Revisiting climate region definitions via clustering

Robert Lund, Bo Li

Research output: Contribution to journalArticle

Abstract

This paper introduces a new distance metric that enables the clustering of general climatic time series. Clustering methods have been frequently used to partition a domain of interest into distinct climatic zones. However, previous techniques have neglected the time series (autocorrelation) component and have also handled seasonal features in a suboptimal way. The distance proposed here incorporates the seasonal mean and autocorrelation structures of the series in a natural way; moreover, trends and covariate effects can be considered. As an important by-product, the methods can be used to statistically assess whether two stations can serve as reference stations for one another. The methods are illustrated by partitioning 292 weather stations within the state of Colorado into six different zones.

Original languageEnglish (US)
Pages (from-to)1787-1800
Number of pages14
JournalJournal of Climate
Volume22
Issue number7
DOIs
StatePublished - May 11 2009

Fingerprint

autocorrelation
climate
time series
weather station
partitioning
method
station
trend
effect
climatic zone
by-product

ASJC Scopus subject areas

  • Atmospheric Science

Cite this

Revisiting climate region definitions via clustering. / Lund, Robert; Li, Bo.

In: Journal of Climate, Vol. 22, No. 7, 11.05.2009, p. 1787-1800.

Research output: Contribution to journalArticle

Lund, Robert ; Li, Bo. / Revisiting climate region definitions via clustering. In: Journal of Climate. 2009 ; Vol. 22, No. 7. pp. 1787-1800.
@article{063df2bd147b47ef8fa91b61b3cebab6,
title = "Revisiting climate region definitions via clustering",
abstract = "This paper introduces a new distance metric that enables the clustering of general climatic time series. Clustering methods have been frequently used to partition a domain of interest into distinct climatic zones. However, previous techniques have neglected the time series (autocorrelation) component and have also handled seasonal features in a suboptimal way. The distance proposed here incorporates the seasonal mean and autocorrelation structures of the series in a natural way; moreover, trends and covariate effects can be considered. As an important by-product, the methods can be used to statistically assess whether two stations can serve as reference stations for one another. The methods are illustrated by partitioning 292 weather stations within the state of Colorado into six different zones.",
author = "Robert Lund and Bo Li",
year = "2009",
month = "5",
day = "11",
doi = "10.1175/2008JCLI2455.1",
language = "English (US)",
volume = "22",
pages = "1787--1800",
journal = "Journal of Climate",
issn = "0894-8755",
publisher = "American Meteorological Society",
number = "7",

}

TY - JOUR

T1 - Revisiting climate region definitions via clustering

AU - Lund, Robert

AU - Li, Bo

PY - 2009/5/11

Y1 - 2009/5/11

N2 - This paper introduces a new distance metric that enables the clustering of general climatic time series. Clustering methods have been frequently used to partition a domain of interest into distinct climatic zones. However, previous techniques have neglected the time series (autocorrelation) component and have also handled seasonal features in a suboptimal way. The distance proposed here incorporates the seasonal mean and autocorrelation structures of the series in a natural way; moreover, trends and covariate effects can be considered. As an important by-product, the methods can be used to statistically assess whether two stations can serve as reference stations for one another. The methods are illustrated by partitioning 292 weather stations within the state of Colorado into six different zones.

AB - This paper introduces a new distance metric that enables the clustering of general climatic time series. Clustering methods have been frequently used to partition a domain of interest into distinct climatic zones. However, previous techniques have neglected the time series (autocorrelation) component and have also handled seasonal features in a suboptimal way. The distance proposed here incorporates the seasonal mean and autocorrelation structures of the series in a natural way; moreover, trends and covariate effects can be considered. As an important by-product, the methods can be used to statistically assess whether two stations can serve as reference stations for one another. The methods are illustrated by partitioning 292 weather stations within the state of Colorado into six different zones.

UR - http://www.scopus.com/inward/record.url?scp=65449166720&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=65449166720&partnerID=8YFLogxK

U2 - 10.1175/2008JCLI2455.1

DO - 10.1175/2008JCLI2455.1

M3 - Article

VL - 22

SP - 1787

EP - 1800

JO - Journal of Climate

JF - Journal of Climate

SN - 0894-8755

IS - 7

ER -