Eritrea – Yemen Arbitration Documents [Archives:2000/04/Law & Diplomacy]
Part 4 in a series
131. It is a generally accepted view, as is evidenced in both the writings of commentators and in the jurisprudence, that between coasts that are opposite to each other the median or equidistance line normally provides an equitable boundary in accordance with the requirements of the Convention, and in particular those of its Articles 74 and 83 which respectively provide for the equitable delimitation of the EEZ and of the continental shelf between States with opposite or adjacent coasts. Indeed both Parties to the present case have claimed a boundary constructed on the equidistance method, although based on different points of departure and resulting in very different lines.
132. The Tribunal has decided, after careful consideration of all the cogent and skilful arguments put before them by both Parties, that the international boundary shall be a single all-purpose boundary which is a median line and that it should, as far as practicable, be a median line between the opposite mainland coastlines. This solution is not only in accord with practice and precedent in the like situations but is also one that is already familiar to both Parties. As the Tribunal had occasion to observe in its Award on Sovereignty (paragraph 438), the offshore petroleum contracts entered into by Yemen, and by Ethiopia and by Eritrea, “lend a measure of support to a median line between the opposite coasts of Eritrea and Yemen, drawn without regard to the islands, dividing the respective jurisdiction of the Parties”. In the present stage the Tribunal has to determine a boundary not merely for the purposes of petroleum concessions and agreements, but a single international boundary for all purposes. For such a boundary the presence of islands requires careful consideration of their possible effect upon the boundary line; and this is done in the explanation which follows. Even so it will be found that the final solution is that the international maritime boundary line remains for the greater part a median line between the mainland coasts of the Parties.
133. The median line is in any event some sort of coastal line by its very definition, for it is defined as a line “every point of which is equidistant from the nearest points on the baselines from which the breadth of the territorial seas of the two States is measured” (Article 15 of the Convention), although the same definition will found in many maritime boundary treaties and also in expert writings. The “normal” baseline of the territorial sea as stated in Article 5 of the Convention – and this again accords with long practice and with the well established customary rule of the law of the sea – is “the low-water line along the coast as marked on large scale charts officially recognised by the coastal State”. There do arise some questions about what is to be regarded as the “coast” for these purposes, especially where islands are involved; and these questions, on which the Parties differ markedly, require decisions by the Tribunal.
134. First, it is necessary to deal with a complication that arises in the present case concerning this general rule of measuring from the low-water line. The domestic legislative definition of the territorial sea of Eritrea is still the 1953 enactment by Ethiopia which fixed Ethiopia’s territorial waters as “extending from the extremity of the seaboard at maximum annual high tide”. This was done even though an Ethiopian customs enactment of 1952 had provided for a customs zone measured from the “the mean low-water mark at neap tides”. The Yemen claim was that, in view of this 1953 legislation, the Tribunal should measure the median line boundary from the high-water line instead of the low-water line along the Eritrean coast (and indeed Yemen’s median line does).
135. In this matter the Tribunal prefers the Eritrean argument that the use of the low-water line is laid down by a general international rule in the Convention’s Article 5, and that both Parties have agreed that the Tribunal is to take into account the provisions of the Convention in deciding the present case. The median line boundary will, therefore, be measured from the low-water line, shown on the officially recognised charts for both Eritrea and Yemen, in accordance with the provision in Article 5 of the Convention. The officially recognised charts used by the Tribunal are BA (British Admiralty) Charts; those Charts use as a Chart Datum approximately the level of the Lowest Astronomical Tide. These Charts were among those relied on by the Parties in the present Stage of the Proceedings.
Northern and Southern Extremities of the Boundary Line
136. There is also a problem relating to both the northern and the southern extremities of the international boundary line. The Tribunal has the competence and the authority according to the Arbitration Agreement to decide the maritime boundary between the two Parties. But it has neither competence nor authority to decide on any of the boundaries between either of the two Parties and neighbouring States. It will therefore be necessary to terminate either end of the boundary line in such a way as to avoid trespassing upon an area where other claims might fall to be considered. It is, however, clearly necessary to consider the choices of the base points controlling the median line first, and then to look at the cautionary termination matter when the line to be thus terminated at its northern and southern ends has been produced.
