Full opinion text
Justice O’Connor delivered the opinion of the Court. This original action presents a dispute between the United States and the State of Alaska over the ownership of submerged lands along Alaska’s Arctic Coast. In 1979, with leave of the Court, 442 U. S. 937, the United States filed a bill of complaint setting out a dispute over the right to offer lands in the Beaufort Sea for mineral leasing. Alaska counterclaimed, seeking a decree quieting its title to coastal submerged lands within two federal reservations, the National Petroleum Reserve-Alaska and the Arctic National Wildlife Range (now the Arctic National Wildlife Refuge). The Court appointed a Special Master. 444 U. S. 1065 (1980). Between 1980 and 1986, the Special Master oversaw extensive hearings and briefing. Before us now are the report of the Special Master and the exceptions of the parties. We overrule Alaska’s exceptions and sustain that of the United States. I Alaska and the United States dispute ownership of lands underlying tidal waters off Alaska’s North Slope. The region is rich in oil, and each sovereign seeks the right to grant leases for offshore exploration and to share in oil and gas revenues from the contested lands. Several general principles govern our analysis of the parties’ claims. Ownership of submerged lands — which carries with it the power to control navigation, fishing, and other public uses of water — is an essential attribute of sovereignty. Utah Div. of State Lands v. United States, 482 U. S. 193, 195 (1987). Under the doctrine of Lessee of Pollard v. Hagan, 3 How. 212, 228-229 (1845), new States are admitted to the Union on an “equal footing” with the original 13 Colonies and succeed to the United States’ title to the beds of navigable waters within their boundaries. Although the United States has the power to divest a future State of its equal footing title to submerged lands, we do not “lightly infer” such action. Utah Div. of State Lands, supra, at 197. In United States v. California, 332 U. S. 19 (1947) (California I), we distinguished between submerged lands located shoreward of the low-water line along the State’s coast and submerged lands located seaward of that line. Only lands shoreward of the low-water line — that is, the periodically submerged tidelands and inland navigable waters — pass to a State under the equal footing doctrine. The original 13 Colonies had no right to lands seaward of the coastline, and newly created States therefore cannot claim them on an equal footing rationale. Id., at 30-33. Accordingly, the United States has paramount sovereign rights in submerged lands seaward of the low-water line. Id., at 33-36. In 1953, following the California I decision, Congress enacted the Submerged Lands Act, 67 Stat. 29, 43 U. S. C. § 1301 et seq. That Act “confirmed” and “established” States’ title to and interest in “lands beneath navigable waters within the boundaries of the respective States.” § 1311(a). The Act defines “lands beneath navigable waters” to include both lands that would ordinarily pass to a State under the equal footing doctrine and lands over which the United States has paramount sovereign rights, beneath a 3-mile belt of the territorial sea. § 1301(a). The Act essentially confirms States’ equal footing rights to tidelands and submerged lands beneath inland navigable waters; it also establishes States’ title to submerged lands beneath a 3-mile belt of the territorial sea, which would otherwise be held by the United States. California ex rel. State Lands Comm’n v. United States, 457 U. S. 273, 283 (1982). The Alaska Statehood Act expressly provides that the Submerged Lands Act applies to Alaska. Pub. L. 85-508, §6(m), 72 Stat. 343 (1958). As a general matter, then, Alaska is entitled under both the equal footing doctrine and the Submerged Lands Act to submerged lands beneath tidal and inland navigable waters, and under the Submerged Lands Act alone to submerged lands extending three miles seaward of its coastline. In hearings before the Special Master, the parties identified 15 specific issues for resolution, which we treat in three groups. First, the parties disputed the legal principles governing Alaska’s ownership of submerged lands near certain barrier islands along the Arctic Coast. Second, the parties contested the proper legal characterization of particular coastal features, including a gravel and ice formation in the Flaxman Island chain known as Dinkum Sands. Third, the parties disputed whether, when Alaska became a State, the United States retained ownership of certain submerged lands located within two federal reservations, the National Petroleum Reserve-Alaska in the northwest and the Arctic National Wildlife Refuge in the northeast. For each reservation, the Master considered both whether the seaward boundary encompassed certain disputed waters and whether particular executive and congressional actions prevented the lands beneath tidally influenced waters from passing to Alaska at statehood. Alaska excepts to three of the