Bild 1

1
Present Status and Future Plans for Climate
Modeling at SMHI-Rossby Centre Colin
Jones Rossby Centre Rossby Centre Group View
(as understood by me!)
2

• Outline
• Present and ongoing Regional Climate Modeling
  Activities
• Present Future Global Climate Modeling
  Activities EC-Earth
• Global Earth System Modeling
• Global (decadal) Climate Prediction
• 3. Future Regional Climate Modeling Directions
• Harmony/Arome/Aladin-climate Regional ESM
• NEMO Ocean Model

3
The Rossby Centre RCM is one of a number of RCMs
within Europe Collaboration allows for an
increased number of common RCM scenarios to be
available to all groups, reducing uncertainty of
future climate conditions Ensembles/PRUDENCE
projects.
For CMIP5/AR5 we aim to have these
RCM/GCM matricies for all populated regions of
the world. Not just Europe/N. America
4
RCA3 European Climate Scenarios Summer 2009
GCM
Emissions
Time period
Domain/resol.
RCA3 is a fast RCM We can produce a large
number of scenarios quickly The quality of the
results is also okay!
In production / planned
5
Winter season mean temperature change
(2070-2100)(1961-1990) RCA3 downscaling of 5
GCMs all employing the same emission scenario.
Areas of commonality. Areas of differences
From Grigory Nikulin (Rossby Centre)?
6
RCA is also coupled to the Rossby Centre Ocean
Model We have 2 coupled Regional Climate
Models (RCAO) Over the Baltic Sea and Arctic
Ocean
Such models allow for specific analysis of
coupled regional climate change and environmental
responses, along with targeted development of
modeling approaches
7
Arctic Sea-Ice extent simulated by RCAO
Spinup phase
8
The standard resolution of RCA is 50km Similar
to most other RCMs
50 x 50 km
9
We are presently moving that standard to be
25km Similar to most other European RCMs
50 x 50 km
10
And have begun testing a future standard of
12km
50 x 50 km
11
Increased resolution can help
RCM simulation of precipitation over the
Alps Mean annual precipitation (mm/day)
From Christensen et al. 2005
12
WINTER DAILY RAINFALL OVER THE ALPS
Resolution yields a much better description of
extremes
13
RCA is (and will increasingly be) applied to
generate regional climate change scenarios in
other areas of the world
ARCMIP
PRUDENCE ENSEMBLES NEWBALTIC
PIRCS NARCCAP SGMIP
RMIP
AMMA AFRMIP AIACC
PLATIN ARC
Transferability project
The International RCM community aims to perform a
coordinated downscaling of CMIP5 GCM scenarios
around the globe in support of climate impact
adaptation
14

• Near-term plans with existing RCM tools
  (2008-2012)
• Increased Resolution (50km 10km)
• 2. Regional climate scenarios over Europe (IPCC
  AR5)
• 3. Regional scenarios in other regions of the
  globe (IPCC AR5)
• 4. Targeted developments
• Non-European Regions
• Specific Regional Phenomena
• Specific (Localized) Environmental
  applications and developments.
• Localized coupling issues.

