Pacific panel and friends

1
Pacific panel and friends
2
Synopsis

• Decadal predictability
• The 1990s shift
• Forced versus internal signals
• SPICE
• Western boundary currents
• SPCZ
• Analysis of hindcast experiments
• Miscellaneous

3
Science issues of Pacific panel

• Scale interactions in the ENSO system
• Guangzhou meeting 2007, review paper
• ENSO response to greenhouse warming
• Eric Guilyardis activities
• Origin and cure for model biases in eastern and
  western Pacific
• NOAA program to reduced model errors, VOCALS
• Predictability on decadal timescales (SST, SSH,
  Z20, winds )
• Decadal predictability working group in PP,
  will prepare web-site with decadal trend patterns
  for key climate variables
• Origin of Multidecadal variability and
  Atlantic-Pacific connections
• Honolulu workshop in 2006
• Western Pacific recharge-discharge processes
• Guangzhou meeting 2007
• Pacific Boundary currents, SPCZ and their
  influence on climate
• SPICE
• Origin of regional climate and sealevel trends
• Decadal predictability working group in PP

4
Scale interactions in the tropical Pacific
(multi) decadal Variability, trends
? research need
Adopted from Slingo
5
Decadal Predictability- internal variability-
externally forced signal
6
Sealevel prediction with simple Rossby wave model
SSH hindcast Skill of 1.5 layer Model forced With
wind History in North Pacific (Qiu, pers.
comm.)

• Decadal sea-level predictions with simple reduced
  gravity models, evaluation of SSH skill against
  CGCM results
• Initialization of decadal predictions Nudging of
  SST history versus data-assimilation?
• Internal versus externally forced variability,
  natural versus anthropogenic effects
• Application list of decadal predictions
  www.clivar.org/pacific

7
A window for decadal predictability in the
tropical Pacific
Sub-surface ENSO-driven off-equatorial decadal
variability, being a low-pass filtered version of
ENSO, is highly predictable
Decadal-long Perturbation Experiments
Power and Colman, 2006Climate Dynamics
13 years
8
Understanding recent trends towards decadal
prediction

• North Pacific climate shift around 1990
• Flattening of the equatorial thermocline (compare
  the post-1976 and pre-1976 conditions)
• More westerlies in the Eq. Pacific, and more
  easterlies over the equatorial IO
• Overall longterm weakening of Walker circulation

GFDL Ocean data assimilation
SLP data
Vecchi et al. 2006, Nature
9
Shift in the North Pacific Results from the GSOP
ocean reanalysis intercomparison

• The North Pacific shows a rapid shift in heat
  content in the late 80s early 90s.
• Consistently reproduced by all the reanalysis
  (observation-only and ocean-model only).

Courtesy of Balmaseda
10
Shift in the North Pacific SST and Heat Content

• It is the largest change in SST and HC in record.
• Reports of effects in fisheries and water mass
  properties
• In the ECMWF ORA-S3 analysis, the SST warming
  leads the HC.
• Tropical origin? How predictable is it?

Courtesy of Balmaseda
11
Subsurface trends on the equator
Levitus, since 1960
Probably due to wind changes, See Vecchi et al.
Nature 2006
Multi-model ensemble mean of 23 IPCC 4AR model
simulations for the 20th century, trend since
1950
Courtesy of Cai
12
20oC isotherm trend 1950-2000
SODA
Average over 23 AR4 models
Forced IPCC models reproduce observed 20th
century thermocline trends
Courtesy of Cai
13
Heat content trend 1950-2000
Upper 700m, observations
Upper 1000m, AR4
Courtesy of Cai
14
The climate of the next decade
Surface air-temperature difference C between
2015-2029 and 2000-2014, average over 25
CMIP3CGCMs and 4 scenarios (SRESA1B, SRESB1,
SRESA2, 1ptCO2)
Decadal prediction group of the CLIVAR-Pacific
panel
15
Observations
Theory
Ocean data assimilation
Coupled data assimilation
State-estimation Ensemble generation Evalu
ation Downscaling Application
CGCM
External forcing
Statistical model
Initial Perturbations, Stochastic optimals
Regional downscaling
Resource management
Forestry

Tourism
Fisheries
16
Discussion

• Climate shift in 1990s great test case for
  decadal predictions
• Regional effects of 1990s climate shift?
• How much ocean initial state is needed for
  decadal forecasts?
• How much of the decadal signal is externally
  forced
• Is the ocean state known well enough for decadal
  global forecast runs?

17
SPICE
18
SPICE a Southwest PacIfic Ocean Circulation and
Climate Experiment

• B. Qiu (U. Hawaii)
• D. Roemmich, J. Sprintall, (Scripps/UCSD)
• M. Williams (NIWA)

A. Ganachaud, S. Cravatte, L. Gourdeau
(LEGOS/IRD) B. Kessler (NOAA/PMEL) K. Ridgway,
W. Cai (CSIRO)
19
Why is this region interesting?
Complicated boundary current Structure, effect on
heat transport Regional (maybe global) climate
Region of large eddy Variability effects on Heat
transport?
rms SSH
20
SPICE scientific questions
21
SPICE field experiments
22
Modeling component of SPICE, Goals

• To improve the realism of the Southwest Pacific
  in large-scale models
• To help design optimal observations
• To simulate small-scale features of boundary
  currents, islands and straits
• To adapt and downscale global climate projections
  into results that are useful to island communities

23
South Pacific climate is poorly understood and
simulated
Observed cloudiness
Simulated NCAR cloudiness

• How does the SPCZ respond to tropical and
  extratropical SST on interannual to decadal
  timescales?
• What influence does the SPCZ wind convergence and
  its modulation have on southwest Pacific
  boundary currents?
• Precisely why is there a SPCZ?
• How does the SPCZ interact with the ITCZ, the
  MJO, SST, ENSO

24
Analysis of hindcast experiments
25
Analysis of hindcast experiments

• Several recent modeling studies (CGCM and HCM)
  have demonstrated that the optical properties of
  phytoplankton have an important influence on the
  tropical Pacific mean state and its variability
• Is this an important factor to consider for ocean
  hindcasting?

26
How important is biology for data-assimilation?
Hindcast with ¼ x ¼ SPFLAME NPDZ, Loeptien et
al. 08
27
Miscellaneous
28
Miscellaneous

• ENSO summer school on the Big Island, June 14th
  23rd, 2008 http//www.clivar.org/organization/paci
  fic/meetings/ENSOsummerschool.php
• Review paper ENSO stochasticity revisited, in
  preparation
• Decadal prediction working group
• SPICE endorsement expected soon
• ENSO Metrics white paper in preparation
• Exploring possibility to equip ARGO with pH
  sensors
• ENSO expert prediction system online
• Next PP meeting likely in March 2009, in
  Australia in combination with the Greenhouse
  conference in Perth

29
ENSO expert prediction

• Interannual timescales
• set-up of ENSO-expert prediction web site
  http//apdrc.soest.hawaii.edu/ENSO enso,
  expert1

30
ENSO expert prediction

• Interannual timescales
• set-up of ENSO-expert prediction web site
  http//apdrc.soest.hawaii.edu/ENSO enso,
  expert1

31
ENSO expert prediction
32
What does the PP need from GSOP

• Recommendations, advice for SPICE
  sustainability, use of hindcast products, flux
  products
• Joint evaluation of the effects of SPCZ
  variations on ocean transports, thermocline,
  watermasses
• Joint evaluation of the 1990 climate shift,
  origin, subsurface manifestation etc.