Decadal Variability and Projection of Future Climate Change Thomas L. Delworth GFDLNOAA August 5, 20 - PowerPoint PPT Presentation

Loading...

PPT – Decadal Variability and Projection of Future Climate Change Thomas L. Delworth GFDLNOAA August 5, 20 PowerPoint presentation | free to download - id: 9eade-NjFjN



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

Decadal Variability and Projection of Future Climate Change Thomas L. Delworth GFDLNOAA August 5, 20

Description:

Decadal Variability and Projection of Future Climate Change Thomas L. Delworth GFDLNOAA August 5, 20 – PowerPoint PPT presentation

Number of Views:73
Avg rating:3.0/5.0
Slides: 50
Provided by: mar211
Learn more at: http://www.cgd.ucar.edu
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: Decadal Variability and Projection of Future Climate Change Thomas L. Delworth GFDLNOAA August 5, 20


1
Decadal Variability and Projection of Future
Climate Change Thomas L. Delworth GFDL/NOAA Augu
st 5, 2009 NCAR Summer ASP Colloquium
  • Outline
  • Relevance and types of decadal variability in the
    climate system
  • Observed decadal variability (emphasis on
    Atlantic)
  • Impact of Atlantic decadal variability
  • Mechanisms of Atlantic decadal variability
  • Predictability and predictions on decadal time
    scales

2
Figure 10.4
3
Global Mean Surface Air Temperature (Source
Climatic Research Unit, Univ of East Anglia)
"If youre 29, there has been no global warming
for your entire adult life. If youre graduating
high school, there has been no global warming
since you entered first grade. There has been no
global warming this century. None. Admittedly the
21st century is only one century out of the many
centuries of planetary existence, but it happens
to be the one youre stuck living in. Mark
Steyn, National Review online, July 4, 2009,as
quoted by syndicated columnist George Will on
July 23, 2009 in the Washington Post
4
Observed North Atlantic Sea Surface Temperature
(Annual Mean)
5
Projected Atlantic SST Change (relative to
1991-2004 mean)
Can we predict which trajectory the real climate
system will follow?
Results from GFDL CM2.1 Global Climate Model
6
Projected changes in annual water budget over SW
North America using the SRES A1B scenario.
American West aridity will likely intensify and
persist due to future greenhouse warming if
model results are correct, this drying is likely
to have already begun.
average of 19 models participating in IPCC AR4.
Red line Precipitation minus Evaporation
averaged over 19 models used in IPCC AR4.
BluePrecipitation GreenEvaporation.
Seager et al. (2007)
7
Substantial decadal to centennial scale
variability of drought in the western U.S.
Cook et al. (2004)
8
Components of decadal variability and
predictability
  • Forced climate change
  • Predictability arising from estimates of future
    changes in radiative forcing agents, and the
    climate system response to those changes.
  • Committed warming from past radiative forcing
  • Internal variability
  • Decadal-scale fluctuations are an important part
    of climate variability
  • Is there meaningful decadal-scale
    predictability in the climate system?
  • Can we realize that predictability?
  • Is the natural variability altered by radiative
    forcing?

9
Figure 10.4
10
Components of decadal variability and
predictability
  • Forced climate change
  • Predictability arising from estimates of future
    changes in radiative forcing agents, and the
    climate system response to those changes.
  • Committed warming from past radiative forcing
  • Internal variability
  • Decadal-scale fluctuations are an important part
    of climate variability
  • Is there meaningful decadal-scale
    predictability in the climate system?
  • Can we realize that predictability?
  • Is the natural variability altered by radiative
    forcing?

11
These results suggest that it is crucial that
we (a) Better document the characteristics of
observed decadal variability (instrumental and
paleo records, and diagnostic analysis) (b)
Better understand the mechanisms and relevance of
decadal variability (theory and models) (c)
Assess whether such fluctuations are
predictable (modeling and observing
systems) (d) If warranted, attempt such
predictions in the context of future climate
change projections (initialize models with the
observed climate state for future climate change
predictions)
12
  • Outline
  • Relevance and types of decadal variability in the
    climate system
  • Observed decadal variability (emphasis on
    Atlantic)
  • Impact of Atlantic decadal variability
  • Mechanisms of Atlantic decadal variability
  • Predictability and predictions on decadal time
    scales

