By: Alex Oude Elferink
Case commented on: ITLOS Special Chamber, Judgment in Dispute Concerning Delimitation of the Maritime Boundary Between Ghana and Côte d’Ivoire in the Atlantic Ocean, 23 September 2017.
The Dispute Concerning Delimitation of the Maritime Boundary between Ghana and Côte d’Ivoire in the Atlantic Ocean (Ghana/Côte d’Ivoire), which has been decided by a Special Chamber of the International Tribunal for the Law of the Sea (ITLOS) is a further addition to a long series of maritime boundary cases. In writing the concluding chapter of the edited volume Maritime Boundary Delimitation: The Case Law; Is it Consistent and Predictable?, the author of this post and his co-editors Tore Henriksen and Signe Veierud Busch made a forecast about the outcome of Ghana/Côte d’Ivoire on the basis of the existing case law – the text of the chapter was finalized on 10 May 2017 (on file with the author). This concerned the following points:
First, there is no reason to assume that […] the Chamber of the ITLOS will not apply the three-stage approach. Second, the coastal geography […] is such that we expect that a strict equidistance line, or an equidistance line very similar to the strict equidistance line, will be adopted as the provisional delimitation line. Third, we do not expect a radical departure from the equidistance line after the second-stage consideration of the relevant circumstances […]. Hydrocarbon licensing is likely to be considered as a potential relevant circumstance […], but it is not expected that this practice requires a shifting of the provisional line.
We also anticipated that the Chamber would “have no difficulty in determining the relevant coasts and the relevant area”. We indicated a couple of options for the relevant coasts of both parties, and for both Ghana and Côte d’Ivoire the Chamber adopted a relevant coast that was among the options we presented (for the Chamber’s selection see para. 379 and Sketch-map No. 2). Similarly, we submitted that:
[t]he relevant area can be expected to be bounded by the relevant coasts and the outer limits of the maritime zones of the parties. Depending on the definition of the relevant coasts, the lateral limits of the relevant area will be either perpendiculars to the general direction of these coast, lines of longitude (meridians) or the (potential) maritime boundaries with neighboring States.
As a matter of fact, the Chamber used the relevant coasts, meridians and the outer limits of the continental shelf beyond 200 nautical miles to define the relevant area (see paras 383-386 and Sketch-map No. 3). We concluded our analysis by noting that:
Our assessment of Ghana/Côte d’Ivoire […] strongly suggests that the law and methodology as developed by the case law result in a degree of predictability. It would not even seem unreasonable to submit that [it] suggest[s] a high degree of predictability.
Although the Special Chamber’s approach to the delimitation of the maritime boundary in Ghana/Côte d’Ivoire to a large degree aligns with the earlier case law and arguably contributes to the objective of predictability of the delimitation process, to which the Chamber explicitly refers (paras 281 and 289), there are a couple of points in the judgment that may seem to raise some concerns in that respect. The current post focusses on those points.
A previous post by Nigel Bankes provides a general commentary on the judgment of the Special Chamber, including the delimitation aspects. A previous post by Signe Veierud Busch provides a commentary on the judgment’s delimitation of the continental shelf beyond 200 nautical miles.
