Please use this identifier to cite or link to this item:
Scopus Web of Science® Altmetric
Type: Conference paper
Title: Constrained spatial clustering of climate variables for geostatistical reconstruction of optimal time series and spatial fields
Author: Dowd, P.
Wang, H.
Pardo-Igúzquiza, E.
Yang, Y.
Citation: Geostatistics Valencia 2016, 2017 / Gómez-Hernández, J., Rodrigo-Ilarri, J., Rodrigo-Clavero, M., Cassiraga, E., Vargas-Guzmán, J. (ed./s), vol.19, pp.879-891
Publisher: Springer
Publisher Place: Switzerland
Issue Date: 2017
Series/Report no.: Quantitative Geology and Geostatistics
ISBN: 978-3-319-46818-1
ISSN: 0924-1973
Conference Name: The 10th International Geostatistics Congress (5 Sep 2016 - 9 Sep 2016 : Valencia, Spain)
Editor: Gómez-Hernández, J.
Rodrigo-Ilarri, J.
Rodrigo-Clavero, M.
Cassiraga, E.
Vargas-Guzmán, J.
Statement of
Peter Dowd, Hong Wang, Eulogio Pardo-Igúzquiza, and Yongguo Yang
Abstract: The purpose of this work is to present a new methodology for identifying geographical regions within which the climatic behaviour of a meteorological variable is coherent. We have chosen temperature as the variable of interest, and thermal coherence is defined here in the sense of having a strong (negative) correlation between terrain altitude and temperature. An improved method of constrained spatial cluster analysis is described in the form of a new constrained clustering algorithm. The methodology includes spatial bootstrap statistical tests to provide a more realistic measure of the uncertainty of the coefficient of correlation together with a spatial test of the correlation of residuals. The results are used as optimal estimates of areal temperature averages. The methodology is illustrated by applying it to the annual mean temperature measured at 1220 temperature stations across Spain.
Rights: © Springer International Publishing AG 2017
DOI: 10.1007/978-3-319-46819-8_61
Appears in Collections:Aurora harvest 8
Civil and Environmental Engineering publications

Files in This Item:
There are no files associated with this item.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.