Geoscience Reference
In-Depth Information
Figure B11.1.2 Fit of the aggregated dead zone model to a tracer experiment for the Colorado River for
the Hance Rapids to Diamond reach (238.8 km).
associated dead zones. This might be due to exchanges into larger dead zones or perhaps
mixing with hyporheic zone waters in the bed and banks. It is worth noting that, as shown
by Reynolds et al. (1991), the lateral velocity shear at the boundary between a dead zone
and the main flow can induce locally efficient mixing. However, once solute or pollutant is
transferred into such a storage, mixing is much slower and the time scale of pollutant retention
much longer. This contributes to the heavy tails of observed tracer concentrations. Examples
of fitting a simple first-order ADZ model to tracer data from the Colorado River is shown in
Figure B11.1.2.
Both the ADE and ADZ models are difficult to use in predicting catchment scale transport
of tracers or solutes. This is because of the complexity of the flow pathways and the possibility
that at the soil profile and hillslope scale, mixing is still in its initial stages so that it is difficult
to justify the assumption of the ADE that mixing is beyond the mixing length (or at least to be
able to estimate the dispersion parameters required at a given scale). Despite this it has been
used in some practical applications; quite widely in the prediction of transport and dispersion
(most hydraulic routing packages have a transport and dispersion option that uses the mean
velocities calculated for the flow); and in groundwater transport (again, many groundwater
modelling packages have an option to predict dispersion based on the ADE, such as MT3D for
MODFLOW, see Appendix A).
The ADE model has also been applied at the catchment scale in an interesting way, by
simplifying the subsurface flow pathways into a series of one-dimensional stream tubes (e.g.
Simic and Destouni, 1999; Persson and Destouni, 2009). The steady mean velocities are then
calculated in each stream tube, taking account of the changing cross-sectional area of each
stream tube. The ADE is then solved for the one-dimensional transport along each stream tube.
The delivery of all the stream tubes to the channels in the catchment provides a prediction of
the complete transport on the hillslopes, given some specified sources of the solute of interest.
Biogeochemical reactions can be added to these calculations for non-conservative solutes.
Search WWH ::




Custom Search