A STATISTICAL METHODOLOGY FOR EXPLORING ELEVATIONAL DIFFERENCES IN PRECIPITATION CHEMISTRY

被引:3
作者
WARREN, WG
BOHM, M
LINK, D
机构
[1] OREGON STATE UNIV,DEPT FOREST MANAGEMENT,CORVALLIS,OR 97333
[2] US EPA,COBVALLIS ENVIRONM RES LAB,WESTERN CONIFERS RES COOPERAT,NSI TECHNOL SERV CORP,CORVALLIS,OR 97333
[3] US EPA,DENVER,CO 80202
来源
ATMOSPHERIC ENVIRONMENT PART A-GENERAL TOPICS | 1992年 / 26卷 / 01期
关键词
ELEVATION EFFECTS; SULFATE; LIKELIHOOD RATIO TESTS;
D O I
10.1016/0960-1686(92)90268-P
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
A statistical methodology for exploring the relationships between elevation and precipitation chemistry is outlined and illustrated. The methodology is only applicable to situations where the precipitation at two (or more) matched sites is correlated. Maximum likelihood estimates and likelihood tests are utilized, with contour ellipses of assumed bivariate log-normal distributions to assist in the interpretation. Data from 12 sites located in the southern Rocky Mountains of the U.S. were used for illustration. The results indicate differences in sulphate concentrations between airsheds, between snow and rain, and between higher and lower elevations in the Rocky Mountains. There are other approaches for investigating these issues, however, the likelihood ratio with contour ellipses is easy to apply and interpret, and we feel that it provides a greater insight than some of the more common analyses of variance or regression techniques. However, it is important to note that the superiority of one method over others depends on the criterion used.
引用
收藏
页码:159 / 169
页数:11
相关论文
共 17 条