13
Three patterns of SST variability derived from
EOF analysis (EOF Empirical Orthogonal
Function)
Atlantic Multidecadal Oscillation (AMO)
or Atlantic Multidecadal Variability (AMV)
Pioneering work of Folland et al (1984,1986) in
elucidating this pattern and its climatic
relevance
Wang et al., 2008
14
Evidence for pattern of enhanced multidecadal
surface temperature variability derived from
multi-century paleo reconstructions (from tree
rings, ice cores, corals)
Delworth and Mann, 2000
15
  • Outline
  • Relevance and types of decadal variability in the
    climate system
  • Observed decadal variability (emphasis on
    Atlantic)
  • Impact of Atlantic decadal variability
  • Mechanisms of Atlantic decadal variability
  • Predictability and predictions on decadal time
    scales

16
North Atlantic Temperature
Why look to the Atlantic for decadal
predictability?
What will the next decade or two bring?
Warm North Atlantic linked to …
Drought
More rain over Sahel and western India
More intense hurricanes
  • Two important aspects
  • Decadal-multidecadal fluctuations
  • Long-term trend

17
Precip
SLP
Simulated response to idealized
AMO-like positive SST anomalies (Summer season)
Sfc Air Temp
Sutton and Hodson, 2005
18
Hybrid coupled model (based on GFDL CM2.1)
Global Atmosphere/Land System
2 deg horizontal resolution 24 vertical levels
Heat Water Momentum
Heat
Heat Water Mom.
Atlantic Slab Ocean
Pacific Dynamic Ocean
Indian Dynamic Ocean
1-0.3 deg resolution, 50 levels,MOM4
Constant Flux Adjustment
Time varying heating to induce AMO-like SST
variations
Zhang and Delworth,2005,2006
19
Modeling Description
GFDL CM2.1 Latest developed fully coupled
ocean-atmosphere global GCM, which produces a
stable, realistic multi-century control
integration (Delworth et al., 2005) To simulate
the impact of AMO, we modified CM2.1 into a
hybrid coupled model the Atlantic basin is
modified to be a slab ocean, all other model
components are the same as CM2.1
Observed AMO Index (HadISST)
Schematic diagram of the hybrid coupled model
10-member ensemble experiments for a duration of
100 years were conducted with the hybrid coupled
model. Each experiment is forced by the same
anomalous qflux in the Atlantic modulated by
observed AMO Index (1901-2000). The anomalous
qflux represents ocean heat transport anomaly
associated with THC variability.
20
(or Atlantic Multidecadal Variability, AMV)
Slide courtesy Rong Zhang
21
Simulated multidecadal JJAS surface air
temperature difference (K) (1931-1960)
(1961-1990)
22
Slide courtesy Rong Zhang
23
  • Outline
  • Relevance and types of decadal variability in the
    climate system
  • Observed decadal variability (emphasis on
    Atlantic)
  • Impact of Atlantic decadal variability
  • Mechanisms of Atlantic decadal variability
  • Predictability and predictions on decadal time
    scales

24
What is the mechanism behind Atlantic
multidecadal variability?
  • Two different, but not mutually exclusive, ideas
  • Atlantic Multidecadal Variability is a product of
    internal variability of the climate system
    through multidecadal scale strengthening and
    weakening of the Atlantic Meridional Overturning
    Circulation (AMOC).
  • Atlantic Multidecadal Variability is a product of
    changing radiative forcing (greenhouse gases and
    aerosols) in the 20th century.

25
Comparing the results to observations, it is
argued that the long-term, observed, North
Atlantic basin-averaged SSTs combine a forced
global warming trend with a distinct, local
multi-decadal oscillation that is outside of
the range of the model-simulated, forced
component and most likely arose from internal
variability. Ting et al, Journal of Climate, 2008
Kravtsov (2007) came to similar conclusion
Pattern of variability after removal of forced
signal.
26
What do coupled models tell us about internal
variability in the Atlantic? Many models
simulate enhanced multidecadal variability
involving Atlantic MOC Similar spatial
structure as observations Differing timescales
in the multidecadal range, differing
mechanisms Large-scale atmospheric impact
  • GFDL R15, R30 40-80 years (Delworth et al.,
    1993, 1997)
  • GFDL CM2.1 20 years
  • HADCM3 25 years (Dong and Sutton, 2005)
  • HADCM3 centennial (Vellinga and Wu, 2004
    Knight et al., 2005)
  • NCAR CCSM3 20 years (Danabasoglu, 2008)
  • ECHAM3 35 years (Timmermann et al., 1998)
  • ECHAM5 70-80 years (Jungclaus et al., 2005)
  • Theoretical work in hierarchy of models te Raa
    et al. (2004)
  • Multiple physical processes influencing the
    Atlantic MOC may contribute
  • to the variety of timescales found.

