Title: Modeling Pacific Decadal Variability:
1Modeling Pacific Decadal Variability Physics,
Feedbacks and Ecosystem Impacts Arthur J.
Miller Scripps Institution of Oceanography Univers
ity of California, San Diego La Jolla, CA
USA U.S. GLOBEC Scientific Steering
Committee Meeting Martin Johnson House SIO, La
Jolla May 7, 2004
2- Modeling Pacific Decadal Variability
- Physics, Feedbacks and Ecosystem Impacts
- Outline
- Physics that organizes the patterns of Pacific
ocean decadal variability - Some interesting new results concerning
mechanisms - Relations to our current research on ecosystem
response - Recent Collaborators
- Physics Schneider, Di Lorenzo, Pierce, Kim,
Bograd, - Alexander, Capotondi, Deser, Lynn, McWilliams,
Mestas-Nunez - Biology Moisan, McGowan, Neilson, Chai, Chiba,
Gabric - Funding NSF, NASA, NOAA, DOE, ONR
3Schematic of Pacific Oceanic Response to Decadal
Forcing by the Aleutian Low
Canonical SST Pattern
Rossby waves
2 - 5 yrs Lagged KOE SST Pattern
(Miller and Schneider, 2000, Prog. Oceanogr.)
4Canonical Pattern of Decadal SST Response
(Aleutian Low Strengthening)
Schematic
SST
Warming
SST Cooling
Driven by surface atmospheric forcing
Canonical Pattern of Decadal SST Response
Equator
Tropical SST Warming
From Miller, Chai, Chiba, Moisan and Neilson
(2004, J Oceanogr.)
5Lagged Pattern of Decadal SST Response
(Aleutian Low Strengthening)
Schematic
sCooling
SST Cooling
Driven by thermocline changes
via wind-stress curl
Lagged Pattern of Decadal SST Response
Equator
From Miller, Chai, Chiba, Moisan and Neilson
(2004, J Oceanogr.)
6Basin-Scale Pattern of Decadal Thermocline
Response (Aleutian Low
Strengtherning)
Schematic
Thermocline
Deepening
Thermocline Shallowing
Lagged response in west due to
Rossby wave propagation
Basin-Scale Pattern of Decadal Thermocline
Response
Equator
Tropical Thermocline Deepening
From Miller, Chai, Chiba, Moisan and Neilson
(2004, J Oceanogr.)
7Sources of North Pacific Decadal Variability
- Tropical Teleconnections (requires tropical
decadal mechanism) - a. Atmospheric (ENSO-like)
- - canonical SST pattern
- - basin-scale thermocline
response - b. Oceanic (ENSO-like)
- - eastern boundary thermocline
response
8Simple ENSO-forced PDO Model (Newman et al.,
2003, J Climate)
PDO Index ENSO Index SST Persistence Noise
50 variance
But observed spectrum has more decadal power
Model spectra
ENSO index
9 Reconstruction of PDO Index based on time
integration of several index time series
low-freq Total
85 75 ENSO
36 20 North Pacific 37
30 Index KOE-1 7
25 KOE-2 nil nil
(Schneider, 2004, in prep)
10Tropics are correlated to North Pacific on
decadal timescales suggesting common forcing or
tropical origin of decadal signal
N Pac
Indian SST
Pacific SST
SPCZ1 SPCZ2
Cloud
Deser et al., J Climate in
press.
Indo-Pac SLP
11Spectral content and coherence between N Pacific
Index (zonal wind) and Tropical Index (composite)
Strong peaks and strong coherency in ENSO and
decadal bands
Deser et al., J. Climate, in press
12Sources of North Pacific Decadal Variability
- Tropical Teleconnections (requires tropical
decadal mechanism) - a. Atmospheric (ENSO-like)
- - canonical SST pattern
- - basin-scale thermocline
response - b. Oceanic (ENSO-like)
- - eastern boundary thermocline
response - Subduction Modes
- Midlatitude Gyre Modes
13Schematic of the Gu-Philander class of decadal
mode
Schematic of the Latif-Barnett class of decadal
mode
Miller et al., 2003, Bull. Am. Meteorol. Soc.
14Response of ENSO to upwelling spiciness anomalies
(Schneider, J Climate, 2004)
15Schematic of the Gu-Philander class of decadal
mode
Schematic of the Latif-Barnett class of decadal
mode
Miller et al., 2003, Bull. Am. Meteorol. Soc.