15
Simulated regional climate change can be very
different depending on the driving GCM boundary
conditions Mean annual change simulated by an
RCM with 2 different GCM boundary conditions
with the same A2 emission scenario (2071-2100) -
(1960-1990)
Räisänen 2003
The Climate System is global
16
Many of the major uncertainties in the response
of the climate system to increasing levels of
CO2 are global Cloud Feedbacks
Carbon-Climate Feedback Aerosol Feedbacks
Sea-Ice/Snow Feedbacks Methane Release and
in particular impact the Global Climate System
through feedback processes.
17
Coupled Carbon Cycle Climate Models indicate a
likely amplification of atmospheric CO2
concentrations (5-20) and climate warming due
to a sensitivity of global carbon uptake to
physical climate change
Friedlingstein et al., 2006
The uncertainty is large due to being dependent
on both the simulated physical climate change and
the carbon cycle response to this change.
18
Change in Ocean and Terrestrial Carbon Uptake as
a function of simulated climate warming
LAND
OCEAN
IPSL_CM4_LOOP
Friedlingstein et al., 2006
IPSL-CM2_C
19
A consortium of European National Met. Services
universities interested in developing and
applying a common Global Climate Earth System
Model
Rationale Use the ECMWF Seasonal Prediction
System for decadal and centennial climate
prediction.
20
Background Important Considerations for
EC-Earth members Members (generally) do not
have direct access to a coupled GCM. This makes
them dependent on other GCM groups for climate
simulations (e.g. as boundary conditions for RCM
climate scenario generation) Many university
research groups in Climate and Earth System
Science have no natural global model into which
to apply their developments The NWS pay for and
thus partially own ECMWF tools. At the NWS,
scientists are familiar with using the ECMWF
System, ECMWF products and ECMWF GRIB
format. ECMWF IFS is a/the world-leading
atmospheric model on medium to seasonal
timescales, has very efficient numerics,
excellent physical parameterisations and an
advanced data assimilation system Use the ECMWF
coupled seasonal prediction system for climate
prediction and as the basis of a Global Earth
System Model ?
21
EC-EARTH Components
Interactive (chemistry) aerosols
22
EC-Earth-status members (signed letter of
intent) DMI Denmark IMAU The
Netherlands Instituto de Meteorologia
Portugal Centro de Geofísica,University of
Lisbon Portugal KNMI The Netherlands Meteoro
logisk Institut Norway Unité ASTR
(UCL) Belgium Met Éireann Ireland Universit
y College Dublin Ireland Universiteit Utrecht
The Netherlands Vrije Universiteit Amsterdam
The Netherlands Stockholm University
Sweden Lund University Sweden Chalmers
University Sweden ICTP Italy SMHI
Sweden INM (AEMET) Spain ETH-Zurich S
witzerland Barcelona Supercomputing Center
Spain Observer ECMWF
23
With its origin as a weather forecast model
ECMWF IFS has a good simulation of weather
variability. Climate is the average of weather
events
T255 ERA40
T159
T95
T. Jung 2006
24
Winter season mean temperature change
(2070-2100)(1961-1990) RCA3 downscaling of 5
GCMs all employing the same emission scenario.
Areas of commonality. Areas of differences
From Grigory Nikulin (Rossby Centre)?
25
Winter season mean temperature change
(2016-2045)(1961-1990) Now greater differences
between simulated climate change signal. Even
between 3 members of the same GCM with same
emission scenario (ECHAM5 A1B) but different
initial conditions.
From Grigory Nikulin (Rossby Centre)?
Short timescales are very relevant for
adaptation
26
Winter temperature change from 3 runs of the
same ECHAM5 GCM (2016-2045) minus
(1962-1991). Differences due to using different
initial conditions can be of the same magntidue
as the simulated climate change signal.
27
Natural Variability is of a similar magnitude to
the near term climate change signal Mainly
related to ocean circulation variability
Reconstructed North Atlantic Sea Surface
Temperature Anomalies
28
When we make climate simulations we spin-up the
ocean over thousands of simulated years forced
only by The Sun, Earths rotation and
pre-industrial concentrations of greenhouse gases.
Northern Hemisphere temperature anomalies
COSMOS/MPI Millenium Experiment
Reconstructed observed values
GCM simulated value
Good climate models can simulate a realistic
amount of natural variability. But there is
absolutely no reason to expect this variability
is occuring at the same time as in reality. i.e
The model calendar is largely imaginary
29
In terms of natural variability we have
absolutely no way of knowing where we are on this
timeseries when we start a future climate
integration
Any greenhouse gas induced signal is
likely/hopefully included
30
To make climate predictions for 2010-2040 We
need to simulate the time evolution of the slowly
varying components of natural variability (the
ocean) starting from a realistic initial state
and the effect of increasing greenhouse gas
amounts.
To address this requires that we
initialize our coupled climate models from
observed conditions Particularly those
bits that have a slow evolution
the ocean This moves climate
prediction towards weather prediction
31
EC-EARTH Components
Interactive (chemistry) aerosols
32
An early example of global climate prediction
from the U.K.
33

• It is now possible for WCRP to address the
  seamless prediction of the climate system from
  weekly weather to seasonal, interannual, decadal
  and centennial climate variations and
  anthropogenic climate change. (WCRP 2005)