Master’s recommendations. First, it claims that the Master erred in concluding that waters between the Alaskan mainland and certain barrier islands were not “inland waters,” the limits of which would form a portion of the State’s coastline for purposes of measuring the State’s 3-mile Submerged Lands Act grant. Alaska argues that, at the time of its statehood, the United States had a clear policy of enclosing waters behind near-fringing islands as “inland waters.” In abandoning that policy in 1971, Alaska argues, the Federal Government impermissibly “contracted” Alaska’s recognized territory. Second, the State challenges the Master’s conclusion that Dinkum Sands is not an “island.” Under the Master’s approach, the low-water line on Dinkum Sands is not part of Alaska’s coastline, and the State cannot claim ownership of submerged lands, covering an area of 28 square miles, surrounding the feature. Alaska argues that the Master erred in construing the relevant definition of an “island” and in applying that definition to Dinkum Sands. Third, the State claims that the Master erred in determining that the United States retained ownership of certain submerged lands within the boundaries of the National Petroleum Reserve at Alaska’s statehood. Alaska argues both that the Executive lacked authority to prevent submerged lands from passing to Alaska, and that any attempt to include submerged lands within the Reserve was not sufficiently clear to defeat Alaska’s title under the equal footing doctrine or under the Submerged Lands Act. The United States excepts to the Master’s recommendation concerning the Arctic National Wildlife Refuge. The Master concluded, among other things, that an administrative application for the Refuge was insufficient to “set apart” submerged lands within the proposed boundaries. As a result, the Master concluded, submerged lands within the Refuge passed to Alaska at statehood. We consider these exceptions in turn. By applying the Submerged Lands Act to Alaska through the Alaska Statehood Act, see Pub. L. 85-508, § 6(m), 72 Stat. 343 (1958), Congress granted the State title to submerged lands beneath a 3-mile belt of the territorial sea, measured from the State’s “coast line.” 43 U. S. C. §§ 1301(a)(2), 1311(a). The Act defines the term “coast line” as “the line of ordinary low water along that portion of the coast which is in direct contact with the open sea and the line marking the seaward limit of inland waters.” § 1301(c). Alaska’s first exception requires us to consider how the presence of barrier islands along its northern shore affects the delimitation of its coastline. The issue is of primary relevance in the Beaufort Sea, between the National Petroleum Reserve-Alaska and the Arctic National Wildlife Refuge. A joint federal-state sale of mineral leases covering this so-called Leased Area, conducted in December 1979, yielded large sums now held in escrow awaiting the outcome of this suit. In cases in which the Submerged Lands Act does not expressly address questions that might arise in locating a coastline, we have relied on the definitions and principles of the Convention on the Territorial Sea and the Contiguous Zone, Apr. 29, 1958, [1964] 15 U. S. T. 1606 (Convention). See United States v. California, 381 U. S. 139, 166 (1965) (California II). Specifically, the coastline from which a State measures its Submerged Lands Act grant corresponds to the “baseline” from which the United States measures its territorial sea under the Convention. The Government argued before the Special Master that the United States measures its territorial sea from a “normal baseline” — the low-water line along the coast, Art. 3, supplemented by closing lines drawn across bays and mouths of rivers, see Arts. 7, 13, Under Article 10(2) of the Convention, each island has its own belt of territorial sea, measured outward from a baseline corresponding to the low-water line along the island’s coast. Although the United States now claims a territorial sea belt of 12 nautical miles, see Presidential Proclamation No. 5928, 3 CFR 547 (1988 Comp.), note following 43 U. S. C. § 1331, we are concerned in this case only with the 3-mile belt of the territorial sea that determines a State’s Submerged Lands Act grant. Under Article 6 of the Convention, the outer limit of that territorial sea belt is a line every point of which is three miles from the nearest point of the baseline. This means of measuring the outer limit of the belt is also known as the “arcs-of-circles” method. Alaska objected to application of the Article 3 “normal baseline” approach to its Arctic Coast. In the Leased Area of the Beaufort Sea, some offshore islands are more than six miles apart or more than six miles from the mainland. If Alaska owns only those offshore submerged lands beneath each 3-mile belt of territorial sea, the United States will own “enclaves” of submerged lands, wholly or partly surrounded by state-owned submerged lands, beneath waters more than three miles from the mainland but not within three