16Predicting Observed SST in the Kuroshio-Oyashio Ex
tension (KOE) Region
Basin-wide wind stress curl drives Rossby wave
model
Rossby waves change the upwelling and currents
(Qiu, 2003) in the KOE during winter
Quantified forecast skill up to 3 years in advance
Schneider and Miller, 2001, J Climate
17Hindcast skill of simple Rossby wave
model predicting winter SST
KOE Pattern is important in its feedback to
the atmosphere and potential in predicting
ecosystem response in KOE
Schneider and Miller, 2001
J. Climate
18Relation between SST and surface heat fluxes in
KOE reveals ocean forcing atmosphere, much like
in tropics
Schneider, Miller and Pierce 2002, J Climate
19KOE SST is correlated with local regional
rainfall over the ocean in this coupled
model, and probably for obs.
Understanding how the downstream atmosphere
behaves in response to this is vital to
deterimining if decadal gyre mode feedback loops
exist.
20Schematic of Qiu (2004) Rossby wave timescale
selection
Rossby wave wavelength depends on latitude
High
Sea Level
Line source Of stocahstic wind Stress curl
KOE
Zonal Flow
Sea Level
Low
For realistic parameters, strongest north-south
gradient of sea level occurs for T 10 years
21Sources of North Pacific Decadal Variability
- Tropical Teleconnections (requires tropical
decadal mechanism) - a. Atmospheric (ENSO-like)
- - canonical SST pattern
- - basin-scale thermocline
response - b. Oceanic (ENSO-like)
- - eastern boundary thermocline
response - Subduction Modes
- Midlatitude Gyre Modes
- Stochastic Forcing
- - oceanic spectral peaks possible
- - predictable components possible
- Deterministic Forcing
- - solar cycles, greenhouse gases
22Summary of Some Regional Ecosystem Impacts
Organized by Pacific Decadal Variability
GOA
Steller Sea Lions Zooplankton Phytoplankton
KOE
Sardines Zooplankton Phytoplankton
CCS
Sardines Zooplankton Phytoplankton
Adapted from Yasuda et al., 1999, Fish. Oceanogr.
23Ecosystem response processes in KOE on long
timescales
Strengthened Aleutian Low
Miller, Chai, Chiba, Moisan and Neilson, J.
Oceanogr., 2004
24Physical-Biological Hindcast of Pacific Ocean
Decadal Variability
First EOF of Combined Thermocline, Phyto-, and
Zooplankton fields
Miller, Chai, Chiba, Moisan, and Neilson, J.
Oceanogr., 2004
25Physical-Biological Hindcast of Pacific Ocean
Decadal Variability
First EOF of Combined SST, Phyto-, and
Zooplankton fields
Miller, Chai, Chiba, Moisan, and Neilson, J.
Oceanogr., 2004
26CalCOFI Observations along the Southern
California Coast
Over 50 yrs 1 deg C warming of SST 70
decline in macro zooplankton
Roemmich and McGowan, 1995
27The Ocean Model Configuration
Regional Ocean Modeling System (ROMS) Primitive
Equations Horizontal Resolution 9 km Vertical
Resolution generalized sigma coordinate with 20
levels Open Boundaries North, West and
South Bathymetry ETOPO5 Sandwell and Smith
(TOPEX)
N
S. Francisco
California
E
W
m
Los Angeles
Latitude N
S
S. Diego
North Pacific Ocean
Longitude W
28Local Atmospheric and Remote Oceanic Forcings
That Can Affect the Regional Oceanic Heat Budget
Mean Advection
Anomalous Advection
Southern California
Alongshore Wind Stress
29An eddy-permitting ocean model hindcast captures
the observed SST and thermocline variations
Interannual and Decadal physical variations
CalCOFI Observed Thermocline
Thermocline depth over the last 50 yrs shows an
overall deepening of 20 m
10
0
m
-10
Model Thermocline
Warming is due to large-scale decadal
surface heat fluxes combined with southward
advection of concomitantly warmed water
Increase in upwelling favorable winds partially
cools water column
Di Lorenzo, Miller, Schneider and McWilliams,
JPO, in press.
