Professor Gabriel Eckstein: Implications of the UN Watercourses Convention for Groundwater Resources

The following post by Professor Gabriel Eckstein is the seventh in the series of essays related to the entering into force of the 1997 UN Watercourses Convention (see links to all of the essays here). Professor Eckstein is a member of the law faculty at Texas A&M University, directs the International Water Law Project, and edits the International Water Law Project Blog. He can be reached at gabriel [at]   or   gabrieleckstein [at]

With Vietnam’s accession to the 1997 UN Watercourses Convention (UNWC), the global community has taken an important step toward cooperative riparian management of transboundary waters. Although most scholars and UNWC parties have focused on the surface water implications of the Convention, there is another critical component of the instrument that should not be neglected. The UWNC applies to many aquifers worldwide. The purpose of this essay is to consider the scope of the Convention in relation to groundwater resources and place it in the context of emerging international law for transboundary aquifers.

Groundwater: a hidden treasure

Groundwater is the world’s most extracted natural resource. It provides approximately 45% of humanity’s freshwater needs for everyday uses, such as drinking, cooking, and hygiene, and 24% of water used in irrigated agriculture (see here).

Transboundary Aquifers of the World - 2012 Source: International Groundwater Resources Assessment Centre (

Transboundary Aquifers of the World – 2012
Source: International Groundwater Resources Assessment Centre



Not surprising, groundwater is highly transboundary. While 276 international watercourses traverse the world’s land areas, an ongoing study identified 448 aquifers and aquifer bodies traversing international political boundaries. In places like the Middle East, North Africa, and the Mexico-United States border, transboundary aquifers serve as the primary or sole source of available freshwater for human and environmental sustenance.

Despite their importance, transboundary aquifers have been comparatively ignored in cross-border water management and treaty development. While thousands of agreements have been forged for transboundary rivers and lakes, only a handful directly apply to aquifers that traverse international frontiers. As one of the few international instruments to address this topic, the UNWC provides critical recognition of the important role groundwater resources play in human progress and development, as well as the need to establish principles of law governing this “hidden” but valuable natural resource.

Applicability of the UNWC to groundwater resources

Although the UNWC clearly applies to many of the world’s groundwater resources, it is important to delineate precisely which aquifers are included and excluded from the rubric of the Convention. The UNWC defines “watercourse” to mean “a system of surface waters and groundwaters constituting by virtue of their physical relationship a unitary whole and normally flowing into a common terminus,” and an “international watercourse” as “a watercourse, parts of which are situated in different States.” Parsing out this phraseology reveals a number of important qualifications.

For an aquifer to fall within the scope of the UNWC, it must be a part of a “system of surface waters and groundwaters.” Use of the “system” criterion in the definition implies an interrelationship between multiple and interlinked water bodies. This assessment is supported and complemented by the subsequent definitional language that emphasizes the “physical relationship” and “unitary whole” of the system, and the “common” characterization of a terminus. Hence, solitary transboundary aquifers – such as independent fossil aquifers and rain-fed aquifers – are presumptively excluded from the scope of the UNWC.

It is noteworthy that subsequent to drafting the principles for the UNWC, the UN International Law Commission (ILC) submitted a Resolution on Confined Transboundary Groundwater in which it commended states to be guided by the principles of its work product in regulating independent and hydraulically unrelated transboundary groundwater resources. This progressive recommendation was not incorporated into the UNGA’s final version of the UNWC.

Model of a Transboundary Aquifer Model, from Puri,, "Internationally shared aquifer resources management, their significance and sustainable management: A framework document," IHP-VI International Hydrological Programme Non-Serial Publications in Hydrology SC-2001/WS/40 (UNESCO 2001)

Model of a Transboundary Aquifer
Source: Puri,, “Internationally Shared (Transboundary) Aquifer Resources Management: Their Significance and Sustainable Management – A framework document,” IHP-VI International Hydrological Programme Non-Serial Publications in Hydrology SC-2001/WS/40 (UNESCO 2001)



In addition, a textual reading of the two definitions suggests that the Convention applies where the transboundary character exists in any part of the system. Hence, a domestic aquifer is subject to the UNWC if it is hydraulically connected to a transboundary river. Similarly, an internal river would be bound by the terms of the Convention if it is linked to a transboundary aquifer. This latter scenario, however, may be subject to debate. In its Thirty-Second Session Report during its preparatory work to the UNWC, the ILC asserted that “the main stem of a river traversing or forming an international boundary” is the “core” of a watercourse. Additionally, Ambassador Chusei Yamada, who served on the ILC during the drafting of the UNWC and later as Special Rapporteur for the ILC’s Draft Articles on the Law of Transboundary Aquifers (Draft Articles), explained to this author in a private conversation that in its deliberations the ILC, and later the UNGA, never contemplated applying the Convention where the transboundary character of the system could not be found in a surface water body. Given that the qualification has yet to be considered in state practice, it remains unclear how broadly it may be interpreted.