Delimitation of the territorial sea
Having concluded that there was no existing agreement between Ghana and Côte d’Ivoire on the delimitation of their maritime boundary, the Chamber turned to the task of delimiting this boundary itself (para. 247). This delimitation concerned the territorial sea, the exclusive economic zone and the continental shelf. In considering the delimitation of the territorial sea, the Chamber observes that the rules for the delimitation of the territorial sea on the one hand, and the exclusive economic zone and the continental shelf on the other are different under the United Nations Convention on the Law of the Sea (LOSC) (para. 260). In particular, the Chamber observed that “in delimiting the territorial sea it has to be borne in mind that the rights of the coastal States concerned are not functional but territorial since they entail sovereignty over the seabed, the superjacent waters and the air column above” (para. 262). Nonetheless, the Chamber considered it appropriate to use the same delimitation methodology for the territorial sea and the maritime zones beyond (para. 262). The Chamber subsequently proceeded with the delimitation of the territorial sea, the exclusive economic zone and the continental shelf as a single exercise (paras 264 and following). The Chamber offered two arguments to support this approach. First, neither party had advanced arguments that were specifically related to the regime of the territorial sea (para. 262). Second, the parties had requested a single maritime boundary delimiting the territorial sea and the maritime zones beyond, which, according to the Chamber, implied that they had “implicitly agreed that the same delimitation methodology be used for these maritime spaces” (ibid.). It may be questioned whether these arguments are completely convincing. It would seem that the Chamber itself should have established, as required by article 15 of the LOSC, that there were no circumstances particular to the delimitation of the territorial sea. Second, as previous case law indicates, a request for a single maritime boundary does not imply that the delimitation of the territorial sea is subsumed in the process of delimiting the exclusive economic zone and the continental shelf. On the contrary, courts and tribunals, where relevant, in general have first addressed the delimitation of the territorial sea and subsequently the delimitation beyond. In its judgment on the merits in Qatar/Bahrain the International Court of Justice (ICJ) specifically observed that in delimiting the territorial sea between the parties:
the Court has to apply first and foremost the principles and rules of international customary law which refer to the delimitation of the territorial sea, while taking into account that its ultimate task is to draw a single maritime boundary [for the exclusive economic zone and the continental shelf] that serves other purposes as well (Qatar/Bahrain, judgment of 16 March 2001, para. 174).
The Court also concluded that article 15 of the LOSC reflects customary law (ibid., para. 176). Although the Chamber’s approach is easily understood from the perspective of procedural economy, from the perspective of predictability it might have been preferable to stick to the standard approach to this matter.
Use of charts
One of the preliminary questions the Chamber had to decide was what nautical charts to derive baseline information from. Côte d’Ivoire had argued for the use of a recent chart and the Chamber acknowledged that this chart might reflect “the most recent data concerning that coast” (para. 342). While agreeing with Côte d’Ivoire that use of a more recent chart in principle would be preferable, the Chamber rejected this approach in the present case because “it is essential that the same methodology be used for the two coasts in question” (ibid.). This led the Chamber to employ two older charts, while observing that the parties had placed reliance on those charts until at least 2014 (para. 342). The Chamber does not state any authority for its approach to the use of charts. Earlier jurisprudence indicates that the most recent baseline information should be employed. In Guyana v. Suriname, Guyana objected to a recently produced chart that had been introduced by Suriname. The tribunal in its Award concluded that it was “not convinced that the depiction of the low-water line on chart NL 2218, a chart recognised as official by Suriname, is inaccurate. As a result, the Tribunal accepts the basepoint on Vissers Bank, Suriname’s basepoint S14” (Award of 17 September 2007, para. 396). The tribunal relied on older charts to determine other basepoints along the coasts of the parties (ibid., para. 393).
Relevant coasts and relevant area
In determining the relevant coasts and the relevant area, the Chamber adopts the commonly accepted principle that “[f]or a coast to be considered relevant in maritime delimitation it must generate projections which overlap with those of the coast of another party” (para. 372). For the relevant coast of Ghana, the Chamber explains its choice for the coast extending from the common land boundary of the parties up to Cape Three Points by observing that Ghana’s coast west of that cape faces the disputed area, while to the east it faces away from the relevant area (para. 376). While this justification is quite unexceptional, the explanation the Chamber offers for the definition of the relevant coast of Côte d’Ivoire is difficult to grasp. That relevant coast is located between the terminus of the common land boundary and Sassandra. According to the Chamber, the projection of that part of the coast of Côte d’Ivoire overlaps with the projection of the relevant coast of Ghana (para. 377). On the other hand, the Chamber submits that the coast west of Sassandra “does not have a projection to the sea in a way that overlaps with the disputed area” (para. 378). Two points are to be noted about this finding. First, the Chamber refers to the disputed area in determining the overlapping projections of the coasts of the parties. This parameter is not included in the principle for identifying the relevant coasts of the parties to which the Chamber subscribes, and the case law amply demonstrates that the relevant area may include areas that are beyond the area between the claim lines of the parties.
