Annual cycle in river water quality: A time series approach

Abstract

Regression, spectral, and cross‐spectral techniques are used to examine the trends and periodicity in a 20‐year record of weekly observations of eight water quality variables in the River Stour, eastern England. Linear regression shows that all except p H. and carbonate hardness have significantly increased over time. A strong annual cycle is present in the data, and cross‐spectral analysis is used to model the relationship between the variables and the water discharge in time. Nitrate and noncarbonate hardness are positively correlated with discharge, whereas carbonate hardness is an inverse function.