New WFD Explorer prototype
In collaboration with other researchers, government authorities and consultancies, Deltares is going to overhaul the WFD Explorer. The WFD Explorer is an important tool that water managers use to determine which measures are needed to improve ecological water quality. The new version is intended to make more informed decisions possible relating to the recovery of plants, fish and other aquatic organisms.
The Netherlands is facing a major challenge to comply with the objectives of the Water Framework Directive (WFD) by 2027. The WFD Explorer is one of the resources we can use to determine the current situation. An assessment was made in an interim evaluation recently, establishing a picture of what is needed further to achieve the objectives. But the WFD water quality objectives do not stand alone and they are linked to broader agendas such as nitrogen emissions and the restoration of natural assets.
The new prototype will contribute to the broader agendas by combining current ecological insights with modern AI techniques. In that way, we can work towards ecological restoration.
The details of the overhaul
The heart of the operation involves using AI to examine cause-and-effect relationships rather than simply correlations. The first step is to look at the different species: by determining which factors (pollution, temperature) are important for the species, we can arrive at a prediction. Only then do we look at the ecological quality ratio or another biodiversity index. Then we also determine the future impact of the measures on water quality. This results in a possible course of action because water managers can work towards the goal.
The updated explorer will also serve to assess the different variables: whether they require modification, whether other variables are needed, or whether variables should be split up. Climate variables are especially important because, in addition to the effects, water managers want to know about climate change. In this way, the scores align with European policy. Finally, AI helps to provide a more robust determination of which factors contribute to deterioration and to predict which measures will be effective in which conditions.
This TKI Delta Technology project will deliver an updated prototype that policymakers and water managers can use to get to work. The prototype will produce reusable datasets, knowledge rules and algorithms that can be used for broader policy questions such as nature and nitrogen analyses.
The new explorer will be a flexible tool that can be easily linked to other tools and software that are already available. One example could be a model train with EBEO2.0 for diagnosis, WFD Explorer for forecasting and LWKM2.0 for an up-to-date schematisation, but also SFTox2.0 for toxicity. This will result in a national information infrastructure.
The added value of artificial intelligence
The use of AI in the water sector is spreading rapidly and that includes the field of ecological assessment. Where many applications get bogged down in pattern recognition and correlations, this project focuses on exposing causal relationships between environmental pressures and ecological response. These relationships are translated into policy indicators such as EQR scores, Natura2000 measures and biodiversity indices.
Combining AI with ecological knowledge rules, field data and expert assessments results in a model that is not only powerful but also explainable and relevant for policy. Thanks to AI, complex, non-linear relationships and interactions between challenges can be identified better, even when data are incomplete or uncertain. Moreover, AI makes it possible to develop models that can follow the growth of new knowledge and data.
This TKI project contributes directly to the key technology for Smart Technology for Agri-Horti-Water-Food by developing an AI-driven prototype that enables data-driven and explainable ecological assessment. In this way, we are using new digital technology for nature, water and biodiversity management.
In this project, we are working with Rijkswaterstaat, Haskoning, Witteveen and Bos, Aveco de Bondt, Arcadis, STOWA and the Netherlands Environmental Assessment Agency (PBL).
