Seasonal Trend Analysis of Monthly Water Quality Data

This paper presents user-friendly SAS macro and SAS programs to check the three parts of a time series: Trend, Cyclical patterns, and Stationarity. The SAS macro KENDALL performs the trend analysis. Seasonal Kendall Trend analysis, including summery statistics; overall Tau and the P-value of the test for trend; monthly Tau values, the corresponding P-values for each month; the Seasonal Kendall estimate with confidence intervals; 3 exploratory plots by months comparative boxplot, comparative histogram, and trend plot against year. Spectral analysis program calculates Periodogram ordinates and Spectral density estimates, Periodogram plots against frequency, plots of Spectral density estimates against frequency and period, 2 tests for White Noise. Dickey-Fuller program performs a test for Stationarity using the test of the unit-root hypothesis. An example is given using l water-quality data from Truckee River, Nevada. The water-quality variable being used is total nitrogen (NTO, mg/l). (key words: Seasonal Kendall Trend Analysis, Spectral Analysis, Stationarity) INTRODUCTION The issue of surface-water quality is of vital importance. Various pollution sources related to natural-resource industries, primarily agriculture and mining, are having a strong effect on rivers and streams. Therefore, trend analysis of water quality fluctuations is essential for short and long term policy making. The components of time series trend are, Time Series Stochastic Models