27
Atlantic meridional overturning circulation
28
SST anomalies associated with interdecadal MOC
fluctuations
MODEL
Modest Tropical Amplitude
EOF 1 HADISST OBSERVED SST
29
Complicating factor Changing radiative forcing
alters not only the thermal structure of the
ocean, but its circulation as well. This
complicates attribution.
Aerosol only forcing
All forcings
106 m3 s-1 (Sverdrups)
Simulated North Atlantic AMOC Index
Greenhouse gas only forcing
30
Simulated late 20th century Atlantic Ocean
Poleward Heat Transport (opposite impacts of
increasing greenhouse gases and aerosols)
Aerosol only
1015 W
GhGs only
Control
Latitude (oN)
31
  • Outline
  • Relevance and types of decadal variability in the
    climate system
  • Observed decadal variability (emphasis on
    Atlantic)
  • Impact of Atlantic decadal variability
  • Mechanisms of Atlantic decadal variability
  • Predictability and predictions on decadal time
    scales

32
Decadal Predictability and Prediction Efforts
  • Characterizing and understanding decadal
    variability, climatic impacts, and interactions
    with radiative forcing
  • Idealized predictability experiments
  • - inherent predictability of AMOC and other
    phenomena
  • - climatic relevance
  • New coupled assimilation methods for reanalysis
    and initialization of predictions
  • Improved models
  • - higher resolution
  • - new physics/numerics
  • - reduced bias
  • 5. Prototype decadal predictions and attribution

? Focus on Atlantic, but methodology is general.
33
Atlantic Meridional Overturning Circulation
(AMOC) in GFDL CM2.1 Model
106 m3 s-1 (Sverdrups)
Spectrum of AMOC
34
AMOC Index
Predictability of Atlantic Meridional Overturning
Circulation (AMOC) in GFDL CM2.1 Climate Model
AMOC Index
35
Surface Air Temperature
Histogram of central US temperature
June-July-August Years 3-10 after common ocean
initialization
JJA
Control
Frequency of occurrence
Wet central US
Rainfall (cm/day)
Dry Sahel
36
Histogram of NH Extratropical Mean Surface Air
Temperature
Two ensembles of projections for 2001-2010 using
same forcing. They differ due to predictable
decadal variability, primarily associated with
Atlantic Meridional Overturning Circulation.
Frequency of occurrence
Temperature
37
GHGNA
forcings
Pioneering development of coupled data
assimilation system
Atmosphere model
u, v, t, q, ps
S. Zhang, M. J. Harrison, A. Rosati, and A.
Wittenberg (2008)
Land model
uo, vo, to
tx,ty
(Qt,Qq)
  • Coupled Ensemble Data Assimilation estimates the
    temporally-evolving probability distribution of
    climate states under observational data
    constraint
  • Multi-variate analysis maintaining physical
    balances between state variables such as T-S
    relationship - geostrophic balance mostly
  • Ensemble filter maintaining the nonlinearity of
    climate evolution mostly
  • All coupled components adjusted by observed
    data through instantaneously-exchanged fluxes
  • Optimal ensemble initialization of coupled
  • model with minimum initial shocks

Sea-Ice model
Ocean model
T,S,U,V
Tobs,Sobs
a)
obs PDF
Prior PDF
Data Assim (Filtering)
Analysis PDF
b)
38
New coupled assimilation system
dramatically improves ENSO prediction skill
NINO3 Anomaly Correlation Coefficient
0.6
Forecast Lead Time (months) 1 3 5
7 9 11
Traditional assimilation system
J F M A M J J A S O
N D Forecast Start Month
Forecast Lead Time (months) 1 3 5
7 9 11
New coupled assimilation system
J F M A M J J A S O N
D Forecast Start Month
39
Assimilation and Observing systems are crucial
  • Prediction skill will depend on how well we can
    observe the climate system (and how well system
    can be initialized)
  • Revolution in ocean observations with advent of
    ARGO network (floats that provide profiles of
    temperature and salinity over the top 2000 m on a
    global scale)
  • Prediction skill better when incorporating ARGO
    implications for interpreting hindcasts

