Decadal-multidecadal Variability of El Niño-Southern Oscillation and its Impacts on the Global Climate
Part I: The Oceans

Carolina Fayos and Vikram M. Mehta

The Center for Research on the Changing Earth System
Columbia, MD
USA

Introduction

The Indo-Pacific Warm Pool (IPWP) is a topic of growing interest in climate research for its role in global climate variability and its societal impacts. The surface temperature, areal extent, and the upper-ocean heat content of the IPWP vary at decadal-multidecadal time scales (Mehta and Mehta, 2004). There is thermally direct variability of the Walker circulation and the equatorial Pacific thermocline associated with the IPWP variability (Mehta and Mehta, 2004).
Our objective is to investigate the possibility that decadal-multidecadal variability of the Walker circulation and thermocline modulate interannual ENSO variability. We address this question with in situ and model- assimilated ocean observations.
Data
  1. Simple Ocean Data Assimilation (SODA) system (Carton et al., (2000a,b))
    Time period: 52 years (1950-2001), monthly data
    Resolution: 1.0º x 0.5º in the tropics and 1.0º x 1.5º in midlatitudes
    Levels in the vertical: 20
  2. HadISST (provided by the Met Office Hadley Centre for Climate Prediction and Research, U.K)
    Time period: 133 years (1870-2002), monthly data
    Resolution: 1º x 1 º
  • Low-pass (= 8 years) filtered western Pacific Warm Pool (5º S-5º N, 145º E-155º E) SST anomaly time series from (a) HadISST, and (b) SODA; used as the reference time series in compositing ocean and atmosphere variables
  • Niño3 (5º S-5º N, 150º W-90º W) SST anomaly time series from
    (c) HadISST, and (d) SODA.
Analysis technique
  • The IPWP SST anomaly was low-pass filtered (=8 years) and used as the reference time series with Niño3 time series in compositing high-pass(=7 years) filtered ocean variables (temperature, currents, upper-ocean heat content, depth of the 20°C isotherm) and surface wind stress.
  • The composite analysis approach was taken to stratify interannual variability of El Niño and La Niña according to the phase of the decadal-multidecadal variability of the IPWP: IPWP warmer than average and IPWP less warm than average.
  • The composite analysis approach was also used to study evolutions of El Niño and La Niña events in the months preceding and following the maximum amplitudes of these events for the two phases of the IPWP variability.
  • La Niña analyses are not shown here since there were only two events in the IPWP warmer-than-average phase in the SODA period.

