Showing 1–2 of 2 results for author: Selten, F M

Decomposing data sets into skewness modes

Authors: Rubén A. Pasmanter, Frank M. Selten

Abstract: We derive the nonlinear equations satisfied by the coefficients of linear combinations that maximize their skewness when their variance is constrained to take a specific value. In order to numerically solve these nonlinear equations we develop a gradient-type flow that preserves the constraint. In combination with the Karhunen-Loève decomposition this leads to a set of orthogonal modes with maxi… ▽ More We derive the nonlinear equations satisfied by the coefficients of linear combinations that maximize their skewness when their variance is constrained to take a specific value. In order to numerically solve these nonlinear equations we develop a gradient-type flow that preserves the constraint. In combination with the Karhunen-Loève decomposition this leads to a set of orthogonal modes with maximal skewness. For illustration purposes we apply these techniques to atmospheric data; in this case the maximal-skewness modes correspond to strongly localized atmospheric flows. We show how these ideas can be extended, for example to maximal-flatness modes. △ Less

Submitted 24 August, 2009; originally announced August 2009.

Comments: Submitted for publication, 12 pages, 4 figures

Journal ref: Physica D, 239 (2010), 1503-1508

Extreme Associated Functions: Optimally Linking Local Extremes to Large-scale Atmospheric Circulation Structures

Authors: Debabrata Panja, Frank M. Selten

Abstract: We present a new statistical method to optimally link local weather extremes to large-scale atmospheric circulation structures. The method is illustrated using July-August daily mean temperature at 2m height (T2m) time-series over the Netherlands and 500 hPa geopotential height (Z500) time-series over the Euroatlantic region of the ECMWF reanalysis dataset (ERA40). The method identifies patterns… ▽ More We present a new statistical method to optimally link local weather extremes to large-scale atmospheric circulation structures. The method is illustrated using July-August daily mean temperature at 2m height (T2m) time-series over the Netherlands and 500 hPa geopotential height (Z500) time-series over the Euroatlantic region of the ECMWF reanalysis dataset (ERA40). The method identifies patterns in the Z500 time-series that optimally describe, in a precise mathematical sense, the relationship with local warm extremes in the Netherlands. Two patterns are identified; the most important one corresponds to a blocking high pressure system leading to subsidence and calm, dry and sunny conditions over the Netherlands. The second one corresponds to a rare, easterly flow regime bringing warm, dry air into the region. The patterns are robust; they are also identified in shorter subsamples of the total dataset. The method is generally applicable and might prove useful in evaluating the performance of climate models in simulating local weather extremes. △ Less

Submitted 25 September, 2007; v1 submitted 22 June, 2007; originally announced June 2007.

Comments: 10 pages, 7 figures, 14 eps figure files; to appear in J. Atmos. Chem. Phys

Journal ref: Atmos. Chem. Phys. Discuss. 7, 14433-14460 (2007)