40
Model development
  • Simulated variability and predictability is
    likely a function of the model
  • Developing improved models (higher resolution,
    improved physics) is crucial for studies of
    variability and predictability
  • New GFDL models CM2.4, CM2.5
  • Ocean
  • Resolution 25Km in tropics to 10 Km high
    latitudes
  • Very energetic, low viscosity, higher order
    advection
  • Atmosphere
  • CM2.4 Global, 1 degree (100 Km grid)
  • CM2.5 Global, 0.5 degree (50 Km grid)

41
Ocean resolution as fine as 10Km in high
latitudes
GFDL CM2.4 Global Coupled Model SST, surface
currents
GFDL CM2.1 Global Coupled Model SST, surface
currents
GFDL CM2.1 model was one of the best in the world
for Atlantic simulations in AR4. Even so,
important processes are not well resolved.
Sfc currents and SST
Key issue How sensitive is simulated decadal
variability and predictability to model
resolution and physics?
42
High resolution coupled model shows realistic
simulation of eddy kinetic energy
Observational estimate (satellite)
Eddy Kinetic Energy Log scale, cm2 s-2
GFDL CM2.4 Model
Courtesy Riccardo Farnetti
43
From presentation by R. Stouffer, GFDL
44
  • IPCC AR5/CMIP5
  • Coordinated Decadal Prediction Experiments
  • Key question
  • If we start climate change simulations with the
    observed state of the climate system, does that
    improve our predictions of future climate change?
  • Major challenges
  • What is true predictability of the climate
    system?
  • Are models good enough to realize that
    predictability?
  • Are current observing and initialization systems
    adequate?
  • How do we handle model bias and drift?
  • How can decadal predictions be used?

45
Discussion
  • Decadal variability is an important part of
    climate spectrum, particularly on regional scales
    that are important for society and ecosystems
  • Decadal prediction/projection is a mixture of
    boundary forced and initial value problem
  • Changing radiative forcing (esp. aerosols) will
    be a key ingredient
  • Some basis for decadal predictability of internal
    variability, probably originating in ocean
  • Some of predictability will arise from unrealized
    climate change already in the system
  • Substantial challenge for models, observations,
    assimilation systems, and theoretical
    understanding

46
  • Crucial points
  • Robust predictions will require sound
    theoretical understanding of decadal-scale
    climate processes and phenomena.
  • Assessment of predictability and its climatic
    relevance may have significant model dependence,
    and thus may evolve over time (with implications
    for observing and initialization systems).

But … even if decadal fluctuations are not
predictable, it is still important to understand
them to better understand and interpret observed
climate change.
47
Current/planned activities at GFDL
  • Ongoing studies with observations and models to
    develop improved understanding of
  • mechanisms of simulated decadal variability
  • decadal scale predictability arising from
    internal variability
  • Development and use of higher resolution coupled
    models.
  • Development of new coupled assimilation system
  • Assessment of observation systems for decadal
    predictability
  • Prototype decadal predictions currently underway

48
Cautionary Notes
  • This field is in its infancy many fundamental
    challenges remain
  • Results from CMIP5 decadal predictions should be
    viewed with caution in light of
  • Model bias and drift and their impacts on
    prediction
  • Varying initialization strategies
  • Unknown level of true predictability of the
    system
  • It is possible that initial decadal prediction
    attempts will show little or no meaningful
    predictability (from internal variability). That
    would lead to at least two possibilities
  • The system is not predictable on decadal time
    scales
  • We are not yet able to realize that
    predictability
  • Will we be able to distinguish between these two
    possibilities?

49
  • Final points …
  • Decadal prediction is a major scientific
    challenge.
  • If scientific progress warrants, decadal
    prediction needs to eventually involve complete
    Earth System Models including biogeochemical
    processes.
  • An equally large challenge is evaluating and
    understanding the possible utility of decadal
    predictions.
  • What can we say about evolution of Atlantic SST?
  • What can we say about the likelihood of North
    American drought over the next 1-10 years?
  • Early Warning System for possibly abrupt
    climate change
About PowerShow.com