TOWARD UNDERSTANDING THE CAUSES OF LOW-FREQUENCY VARIABILITY - THE INTERANNUAL STANDARD-DEVIATION OF MONTHLY MEAN 700-MB HEIGHT

Citation
Ag. Barnston et Hm. Vandendool, TOWARD UNDERSTANDING THE CAUSES OF LOW-FREQUENCY VARIABILITY - THE INTERANNUAL STANDARD-DEVIATION OF MONTHLY MEAN 700-MB HEIGHT, Journal of climate, 6(11), 1993, pp. 2083-2102
Citations number
41
Categorie Soggetti
Metereology & Atmospheric Sciences
Journal title
ISSN journal
08948755
Volume
6
Issue
11
Year of publication
1993
Pages
2083 - 2102
Database
ISI
SICI code
0894-8755(1993)6:11<2083:TUTCOL>2.0.ZU;2-S
Abstract
The field of standard deviation of monthly mean 700-mb geopotential he ight in the Northem Hemisphere for each of the 12 months over the 1950 -1991 period, among other auxiliary statistics, is compiled in an atla s to which this paper is companion. Some of the major features found i n the atlas are highlighted and extended here. A comparison is also ma de to the same statistics derived from a 10-year run of the NMC model. There are three distinct regions of peak standard deviation (up to 85 geopotential meters in winter), all of which are located over water. Two of them remain positionally relatively stationary throughout the y ear in the high-latitude Pacific and Atlantic oceans, respectively. A portion of the Pacific region's winter variability comes from interdec adal fluctuations. The third region is over the Arctic Ocean and exhib its some large seasonal changes in location. A roughly north-to-south troughlike minimum in standard deviation (down to less than 20 geopote ntial meters in summer) is found in west central North America through out most of the year. The standard deviation maxima (minima) coincide largely with areas with a high (low) frequency of occurrence of height anomaly centers of both signs. Many of these anomaly centers occur in spatial coherence with other centers, forming familiar teleconnection and principal component patterns. While the high (low) standard devia tion areas invest greater (lesser) amounts of variance in these cohere nt variability clusters than the surrounding regions, their involvemen t in terms of the strength of the relationships is not substantially g reater (smaller). The standard deviation field does not move north and south with the changes in season as do the jets, storm tracks, and th e mean flow. In summer the standard deviation peaks are largely detach ed from spatially coherent variability patterns, suggesting that they may be caused in large part by local interactions related to permanent (spatially fixed) features of the lower boundary at all times of the year. The observed monthly mean 700-mb flow and the quasi-stationary l ocations of its interannual standard deviation maxima and minima are r eproduced in approximate form in a 10-year run of the NMC medium-range forecast model. This helps provide evidence that the field of standar d deviation is related, directly or indirectly, to some of the geograp hically fixed boundary conditions across the globe such as SST, ocean- land interfaces, and terrain.