Download Quasi-realistic climate models

Generation of added value with
models
Hans von Storch,
GKSS Research Centre, Geesthacht, and
KlimaCampus „clisap“, University of Hamburg
Germany
Folie 1
Overview:
1. Quasi-realistic climate models („surrogate reality“)
2. Free simulations and forced simulations for
reconstruction of historical climate
3. Climate change simulations
4. Downscaling - Regional climate modelling
Folie 2
Conceptual aspects of
modelling
Folie 3
Conceptual aspects of modelling
Hesse’s concept of models
Reality and a model have attributes, some of which are consistent and
others are contradicting. Other attributes are unknown whether reality and
model share them.
The consistent attributes are positive analogs.
The contradicting attributes are negative analogs.
The “unknown” attributes are neutral analogs.
Hesse, M.B., 1970: Models and analogies in science. University of Notre Dame Press, Notre Dame 184
Folie 4
pp.
Conceptual aspects of modelling
Validating the model means to determine the positive and negative
analogs.
Applying the model means to assume that specific neutral analogs are
actually positive ones.
The constructive part of a
model is in its neutral analogs.
Folie 5
Conceptual aspects of modelling
Folie 6
Conceptual aspects of modelling
Folie 7
Conceptual aspects of modelling
Folie 8
Conceptual aspects of modelling
• Models represent only part of reality;
• Subjective choice of the researcher;
Certain processes are disregarded.
• Only part of
contributing spatial and
temporal scales are
selected.
• Parameter range
limited
Folie 9
Conceptual aspects of modelling
Models can be shown to be consistent with
observations, e.g. the known part of the
phase space may
reliably be
reproduced.
Folie 10
Conceptual aspects of modelling
Models can not be verified because reality is
open.
Coincidence of modelled and observed state
may happen because of model´s skill or
because of fortuitous (unknown) external
influences, not accounted for by the model.
Folie 11
Conceptual aspects of modelling
Purpose of models
• reduction of complex systems
understanding
• surrogate reality
realism
Folie 12
Conceptual models for the
reduction of complex systems
Folie 13
Models for reduction of complex systems
Models for reduction of complex systems
• identification of significant, small subsystems and key processes
• often derived through scale analysis
(Taylor expansion with some characteristic numbers)
• often derived semi–empirically
• constitutes “understanding”, i.e. theory
• construction of hypotheses
characteristics:
simplicity
idealisation
conceptualisation
fundamental science approach
Folie 14
2. Models for reduction of complex systems
Idealized energy balance
Folie 15
2. Models for reduction of complex systems
E =bE +aA
with
b = albedo
a = transmissivity
E = short wave solar radiation
A = long wave thermal radiation = sT4
 Teq
 1 b E 
 

 as 
1
4
equilibrium
without atmosphere
a=1,
with present atmosphere
a=0.64, b= 0.30 : Teq = +15°C
Folie 16
b= 0 :
Teq = - 4°C
Models for reduction of complex systems
Temperature dependent albedo (reflectivity)
Folie 17
Noise or deterministic chaos?
Mathematical construct of
randomness adequate concept
for description of features
resulting from the presence of
many chaotic processes.
Folie 18
Integration of a zero–dimensional energy balance model
no
noise
with constant transmissivity and
temperature dependent albedo
evolution from different initial values
with
noise
evolution with slightly randomized
transmissivity
Folie 19
Models for reduction of complex systems
Numerical experiment with ocean model: standard simulation with steady
forcing (wind, heat
and fresh water
fluxes) plus random
forcing
zero-mean
precipitation overlaid.
Example for
Stochastic Climate
Model at work.
Folie 20
response
Quasi-realistic modelling
Folie 21
Models as surrogate reality
• dynamical, process-based models,
•
•
•
•
experimentation tool (test of hypotheses)
design of scenario
sensitivity analysis
dynamically consistent interpretation and extrapolation
of observations in space and time (“data
assimilation”)
• forecast of detailed development
(e.g. weather forecast)
characteristics:
Folie 22
complexity
quasi-realistic
mathematical/mechanistic
engineering approach
Components of the climate system. (Hasselmann, 1995)
Folie 23
Quasi-realistic climate models …
… are dynamical models, featuring discretized equations of the type
dΨ k
  Pi ,k (k )
dt
i
with state variables Ψk and processes Pi,k.
The state variables are typically temperature of the air or the ocean, salinity and humidity,
wind and current.
… because of the limited resolution, the equations are not closed but must be closed by
“parameterizations”, which represent educated estimates of the expected effect of nondescribed processes on the resolved dynamics, conditioned by the resolved state.
Folie 24
atmosphere
Folie 25
Dynamical processes in the atmosphere
Folie 26
Dynamical processes in a
global atmospheric general circulation model
Folie 27
Dynamical processes in the ocean
Folie 28
Dynamical processes in a global ocean model
Folie 29
Bray and von Storch, 2010
Results of a survey among climate modellers in 1996, 2003 and 2008
Folie 30
Folie 31
Folie 32
validation
1880–2049
ECHAM3/LSG
1973–1993
ERA ECMWF
Folie 33
Climatic Zones
Modell
Observed
Classification following Koeppen
Folie 34
Erich Roeckner,
pers.
communication
Observed
Simulated
Winter
(DJF)
Cyclogenesis
Density of
stromtracks
Erich Roeckner,
pers. Comm.