Second, the reasoning of the Chamber in support of its definition of the relevant area is not in accordance with the relevant facts. The disputed area, which is defined para. 25 of the Order of 25 April 2015 indicating provision measures, is bounded by an equidistance line and the claim line of Côte d’Ivoire, which is an azimuth with a bearing of 168.7° (the disputed area is included in the figure contained in Annex 1 to this post). The Chamber does not define how it has determined the seaward projection of the coasts of the parties. However, any reasonable definition of the term seaward projection indicates that the point at Sassandra cannot be the point at which the relevant coast of Côte d’Ivoire should terminate. This may be illustrated by two scenarios. If seaward projection were to be understood as a projection at an angle of 90° to the direction of the coast – an approach that is used by the Special Chamber in considering coastal convexity as a relevant circumstance (paras 421-426 and Sketch-map No. 5), a large part of the relevant coast of Côte d’Ivoire as defined by the Chamber does not project into the disputed area. Moreover, the seaward projection of a part of that relevant coast at the 200-nautical-mile limit is further from the disputed area than a part of the coast west of Sassandra that the Chamber does not include in the relevant coast. On the other hand, if we accept the Chamber’s position that the relevant coasts project into the disputed area, seaward projection has to include a projection at an acute angle of around 75°. If all coasts that project into the disputed area at that angle are taken into account, the conclusion has to be that a part of the coast of Côte d’Ivoire, which was disregarded by the Chamber in defining the relevant coast, is part of the relevant coast.
Coastal concavity and convexity
Perhaps the most problematic aspect of the Chamber’s approach to the delimitation of the maritime boundary is its handling of the argument concerning coastal concavity and convexity. Côte d’Ivoire had submitted that the concavity of its coast in combination with the convexity of the coast of Ghana constituted a relevant circumstance that required the adjustment of the provisional equidistance line in favor of Côte d’Ivoire (paras 408 and 410-411). The Chamber accepted that this combined concavity and convexity existed, but did not find it of such a nature to warrant an adjustment of the provisional equidistance line (paras 424-425). Interestingly, the geography of Côte d’Ivoire and Ghana shows no resemblance to previous cases in which concavity and or convexity were found to be relevant circumstances. The ICJ in its judgment in North Sea Continental Shelf observed in this connection:
the effect of the use of the equidistance method is to pull the line of the boundary inwards, in the direction of the concavity. Consequently, where two such lines are drawn at different points on a concave coast, they will, if the curvature is pronounced, inevitably meet at a relatively short distance from the coast, thus causing the continental shelf area they enclose, to take the form approximately of a triangle with its apex to seaward and, as it was put on behalf of the Federal Republic, “cutting off” the coastal State from the further areas of the continental shelf outside of and beyond this triangle (judgment, para. 8).
In Guyana v. Suriname, the tribunal in its award observed that:
The relevant coastlines do not present any marked concavity or convexity. After careful examination the Tribunal accordingly concludes that the geographical configuration of the relevant coastlines does not represent a circumstance that would justify any adjustment or shifting of the provisional equidistance line in order to achieve an equitable solution (award, para. 377; emphasis provided).
In the case of Côte d’Ivoire, equidistance lines come nowhere near producing the effect described by the Court in North Sea Continental Shelf. The seaward extension of the relevant coast of Côte d’Ivoire as defined by the Chamber projects seaward up to the 200-nautical-mile limit and beyond. One would moreover have a difficult task to credibly argue that the coasts of Côte d’Ivoire and Ghana show marked concavity or convexity.
The Chamber’s explanation as to why it considers that a certain cut-off exists in the case of Côte d’Ivoire actually further presses home the point that its assessment of the presence and impact of concavity and convexity is off the mark. The Chamber observes in this respect:
The Special Chamber also acknowledges that the coast of Ghana is convex, which enhances the effect of the concavity of the coast of Côte d’Ivoire. Owing to this concavity combined with the convexity of the coast of Ghana, some cut-off effect exists to the detriment of Côte d’Ivoire. Such cut-off effect affects only the projection of Côte d’Ivoire’s coast east of Abidjan and this cut-off only comes into being 163 nm from [the starting point of the maritime boundary]. The seaward projection of the relevant coast of Côte d’Ivoire [extending west] from Abidjan to Sassandra, however, extends beyond 200 nm, as claimed by Côte d’Ivoire (para. 424).
This observation is problematic on three counts. The Chamber submits that the convexity of Ghana’s coast enhances the effect of the concavity of Côte d’Ivoire coast. The parties were in agreement that the convexity of Ghana’s coast was located at Cape Three Points (paras 294 and 298). However, there is no impact of this convexity on the provisional equidistance line. As was also observed by Ghana (para. 298), all Ghanaian basepoints used in constructing the equidistance line are located on the straight line coast of Ghana west of the convex coast at Cape Three Points. In other words, that convexity is immaterial in determining whether the provisional equidistance line leads to a cut-off of the seaward projection of the coast of Côte d’Ivoire.
