Mexico's summer rainfall patterns: an analysis of regional modes and changes in their teleconnectivity

P. J. ENGLEHART, A. V. DOUGLAS

Abstract

This study explores the interannual variability in Mexico's warm season rainfall. It is based on historical rainfall data (1927-1997) from a grid of 130 long-term stations. The grid provides reasonable spatial coverage over most of the country. Using Principal Components Analysis, this study delimits five relatively large sub-areas of the country. Within each of these regions, monthly station rainfall exhibits high levels of spatial coherence. The regional structures are distinct and physically plausible with respect to the climatology of Mexico. The regional time series are expressed as departures from normal. In most of these series the intraseasonal, that is, month to month, persistence is minimal; in fact, only a few of the series display significant tendencies toward nonrandom serial behavior. This study evaluates the teleconnectivity between the regional rainfall series and several different indices of large-scale ocean and atmosphere variability. These indices include an El Nine-Southern Oscillation index, indices that describe the strength and position of the subtropical anticyclone belt, as well as local SSTs in the Eastern Pacific and the Gulf of Mexico. All of the teleconnection analyses consider the quasi-decadal mode of variability that is commonly referred to as the Pacific Decadal Oscillation (PDO). The study results generally agree with the findings of other researchers to the extent that they indicate differential teleconnectivity as a function of PDO phase. However, unlike other studies, we see no compelling evidence for non-linear behavior in the teleconnections within PDO phase. Instead, the more prominent feature of the teleconnections is simply that they tend to be both more spatially extensive and stronger in the positive PDO phase.

 

Keywords

México; summer rainfall; mexican monsoon; principal component analysis; teleconnections

Full Text:

pdf