Water challenges — ranging from lack of access to safe drinking water and sanitation services, to hydrological uncertainty and extremes such as floods and droughts, to chronic water scarcity — are perceived as some of the greatest threats to global prosperity and stability (Sadoff et al. 2014). Many of these challenges are expected to intensify as climate change unfolds and population continues to grow (World Bank, 2017). Water resources are a critical asset in any country. Therefore, their monitoring, maintenance, preservation, and use abide strict rules and regulations enforced and executed by specialized personnel.
Water resources managers use daily collected information about the condition of the dams and rivers, water economic data, meteorological data, weather forecasts, geospatial information, and the surrounding nature environment. They need to have an overview of these heterogeneous data and their interconnectedness to analyze, evaluate the situation, plan water discharge activities, trace fairways and make informed decisions for maintenance, routine exploitation, and emergency situations.
These data come from different sources, e.g., human input, sensors, satellites, and are usually dispersed on servers or computers in different buildings, and in different formats, e.g., excel sheets, files, databases. They are looked at separately by experts who analyze, summarize the evidence, and come up with adequate conclusions and action plans. Conversations with water resources managers and decision makers suggest that a one-stop instrument that gathers all data to be inspected and reviewed providing visualizations and analytics are needed. On thing can be concluded: Cutting-edge water resources management is required for the efficient operation of rivers and dams.
Linked data technologies (W3C, 2015) offer capabilities that enable the creation of knowledge value chains mixing heterogeneous kinds of data, including earth observation (EO) data, reasoning over them in explainable manner, and converting the data into insight. This is a paradigm change in data management.
In this article we compare solutions based on linked data with leading mainstream Geographic Information Systems (Zhong et al. 2022), (ESRI, 2023) and Supervisory Control and Data Acquisition (SCADA)-based (Inductive Automation, 2018) solutions for the benefit of water resources management.
Linked data technologies present a standardized framework to address the above-described major challenge: having to work with heterogeneous data. They provide the ability for semantic data integration, optimal usage of computational and hardware resources and seamless extendibility of the required information. Unlike relational databases, linked data knowledge bases are schema-less. Both the data and the data model that describes them are expressed in the same language. That is why the cost of maintenance of semantic knowledge bases, produced with ontology driven process, gets increasingly reduced over time (see Figure 1), as there is no need of re-engineering of the data base in case of extension of the functionality of the information system using it.
Unlike the relational databases, they record only available information and do not require to store empty cell values in database tables. In this sense, they allow for an economic use of computational and hardware resources. Moreover, they provide semantic interoperability between different in nature data, e.g. digits, maps, images, symbols, and thus enable the generation of new information, based on the formal logic, underlying the principles of linked data technologies. Linked data produce a network of connected objects harvested from different sources (e.g., sensors, databases, files, web-based data entry) in a graph and allow the application of methods to search, compare and browse the graph in an explainable manner (see Figure 2), e.g., it is possible to follow and review the path of how the graph has been traversed and how a goal has been reached or a result found. This contributes to the trustworthiness of the responses to queries or to the conclusions drawn automatically by the system, designed with linked data technologies. Finally, Linked data allow for distributed knowledge discovery, because they enable access to information located in different locations.
Linked data for water resource management
An information system that successfully addresses the need of daily and effective monitoring of hydropower reservoirs and informed decision making for maintenance, routine exploitation and emergency situations requires capability of federated and integrated representation of spatial, numeric and symbolic data as well as of images. It further requires providing a way to easily interlink heterogenous data in an open, easy to maintain, up-to-date way. Linked data is a suitable technology to help solve this problem.
Linked Data can enable a powerful e-infrastructure to monitor water resources of dams and rivers as well as their exploitation, that makes use of all necessary kinds of information, e.g., discharge, water level, available volume, inflow, forecasts, earth observation information, geo-located objects, satellite images, etc. all in one place and semantically integrated, working in concert. A system based on linked data can provide forecasts, formulate queries and throw alerts. Spatial information, remote sensing information, symbolic and numeric data are put together via an ontology that allows for knowledge discovery in explainable manner (see Figure 3).
