The tenth IMSC, Beijing, China, 2007 - International Meetings on ...

Joint Program on the Science and Policy of Global Change,Massachusetts Institute of Technology,
Cambridge, MA 02139, USA
Peter H. Stone
Joint Program on the Science and Policy of Global Change,Massachusetts Institute of Technology,
Cambridge, MA 02139, USA
Two major challenges in estimating climate system properties (e.g., climate sensitivity
and rate of deep-ocean heat uptake) from climate observations include: (a) the uncertainty in
the historical climate forcings and climate observations, and (b) the uncertainty in the unforced
variability of the climate system. Each of these poses a major challenge due to both the limited
data available from either observations or climate model output and the required accuracy in
the statistical algorithms. To address these issues, we have implemented a new estimation
algorithm based on Bayesian methods to estimate the probability density functions for climates
system properties. This is similar to model calibration algorithms in the statistics literature
although here we use multi-variate patterns rather than scalar diagnostics to estimate the
likelihood functions. We use a Bayesian approach that allows for the use of scientifically based
information on the climate system properties to be used in the calibration process. <strong>The</strong>
statistical model tackles the problem of dealing with multivariate diagnostics and incorporates
all estimation uncertainties into the posterior distributions of the climate system properties.
Additionally we obtain estimates of the covariance structure of the unforced variability of
temperature change patterns. <strong>The</strong>se results are critical for understanding uncertainty in future
climate change and provide an independent check that the information contained in recent
climate change is robust to statistical treatment. <strong>The</strong>se results include uncertainties in the
estimation of the multivariate covariance matrices for the first time.
Quantification of uncertainty in global temperature projections over the twenty-first century: A
synthesis of multiple models and methods
Speaker: Reto Knutti
Reto Knutti
Institute for Atmospheric and Climate Science, ETH Zurich, Switzerland
reto.knutti@env.ethz.ch
<strong>The</strong> quantification of the uncertainties in future climate projections is crucial for the
implementation of climate policies. Here I provide projections of global temperature change
over the twenty-first century for the six illustrative, non-intervention SRES emission scenarios
based on the latest generation of coupled general circulation climate models, and assess
uncertainty ranges and probabilistic projections from various published methods and models.
Short-term trends in global temperature are comparably well constrained and similar across all
3