A general trend analysis methodology is developed for detecting and modelling trends in water quality time series measured in rivers and streams. The procedure is specifically designed for use with typically ill-behaved river quality series characterized by problematic features such as non-normal positively skewed populations, irregularly spaced instantaneous observations, seasonal periodicities, and dependence among water quality variables and riverflows. In order to analyze these "messy" environmental data in a systematic and rigorous fashion, the overall trend analysis approach is divided into the two main categories of graphical studies and trend tests. Within these two main steps, specific graphical, parametric and nonparametric statistical techniques are utilized. Graphical methods used in the procedure include time series plots, robust regression smooths, as well as box and whisker graphs. Nonparametric techniques include the Mann-Kendall and Kruskal-Wallis tests. Additionally, a test based on Spearman's partial rank correlation is introduced as a means for eliminating seasonal effects when testing for the presence of a trend. The efficacy of the trend analysis methodology is explained and demonstrated by applying it to water quality time series observed in the Saugeen and Grand Rivers of Southwestern Ontario, Can ad a.

No abstract

Constrained by computational feasibility, attempts to describe random natural phenomena• of complex origin analytically can lead to a multiplicity of simplistic potentially representative model forms, as has occurred in the case of flood frequency analysis. Classical statistical methods inadequately confront this model uncertainty. Likelihood and Bayesian methods are presented and shown to permit inference concerning the relative goodness of several potential model candidates with respect to a given set of flood events.. The Bayesian inferences are further combined within a decision theoretic structure for examiriation of the anticipated economic consequences of model uncertainty regarding decisions concerning flood protection levels. Results show that Bayesian methods supply more precise informa{ion butrequire greater effort. Since a model world of simplistic forms may never be defined absolutely, both decision and inference remain subject to the astute judgment of the analyst. the Bayesian approach to statistical inference has been developed to the stage at which the methods are capable of entertaining those uncertainties iritroduced by modelings. Work in the area of regression models [Box and Hill, 1969; Reilly, 1970; Hsiang and Reilly, 1971] has established Bayesian inference as a sound basis for discrimination among potential model candidates. In a hydrologic application, Wood et al. [1974] adopted Bayesian methods to ascertain the relative goodness of three flood frequency models that have generated a given flood sequence. Beyond inference the engineer is primarily concerned with decision, which in the case of flooding might involve selecting some optirrial design level of protection works from an array of possible alternatives. Generally, these decisions are based on some form of economic analysis conducted as a separate step following the inference. Such standard approaches are somewhat arbitrary and inefficient, sinc e they do not logically relate the hydrologic uncertainties to the economic variables of the decision problem. Davis et al. [1972], using a flood protection design problem as an example, successfully demonstrated that decision theory provides a rational framework within which the probabilistic and economic aspects of a decision problem may be jointly considered, a process leading to more effective design. Their study assumed only one model and centered on the uncertainties associated with the modei's parameters; however, Wood et al. [1974] applied the general-Copyright ¸ 1976 by the American Geophysical Union. ized decision theoretic formulation of Smallwood [1968] to treat an essentially identical problem, accomodating model uncertainty as well [see also Wood and Rodr[guez-lt•trbe, 1975a, b].The intent of this investigation is t 9 review once more the flood protection problem with a slightly different emphasis.First, flood frequency analysis is reexamined from a pdrely inferential perspective so tha t Bayesian method• of model discrimination may be developed by way of comparison with some o...

Aqueous concentration time series of the herbicide atrazine and its phytotoxic metabolite desethylatrazine obtained from 1981 to September 1990 in the Thames, Grand, and Saugeen rivers draining agricultural southwestern Ontario were analyzed to ascertain temporal trend, seasonal variation, and mass‐discharge patterns Substantial year‐to‐year changes were evident. Data suggest that the greatest applications of the decade might have occurred in 1984; however, in 1985 significant decreases in levels were observed at all three sites, apparently due to an abrupt shift away from corn cultivation. Mass‐discharge estimates generally parallel the concentration trends but also reflect hydrometeorologic trends. In 1984 for the Thames basin, extremely high concentrations together with high flows combined to export about 13 tonnes (t) atrazine plus metabolite, the largest single annual loss to surface‐water systems observed over the decade. Driven by the annual spring‐summer application cycle, aqueous concentrations surge in late May to reach annual peaks in late June and early July Despite the declining levels of recent years, Thames River data suggest that the probability of exceeding the Canadian water quality guideline of 2 μg/L for protection of aquatic life, primarily algae and vascular plants, may yet approach 65%. The greatest risks to susceptible species are likely associated with large summer storm runoff in areas of high soil to surface water transfer potential. Herbicide usage in general continues in large quantities, and continued surveillance appears warranted. Further research is required to ascertain the ultimate fate of these compounds in end receiving waters such as the Great Lakes