The Vision/Objectives


1. Characterization of emission source and load through modeling and monitoring

 
Sources of contaminants and nutrients have to be identified, quantified and ranked  according to their impact in order to take measures that improve the ecological status. These sources vary locally in their importance, depending on hydrological and land use characteristics of the catchment. Diffuse pollution sources need to be modeled with input data or qualified assumptions and should be validated by targeted monitoring campaigns taking into account the transport dynamics of the pollutant. 


2. Uncertainty assessment in river basin mass flow modeling and monitoring
 
Emission models rely heavily on spatial information (GIS-data) and pollution data that are aggregated from limited measurement sets or that have been derived from the literature. All these input parameters carry uncertainties that propagate through the models and will yield some variance in the predicted mass flows. These uncertainties need to be quantified to reach the necessary confidence in the model predictions for decision support. Action 8 will implement recently developed procedures to assign uncertainties to all input data in a hierarchical way and will generate data realizations for Monte Carlo simulations in the emission models. Monitoring data from rivers and the derived mass flows for pollutants are also prone to errors, which are often due to non-representative or ill-weighted sampling event distributions. Interpolation or extrapolation procedures (with discharge or other regressors) therefore introduce errors in mass flow calculations at river gauges. As these data are used to validate the emission model predictions their reliability also needs to be assessed. 


3. Best practice guideline for monitoring design
 
If programs of measures are to be implemented to improve the ecological status, the base scenario needs to be determined by a reference situation characterization and post-measure effects need to be assessed by further monitoring. The data needs to characterize reference and post-measure situation vary according to the dynamics of the pollutant transport. Steadily-flowing inputs such as Waste Water Treatment Plant outlets are easier to assess than dynamic and highly variable sources like combined sewer overflows or soil erosion. The monitoring campaigns need to be adapted to yield mass flows that reflect a certain statistical confidence and a hydrological representativeness for the catchment.


4. Evaluation of accuracy and cost-efficiency of modelling and monitoring approaches

 
Modeling and monitoring approaches are both resource-intensive, as they need substantial data collection to yield trustworthy predictions. Depending on the type of pollution that needs to be assessed and the hydrological characteristics of the catchment the choice of the model complexity  and the data needs provided by monitoring need to be optimized to achieve maximum prediction confidence with minimal effort.


5. Characterization of immission situation through water quality modeling
 
Mass flow calculations are extremely important to identify and rank sources of contaminants as well as to quantify catchment exports. Yet most of the load, especially for solid-phase contaminants is just on passage through the stream beds and the fraction of contaminants with exposures in rivers that yield an effect on ecological status need to be evaluated with river quality models. These models can cover nutrient, but also heavy metal and organic micro-pollutant fate and effects in river systems. M3 will apply DWA Water quality models and SOBEK WQ to simulate the immission situation in rivers of the 3 regions.  


6. Scenario building and effect prediction for river basin management measures

 
Programs of measures can be evaluated a priori by simulating different scenarios to assess the efficiency of the measures in relation to their economic costs. Realistic scenario building strongly depends on the models accuracy to picture the management measure. Here, specific information on natural processes and managerial execution need to be assessed. These estimations or measurements carry uncertainties and the structural ability of a model to simulate a specific measure as well as the confidence of the predictions need to be assessed.