Riverine modeling

The state-or-the-art of chemical riverine modeling is captured in the applications ongoing with PCB in the Hudson River (4,5,6,). This site has received extraordinary study over several decades by federal and state authorities as well as the General Electric Company. The basic elements of the CFaT models presented here are derived from these sources using conventional nomenclature employed by practitioners in the field. 

Together equations 1 and 2 are the CFaT portion of the riverine model. It is dependent open a solids mass balance module for determining the suspended solids concentration plus the particle deposition and resuspension fluxes. The hydrodynamic, particle balance as well as the biota up-take modules are beyond the scope of this study. In addition to the noted coupling to other models, equations 1 and 2 are coupled at the sediment-water interface through the flux expressions. 

The quasi-steady state assumption. The riverine models described above are solved simultaneously in the transient case. A simpler approach will now be used. Although over long time-periods the accumulation process is necessary in both equations for shorter time-periods the quasi-steady-state (QSS) assumption applies (8). For a large bed-source mass the small chemical releases will not decrease it in the weeks to month time-period so that a steady-state conditions is closely approximated. For a stream of width w (L) of volume section whx and area A=wx (L ), Equation 1 with the particle process terms omitted can be transformed to... 

More recent riverine modeling approaches attempt to disconnect the particle and chemical data calibration steps. Specifically, chemical data is not used to adjust parameters for the solids mass balance module (6). What follows is a review of the approaches taken by recent CFaT modelers and the results obtained using Eqns. 3 and 4 for calibrating the release of soluble PCB fractions on the Hudson and Grasse Rivers. 

CFaT riverine models were presented for both the water column and bed sediment. They were then simplified to focus onto the non-flow resuspension soluble fraction using the quasi-steady state assumption to isolate the key water-side and sediment-side process elements. Field evidence of soluble release based on CFaT model derived data was reviewed for three rivers. Both the traditional particle background resuspension process and more recent soluble fraction process algorithms data interpretation were covered. Numerical field calibrated resuspension velocities and soluble mass-transfer coefficients were presented. Candidate water-side and sediment-side transport processes, selected from the literature were reviewed. Those that provided the best theoretical explanation and contained laboratory and/or field data support were selected. Finally, the flux and the overall transport coefficient which captures the essential features of the framework were presented. Following this the theoretical mass-transfer coefficients were applied to a site on the Fox River below De Pere Dam. Numerical calculations were made for the transport coefficients for both individual and combined processes. 

In a continual biocoenosis test (Riverine biocoenosis model) for LAS the biocoenotic NOEC value of >0.65 mg/L has been determined. The test was carried out with a starting concentration of 1.1 mg/L LAS/L over a period of 5 weeks. The NOEC value represents the effect caused by the original detergent and the effects caused by the catabolites as well [297]. The NOEC value shows that biocoenosis reacts more sensitively to compounds than the most sensitive laboratory species, which according to general opinion should not be generalized. 

Figure 4. A conceptual model of the ecological risk assessment of persistent organochlorinated compounds (POCs) contamination due to riverine inputs into the Caspian Sea.

This simple two component model for the Fe isotope composition of seawater does not consider the effects of the Fe isotope composition of dissolved Fe from rivers or from rain. Although the dissolved Fe fluxes are small (Fig. 19) the dissolved fluxes may have an important control on the overall Fe isotope composition of the oceans if they represent an Fe source that is preferentially added to the hydrogenous Fe budget that is ultimately sequestered into Fe-Mn nodules. In particular riverine components may be very important in the Pacific Ocean where a significant amount of Fe to the oceans can be delivered from rivers that drain oceanic islands (Sholkovitz et al. 1999). An additional uncertainty lies in how Fe from particulate matter is utilized in seawater. For example, does the solubilization of Fe from aerosol particles result in a significant Fe isotope fractionation, and does Fe speciation lead to Fe isotope fractionation ... 

As shown in Figure 23.7, the continuous lake model nicely describes the concentration maximum, which slowly moved to greater depth due to the deepening of the surface mixed layer. From the model calculation we can conclude that the processes involved in producing this maximum were the combination of riverine PCE input into the surface mixed layer and loss to the atmosphere by gas transfer. The extra input of PCE into the lake between May 6 and July 1, 1985 had to be about 360 moles. The model calculations suggest that the input had dropped to virtually zero after July 1. Part of the compound was quickly and continuously lost to the atmosphere so that the PCE content of the lake never increased much beyond 200 moles. 

In the case of nutrients,source components have been modeled as conservative mixtures (Cmix) of riverine and marine end-members using the following equation ... 

Figure 9.2 Freshwater-marine mixing models for hydrogen, oxygen, sulfate, and dissolved inorganic carbon (DIC) and nitrogen (DIN) across salinity gradients. Modeled gradients for sulfate are based on riverine-marine ratios of 60 pm versus 28 pm (data from Kendall and Coplen, 2001). (Modified from Fry, 2002.)...

Figure 36.8 Comparison of estimates of total riverine N export (Tg N year ) from the continental land masses to coastal ocean basins during the mid-1990 s, from modeling studies reported in recent literature.

Figure 24 Bulk partitioning coefficients between dissolved and adsorbed Be on particles measured in tbe Orinoco for Be (fiUed circles) and °Be (open circles). Open squares are experimental distribution coefficients determined by You et al. (1989) by adsorption of Be onto riverine particles. The line is deduced from a speciation-dependent model...

---