Statistical performance evaluation and inference on bounded and seasonally distributed water quality variables

Water quality variables behave differently in time and space. The concentrations of pollutant are dispersed and assimilated by streamflow quantity, and variations are driven by flow rate distribution. Time trends of deseasonalized concentrations showed marked differences in percentage variation due to trends and by seasonal series. Significant impacts on the system are due to point and diffuse sources that influenced trends. Laboratory analytical tests of sample detection were used to sort and bound data sets in providing reliable series for analysis. These series were used to derive statistical performance evaluators for specific water quality variables using reliability concepts. Valuable inference on recurrence of failures was deduced that added causes to be concern for the vulnerable performances of the system. Typical examples of assessment include nutrient, fecal coliform, metals, unionized ammonia and dissolved oxygen. It was shown that the independent series provided performance evaluators that are applicable for decision making in attending to in-stream pollutant control at industrial and urbanized areas