Transcript Strategies for assessing natural variability

Strategies for
assessing natural variability
Hans von Storch
Institute for Coastal Research, GKSS Research Center
Geesthacht, Germany
Lund, 20.11.2006, ENSEMBLES assembly, RT2B meeting
The 300 hPa geopotential height fields in the Northern Hemisphere: the mean 1967-81
January field, the January 1971 field, which is closer to the mean field than most others,
and the January 1981 field, which deviates significantly from the mean field. Units: 10 m
Natural variability
• Global: Variability due to external
natural factors
• Regional: Variability inherited from
large-scale variability.
• Global AND regional: Stochastic
variability due to internal dynamical
processes
Variability in RCM simulations
• Inherited from large-scale structure
• But: IDPS - Intermittent divergence in
phase space (not a problem, when
spectral nudging or other forms of
large-scale constraints are applied).
„Natural uncertainty“
in empirical
downscaling
approaches.
- Is the variability,
best described by
the analog approach,
“natural” or a deficit
of the predictors?
- I guess, mostly: yes.
Because: large-scale
constrained RCMs do
not show this
uncertainty.
Where does the stochasticity
found in data come from?
• Observational data: irregular spatial coverage,
observational errors, limited observation time
span.
And natural unforced variability. Dynamical
“cause” for natural unforced variability as in
simulation models.
• Simulation data: internally generated by a very
large number of chaotic processes.
• Stochasticity as mathematical construct to allow
an efficient description of the simulated (and
observed) climate variability.
Noise or
deterministic
chaos?
Mathematical
construct of
randomness – an
adequate concept
for description of
features resulting
from the presence
of many chaotic
processes.
Determining the characteristics
of natural variability
• Re-analyses: limited time, internally consistent, mostly
homogeneous; may contain anthropogenic signals.
• Reconstructions based on instrumental data: available
only for few variables, possibly contaminated by
anthropogenic signals; sometimes inhomogeneous.
• Paleo-reconstructions: may have problems in
estimating variability on different time scales.
• Rare long instrumental records may be
useful to validate model-based estimates; recent data
may be contaminated by anthropogenic signals.
• Millennial global simulations – possibly augmented with
suitable (preferably) dynamical and empirical
downscaling.
Temporal development of
Ti(m,L) = Ti(m) – Ti-L(m)
divided by the standard
deviation (m,L) of the
considered reconstructed temp
record
for m=5 and L=20 (top), and
for m=30 and L=100 years.
The thresholds R = 2, 2.5 and 3
are given as dashed lines.
Gouirand et al., 2006, in press
Low-pass filtered (>30-year scales)
temperatures from the simulation (black),
from reconstructions based on proxy data
(grey) and instrumental data (dashed) for
April-August (a) and December-March (b).
The reconstruction in (a) is based on treeringwidth and densities from northern
Fennoscandia. The reconstruction in (b) is a
combination of documentary evidence for ice
break-up dates and instrumental observations
from Tallinn, Estonia. The instrumental data
are from Uppsala, southern Sweden. All series
are given as anomalies from their respective
long-term means.
Gouirand et al., 2006, in press
Scandinavian temperatures from the simulation during 1000-1990 and
observations during 1874-1996 in summer (JJA) (a-b) and winter (DJF) (c-d).
Black lines show variability at timescales longer than 10 years.
Grey lines show shorter timescales.
All data are shown as anomalies from the respective long-term means.
The CoastDat-effort at the Institute for
Coastal Research at GKSS (ICR@gkss)
 Long-term, high-resolution reconstuctions (50 years) of present and recent
developments of weather related phenomena in coastal regions as well as
scenarios of future developments (100 years)
 Northeast Atlantic and northern Europe
 “Standard” model systems (“frozen”)
 Assessment of changes in storms, ocean waves, storm surges, currents and
regional transport of anthropogenic substances.
 Data freely available.

www.coastdat.de