A second issue concerning the Chamber’s observation is the assertion that cut-off only comes into being 163 nautical miles from the starting point of the maritime boundary. That cut-off only comes into being on the assumption that seaward projection means projection at an angle of 90° to the direction of the coast. If that assumption is relaxed, as the Chamber did in determining the relevant coasts of the parties (see above), the provisional equidistance line produces no cut-off of the seaward projection of any part of the coast of Côte d’Ivoire either within or beyond 200 nautical miles. A final problem with the Chamber’s assessment of convexity and concavity is that, if the coast of Ghana is evaluated using the same approach as the Camber did for Côte d’Ivoire, the equidistance line produces a more significant cut-off of the coast of Ghana. As is illustrated by Sketch-map No. 5 included in the judgment (reproduced in Annex 2 to this post), the Chamber has determined the amount of cut-off by looking at the frontal, 90°, projection of different parts of the coast of Côte d’Ivoire. For Ghana, its relevant coast may be divided into two segments in this connection: the almost straight segment running from the land boundary to the point at which it changes direction and from that point to Cape Three Points. A frontal projection of the latter segment of the coast starts to be cut off by the equidistance line at approximately 95 nautical miles from the starting point of the maritime boundary, as compared to the alleged cut off of part of Côte d’Ivoire coast at 163 nautical miles. The larger (but still limited) cut-off effect of the equidistance line on Ghana’s coast is explained by the fact that Ghana’s coast, if assessed on the same plane as that of Côte d’Ivoire, actually is more concave than that of Côte d’Ivoire. A measure for establishing the extent of a concavity is determining the ratio between the extent of the concavity measured along a straight line between its extremities and the length of a perpendicular to that straight line at the point at which it is furthest from the coastline of the concavity. When this ratio is larger, the concavity is less pronounced. For Ghana the relevant ratio is 11.6:1 and for Côte d’Ivoire 15.2:1.
The quest for predictability has been one of the core concerns of the case law on maritime delimitation. As was also observed above, the Special Chamber of the ITLOS in Ghana/Côte d’Ivoire confirmed its adherence to the objective of predictability (paras 281 and 289). Although the delimitation effected by the Chamber could be said to fully comply with this objective, the above discussion indicates that the argumentation employed in arriving at this outcome leaves something to be desired. Seemingly plausible explanations might be offered in justification. As was also submitted above, the Chamber’s approach to the delimitation of the territorial sea might be explained by reasons of procedural economy. The Chamber’s acknowledgement that there was some merit to Côte d’Ivoire’s argument on convexity and concavity could be seen as serving the purpose of conveying that the arguments of both parties were carefully assessed. However, it is submitted that from the perspective of predictability – and its twin objective of consistency – it would have been preferable that the judgment on the points discussed above to a larger extent had been in line with the preceding case law. Such reliance might also have avoided the Chamber’s problematic treatment of the relevant coasts and the issue of convexity and concavity.
Thanks to Nigel Bankes for his careful review of a draft of this post.
Annex 1 – The disputed area*
* Source: Request for the Prescription of Provisional Measures submitted by the Republic of Côte d’Ivoire under Article 290, Paragraph 1, of the United Nations Convention on the Law of the Sea, 7 February 2015 (available at https://www.itlos.org/fileadmin/itlos/documents/cases/case_no.23_prov_meas/C23_Request_prov_measures_translation_Reg.pdf)
Annex 2 – Sketch-map No. 5 included in the judgment of the Special Chamber
This post may be cited as: Alex Oude Elferink, “Maritime Delimitation in Ghana/Côte d’Ivoire: Predictability … with an Occasional Glitch”, 2018, online: http://site.uit.no/jclos/files/2018/02/6.2.2018_JCLOS-Blog_Alex-Oude-Elferink_Maritime-Delimitation-in-Ghana-Cote-dIvoire-Predictability-with-an-Occasional-Glitch-1.pdf
To subscribe to the JCLOS Blog by email, please go to http://site.uit.no/jclos/