Water, energy, and food security are inextricably linked, especially now that we are at a crossroads in history to ensure water security for the planet. In that regard, the increasing demand by citizens, environmental organizations, and many other vital importance sectors for cleaner inland and coastal waters has been evident.

The Water Framework Directive (WFD) (2000/60/EEC) [1] has been trying since 2000 until today to push the European Union (EU) Member States to improve their water monitoring networks, to restore their sensitive water bodies, as well as to respond to the pressures that the water systems face in a timely manner before they reach levels of total degradation. This overall effort of the Directive has inspired the following very apt phrase, that the WFD is “the most ambitious and complex piece of legislation on the environment ever enacted in the EU” [2], having simultaneously a multifaceted beneficial character for the nexus among socio-economic and environmental dimensions. These growing demands for continuous information and data on surface water quality are impossible to achieved using only conventional in situ techniques [3]. In general, these techniques are time consuming and costly, slowing down the Member State obligations to the WFD, with consequential penalties. Thus, researchers focused on the development of alternative approaches that can be implemented at different water bodies and conditions ^[4][5][6][7][4,5,6,7]. Earth Observation (EO) has proved as a valuable tool to address the challenges related to the provision of reliable, timely, and accurate information of water resources [8]. Hence, significant opportunities arise from the new EO means paving the way for an ambitious and realistic future in terms of monitoring capacities.

European nations today depend on monitoring networks, with unprecedented and increasing complexity and several times forming a fragmented in situ component. Vitally, a series of novel monitoring systems have brought insight in the reporting and verification of WFD. In this new reality, if we are to ensure continued sustainment of Europe’s societies and vital natural and industrial ecosystems, comprehensive, yet installation-specific methodologies and tools are essential to allow an efficient and harmonized monitoring framework (common protocols). With collaboration of organizations from throughout Europe and worldwide, highlighted as a crucial step to respond to the recognized challenges with an innovative, user-facing AI-enhanced EO-based monitoring framework, enabling a step-change in RBMPs from all the Member States. Hence, the research community should focus on the development, integration, and utilization of best-in-class technological capabilities to strengthen Europe’s capacity through the detection and monitoring of surface water quality. In that regard, important capabilities and priorities are presented below.

While for multispectral imagery data the methods to predict water parameters are already matured substantially, the coupling of these methods to hyperspectral data streams is still largely lacking. Nonetheless, it is widely recognized that imaging spectroscopy can greatly increase the reliability of water-related indicators monitoring. Future studies can focus on the exploitation of PRISMA, DESIS, and EnMAP archives to develop and test new proxies for water WFD-related indicators. Further, improvement of the hyperspectral imagery data will be achieved through the introduction of deep learning-based super-resolution modeling (e.g., SR-GAN) aiming to augment the spatial resolution of hyperspectral data (30 m). This could be achieved by establishing a relationship between the predictor high spatial resolution Sentinel-2 data (10 m) aggregated to low-resolution data pixels and the low-resolution variables that need to be sharpened, and thus allow enhanced monitoring of low-scale features and relevant water bodies. The utilization of multi-infrastructure monitoring is recommended to improve the spatial and temporal resolution by leveraging the emerging technologies of in situ low-cost platforms, as well as the recently advanced spaceborne platforms. For instance, the deployment of UAVs seems to address key tasks (cloud coverage, mapping of small water bodies, etc.). This is imperative to pave the way for the synergistic use of UAVs and spaceborne capacities to optimize advanced sensors (e.g., thermal) and imaging capabilities as new ways for stakeholders to increase the comprehensiveness and credibility of the monitoring process. Moreover, data fusion with unlimited scaling between EO systems: Exploitation of multiple datasets for water bodies monitoring, including EO and terrestrial data is necessary in order to enhance the predictive performance of the model (e.g., active learning) and the spatial and temporal resolution. Furthermore, AI-enabled services and distributed actionable analytics can also be considered to streamline the execution of distributed analytics workflows through semi-automatic orchestration of distributed execution engines and utilization of rich datasets residing in federated computing infrastructures. This will eliminate the hesitation of the relevant authorities to exchange sensitive data, allowing the operation of advanced decision support tools by applying AI to deploy fit for purpose algorithms. All the aforementioned can be achieved following an interdisciplinary approach and by exploiting technological components and results produced by a series of research projects that focus on a variety of aspects related to the WFD concept and objectives. However, the proposed methodologies and tools should be validated in more operational environments, along with the active involvement of the relevant stakeholders. Hence, a framework for better communication and cooperation between the actual actors involved for the WFD implementation can be achieved through a co-design approach in order to reflect and solve their real needs.

To address the challenges in water resources monitoring, several novel approaches and methodologies, utilizing diverse yet complementary technological platforms, can be adopted and combined with the EO component. Among others, the potential of ICT (Information and Communication Technology)-enabled citizen observatories have been recognized ^[9][131]. Herein, citizen observatories refer to an environment and infrastructure supporting an information environment for WFD-relevant communities and decision makers to discuss, monitor, and intervene in situations, places, and events. Despite their increasing popularity and acclaimed potential, citizen observatories are not “plug and play” solutions or simple technical fixes for citizen-based in situ data collection, stakeholder engagement, or participation in the decision-making process ^[10][132]. Kelly-Quinn et al. (2022) ^[11][133] proposed a unique framework to leverage the enabling technologies for citizen observatories in order to fill small water body data gaps. This may foster tremendous opportunities considering that a significant number of surface water bodies are not characterized due to their small coverage. Similarly, Hegarty et al. (2021) ^[12][134] utilized citizen science to evaluate water quality in river bodies and simultaneously fill the recognized data gaps in an effort to support United Nations Sustainable Development Goal 6 objectives. The need to embed the enabling technologies for citizen observatories into their social dimensions should be highlighted, ensuring a continuous uptake of resulting solutions, always respecting privacy of personal data.

Furthermore, the recent advances, during the last decade, in key technologies for the development of Internet of Things (IoT) have made them a valuable solution for water quality monitoring. The main advantage of IoT sensor networks is that they provide continuous measurements; however, they are able to geographically cover a limited area. This is well recognized by the EO community, using them for calibration/validation of satellite-based data and for covering the gap of data streams between two EO data observations. Wide-scale networks have been put in place for water quality monitoring in rivers ^[13][135], lakes ^[14][136], and marine environments ^[15][137].