Piloting a Water Data Management Ecosystem to Enable an Efficient and Resilient Decision Support System for the IJsselmeer

Abstract

The water sector faces great challenges and stresses on the major water system due to climate change and increasing population. As a result, water utilities are increasingly undergoing a digital transformation, to achieve more resilient and sustainable water services while implementing more data-driven decisions. To tackle challenges such as cybersecurity, data ownership and poor quality of data, the European Commission proposes the creation of Data Spaces, as part of the European strategy for Data. Within the Horizon Europe project called WATERVERSE, a holistic approach is being developed to drive the development of data spaces for water utilities. The project involves the development of a Water Data Management Ecosystem (WDME) to enhance the adoption of data management practices that are affordable, accessible, secure, fair and easy to use, while improving the usability of data. In this work, the piloting of a WDME for the Netherlands case study will be presented. The lake IJsselmeer, is used by the water company PWN as a crucial source of drinking water supply for almost 2 million customers in the North-West region of the Netherlands. However, due to population growth, sea level rise, and climate change, the lake IJsselmeer faces extreme variability in water quality in the future. Furthermore, the lake IJsselmeer is at the end of the Rhine Delta, and therefore faces varying water quality challenges from upstream users and stakeholders and saltwater intrusion from the Wadden Sea. Therefore, the development of a digital twin for the lake IJsselmeer is needed to predict chloride (Cl-) and other important water quality parameters for operational (daily basis) and strategic (coming decades) decision making. Such a digital twin requires various data as input from heterogeneous sources. Therefore, to enable the deployment and efficient use of the digital twin as part of a decision support system, the Cl- source prediction model is being piloted within the WDME. An open-source data exchange system called FIWARE is deployed within the pilot. FIWARE serves as the primary broker to exchange contextual information between the various components. Raw data from various sources such as – PWN’s internal data on water quality, data from the national weather agency, water level data of lake Ijsselmeer from the governmental water management agency, are accessed in real-time and fed into the WDME. The data is then processed and prepared as input to the digital twin, which provides predictions over multiple forecasting horizons. Finally, all relevant data, including the predictions, are relayed to a dashboard.