Folie 35
Typical different
atmospheric model
grid resolutions
with corresponding
land masks. T42
used in global
models. (courtesy:
Ole BøssingChristensen)
Folie 36
variance
global model
Insufficiently
resolved
Well resolved
Spatial scales
Folie 37
Free and forced simulations
for reconstruction of
historical climate
Folie 38
4. Free and forced model simulations
Different ways of running the
model
" Free Simulation ":
t 1  F(t )
" Forced Simulation ":
t 1  F(t ; t )
with t  greenhouse gas concentrat ions
Folie 39
or
aerosol concentrat ions
or
or
or
solar output (incl. orbital configurat ion)
topography (e.g., ice sheets)
vegetation
Free Simulation: 1000 years
Folie 40
Model as a constructive tool
Zorita, 2001
Temperature (at 2m) deviations
from 1000 year average [K]
no solar variability, no changes in
greenhouse gas concentrations, no
orbital forcing
Folie 41
validation
Reconstruction from historical
evidence, from Luterbacher et al.
Late Maunder Minimum
Model-based
reconstuction
1675-1710
vs. 1550-1800
Folie 42
Global 1675-1710 temperature anomaly
Folie 43
Model
as a constructive tool
Free and forced model simulations
• Free simulations are routinely done with GCMs;
They reproduce most large-scale features of present
climate in a satisfactorily manner.
They exhibit a rich spectrum of variability.
• Forced simulations, with fully coupled atmosphere-ocean
models, are also done. Changed factors are greenhouse
gases, aerosols, vegetation, topography, orbit parameters
...
A simulation generates one of infinitely many consistent
realizations of the forced state.
Folie 44
Laboratory to test
conceptual models
Folie 45
Example: Stommel model of the North Atlantic overturning
Laboratory to test conceptual models
Ft, Ht freshwater and heat flux Fp, Hp
Subtropical
Atlantic
Tt,St
Subpolar
Atlantic
Tp, Sp
Transport
 t T  H  2 m T
 t S  F *  2 m S
m  k aT  bS 

H   T  T *
Folie 47

Laboratory to test conceptual models
Rahmstorf‘s model
Stommel‘s theory
F *
F *
Rahmstorf, 1995
Folie 48
Testing the of multimodality of large scale atmospheric dynamics
Berner
Folie
49 and Branstator,
pers. comm
Roeckner & Lohmann,
1993
detailed
parameterization
Latitude-height
distribution of
temperature (deg C)
Effect of black cirrus
Difference “black
cirrus” - detailed
parameterization
No cirrus
Model as a constructive tool
Folie 50
Difference “no cirrus”
- detailed
parameterization
Deconstruction of recent climatic development
°C
simulation without anthropogenic drivers
simulation with anthropogenic drivers
vs. „observation“
(centered on 1960-1990 mean)
Folie 51
Climate change simulations
Folie 52
5. Climate Change simulations
Folie 53
Scenario building
• Construction of scenarios of emissions.
• Construction of scenarios of concentrations of
radiatively active substances in the
atmosphere.
• (Ok – not quite exact; aerosols …)
• Simulation of climate as constrained by
presence of radiatively active substances in
the atmosphere (“prediction” of conditional
statistics).
Folie 54
“SRES” Scenarios
SRES = IPCC Special Report on Emissions Scenarios
A1
A world of rapid economic growth and rapid
introduction of new and more efficient technology.
A2
B1
A very heterogeneous world with an emphasis on
family values and local traditions.
A world of “dematerialization” and introduction of
clean technologies.
IS92a
A world with an emphasis on local solutions to
economic and environmental sustainability.
“ business as usual ” scenario (1992).
Folie 55
IPCC, 2001
B2
Scenario A2
Annual
temperature
changes [°C]
(2071–2100)
(1961–1990)
–
Scenario B2
Folie 56
Danmarks Meteorologiske Institut
precipitation
Folie 57
Agreement among 7 out of a total of 9 simulations
Giorgi et al., 2001
TAR (2001) „sub-continental development“
scenarios A2 and B2.
Downscaling
Folie 58
Regional and local conditions –
in the recent past and next
century
Simulation with barotropic
model of North Sea
Globale development
(NCEP)
Tide gauge St. Pauli
Dynamical Downscaling
CLM
Cooperation with a variety of
governmental agencies and with a
number of private companies
Folie 59
Empirical
Downscaling
Typical regional
atmospheric
model grid
resolutions with
corresponding
land masks.
