Transcript The modelling of the climate system

The modelling of the climate system
Professor Lennart Bengtsson
ESSC, University of Reading
Max-Planck-Institut für Meteorologie, Hamburg
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Zur Anzeige wird der QuickTime™
Dekompressor “Foto - JPEG”
benötigt.
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Temperature change
1854 - 2004 ( land areas only)
Zur Anzeige wird der QuickTime™
Dekompressor “GIF”
benötigt.
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Temperature change
1854 - 1904 ( land areas only)
Zur Anzeige wird der QuickTime™
Dekompressor “GIF”
benötigt.
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Temperature change
1904 - 1954 ( land areas only)
Zur Anzeige wird der QuickTime™
Dekompressor “GIF”
benötigt.
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Temperature change
1954 - 2004 ( land areas only)
Zur Anzeige wird der QuickTime™
Dekompressor “GIF”
benötigt.
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Temperature change
1954 - 2004
Zur Anzeige wird der QuickTime™
Dekompressor “GIF”
benötigt.
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
The modelling of the climate system
What is climate?
Climate variations on different time-scales
The modelling of climate
Atmospheric modelling and weather prediction
Modelling of the Earth climate system
How predictable is climate?
Model simulation of the present climate
Why is the climate changing?
Climate change prediction
Concluding remarks
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
The classical view on climate
Climate as a stationary concept
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Köppen climate zones
•
•
•
•
•
•
Main groups
A: Tropical rainy climate, all months > +18 C
B: Dry climate, Evaporation > Precipitation
C: Mild humid climate, coldest month +18 C - -3 C
D: Snowy - forest climate, coldest month < -3C but warmest > +10
E: Polar climate , warmest month < +10 C
ET: Tundra climate, warmest month > 0 C
•
•
•
•
Subgroups
f : Moist, no dry seasons
w: Dry season in winter
s : Dry season in summer
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Köppen climate zones
ECHAM5
GPCP (Prec)
CRU2 (Temp)
T159
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
What is climate?
• Climate is nowadays generally defined as a comprehensive
statistical description of weather ( including extremes) over
a sufficiently long period of time (30-100 years)
• There is no sharp distinction between weather and climate
• There are free atmospheric modes of circulation that
have time-scales of up to about two years ( quasi-biennal
oscillation)
• There are coupled ocean-atmospheric modes that have
time-scales from weeks to several decades. A dominant
feature is the El-Nino phenomenon in the eastern tropical
Pacific ocean with a time-scale of about four years.
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
The present view on climate
Climate as a dynamical system
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Climate variations on different time-scales
• Climate variations are dominated by chaotic weather
events
• Other variations are due to coupled ocean-atmosphere
processes which could cover longer periods of time. They
are probably also mainly chaotic.
• Climate would also vary due to changes in solar
irradiation( regular or otherwise) and under the influence
of volcanic aerosols
• Climate also varies due to changes in the composition of
the atmosphere such as greenhouse gases and aerosols
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Shorter term climate variations and their
likely causes
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Natural temperature variations
ENSO influence
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Natural temperature variations
Pinatubo and ENSO
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
The modelling of climate
• Prior to the 1950s climate was essentially a descriptive science, but
general ideas of the general circulation of the atmosphere and the
oceans existed
• Over the past 50 years the direction of climate research has changed
driven by space based observations and mathematical modelling of
the climate system.
• Climate modelling has occurred along three lines:
• - increased numerical resolution and more accurate treatment of
individual physical and chemical processes
• - coupling of individual model components of the climate into Earth
system models including aspects of the biosphere
• - ensemble predictions to be able to separate signal from noise.
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• What is a CLIMATE MODEL?
• A COMPUTER PROGRAM which numerically solves the
MATHEMATICAL EQUATIONS that represent the LAWS OF PHYSICS.
• Components include the: ATMOSPHERE, OCEAN, LAND, CRYOSPHERE
and BIOSPHERE and all the dynamics, physical processes and interactions
between them.
• The most comprehensive climate models include:
GENERAL CIRCULATION MODELS (GCMs) as atmospheric and oceanic
components.
• An AGCM follows the evolution of all the weather systems, clouds, and rain,
and the interactions with the land and ocean.
