Revista Española de Salud Pública
Print version ISSN 1135-5727
ROLDAN GARCIA, Esther et al. Determination of isothermal areas and selection of representative weather stations in Aragon as a basis for estimating the impact of climate change on the possible relationship between mortality and temperature. Rev. Esp. Salud Publica [online]. 2011, vol.85, n.6, pp.603-610. ISSN 1135-5727. http://dx.doi.org/10.1590/S1135-57272011000600009.
Background: In extensive and diversified regions, such as Aragon, it is believed the need to divide them into areas in terms of the available atmospheric variables with a view to select a representative weather station. The objective of this study was to determine the existence of isothermal regions and select representative stations for Aragon in order to carry out further study on the correlation between variables of temperature and daily mortality. Methods: Daily data on maximum and minimum temperature for the period between January 1987 and December 2006 was selected. In order to determine the isothermal areas a cluster analysis and a discriminate factor analysis were carried out along with a data pretreatment of filled gaps and detection of inhomogeneities in the climatic series. We analyzed data from 93 stations (44 in Huesca, 15 in Teruel and 34 in Zaragoza). Results: The results of the analysis for the regionalization of Aragon lead us to conclude that a unique factor explains the variance of each series; at high temperatures one factor explains 93.43% of the variance and the station with the highest correlation factor is Monflorite-Huesca (correlation = 0.984). At low temperatures one factor explains 90.88% of the variance, with Monegros-Pallaruelo being the station that presents the greatest correlation factor (correlation = 0.976). Conclusions: It was felt that Aragon was a unique isothermal region with one unique representative station of the temperature variability, Zaragoza-Airport with a correlation of 0.980 in maximum temperatures and 0.974 minimum.
Keywords : Temperature; Mortality; Climate change; Time series studies; Cluster analysis.