The main goal of our work is to define the dynamic through
which pollutants from inundated floodplain water bodies are transported
from their original location (i.e., the contaminated water
bodies) along the river to the downstream drinking water intake.
An accurate description of this phenomenon, which accounts for
density stratification effects related to concentration inhomogeneities
and concentration variations associated with density,
can only be obtained using three-dimensional models (see,
Lyubimova et al., 2014). However, even for modern computers it
is impossible to simulate chemical species transport by turbulent
flows using 3D models that take into account density stratification
effects for a sufficiently large water bodies since such simulations
require a fine mesh. Additionally, to implement 3D models, we need
to set the boundary conditions and, due to the limitations of watercourse
observation system, such information can only be obtained
by performing calculations based on models of a lower order that
require much less computational resources. We demonstrate, however,
that 2D models applied to a defined pollution source, and evaluated
on the basis of 3D calculations, allow detailed information to
be obtained on (1) the time needed for contamination to reach the
drinking water intake, and (2) the duration of the increased concentration
of pollutants in water at the drinking water intake. One dimensional models allow us to calculate only characteristics averaged
over a cross-section and do not allow for estimates of the role
of inundated floodplain in the formation of pollution fields. At the
same time, 1D models require a minimum of computational
resources and can be successfully built and implemented for very
large sections of a river system.