Geoscience Reference
In-Depth Information
Some Observations on the Similarity of Tracer
Data from a Small River
S.G. Wallis and J.R. Manson
1
Introduction
The water industry's understanding of pollutant transport in rivers is aided by
undertaking tracer experiments. These usually consist of the discrete release of
tracer followed by the measurement of tracer concentration-time profiles at several
locations downstream of the tracer release site. The commonest analysis of such
data uses the method of moments to estimate the mean flow velocity (
U
) and the
dispersion coefficient (
D
) for the hydraulic conditions pertaining at the time of the
experiment. By undertaking experiments over a wide range of flow rates, the results
can be used to predict
U
and
D
at any flow rate by a simple regression analysis. Such
predictions of
U
and
D
are a necessary requirement for implementing numerical
solutions of the advection-dispersion equation, which are often the cornerstone of
water quality modelling software. For several decades it has been realised that the
method of moments is unreliable when used with incomplete concentration-time
profiles, and the method has become out of favour because it appears to require
extremely good-quality tracer data to avoid its inherent weakness. However, the
aim of this chapter is to show that such pessimism is misplaced. To achieve this,
some insights are provided from a cumulative analysis applied to non-dimensional
tracer concentration-time profiles derived from a recent series of tracer experi-
ments. Furthermore, by exploiting the fact that the non-dimensional profiles have
very similar shapes, a pooling of information from only a few high-quality tracer
experiments enables reliable estimates of
U
and
D
to be obtained from poor-quality
tracer experiments.
