Exploring the use of data and models in transboundary water governance

A balancing act

Good and informed water management ideally needs timely, targeted, relevant, sufficient, valid and reliable data. As water does not stop at administrative boundaries, the sharing of data and information across institutional boundaries is often considered an imperative. Yet, collecting, developing, and applying data and data-driven tools for governing and managing transboundary water is a difficult balancing act. It requires a consideration of if and how opportunities and limitations provided through specific data and data-driven tools connect with the politics of the places the data and tools are implemented in. The balancing act asks reflexivity of those who are commissioning, developing and applying the tools (Milman & Gerlak, 2020).

This is even more true today, as we live in a time of digital exploration. Data collection and data-driven tools are increasingly being used for science and policy development for transboundary water (Bukhari & Brown, 2022). Newer methods of data collection and processing, including remote sensing and artificial intelligence, are already influencing knowledge development and decision-making over transboundary waters. These developments may be complex for governments that consider data and other information on water issues of national security. This requires a conscious engagement with new tools in politically sensitive situations, for instance, by assessing how formerly unobtainable information can influence interactions between countries, or by rethinking legal and institutional frameworks around data and information exchange (Ibrahim, 2020; Leb, 2019). These are questions that need to be answered through both science and practice. Surprisingly, there are very few examples of scientific articles that detail the everyday practice of creating and using data and data-driven tools for transboundary water governance.¹

We have initiated this special issue to bring about more case studies, and therefore the large majority of papers in this special issue are practice based. While working on the special issue, the authors and guest editors experienced specific challenges that may partly explain why only a few articles have so far been written on this topic. We have seen cases that were deemed too sensitive to publish, due not only to the context of transboundary water conflict and cooperation, but also the sensitivities related to the roles of the funders and developers of the tools that are used. There are also practical challenges; writing such analyses requires dedicated time which is not often available for practitioners, and it requires trans- or interdisciplinary approaches that may not be widely accepted. The aim of this special issue is therefore twofold: to showcase how such research can potentially be done to motivate more stories to be shared on this topic; and to inspire joint learning based on the cases.

To share lessons from real-world case studies, we have moved the discussion beyond an academic debate. We specifically reached out to potential authors who had practical experience with developing and applying data and data-driven tools, and who also were required to deal with potential sensitivities. This is why this special issue includes four technical notes, which are shorter articles that focus on case descriptions, in addition to research articles that unpack both historical and recent cases that emerge from real-world negotiation processes. The contributions to this special issue have a large global range, including examples from North and South America, Western Europe, Asia, and Africa. We first introduce the papers included in this issue and draw connections between them, after which we draw specific lessons that arise from the contributions.

Highlights from the papers

The special issue is broadly structured in four topics, including the role of in-situ data, the opportunities and challenges related to ex-situ data, data-sharing practices and frameworks, and the development and implementation of water models to promote transboundary cooperation. With a focus on data, Kipyegon Bosuben et al. (2022, in this issue) analyse the controversy surrounding declining water levels of Lake Victoria in the period 2000–06. They show how the absence of measured data can give space for drawing attention away from key issues and potentially avoiding accountability. Yet, data collection and exchange on transboundary waters such as Lake Victoria is not assured. Mukuyu et al. (2023, in this issue) analyse data exchange in 11 African river basins based on surveys. The authors reaffirm that downstream countries have a higher need for data, and that the highest rated need for water-related data from other riparian states relates to urban water supply. Yet, the surveys show that high data demand from one country does not lead to higher data supply from another, especially when it concerns groundwater and wastewater.

Remote sensing and open data are often seen as ways to mitigate a lack of data-sharing on transboundary water. Yalew et al. (2023, in this issue) provide an overview of the current possibilities to use remote sensing for transboundary water governance, and reflect on the potential of these data to contribute to moving from unilateral action to cooperation between countries. They show that there are limited opportunities related to groundwater and water quality, and that most promising usages relate to monitoring of surface water and ecosystem health and biodiversity. Hassan et al. (2024, in this issue) confirm the usability of a remote sensing-based analysis and showcase how it has been applied to forecast critical water supplies to Sudan during the filling of the Grand Ethiopian Renaissance Dam, which is located immediately upstream of the international border. Such measures were put into place to overcome a lack of information exchange between Ethiopia and Sudan. The authors nevertheless note that ground-measured data and transparent information exchange between the countries remains the preferred source of knowledge, especially to assess the expected risks and benefits of the incoming water flows and to operate the Sudanese infrastructure accordingly.

