FINDING CAUSES OF OUTLIERS IN MULTIVARIATE ENVIRONMENTAL DATA     

FOREST C.GARNER MARTIN A.STAPANIAN KIRK E.FITZGERALD Lockheed Engineering  Sciences Co.  E.Flamingo R  Las Vegas  NV  U.S.A. 《地理学报(英文版)》1991,(3)

Multivariate outliers in environmental data sets are often caused by atypical measurement error in a singlevariable.From a quality assurance perspective it is important to identify these variables efficiently so thatcorrective actions may be performed.We demonstrate a procedure for using two multivariate tests toidentify which variable‘caused’each outlier.The procedure is tested with simulated data sets that havethe same correlation structure as selected water chemistry variables from a survey of lakes in the WesternUnited States.The success rates are evaluated for three of the variables for sample sizes of 50 and 100,significance levels of 0.01 and 0.05 and various amounts of mean shift.The procedure works best forhighly correlated variables.  相似文献   

Anthropogenic Sulfate and Nitrate Signals in Snow from Bogda Glacier, Eastern Tianshan     

张晓宇  Ross Edwards 《中国地质大学学报(英文版)》2011,(4)

A 1.2 m snow pit was recovered on July 29th, 2009 from the Bogda Glacier, eastern Tianshan (天山). The sample site temperature of -9.6 ℃ indicates that the unique glaciochemical re-cord was well preserved and suitable for the reconstruction of air pollution levels in this previously un-explored region. Samples were analyzed for major ions (Na+, K+, Ca2+, Mg2+, NH4+, Cl-, SO42-, NO3-, HCOO-, and CH3COO-). NO3- and SO42- were characterized by significant high levels of pollution con-centration. Most air masses ...  相似文献   

ON CONFIDENCE INTERVALS FOR THE PRODUCT OF THREE MEANS OF THREE NORMALEY DISTRIBUTED POPULATIONS     

E.A.YFANTIS Computer Science Department  University of Neva  Las Vegas  NV  U.S.A.G.T.FLATMAN U.S.Environmental Protection Agency  E.Harmon  Las Vegas  NV  U.S.A. 《地理学报(英文版)》1991,(3)

In many environmental applications,such as exposure assessment and risk modelling,the desiredestimate is a random variable computed as the product of three independently distributed randomvariables.These variables may not necessarily have the same mean and variance.The method for findingthe 100(1-α)% confidence interval for the mean of the product random variable has been proposed bysome practitioners as the product of the 100(1-α)% confidence interval of the three means.In this paperwe show that the distribution of the product of three independent normal random variables is not normal.We find the mean and variance of the product distribution.Further,we show that although the meanof the product is equal to the product of the means,the product of the three confidence intervals is nota good approximation of the confidence intervals for the mean of the product variable.The confidenceinterval of the mean of the product variable may be estimated by computer simulation.An algorithmfor estimating the confidence interval for the mean of the product random variable is given.The programimplementing this algorithm is given as an appendix.  相似文献