Step C. Temporal evolution of contamination and normalisation. The contaminant contents expressed on a dry weight basis (μg g
À1 dw) are presented on a
temporal scale. If the core slices are relatively thick, this evolution is presented as
a succession of average contents in time slices. At this stage, there are two different
options to interpret these profiles: (1) in the geochemical approach for metals,
enrichment factors (EF) represent the excess contamination, which are the contents
at time t doubly normalised to the background contents determined in natural
conditions and to a tracer of the finer fraction. A variation of EFs is the Igeo index
[23], which has a geometric progression to account for the increase of EFs, up to two
orders of magnitude; (2) the regulatory approach is based on environmental quality
standards for sediment quality or predicted no-effect concentrations [24]. Before the
1990s, these criteria were not available in France.
Once dated and normalised, the results can be used to compare how various
pollutants have evolved at a given station (Step D), to compare stations within a
given basin (Step E) and finally to compare basins (Step F). The understanding of
these pollution trajectories needs historical socio-economic data (Steps G–K), as will
be discussed in the following sections.
2.2 Contamination Fluxes Estimated from Floodplain Cores
In this approach, the average composition of the core at a given period is multiplied
by the sediment flux for the same period to generate the contaminant flux. Then
excess river fluxes are calculated [25].
This approach has several limitations, however. The extension of core analysis to
river contaminant fluxes is based on the following hypothesis: (H 1 ) these deposits
represent the flux-averaged quality of the particles carried by the river; (H 2 ) once
deposited (without hiatus, which makes the age model uncertain), these particles are
not (bio)turbated; and (H 3 ) the targeted compounds do not decompose (H 4 ) or
migrate in the profiles. This set of assumptions is applied only over a long period,
i.e. decades and more, considering that the contaminant fluxes are first controlled by
the evolution of the chemical composition of river suspended particulate matter
(SPM) much more than sediment fluxes. Note that H 1 is discussed for metals and
radionuclide trends in [26]. The interpretation should also take into account the
possible retention of contaminated particulates in the river basin [27], as in reservoirs, which could result in a decreased leakage not connected to environmental
measures (e.g. reduction of point pollution sources). In some basins, the sediment
loads can be reduced up to 95% as a consequence of large dam or reservoir
cascades [28].
In the Seine River, some of these hypotheses have been checked. Floodplain
particles are generally deposited from the peak discharge throughout the receding
274
S. Ayrault et al.
À1 dw) are presented on a
temporal scale. If the core slices are relatively thick, this evolution is presented as
a succession of average contents in time slices. At this stage, there are two different
options to interpret these profiles: (1) in the geochemical approach for metals,
enrichment factors (EF) represent the excess contamination, which are the contents
at time t doubly normalised to the background contents determined in natural
conditions and to a tracer of the finer fraction. A variation of EFs is the Igeo index
[23], which has a geometric progression to account for the increase of EFs, up to two
orders of magnitude; (2) the regulatory approach is based on environmental quality
standards for sediment quality or predicted no-effect concentrations [24]. Before the
1990s, these criteria were not available in France.
Once dated and normalised, the results can be used to compare how various
pollutants have evolved at a given station (Step D), to compare stations within a
given basin (Step E) and finally to compare basins (Step F). The understanding of
these pollution trajectories needs historical socio-economic data (Steps G–K), as will
be discussed in the following sections.
2.2 Contamination Fluxes Estimated from Floodplain Cores
In this approach, the average composition of the core at a given period is multiplied
by the sediment flux for the same period to generate the contaminant flux. Then
excess river fluxes are calculated [25].
This approach has several limitations, however. The extension of core analysis to
river contaminant fluxes is based on the following hypothesis: (H 1 ) these deposits
represent the flux-averaged quality of the particles carried by the river; (H 2 ) once
deposited (without hiatus, which makes the age model uncertain), these particles are
not (bio)turbated; and (H 3 ) the targeted compounds do not decompose (H 4 ) or
migrate in the profiles. This set of assumptions is applied only over a long period,
i.e. decades and more, considering that the contaminant fluxes are first controlled by
the evolution of the chemical composition of river suspended particulate matter
(SPM) much more than sediment fluxes. Note that H 1 is discussed for metals and
radionuclide trends in [26]. The interpretation should also take into account the
possible retention of contaminated particulates in the river basin [27], as in reservoirs, which could result in a decreased leakage not connected to environmental
measures (e.g. reduction of point pollution sources). In some basins, the sediment
loads can be reduced up to 95% as a consequence of large dam or reservoir
cascades [28].
In the Seine River, some of these hypotheses have been checked. Floodplain
particles are generally deposited from the peak discharge throughout the receding
274
S. Ayrault et al.