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• Climate system in pictorial form
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• Climate system as a principle system
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• Climate model components
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• The strategy of climate research
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• Significant atmospheric processes
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• Aspects of numerics & physical
parameterizations for AGCMs
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• Resolution issues
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Modelling error as a function of horisontal and
vertical resolution
(ECHAM climate model 2005)
T21_L19
100
T31_L19
T42_L31
80
T42_L19
T85_L19
60
T106_L19
T63_L19
40
T63_L31
T85_L31
20
T106_L31
T159_L31
0
Normalized RMSE [ % ]
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
How well can model simulate
present climate?
Some examples
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Typical cyclone storm tracks
Tracks
16 January 2005
Lennart Bengtsson
Intensities
Celsius lecture 2005
Uppsala Unversity
Storm track density ERA40 (left, 1979 - 2002)
ECHAM5 ( right, 1979 - 1999, atmos. model run)
for the relative vorticity at 850 mb
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Number of extra-tropical storms at the Northern
Hemisphere as a function of intensity during winter
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Köppen climate zones
ECHAM5
simulated
ERA40
determined
from analyses.
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
How predictable is climate?
• Do we have a unique climate?
• Predictability of weather
• and predictability of climate
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• CLIMATE PREDICTION AND CHAOS
• “ For want of a nail, the shoe was lost;
• For want of a shoe, the horse was lost;
• For want of a horse, the rider was lost;
• For want of a rider, the battle was lost;
• For want of a battle, the kingdom was lost “
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• Predictability of weather
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
ECMWF EPS: Forecast Started 8th January
00UTC
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
ECMWF EPS: Forecast Started 6th January
00UTC
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• Predictability of snow in Germany
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Why is the climate changing?
• There are natural changes on a variety of time scales.
• Some of these changes are chaotic and unpredictable.
• Sometimes chaotic events are inadvertently interpreted as
due to specific external events ( e.g. solar forcing, volcanic
eruptions, human influences etc.).
• However, increasing greenhouse gases exercise a real
influence on climate and observations and model results
are supporting each other
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• The greenhouse effect
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• The greenhouse effect
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Carbon dioxide increase
1957-2003
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• CO2 and CH4 Concentrations Past, Present and
Future
Compiled by K. Alverson
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• Svante Arrhenius 1859-1927
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• The carbon cycle
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Annual increase in GHG forcing
1958-2003
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
The problem of climate change prediction
• How credible are climate change predictions?
• How will climate forcing change?
• What aspects of climate change is predictable?
• What is unpredictable?
• Some important processes are not yet generally considered
in climate models. This include feedback with the
biosphere
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Climate feedback
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• The feedback problem
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• The feedback problem
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• The feedback problem
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
• Feedback results from different models
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Delworth and Knutson, 2000
Monte-Carlo simulations with a coupled AO GCM: one out five simulations almost
perfectly reproduced the observed global temperature variability.
obs
16 January 2005
Lennart Bengtsson
exp 3
Celsius lecture 2005
Uppsala Unversity
Ensemble climate trends averaged for
different time-periods
(T/decade)
1-30 years
16 January 2005
Lennart Bengtsson
1-80 years
Celsius lecture 2005
Uppsala Unversity
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Storm track density difference between scenario A1B ( aver. cond. 2071-2100 and ( aver.
cond. 1961-1990) for the ECHAM 5 model. NH left and SH right.
Note the poleward change of the storm track at the SH
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Coupled
Model
T63L31
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity
Concluding remarks
• Climate model can reproduce most features of the general circulation
of the atmosphere( and to a lesser degree) of the oceans. Results
generally improves with higher numerical accuracy (resolution)
• The treatment of small scale physical processes, turbulence, clouds,
boundary layer fluxes etc. are parameterised ( only expressed in terms
of the resolvable parameters) and thus to some extent subject to ad hoc
assumptions. However climate change feedbacks are likely to be
influenced by such assumptions
• Climate change predictions over larger areas and longer time-scales are
dynamically robust (albeit model dependent).
• Regional climate change predictions of a few decades and shorter are
most likely unpredictable, but an ensemble over many cases could
indicate a change in the overall probability distribution of climate and
weather events.
16 January 2005
Lennart Bengtsson
Celsius lecture 2005
Uppsala Unversity