analysis of the 18month variability in the Indian Ocean

1

• Analysis of the 18-month variability in the
  Indian Ocean
• based on historical data and proxy climate
  records

Irina V. Sakova, N.J.Abram, G.Meyers, R.Coleman,
and M.K.Gagan
4 August 2007 AOGS2007, Bangkok (Thailand)
UniTas Quantitative Marine Science PhD
Program CSIRO Wealth from Ocean Flagship
2

• Talk outline
• Background 18-month signal in the Indian Ocean
• 1. Interannual variability in the Indian Ocean
  Does the 18-month signal exist?, Sakova at
  al.(2006), Geophys. Res. Lett. (selected to AGU
  Highlights)
• 2. On the low frequency variability in the
  Indian Ocean, (In) Dynamic Planet, Sakova at al.
  (2006), IAG Symposia Series.
• New Analysis of historical data sets
• Main findings
• 18-month signal did exist in the past
• Periods of well developed 18-month signal
  coincide with the periods of IOD activity

3

• Data used
• Reconstructed Sea Level data (SL), J. Church
  (1950 2001)
• Kaplan Extended sea surface temperature (SST)
  (1856 present)
• Coral Oxygen Isotope Data, Bali (1782 1989)
• Coral Oxygen Isotope Data, Mentawai Islands
  (1858 1997)
• Satellite altimetry sea surface height (SSH)
  ERS/Envisat/TOPEX/Jason-1 (1992-2004)
• expendable bathy-thermograph (XBT) temperature
  data (1989-2002)
• Main analysis methods
• Fourier analysis

4
Spectral analysis of XBT and altimetry SSH data
in the Java upwelling region (7.5S, 105E)
during 1989 2002 years

• XBT temperature profile, where black
• line the depth of 20C isotherm (D20)
• Power spectrum of the depth of the
• D20 shows three strong maxima with
• periods of 6, 18.7, and 34 months

5
Spectral analysis of XBT and altimetry SSH data
in the Java upwelling region (7.5S, 105E)
during 1989 2004 years

• XBT temperature profile, where black
• line the depth of 20C isotherm (D20)
• Power spectrum of the depth of the
• D20 shows three strong maxima with
• periods of 6, 18.7, and 34 months
• SSH
• Power spectrum of SSH shows three
• strong maxima with periods of 6, 18,
• and 36 months

6
Spectral analysis of SSH in different locations
in theIndian Ocean

• There are five well
• separated frequency
• bands over the hole IO
• that carry most of the
• low frequency signal
• semi-annual
• annual
• 17-21 months
• 3 years
• 4-7 years

e
7
7 years (point N 3)
16.7 months (point N 11)
4.5 years (point N 4)
15.2 months (point N 12)
3.5 years (point N 5)
13 months (point N 14)
34 months (point N 6)
12 months (point N 15)
28 months (point N 7)
11 month (point N 16)
24 months (point N 8)
6.2 months (point N 28)
21 months (point N 9)
6 months (point n 29)
18.6 months (point N 10)
5.8 months (point N 30)
8
Why

• We consider the 18-20 month signal related to the
  IOD ?

9
Wavelet analysis of XBT and altimeter SSH data
(7.5S, 105E)

• Analysis of the depth of D20
• (1989 2002)
• 18 month signal exists between
• 1993 and 2000
• 34 month signal exists between
• 1991 and 2001
• Analysis of SSH
• (1992 2002)
• Clear existence of 18- and 34-
• month signals between 1993 and
• 2000

10
Spectral analysis of XBT data in the Java
upwelling region (7.5S, 105E) during 1989
2002 years

• a. Temperature profile. Black line the depth of
  20C isotherm (D20)
• b. Low-pass filtered of D20 with the cut-off
  frequency at 1/14 months
• c. 34-month (green line) and 18-month (red line)
  signals and the sum of these two signals (black
  line)

11
(No Transcript)
12
Hypothesis

• 18 month signal develops during IOD
• IOD is an oscillatory mode and appears as several
  events in the row
• These sequences of IOD are separated by
  substantial periods of time
• To check this we analysed available historical
  data sets, mostly in Java upwelling region

13
We propose the Indian Ocean Dipole activity
index
Calculation based on IOD classification by
Rao at al., (2002) Yamagata et al.,
(2004) Meyers et al., (2006)
the blue shading is positive and red negative
IOD
The graph represents 5 years running averaging
14
Kaplan SST at 7.5S, 102.5E, years 1948-2006
15
Kaplan SST at 7.5S, 102.5Eyears 1948-2006
16
Wavelet analysis of altimeter SSH data (7.7S,
104E)
17
Reconstructed sea level at 10S, 109E (J.Church
), years 1950-2002
18
Coral Oxygen Isotope Data, Bali (8S,
115E)years 1950-1965
19
Spatial variability of Reconstructed Sea Level
for individual frequencies (J.Church)
1950 - 1965
1965 - 1979
23 months
20 months
18 months
16 months
15 months
20
Spatial variability of Reconstructed Sea Level
for individual frequencies (J.Church)
1975 - 1989
1986 - 2000
23 months
20 months
18 months
16 months
15 months
21
Have we got the 18-month signal before 1950?
22
Coral Oxygen Isotope Data, Bali (8S,
115E)years 1782 1989
23
Coral Oxygen Isotope Data, Bali (8S, 115E)
24
Coral Oxygen Isotope Data, Mentawai Islands
(0.3S, 98.5E),
25
Kaplan Sea Surface Temperature at 7.5S, 102.5E
26

• Conclusions
• The 18-month signal is not a recent phenomena.
• Historical coral data, reconstructed Sea level
  and SST
• show 18-month signal.
• The 18-month signal achieves maximum amplitude
  in
• period of IOD activity, at least after 1950.
• The 18-month signal appears before 1950 but its
• connection with IOD Is not clear. This could be
  due to
• limitation in the proxy and extended SST
  records.

27
Acknowledgments

• The first author is supported by
• joint CSIRO-UTAS PhD scholarship in quantitative
  marine science (QMS)
• top-up CSIRO PhD stipend (funded from Wealth
  from Oceans National
• Research Flagship)
• Data
• Reconstructed Sea Level, Church and White (2006),
  GRL
• Kaplan SST V2 data provided by the NOAA/OAR/ESRL
  PSD, Boulder, Colorado, USA, from their Web site
  at (http//www.cdc.noaa.gov)
• Bali Coral Oxygen Isotope Data, Charles et al.,
  (2003), Marine Geology
• Mentawai Coral Oxygen Isotope Data, Abram et
  al., (in prep)
• SSH data (www.Jason.oceanobs.com/html/donnees/prod
  uits/msla_uk.html)
• XBT temperature data (www.marine.csiro.au/pigot/R
  EPORT/overview.html)
• Meyers and Pigot (1999)
• Wavelet software was provided by C. Torrence and
  G. Compo (http//paos.colorado.edu/research/wavele
  ts)