Earth System Physics Seminar

[Pandora Malchose Pieri]

EARTH SYSTEM PHYSICS SEMINAR

Tuesday, 12 December 2006
15:00 hrs, MB ROOM 140

A REPRESENTATION OF CLIMATE DYNAMICS USING NETWORK THEORY

Giulia De Masi
Economics Department,
University of Ancona, Italy

Abstract
In recent years complex networks have been more and more applied for 
many applications of statistical physicists. Several systems in 
different fields can be described as a network.  In this study we apply 
network theory to the study of climatic system.
Preliminary results show how network theory could represent useful 
tools to study correlations of climate parameters. The basic idea is 
that climate network can be defined using the average geopotential 
height defined on a regular grid over the globe: the vertices of such 
network are the grid points, while links are represented by the 
correlations between different grid points. Following these rules a 
weighted complete graph can be built defining a link between two points 
when the correlation coefficients of Geopotential Height time series is 
greater than a threshold value. Other authors have shown that the 
properties of small worlds complex network emerge from this kind of 
approach. In this work we focus on the topology of climate networks, 
with a detailed study of architectural properties, focusing on 
robustness of results changing main parameters. We compare networks 
obtained from different atmospheric measures and different datasets. A 
tentative of study of climate trends using network theory is shown.