30 Modeling Observed California Current System
Eddies and the Ecosystem Response
(0)
(1)
(2)
Satellite Observations
Model Hindcast
In Situ Observations
AVVISO TOPEX/ERS
CalCOFI 98-02
SSH
SSH
TOPEX
E2
E2
m
m
E1
E1
E1
SSH
Independent Validation
Data Assimilation
Independent Validation
?M N/m3
?M N/m3
SeaWiFS
E2
E2
Chl-a
Chl-a
E3
E3
E3
E1
E1
Chl-a
Di Lorenzo, Miller, Cornuelle, Neilson and Moisan
(IJRS, 2004)
31The Climate-Ocean Regime Shift Hypothesis of the
Steller Sea Lion Decline Relating temporal
variability in the physical system to ecosystem
changes Collaborators E. Di Lorenzo, D.
Neilson, H.-J. Kim (SIO) S. Bograd, F. Schwing,
R. Mendelssohn (PFEL) A. Capotondi, M. Alexander
(CDC) K. Hedstrom, D. Musgrave (UAF) A. Trites
(UBC)
32 Steller Sea Lion Population Distributions
NMFS
33SSL population declines since 1976-77 Climate
ShiftWestern Gulf of Alaska population
droppedEastern Gulf of Alaska population was
stable
NMFS
34Basic Issues in SSL Decline
- Temporal change
- - Decline after mid-1970s
- Spatial asymmetry
- - Decline in western Gulf of Alaska
- - Stable populations in eastern Gulf
35(Trites, 2004)
36Of course, there are some other
less appealing theories.. Overfishing of
Favorite ForageDiseasePollution Shot by
Fishermen.Increased Predation
37Observed Changes in Ekman pumping
Pre-shift mean conditions 1960-75 Dec-May
Change after shift (1977-97) -(1960-75)
Capotondi, Alexander, Deser and Miller
(JPO, sub judice)
38Climate shift in wind-stress curl
Theoretical solution (steady state) to
wind-stress curl forcing yields a weakening
of the Alaskan Stream but this assumes
Rossby waves equilibrate the western bouindary
currents. Cummins and Lagerloef (2003) show that
Rossby waves are not important in
open-ocean interannual/decadal variability. (Mode
ls suggests topographic Rossby waves along
shelf-slope may be important in establishing WBC
response.)
Theoretical streamfunction response
39Coarse Resolution Model Pycnocline (26.4??) depth
changesPeriod2 (1977-97) Period1 (1964-75)
Capotondi, Alexander, Deser, and Miller (JPO, sub
judice)
Strengthened Alaskan Stream!
Model vs. Obs PAPA
Model vs. Obs GAK
1970
1980
1990
2000
1980
1970
40Eddy Permitting Primitive Equation Model Hindcast
Regional Ocean Modeling System
(ROMS) 1950-1999 NCEP Winds 16km
resolution Relaxation to climatological SST,
BCs, SSS
Topography
Miller. Di Lorenzo, et al. (GRL, sub judice)
41Eddy-Permitting Model Eddy Surface Currents
Before 76-77 Shift
After 76-77 Shift
Difference
Kodiak Is.
- Stronger Stream north of Kodiak - Weaker Stream
southwards
Miller, Di Lorenzo, et al. (GRL, sub judice)
42Eddy-Permitting Model Eddy Surface Currents
Before 76-77 Shift
After 76-77 Shift
Difference
Kodiak Is.
- More eddies north of Kodiak - Fewer eddies
southwards
Miller, Di Lorenzo, et al. (GRL, sub judice)
43Trites, 2004
44Schematic of the Gu-Philander class of decadal
mode with DMS aerosols and phytoplankton heat
absorbtion effects
Schematic of the Latif-Barnett class of decadal
mode with DMS aerosols and phytoplankton heat
absorbtion effects
Miller et al., 2003, Bull. Am. Meteorol. Soc.
45Directions.
Atmosphere - Details of atmosphere response over
KOE region - Sensitivity to ocean biology DMS
aerosols - Regional downscaling over mountains
and coasts
Ocean - Physical mechanisms of adjustment to
forcing - Lags and predictable compoments -
Changes in eddy statistics - Sensitivity to
ocean biology phytoplankton absorbtion
Biology - Organization of response by ocean
patterns - Lags and predictable components -
Distinguishing forced from intrinsic variations
and Global Change effects on all these.
46Effects of anthropogenic forcing on biological
activity
Biological Model Phytoplankton mmol C/m3
Ratio, Year 2100 / Year 2000
Pierce, Climate Change, 2004
47Thanks!