Two nonparametric procedures to test for trends in water-quality data (SEASKEN AND SEASRS) have been developed for the Statistical Analysis System* (SAS). The procedure SEASKEN tests for a monotonic trend in time by a modified form of Kendall's tau, the Seasonal Kendall test. The procedure SEASRS tests for a step trend between two different periods in a time series using a modified form of the Wilcoxon (Mann-Whitney) rank sum test, the Mann-WhitneyWilcoxon rank sum test for seasonal data. Examples are presented using the two procedures. The source code and user's guide for each of the two procedures are also presented.
Procedures for flow adjusting water-quality data by the SAS procedures REG and SYSREG and techniques for plotting water-quality data as a time series by the SAS procedure PLOT are presented.
Additionally, examples are presented to demonstrate the use of the U.S. Geological Survey procedures QWRETR, DVRETR, QWSAS, and DVINPUT to retrieve data from the Geological Survey WATSTORE system and make it available to SAS.
|Title||Nonparametric tests for trends in water-quality data using the statistical analysis system|
|Authors||Charles G. Crawford, James R. Slack, Robert M. Hirsch|
|Publication Subtype||USGS Numbered Series|
|Series Title||Open-File Report|
|Record Source||USGS Publications Warehouse|