EC-Earth aims to contribute to the CMIP5 decadal
runs (2010)
34
(No Transcript)
35
Methane Release from Arctic Oceans and
Permafrost Changes in sea-ice cover and global
albedo
Many of these phenomena have global implications
but they are often occurring (and being observed
and studied) on small scales
36
to study and understand such phenomena. to
develop suitable descriptions of the processes
for use in Global Earth System Models. may
require getting down to the spatial and
temporal scales of the processes. This will
allow a better link with observationalists and
scientists working at the process level. Needed
A high-resolution Regional Earth System
Model? This is a major task (way too big for the
Rossby Centre alone) RCA is resolution-limited
due to lack of non-hydrostatic dynamics RCA was
developed from the HIRLAM NWP dynamical core
through a collaboration. Can this be repeated
on a larger European scale ?
37
HIRLAM NWP Vision for 2010 as of 2005
Best available 2.5 km meso-scale modelling
system operational in most of the HIRLAM
countries and applicable at 1-2 km Should
include Non-hydrostatic dynamics (non
approximated equations) Efficient dynamics
(long time steps possible) Advanced physics for
clouds, precipitation and turbulence (convection
mainly resolved) Transparent boundary
treatment Externalized and advanced
land-surface treatment. Synoptic 10km model
coupled to ECMWF for regional forecasting Of
equal quality to global model and tightly coupled
to ECMWF Consistent physics with meso-scale
model To provide coupling to meso-scale model
and other applications
38
Physical Processes in Atmospheric Models
AROME
Physics No Mans Land
RCA/HARMONY EC-Earth
HARMONY
Hydrostatic Approximated Dynamics
Full equations
Non-Hydrostatic
Klemp (2002) modified Jones (2008)
39
How (as asked in HIRLAM community 2003-2004) ?
Develop own meso-scale model? No Not enough
dynamics staff Large effort years of
development Duplication of work in Europe
New developments in assimilation and physics
also Still resources needed for synoptic
model as well Collaborate Yes Quickly
re-gain lost time of development Sharing
research resources with partner(s) Actively
contributing in many key areas Relying on
partner(s) for other things Enough critical
mass in all areas
40
Why ALADIN-AROME-Meteo-France (HARMONY) ?
HIRLAM and ALADIN are closely related
scientifically ECMWF code commonality in
ALADIN (EC-Earth compatibility) ALADIN
efficient non-hydrostatic dynamics Advanced
meso-scale physics through Meso-NH -gt
AROME Good experience of MF collaboration in
HIRLAM Likely to be a true collaboration from
both sides Many partners of different size
Yellow countries are HIRLAM Red countries are
Aladin/Arome Many of these countries are
EC-Earth consortium members also Harmony/Arome/A
ladin climate ? A similar European consortium
exists for NEMO ocean modeling
41
Many Environmental (ESM) processes require a very
accurate physical climate input before they can
be simulated correctly. It may be beneficial to
develop ESM approaches in high-resolution
models NWP and climate can then collaborate
in a unifed system again
24 hour simulated precipitation
Both for model development evaluation
From E. Ebert BMRC
42
Rossby Centre Modelling Plans Summary
RCA/RCAO (2008-2012) Increased resolution
towards 10km Extensive Regional Scenario
Generation IPCC AR5, SWECIA Europe and an
internationally coordinated global
downscaling Targeted developments (increased
resolution, ESM, new regions) EC-Earth
(International Swedish consortium) Coupled
GCM for initialized decadal prediction
(2008-2012) Global Earth System Model
(2009-2016) Harmony/Arome/Nemo Climate
(International consortium) Nested (Synoptic to
Mesoscale) RCM/R-ESM (2009-2016) For
process studies Scheme development Specific
application driven scenarios at mesoscale
resolutions
43
Collaboration on common tools is the key Across
SMHI With other Swedish groups
Universities, research institutes On a European
Level EC-Earth Harmony/Arome/Aladin
climate NEMO Internationally Arctic
R-ESM WCRP coordinated climate scenarios
(CMIP5 RMIP) Faster progress and more fun !