Another criterion affecting the UNWC’s applicability to certain groundwater resources is the phrase “flowing into a common terminus.” The criterion intimates that the interlinked water resources must flow toward the same end point to be subject to the Convention’s regime. The directional flow of rivers and lakes, however, is generally described in two dimensions (from point X to point Y). In contrast, groundwater flow is defined in three dimensions because its movement is dependent on local geological conditions, which can vary throughout the aquifer. As a result, groundwater can flow toward a disparate terminal point from that of a related surface water body. Moreover, while aquifers do sometimes terminate at a single point, such as at a spring, it is more common for aquifers to discharge over an extended geographical area along the entire edge of the aquifer. Accordingly, the Convention does not apply to aquifers that do not share a common terminus with hydraulically connected rivers and lakes.

In summary, the UNWC does apply to groundwater resources. However, the Convention’s definitions narrow its relevance to domestic and transboundary aquifers that are hydraulically linked to a transboundary river or lake and that flow into a common terminus. They may also apply to transboundary aquifers that are hydraulically linked to an internal water body, so long as the interrelated surface and subsurface waters flow into a common terminus. All other aquifers are excluded from the Convention’s regime (for a more detailed analysis, see here).

UNWC, transboundary aquifers, and international law

While the UNWC is widely regarded as codifying customary international law, it draws almost exclusively from state practice related to the management and allocation of transboundary rivers and lakes. This perspective is understandable as the bilateral and multilateral cooperative experience over transboundary groundwater resources is scant in comparison. Nevertheless, many of the norms contained in the UNWC are equally (or, at least, similarly) applicable to transboundary aquifers.

A 2011 study suggests that the customary responsibilities most conspicuous in state practice include the substantive obligations of equitable and reasonable utilization and of no significant harm. The study also recognized the existence of accepted procedural duties, including: regular exchange of data and information, generation of supplemental data and information through continuous monitoring and related activities, and prior notification of planned activities. The latter obligation is considerably more general and less developed procedurally than what is contained in the UNWC. Principles contained in the UNWC, but which have yet to arise in state practice for transboundary aquifers, include norms related to ecosystem protection and pollution prevention, cooperative management mechanisms, and the settlement of disputes. In addition, the study identified groundwater-specific concepts that, while logical, have yet to emerge in state practice, including obligations related to protecting recharge and discharge zones.

In 2002, the UNGA tasked the UNILC with drafting principles of law for transboundary aquifers based on trends in state practice and customary norms. The resulting Draft Articles are now before the UNGA (see here). While the Draft Articles were modeled largely on the UNWC, there are a number of noteworthy differences. The UNWC applies to certain transboundary and some domestic aquifers as discussed above. In contrast, the Draft Articles apply to all transboundary aquifers, regardless of whether they are hydraulically linked to any other water body (surface or subsurface), and to domestic aquifers that are hydraulically related to a transboundary aquifer. In addition, the Draft Articles are tailored specifically for transboundary aquifers and include references and principles related to protecting recharge and discharge zones, ensuring the functioning of aquifers, and aquifer-related monitoring activities. If the Draft Articles proceed toward an independent legal instrument, which is yet uncertain (see here), the Draft Articles and UNWC will have to be harmonized.


The coming into force of the UNWC is a significant milestone in the evolution of international water law. While the Convention’s applicability to certain of the world’s groundwater resources may be limited, its growing acceptance and implementation signifies the global community’s broadening commitment to manage and utilize transboundary freshwater resources through peaceful and cooperative means. It also recognizes and affirms transboundary groundwater resources as a legitimate topic of international law.

6 Responses to “Professor Gabriel Eckstein: Implications of the UN Watercourses Convention for Groundwater Resources”

  1. Dear Gabriel,
    Thanks for sharing your analysis, and for your great initiative in hosting this essay series on the UNWC. In keeping discussions alive, I would only make two observations on your piece.
    Regarding the requirement of “flowing into a common terminus”, the relevant provision contains the expression “normally” qualifying that requirement. Therefore, I wouldn’t agree with your conclusion, as stated in absolute terms, that “the Convention does not apply to aquifers that do not share a common terminus with hydraulically connected rivers and lakes.” Arguably, an aquifer that discharges into an international watercourse, even if it also discharges across a larger area, could meet the requirement of “normally flowing…” and be subject to the convention.
    In relation to the ILC Draft Articles, my understanding is that the commentaries that accompany them make clear that domestic aquifers connected to international watercourses are outside their scope.
    Of course, these are just thoughts off the top of my head. So I’d be interested in reading your reaction and considerations other colleagues might like to add. Cheers,

  2. Thanks Gabriel. I agree with Flavia that more research needs to be done on the exact meaning of “normally”. It could mean “under normal circumstances”, i.e. if the aquifer is not full and does not therefore flow into a common terminus it is still included. However, what about the aquifers which remain domestic at low replenishment and international when full ? Would they be included all the same ?