Several articles discuss initiatives that aim to facilitate data and information exchange between countries. Espíndola and da Silva (2024, in this issue) analyse a decision support system for the La Plata Basin that connects Brazil, Bolivia, Uruguay, Paraguay and Argentina. They show that even when agreements are in place at the transboundary level, the underlying socio-technical systems may not be ready or supportive. Similar lessons are drawn by Zielinski et al. (2024, in this issue) and Ter Horst et al. (2024, in this issue). Based on work done to develop water allocation plans for the Mara River shared by Kenya and Tanzania, Zielinksi et al. (2024, in this issue) show that there was support for their work at the international transboundary level, but this was perceived differently by national institutions. While reflecting, the authors conclude that national and local actors should be actively involved from the start of the project, despite – or perhaps especially – when it concerns sensitive issues such as data-sharing beyond national borders. Ter Horst et al. (2024, in this issue) discuss a similar dynamic in the case of the Cauvery River, which is an Indian federal river shared by Karnataka and Tamil Nadu states. Supported by international donors and the national government, a remote sensing-based model was applied to account for the water uses and flows in the basin. The authors analyse why this approach was a mismatch with the highly politicized local dynamics in multiple ways, concluding that major factors were the late inclusion of local actors, adopting a basin approach for the model while the water discussed at the transboundary scale includes only the main river and dams, and lastly that the model promoted an integrated approach to water management while local institutions work along disciplinary lines.

The last set of articles discuss how modelling can promote cooperation over water, by explicitly highlighting the interactions between the modelling and negotiation processes. Slinger (2023, in this issue) details how eco-morphological modelling has supported the development of a long-term transboundary vision for the River Scheldt that connects Belgium and the Netherlands. The model allowed for constructive negotiations as it served as a boundary object that allowed different stakeholders to communicate their understanding of what the river is and should be. Wheeler et al. (2024, in this issue) share their personal experiences with the development and application of a Colorado River Simulation System, and show how modelling has supported transboundary negotiations over time. They discuss how the system was initially developed by the United States, and how efforts were undertaken to ensure it could be used by Mexico prior to the beginning of critical negotiations over sharing water shortages. Now the challenge is to make the complex modelling tool available for a wider group of stakeholders that includes researchers and non-governmental organisations (NGOs). The authors highlight that the development of such a system is not static; keeping it relevant requires revisiting and updating assumptions that once seemed accurate, which is not simple in an inherently political transboundary context.

Drawing lessons: towards a grounded use of data and models

The studies included in this special issue showcase how introducing and applying data and technologies for transboundary water management entails often a messy non-linear process. It necessarily needs to be mindful of local specificities, such as histories, cultures, values, available resources and expertise, and power relations (see also Timmerman & Langaas, 2005). Success relies heavily on investing in relations between people and institutions, data and models along with critical technologies. We broadly draw four lessons from the case studies to help navigate relations between people and institutions and the use of data and models in transboundary waters:

• Layer up: Agreements on data-sharing and modelling that gain support at the transboundary level may not be supported at national and local levels. This lack of support can be political or based on the availability of resources such as time, specific expertise or practical tools. When engaging with data and data-driven tools for transboundary water management, it is recommended to include different levels of decision making from the start of the process.

• Communication is key: There are expectations that the rapid developments in remote sensing and artificial intelligence can impact power balances in a basin, since information is power. Studies in this special issue show that data collected through remote sensing can alleviate the dependency on sharing of in-situ data between countries, but cannot replace it. Investing in communication and data-sharing remains essential, while recognizing that data and data-driven tools are no easy fixes but require careful implementation.

• Engage people first, and model second: From the variety of case studies we learn that there is a tendency first to dedicate attention to the technical aspects, including data collection and modelling, and only thereafter engage stakeholders and present results. This is especially the case in situations where data are sparse and data collection and exchange are thought to be sensitive issues. Early engagements and developing agreements with stakeholders are investments towards success in the longer run and can prevent a future breach of trust.

• Take time for transdisciplinarity: Transboundary water governance requires transdisciplinary approaches that link practice and science. Thus, it needs people who understand relations between people, countries, nature and water, as well as the methods and processes needed for data collection and model development. We note that the authors include practitioners and academics who have gained such insights over time, and who have learned to cross disciplinary boundaries through their practice. Such expertise is a requirement for success in projects on data exchange and modelling for water cooperation.

As a final note, we want to thank all contributors who shaped this special issue, and in particular those who have worked on these articles in difficult times, including the civil war in Sudan and through the COVID pandemic.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This special issue has been initiated with support of a financial contribution from the Water and Development Partnership Programme of IHE Delft and the Netherlands Ministry of Foreign Affairs and also received a financial contribution from the Oxford Martin School Programme on Transboundary Resource Management.

Notes

1. Three examples we are aware of include: Plengsaeng et al. (2014) and Thu and Wehn (2016), who detail how data-sharing processes in the Mekong River Commission happen in practice from Vietnamese and Thai perspectives; and Wheeler et al. (2018) on the use of water resource models to facilitate interactions between scientists, decision-makers and stakeholders in the Murray–Darling and Colorado River basins.