This session is positioned at the interface between statistics and climatology. Climate time series analysis takes into account the uncertainties in the climate system, measurements, proxy quality and dating. It provides confidence intervals and error bars of estimates, or it gives p-values of hypothesis tests. The challenge is to develop methods and obtain results that are robust and reliable also in the presence of following difficulties:

(1) non-normal distributions,
(2) autocorrelation,
(3) uneven time spacing,
(4) uncertain timescales.

We welcome novel contributions that address those difficulties, show a sound statistical basis of the methodology and demonstrate the feasibility of the method in real-world climatological applications. Methods may include, but are not restricted to:

(1) regression,
(2) correlation,
(3) spectral analysis,
(4) wavelets,
(5) nonlinear dynamical system analysis
and
(6) extreme value analysis.

Data may come from:

(1) instrumental observations,
(2) proxy measurements,
(3) historical documents
and
(4) climate models.

Confirmed contributions:

1. Anthony Davison (Univ. Lausanne, Dept. Statistics, Switzerland)
Nonstationary risk analysis of climate extremes
[tentative title]

2. Aslak Grinsted (Niels Bohr Institute, University of Copenhagen, Denmark)
Exploring linkages between time series: Wavelet coherence and phase-aware teleconnections

3. Jürgen Kurths (Potsdam Institute for Climate Impact Research, Germany)
Nonlinear dynamical system theory applied to climate time series
[tentative title]

4. Jean-Pierre Urbain (Univ. Maastricht, Dept. Economics, Netherlands)
Nonlinear regression with bootstrap confidence intervals applied to climate records
[tentative title]