137. The construction of the international single boundary decided upon by the Tribunal, working generally from the north to the south, will now be described.
The Northernmost Stretch of the Boundary Line
138. In this stretch, where the two lines claimed respectively by Eritrea and Yemen differed so markedly in their courses, there were three main problems: what to do about the Dahlak islands on the Eritrean side; what to do about the lone mid-sea island of Jabal al-Tayr and the mid-sea island group of Jabal al-Zubayr; and what to do about the cluster of islands and rocks off the northern coast of Yemen. These three questions will now be considered in that order.
139. This tightly knit group of islands and islets, or “carpet” of islands and islets as Eritrea preferred to call it, of which the larger islands have a considerable population, is a typical example of a group of islands that forms an integral part of the general coastal configuration. It seems in practice always to have been treated as such. It follows that the waters inside the island system will be internal or national waters and that the baseline of the territorial sea will be found somewhere at the external fringe of the island system.
140. A problem that arises here, however, is that the Dahlak fringe of coastal islands is also suitable for the application not of the “normal baseline” of the territorial sea, but of the “straight baselines” described in Article 7 of the Convention (as there distinguished from the “normal” baseline described in Article 5). The straight baseline system is there described as “the method of straight baselines joining appropriate points”. Yemen appears to have little difficulty in agreeing that the Dahlaks form an appropriate situation for the establishment of a straight baseline system.
141. Eritrea for its part claimed that it has such a system already established. In answer to a question from the Tribunal, Eritrea did give the coordinates for the base points on the Eritrea side for both versions of its claimed “median line”. But these base points in the region of the Dahlaks appear to have been located on a line touching two or perhaps three of the outer islands and the Negileh Rock (for which see below paragraphs 146-147) and then continuing in a more or less straight line out to sea in a south-easterly direction. This scheme is probably part of the “quadrilateral” straight baseline system to which Eritrea referred in argument.
142. The reality or validity or definition of this somewhat unusual straight baseline system said to be existing for the Dahlaks is hardly a matter that the Tribunal is called upon to decide. The Tribunal does however have to decide on the base points which are to control the course of the international boundary line. In plotting its own claimed median line boundary, Yemen has employed as its western base points the high-water line of the small outer islets of Segala, Dahret Segala, Zauber and Aucan.
These islets could reasonably be included in a straight baseline system of the ordinary and familiar kind.
143. Eritrea, however, has in particular suggested a feature called the “Negileh Rock” which lies further out than these larger but still small and uninhabited islets. Yemen objected to the use of this feature by reason of the fact that on the BA Chart 171 this feature is shown to be a reef and moreover one which appears not to be above water at any state of the tide. A reef that is not also a low-tide elevation appears to be out of the question as a base point, because Article 6 of the Convention (which is headed “Reefs”) provides:
In the case of islands situated on atolls or of islands having fringing reefs, the baseline for measuring the breadth of the territorial sea is the seaward low-water line of the reef, as shown by the appropriate symbol on charts officially recognized by the coastal State.
144. This difficulty about the Negileh Rock is reinforced if there is indeed a straight baseline system in existence for the Dahlaks, for paragraph 4 of Article 7 provides:
Straight baselines shall not be drawn to and from low-tide elevations, unless lighthouses of similar installations which are permanently above sea level have been built on them or in instances where the drawing of straight baselines to and from such elevations has received general international recognition.
145. Although Eritrea is not a party to the Convention; nevertheless it has agreed to its application in the present case; and since Eritrea claims the existence of a straight baseline system, that claim seems to foreclose any right to employ a reef that is not proud of the water at low-tide as a baseline of the territorial sea.
146. As will appear more particularly below, the Tribunal has decided that the western base points to be employed on this part of the Eritrean coast shall be on the low-water line of certain of the outer Dahlak islets, Mojeidi and an unnamed islet east of Dahret Segala.
Next, it is necessary to deon the treatment of the mid-sea islands of al-Tayr and Zubayr, for on this decision depends the question of whether it will be necessary to consider base points on the coast of Yemen.