El Niño and La Niña events in high-pass (=7 years) filtered, annual-average SST data: HadISST 1870-2002 (Figure A) and SODA 1950-2001 (Figure B); composited according to decadal-multidecadal phase of the IPWP SST; differences in the tropical Pacific significant at the 0.05 level.
Figure A: More in number but less warm El Niño events when IPWP warmer at decadal-multidecadal timescales; fewer but warmer El Niño events when IPWP less warm at decadal-multidecadal timescales.
Fewer but colder La Niña events when IPWP warmer; many more but less cold La Niña events when IPWP less warm at decadal-multidecadal timescales.
Figure B: Same number but less warm El Niño events when IPWP warmer at decadal-multidecadal timescales. Warmer El Niño events when IPWP less warm at decadal-multidecadal timescales.
Fewer but colder La Niña events when IPWP warmer; many more but less cold La Niña events when IPWP less warm at decadal-multidecadal timescales
    Composites of monthly evolutions of SST (°C) and Wind Stress (N/m2) anomalies in El Niño events during the two phases of the decadal-multidecadal IPWP variability show that when the IPWP less warm than average
  • El Niño events 1.5°-2°C stronger
  • Growth of events begins more than six months before the maximum Niño3 SST
    Events dissipate slowly after reaching the maximum Niño3 SST
  • Greater meridional extent of anomalies, including along the west coast of the Americas
  • Associated midlatitude SST anomalies in central North and South Pacific, stronger circulation around the Aleutian low two months before the maximum Niño3 SST
Composite vertical structures of high-pass filtered, El Niño and La Niña temperature (ºC) anomalies, averaged between 5ºS and 5ºN in the Pacific from SODA in Dec-Jan-Feb
Sub-surface, positive anomalies in the eastern Pacific stronger and deeper, also extend west of 180 when the IPWP less warm than average; negative anomalies under the IPWP stronger
Significant differences in La Niña anomalies also, but inconclusive because of small sample size (2) when the IPWP warmer than average
Composite zonal and meridional (x 2) surface current anomalies (cm/s) during El Niño events from SODA in Dec-Jan-Feb
Zonal convergence near the eastern edge of the IPWP in response to surface wind stress convergence when the IPWP less warm
Slower STC (meridional convergence near the equator) in western and central Pacific in response to westerly surface stress anomalies when the IPWP less warm; this would reduce the poleward heat transport and provide a positive feedback to SST anomalies
Composite evolutions of monthly, equatorial (5ºS-5ºN) upper-ocean temperature (°C) anomalies in El Niño events during the two phases of the decadal-multidecadal IPWP variability show that when the IPWP less warm than average
Temperature anomalies travel eastward and upward from the central Pacific, and begin to strengthen more than six months before an event peaks
Positive anomalies followed eastward and upward by negative anomalies from the western Pacific after an event peaks
    Composites of monthly evolutions of SST (°C) and Wind Stress (N/m2) anomalies in El Niño events during the two phases of the decadal-multidecadal IPWP variability show that when the IPWP less warm than average
  • El Niño events 1.5°-2°C stronger
  • Growth of events begins more than six months before the maximum Niño3 SST
  • Events dissipate slowly after reaching the maximum Niño3 SST
  • Greater meridional extent of anomalies, including along the west coast of the Americas
  • Associated midlatitude SST anomalies in central North and South Pacific, stronger circulation around the Aleutian low two months before the maximum Niño3 SST
Modulation of interannual ENSO variability by the IPWP variability at decadal-multidecadal timescales

A hypothesis

  1. IPWP warmer than average at decadal-multidecadal timescales
    Increased warming in the western Pacific Warm Pool strengthens the Walker circulation, leading to a shallower thermocline in the eastern Pacific; the shallower thermocline conducive to more frequent ENSO events, but the stronger Walker circulation restricts coupled ocean-atmosphere interactions, resulting in weaker ENSO events

  2. IPWP less warm than average at decadal-multidecadal timescales
    Decreased warming in the western Pacific Warm Pool weakens the Walker circulation, leading to a deeper thermocline in the eastern Pacific; the deeper thermocline conducive to less frequent ENSO events, but the weaker Walker circulation allows coupled ocean-atmosphere interactions to grow, resulting in stronger ENSO events
Summary
  • Frequencies of interannual El Niño and La Niña events significantly different in two opposite phases of decadal-multidecadal IPWP variability
  • El Niño events stronger, spatially larger, and longer lived when the IPWP less warm than average at decadal-multidecadal timescales
  • Clear eastward and upward travel of interannual, upper-ocean temperature anomalies when the IPWP less warm than average
  • Surface stress anomalies in the tropical and midlatitude Pacific, associated with El Niño events, much stronger when the IPWP less warm than average at decadal-multidecadal timescales
  • IPWP variability at decadal-multidecadal timescales appears to modulate interannual attributes of El Niño and La Niña events via its influence on the Walker circulation and the equatorial Pacific thermocline
CRCES is a non-profit organization in Columbia, Maryland. The major research themes of CRCES are decade-to-century scale variability and change in Earth system components; societal impacts, especially on water availability, of such variability and change; development and use of Virtual Centers such as the Virtual Center for Decadal Climate Variability; and development of inputs in observing-system design for decade-to-century scale variability and change. CRCES endeavors to develop a "fusion" discipline from this research to guide the formulation of public policy regarding adaptation of societies to such variability and change. CRCES is currently funded by NASA-Oceanography Program and NOAA's Office of Global Programs. The Center for Research on the
Changing Earth System (CRCES)
10211 Wincopin Circle, Suite 240
Columbia, Maryland 21044-3405
www.crces.org
info@crces.org