  3. Dear Gabriel,
    Thank you very much for the essay series and this particular piece of analysis related to the connection between the UNWC and transboundary aquifers (TBAs). I am sure the debate on the applicability of the UNWC with regard to individual TBAs will still continue over the next years.
    At the same time we can note that the global and regional debates on international legal frameworks for addressing cooperation over transboundary surface and groundwater have already left its “footprint” in many specific bi- or multilateral cases. As such the UNWC as well as the UN ILC Draft Articles on TBAs have already contributed to improved cooperation and informed exchange on such transboundary surface and groundwater – even before the UNWC actually comes into force.
    You already mentioned that TBAs are of particular importance in a number of regions, incl. the Middle East and North Africa. This is esp. true for these arid and semi-arid environments in which most rivers are largely depending on groundwater inflows.
    In this context it might be worth commending the efforts made over the last years by the League of Arab States supported by a number of regional institutions incl. UN ESCWA to codify a regional legal framework for surface and groundwater / aquifers taking into account specific regional considerations. While the official exchanges are still on-going and the final character of this regional legal framework is still under debate we can see the on-going dialogue already as an important sign for more intra-regional dialogue and informed decision making.
    You also highlighted that “bilateral and multilateral cooperative experience over transboundary groundwater resources is scant in comparison (to surface water)”. This is one of the reasons that UN ESCWA in cooperation with the German Federal Institute for Geosciences and Natural Resources (BGR) developed the Inventory of Shared Water Resources in Western Asia ( ); providing a comprehensively assessment of the state and evolution of transboundary surface and groundwater resources in Western Asia. I am optimistic that this will contribute further to improved knowledge about complex transboundary aquifers in the region and lead to more informed dialogues between scientists, NGOs, civil society, governmental officials and policy makers.

  4. Dear Flavia and Dominique. Thank you both for your valuable comments! I think I need to better explain my statement. When I wrote that “the Convention does not apply to aquifers that do not share a common terminus with hydraulically connected rivers and lakes”, I referred to an aquifer and a hydraulically-linked river with completely different termini. I completely agree that the example Flavia provided (“an aquifer that discharges into an international watercourse, even if it also discharges across a larger area”) is covered by the UNWC. In contrast, however, I was suggesting an aquifer that is fed by (and, therefore hydraulically linked to) River Y, but which flows toward River X. While a rare occurrence, the well-known Donauversinkung case is somewhat of an example – groundwater linked to the upper region of the Danube River partially flows toward the Rhine River and then toward North Sea rather than toward the Black Sea (the terminus of the Danube) – see here. I should note that this scenario is rather rare. Nevertheless, lawyers are charged with conceptualizing all possibilities and IMHO, the UNWC would not cover this unique scenario.

    Also, be aware that the UNILC’s commentaries to the draft articles on the non-navigational uses of international watercourses (p. 90) discussed the use of the term “normally” indicating that it was introduced to bridge “the gap between, on the one hand, those who urged simple deletion of the phrase `common terminus’ on the grounds, inter alia, that it is hydrologically wrong and misleading and would exclude certain important waters and, on the other hand, those who urged retention of the notion of common terminus in order to suggest some limit to the geographic scope of the articles. Thus, for example, the fact that two different drainage basins were connected by a canal would not make them part of a single `watercourse’ for the purpose of the present articles.” The commentary also raises the Donauversinkungs scenario but only in the context of negating the possibility that the Danube and Rhine would be considered as one unitary whole. Hence, inclusion of the “normally” language seemed to focus more on differentiating between different watercourses rather than on the relationship between hydraulically related surface and ground water resources.

    With regard to your second point – that “domestic aquifers connected to international watercourses are outside the scope” of the UNWC – I respectfully disagree. My reading of the commentaries (see especially p. 90) suggest that just as a domestic tributary to a transboundary river would fall under the UNWC, a hydraulically linked domestic aquifer also would be subject to the Convention. Additionally, the eminent Professor Julio Barberis included this exact scenario in the models he proposed in his 1986 UNFAO study as subject to international water law.

  5. Hi there,
    Just to clarify, in the 2nd part of my comments, I was referring to the ILC Draft Articles on Transboundary Aquifers and what you conclude about their scope. When it comes to the UNWC, yes, I agree completely: domestic aquifers linked to international watercourses would be within the convention’s scope. Apologies for the confusion coming from my reference simply to “ILC Draft Articles”. Thanks,

  6. In fact, Gabriel, now I see I misread this section of your blog. You’re actually referring to domestic aquifers linked to transboundary aquifers, i.e., the idea of an aquifer system. We’re in agreements on this point, of course. My apologies, once again!