Jabal al-Tayr and the Zubayr Group
147. Yemen employed both the small single island of al-Tayr and the group of islands called al-Zubayr as controlling base points, so that the Yemen-claimed median line boundary is “median” only in the area of sea west of these islands. These islands do not constitute a part of Yemen’s mainland coast. Moreover, their barren and inhospitable nature and their position well out to sea, which have already been described in the Award on Sovereignty, mean that they should not be taken into consideration in computing the boundary line between Yemen and Eritrea.
148. For these reasons, the Tribunal has decided that both the single island of al-Tayr and the island group of al-Zubayr should have no effect upon the median line international boundary.
Base Points on the Coast of Yemen
149. Since Jabal al-Tayr and the Zubayr group are not to influence the drawing of the median line boundary, it is necessary to decide upon the base points to be used for this part of the coast of Yemen. For here again there is, if not a carpet, at least a considerable scattering of islands and islets which are the beginning of a large area of coastal islands and reefs which, extending northward, ultimately form part of a large island cluster or system off the coast of Saudi Arabia.
150. There is also the relatively large, inhabited and important island of Kamaran off this part of the Yemen coast. This island, together with the large promontory of the mainland to the south of it, forms an important bay and there can be no doubt that these features are integral to the coast of Yemen and part of it and should therefore control the median line. One significant controlling base point is therefore on the westernmost extremity of Kamaran. It seems reasonable also to use as base points the very small islands immediately south of Kamaran and west of the promontory headland mentioned above.
151. The question remains as to the islands to the north of Kamaran. The relatively large islet of Tiqfash, and the smaller islands of Kutama and Uqban further west, all appear to be part of an intricate system of islands, islets and reefs which guard this part of the coast. This is indeed, in the view of the Tribunal, a “fringe system” of the kind contemplated by Article 7 of the Convention, even though Yemen does not appear to have claimed it as such. Indeed the Tribunal does not have the advantage of any views of Yemen about this part of its coast because it chose to deploy its arguments differently. It is however the view of the Tribunal that it is right to use as median line base points not only Kamaran and its satellite islets which appear in the Yemen Map 12.1, but also the islets to the northwest named Uqban and Kutama.
152. The above decisions having been made, it is now possible to compute and plot the northern stretch of the boundary line between turning points 1 and 13 (the list of the coordinates of the turning points is given below; see also the illustrative Charts 3 and 4). For this entire part of the line, the boundary should be a mainland-coastal median, or equidistance, line.
153. At turning point number 13, however, a simple mainland/coastal median line approaches the area of possible influence of the islands of the Zuqar-Hanish group, and clearly some decisions have to be made as to how to deal with this situation.
The Middle Stretch of the Boundary Line
154. It will be convenient for obvious reasons if the Tribunal first decides the question of the boundary in the narrow seas between the south-west extremity of the Hanish group on the one hand and the Eritrean islands of the Mohabbakahs, High Island, the Haycocks and the South West Rocks on the other. In this part of the boundary there is added to the boundary problem of delimiting continental shelves and EEZ the question of delimiting an area of overlapping territorial seas. This comes about because Zuqar and Hanish, attributed to the sovereignty of Yemen, both generate territorial seas which overlap with those generated by the Haycocks and South West Rocks, attributed to the sovereignty of Eritrea. It would appear from Yemen Map 12.1 that Yemen assumed that Eritrea is entitled only to a strictly 12 mile territorial sea extending from the Eritrean base points chosen by Yemen along the high-water line on the Eritrean coast; the outcome would be, according to Yemen, that the Haycocks and South West Rocks are thus left isolated outside and beyond the Eritrean territorial sea proper.
155. This proposition is questionable, quite apart from the obvious impracticality of establishing limited enclaves around islands and navigational hazards in the immediate neighbourhood of a main international shipping lane. There is no doubt that an island, however small, and even rocks provided they are indeed islands proud of the water at high-tide, are capable of generating a territorial sea of up to 12 miles (Article 121.2 of the Convention). It follows that a chain of islands which are less than 24 miles apart can generate a continuous band of territorial sea. This is the situation of the Eritrean islands out to, and including, the South West Rocks.
