[strat_list] JCLIM paper: "Spatial weighting and iterative projection methods for EOFs"

[Mark Baldwin]

A revised version of Baldwin, Stephenson, and Jolliffe, which is  
under review at JCLIM,  is available via the link below. All comments  
are welcome.

http://www.nwra.com/resumes/baldwin/pubs/ 
Baldwin_Stephenson_Jolliffe_revised_JCLIM_2008.pdf

IDL code to calculate EOFs, using the method in the paper, is  
available at the link below. Given a data matrix, and weighting  
matrix, it calculates the number of EOFs specified.

http://www.nwra.com/resumes/baldwin/eofcalc.pro

I would appreciate any suggestions for improvement. I have another  
version that is specifically designed for large data matrices. If you  
run into memory limits, please email me for this version.

Cheers,

-Mark

--------------

Abstract:

   Often there is a need to consider spatial weighting in methods for  
finding spatial patterns in
climate data. In this paper we focus on techniques that maximize  
variance, such as Empirical
Orthogonal Functions (EOFs). We introduce a weighting matrix into a  
generalized framework
for dealing with spatial weighting. One basic principal in the design  
of the weighting matrix is
that the resulting spatial patterns are independent of the grid used  
to represent the data. A
weighting matrix can also be used for other purposes, such as to  
compensate for the neglect of
unrepresented sub-grid scale variance or, in the form of a  
prewhitening filter, to maximize the
signal-to-noise ratio of EOFs. Our methodology is applicable to other  
types of climate pattern
analysis, such as extended EOF analysis and Maximum Covariance  
Analysis. The increasing
availability of large datasets of 3-dimensional gridded variables  
(e.g., reanalysis products and
model output) raises special issues for data reduction methods such  
as EOFs. Fast, memory
efficient methods are required in order to extract leading EOFs from  
such large data sets. This
study proposes one such approach based on a simple iteration of  
successive projections of the
data onto time series and spatial maps. We also demonstrate that  
spatial weighting can be
combined with the iterative methods. Throughout the paper, we use  
multivariate statistics
notation that can be implemented as simple matrix commands in high- 
level computing
languages.

To subscribe or unsubscribe from strat_list, please visit http:// 
www.nwra.com/resumes/baldwin/howtojoin.php

--------------------------------------------
Mark P. Baldwin
Vice President & Senior Research Scientist

Northwest Research Associates
4118 148th Ave. NE
Redmond, WA 98052 USA

phone +1 425.785.8751
(If I am in Europe) +44 771.764.3622
(fax) +1 425.556.9099
mark at nwra.com

www.nwra.com/baldwin
--------------------------------------------



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.nwra.com/pipermail/strat_list/attachments/20080707/f6001abd/attachment.html