50 km grid used
in regional
models
(courtesy: Ole
BøssingChristensen)
Folie 60
variance
regional model
Insufficiently
resolved
Well resolved
Spatial scales
Folie 61
Added value
Concept of Dynamical Downscaling
RCM
Physiographic
3-d vector of state
detail
State space equation
Ψ t 1  F(Ψ t ;ηt )  εt
Observatio n equation
d t  G(Ψ t )  δt
Known large scale
statewit h
 t ,  t  model and observatio n errors
F  dynamical model
projection of full state on
G  observatio n model
large-scale scale
Ψ t*1  F(Ψ t ;ηt )
Forward integratio n :
d t*1  G (t*1 )
 Ψ t 1  Ψ t*1  K(d t*1  d t 1 )
with a suitable operator K .
Folie 62
Large-scale
(spectral) nudging
Example Extreme Events (Wind & Waves)
Wind [m /s]
Years
SON
EUR
K13
Hipocas
xr90
2
5
25
2
5
25
2
5
25
24.38
25.86
28.44
22.50
23.76
25.67
23.29
24.89
26.68
xr
25.17
27.28
31.33
23.16
24.82
28.00
24.15
26.32
30.70
Waves [m ]
Hipocas
Observed
Observed
xr90
xr90
25.96
28.70
34.22
23.82
25.88
30.33
25.01
27.75
34.72
24.05
25.75
28.09
23.16
24.33
26.43
23.11
24.15
26.42
xr
25.21
27.64
32.77
24.03
25.94
29.75
24.03
25.94
29.75
xr90
26.37
29.53
37.45
24.90
27.55
33.07
24.95
27.73
33.08
xr90
xr
xr90
7.12
7.49
7.86
7.84
8.44
9.04
8.99 10.35 11.71
5.89
6.15
6.41
6.34
6.83
7.32
6.90
8.20
9.50
6.78
7.06
7.34
7.37
7.79
8.21
8.04
9.03 10.02
xr90 xr
6.41
6.93
7.52
5.52
5.89
5.99
5.60
5.97
6.34
xr90
6.77
7.13
7.54
8.15
9.21 10.90
5.84
6.16
6.46
7.03
7.88
9.77
5.84
6.08
6.46
6.95
7.88
9.42
2, 5, and 25-year return values with 90% confidence limits based on 10.000 Monte Carlo simulations
each.
(Weisse and Günther. 2006)
Folie 63
What is coastDat?
A set of model data of recent, ongoing and possible future coastal
climate
(hindcasts 1948-2008, reconstructions and scenarios for the future, e.g., 2070-2100)
Based on experiences and activities in a number of national and
international projects (e.g. WASA, HIPOCAS, STOWASUS, PRUDENCE)
Presently contains atmospheric and oceanographic parameter
(e.g. near-surface winds, pressure, temperature and humidity; upper air meteorological data
such as geopotential height, cloud cover, temperature and humidity; oceanographic data
such as sea states (wave heights, periods, directions, spectra) or water levels (tides and
surges) and depth averaged currents, ocean temperatures)
Covers different geographical regions
(presently mainly the North Sea and parts of the Northeast Atlantic; other areas such as the
Baltic Sea, subarctic regions or E-Asia are to be included)
http://www.coastdat.de, contact: Ralf Weisse ([email protected])
Folie 64
Some applications of
- Ship design
- Navigational safety
- Offshore wind
- Oils spill risk
- Interpretation of
measurements
- Chronic Oil Pollution
- Ocean Energy
Currents Power [W/m2]
Folie 65
Wave Energy Flux [kW/m]
Scenarios for Northern Germany
Folie 66
Conclusions
• “Model” is a term with very many different meaning in
different scientific and societal quarters.
• The constructive part of models is in their neutral analogs with
“reality”.
• Validation of models means to check positive and negative
analogs.
• In climate science we have conceptual models – constituting
understanding – and quasi-realistic models, allowing for
numerical experimentation.
• Quasi-realistic models may be used for testing hypothesis, for
developing hypothesis, for the construction of a full 4-d state,
forecasts and for scenarios.
Folie 67
Conclusions
• Global climate modeling allows the representation of global,
continental and sub-continental scales. Global models fail on the
regional and local scale.
•Scenarios of future climate change hinge on the validity of
economic scenarios.
• Simulation of regional climate is a downscaling problem and
not a boundary value problem.
• Marine weather (winds, waves) have been successfully
reconstructed for Northern Europe for the years 1958-97 with a
1-hourly resolution. (CoastDat@GKSS). Also, scenarios are
available to this end.
Folie 68
Background information on this issue:
von Storch, H., S. Güss und M. Heimann, 1999: Das
Klimasystem und seine Modellierung. Eine Einführung. Springer
Verlag ISBN 3-540-65830-0, 255 pp
von Storch, H., and G. Flöser (Eds.), 2001: Models in
Environmental Research. Proceedings of the Second GKSS
School on Environmental Research, Springer Verlag ISBN 3-54067862, 254 pp.
Müller, P., and H. von Storch, 2004: Computer Modelling in
Atmospheric and Oceanic Sciences - Building Knowledge.
Springer Verlag Berlin - Heidelberg - New York, 304pp, ISN
1437-028X
Folie 69