156. The point that the Yemen suggestion omits to take into account is that the effect of what has been referred to as “leap-frogging” the Eritrean islands and islets in this area is to extend the mainland coast territorial sea beyond the limit of 12 miles from the mainland coast. According to Article 3 of the Convention, the territorial sea extends “up to a limit not exceeding 12 nautical miles, measured from the baselines determined in accordance with this Convention”. This is permissible because each island, however small or unimportant of itself, creates a further low-water baseline from which the coastal territorial sea is to be measured. This “leap-frogging” point was invoked strongly in support of Eritrea’s claims to sovereignty. This reasoning was not accepted by the Tribunal in its Award on Sovereignty, it nonetheless has relevance in the present context.
157. If any further reason were needed to reject the Yemen suggestion of enclaving the Eritrean islands in this area beyond a limit of 12 miles from the high-water line of the mainland coast, it may be found in the principle of non-encroachment which was described by Judge Lachs in the Guinea/Guinea-Bissau Award(15) in the following terms:
As stated in the award, our principal concern has been to avoid, by one means or another, one of the Parties finding itself faced with the exercise of rights, opposite to and in the immediate vicinity of its coast, which might interfere with its right to development or put its security at risk.
158. It will be seen that the international boundary line must therefore lie somewhere in a belt of sea no more than four or five miles wide. Once it is established that there is an area of Eritrean mainland coast territorial sea, potentially extending beyond the South West Rocks and the Haycock group of islands on the one hand and overlapping the territorial sea generated by the Yemen islands of the Hanish group on the other, the situation suggests a median line boundary. Under Article 15 of the Convention the normal methods for drawing an equidistant median line could be varied if reason of historic title or other special circumstance were to indicate otherwise. However, the Tribunal has considered these reasons and circumstances and finds no variance necessary.
159. Further bearing in mind its overall task of delimitation, the Tribunal also finds this line to be an entirely equitable one. The decision of the Tribunal is therefore that the median line is the international boundary line where it cuts through the area of overlap of the respective territorial seas of the Parties.
There remains, however, the part of the boundary line which is to connect the mainland coastal line and the line delimiting the overlapping territorial seas. To the description of this line the Award now turns.
The Boundary Line Which Connects Turning Point 13 and Turning Point 15
160. If the mainland coastal median were continued south of turning point 13, it would cut first the territorial sea of Zuqar and then the territorial sea of Hanish, and then cut through the land territory of the island of Hanish. It must therefore divert to the west round the Zuqar-Hanish group, also respecting the territorial seas of these islands if they are to be regarded as generating a territorial sea. That they ought be regarded as having a territorial sea seems reasonable.
161. Various possibilities were considered by the Tribunal. If therefore the international boundary is, after turning point 13 where it meets a 12 mile territorial sea extending from the island of Zuqar, to be diverted in order to respect that area of territorial sea, it could trace the sinuosities of the Zuqar territorial sea boundary until it has to turn southward again in order to join the Article 15 boundary. The Tribunal has decided, however, that it would be better that the line here should be a geodetic line joining point 13 with point 14, making the necessary southwestwards excursion to join the territorial sea median line described above. Moreover, the Tribunal’s task is, as mentioned above, to determine the maritime boundary; this does not include setting the limits of the territorial seas.
162. From turning point 14, again with a simple line in view, the southward excursion of the international boundary is a geodetic line joining points 14 and 15 where it becomes the Article 15 median. This boundary decided upon by the Tribunal betweenturning points14 and 15 is also very near to the putative boundary of a Yemen territorial sea in this area, but makes for a neater and more convenient international boundary.
The Southern Part of the International Boundary Line
163. From turning point 20, which is the southernmost turning point on the overlapping territorial seas median line, the boundary needs to turn generally south-eastwards to rejoin the mainland coast median line. This it does through a geodetic line which connects turning point 20 and point 21, the latter being the intersection of the extended overlapping territorial seas median line and the coastal median line. Thence the international boundary line resumes as a median line controlled by the two mainland coasts. The Bay of Assab is internal waters, so the controlling base points of the boundary line are seaward of this bay.