That allows users of software based on linked data to navigate through historic and forecast data in various data correlations, where in-situ measurements are interlinked with domain knowledge, geospatial information, forecasts, river dynamics and many more. For example, one can ask questions about the amount of the precipitations and the soil moisture or snow cover, when the water volume of a particular dam was at its maximum in the past 10 years, when the water level of a given cascade was above a certain level, about the future riverbed in a particular river area, if the discharge will be higher than a certain amount in the next 10 days, about the expected discharge for a particular date, etc. The system can render synchronized figures on the dam or river area asked about by the user (see Figure 4) in a table, on a graph and in a map view of the satellite data.
It is possible to have access to forecast data and river dynamics. This is due to the fact that the linked data infrastructure integrates the forecast data and the river dynamics into the overall data flow. The approach provides convenient access to a variety of data and methods in numerous contexts. An example of such a forecast can be seen in Figure 5.
A similar approach has already been used in WATERNOMICS project aiming at providing a wholistic view of water resources and helping citizens to be aware of the amounts of water available to them (Curry et al., 2014).
Geographic Information Systems (GIS)
Geographic Information Systems (GIS) technology also provides web-based workflows, as a component of the broader concept of spatial data infrastructures. GIS technology is instrumental for visualization of geolocated information and content mapping. Different geolocated objects can be shown on an interactive map simultaneously, allowing users to view them together and to observe their interdependencies. For example, GIS can be used to visualize information about population, landscape, vegetation, soil, farms, factories, satellite images (see Figure 6). Some GIS software also supports content modelling, like hydrological models or weather forecast models, with the ability to display the geolocated results on the map.
While powerful and mainstream technology for working with geospatial information, GIS remains predominantly an information visualization and content mapping tool. It is designed on the principles of relational database models, or allows the upload of text files with geospatial information, which makes their support and maintenance expensive and at times inefficient, as modifications in the models would require redesign of the GIS application.
GIS technology does not allow to link and visualize non-spatially positioned data with spatially positioned data, something that is usually required in the context of water resources management, for example the operation of equipment, dam relief facilities (which are located within a dam) and the measurements from the hydrometric stations. Human experts and engineers are required to view, observe and interpret the visualized evidence. GIS technology is well suited for monitoring purposes, in case geospatial location and geospatial visualization are important.
Supervisory Control and Data Acquisition Systems (SCADA)
Supervisory Control and Data Acquisition (SCADA) Systems address the need for technology that can help operations and decision making in the water resources management domain by provide them with one single user interface. SCADA systems combine software and hardware elements and allow the monitoring of industrial processes through collection of real-time data from remote devices such as sensors, valves, pumps, motors via human-machine interface and record events in log files. While very common and renowned, SCADA systems are suited for the observation of process pipelines by operators, but do not provide functionalities that allow for interaction with the data for the purpose of analysis, reporting, forecasting or predictive analytics. They can also be seen as web-based workflows, that allow the inspection of the different sources of information from a single interface and provide analytics based on numeric data from different sources (see Figure 7).
Compared to SCADA, linked data technologies have functionalities that allow for superior interaction with the data. Analysts or responsible operators can search for different data correlations at their convenience, request to view past statuses or analytics. Linked data-based infrastructures can combine logical with numeric reasoning and integrate well with GIS visualizations.
Potential of solutions based on linked data
Linked data technologies enable the creation of a system with numerous functionalities compared to systems based on GIS or SCADA technologies. A comparison of an example implementation with common mapping and SCADA systems is provided in Figure 8.
“Data is the world’s great new natural resource”, claims Ginni Rommetty, the president and CEO of IBM (Deutscher 2013). That is to say, data will bring similar disruptions to the 21st century as the steam power brought to 18th century, electromagnetism – to 19th century and fossil fuels – to the 20th century. Linked data technologies provide a very powerful instrument to master wealth of heterogeneous data in explainable manner allowing computers and humans to collaborate in optimal way. This also pertains to water resources managers and policy and decision makers in the water resources management domain. And it is not necessary to look far.