miles of an island. Two such federal enclaves exist in the Leased Area between the mainland and the Flaxman Island chain, beneath the waters of Stefansson Sound. To eliminate these enclaves, Alaska offered alternative theories for determining the seaward limit of its submerged lands in the vicinity of barrier islands. Alaska principally contended that the United States should be required to draw “straight baselines” connecting the barrier islands and to measure the territorial sea from those baselines. Article 4 of the Convention permits a nation to use straight baselines to measure its territorial sea “[i]n localities where the coast line is deeply indented and cut into, or if there is a fringe of islands along the coast in its immediate vicinity.” The parties agree that Alaska’s coastline satisfies this description. Under this approach, waters landward of the baseline would be treated as “inland” waters, and Alaska would own all submerged lands beneath those waters. The Master rejected this approach, finding that the use of straight baselines under Article 4 is permissive, not mandatory, and that the decision whether to use straight baselines is normally one for the Federal Government. Report of the Special Master 45 (hereinafter Report). The United States has never opted to draw straight baselines under Article 4. See California II, supra, at 167-169; United States v. Louisiana, 394 U. S. 11, 72-73 (1969) (Louisiana Boundary Case); United States v. Louisiana, 470 U. S. 93, 99 (1985) (Alabama and Mississippi Boundary Case); United States v. Maine, 475 U. S. 89, 94, n. 9 (1986) (Massachusetts Boundary Case). As a variant of its straight baselines argument, Alaska claimed that the United States has historically treated waters between the mainland and fringing islands as “inland waters,” so long as the openings between the off-lying islands are no more than 10 miles wide. Alaska did not argue that the United States had ever specifically asserted, in its dealings with foreign nations, that the waters of Stefansson Sound are inland waters. Rather, Alaska attempted to identify a general but consistent “10-mile rule” invoked by the United States in its domestic and international affairs. If applied to Alaska’s Arctic Coast, the State argued, this rule would require treating the waters of Ste-fansson Sound as inland waters. The Master examined the boundary delimitation practices of the United States and concluded that the United States did not have a well-established rule for treating waters between the mainland and fringing islands as inland waters. The Master recognized that, in the Alabama and Mississippi Boundary Case, we suggested that between 1903 and 1961 the United States had “enclosed] as inland waters those areas between the mainland and off-lying islands that were so closely grouped that no entrance exceeded 10 geographical miles.” 470 U. S., at 106-107. Observing that this statement was not “strictly necessary” to the decision in the Alabama and Mississippi Boundary Case, the Master declined to rely on it here. The Master therefore concluded that, for purposes of measuring Alaska’s submerged lands, the State’s coastline should correspond to a normal baseline under Article 3 of the Convention. For the reasons discussed below, we find no error in the Master’s approach. A Under the Convention, a nation’s past boundary delimitation practice is relevant in a narrow context: specifically, when a nation claims that certain waters are “historic” inland waters under Article 7(6) of the Convention. If certain geographic criteria are met, Article 7(4) of the Convention permits a nation to draw a “closing line” across the mouth of a bay and to measure its territorial sea outward from that line. Waters enclosed by the line are considered internal waters. Article 7(6) also permits a nation to enclose “historic” bays, even if those waters do not satisfy the geographic criteria of Article 7(4). For a body of water to qualify as a historic bay, the coastal nation “must have effectively exercised sovereignty over the area continuously during a time sufficient to create a usage and have done so under the general toleration” of the community of nations. Id., at 102 (citing Juridical Regime of Historic Waters, Including Historic Bays 56, U. N. Doc. A/CN.4/143 (1962)) (internal quotation marks omitted). Accordingly, where a State within the United States wishes to claim submerged lands based on an area’s status as historic inland waters, the State must demonstrate that the United States: (1) exercises authority over the area; (2) has done so continuously; and (3) has done so with the acquiescence of foreign nations. See Alabama and Mississippi Boundary Case, supra, at 101-102. Recognizing these strict evidentiary requirements, Alaska does not contend that the waters of Stefansson Sound are historic inland waters. Alaska does not purport to show any specific assertion by the United States that the waters of Stefansson Sound are inland waters. Rather, Alaska argues that, at the time it was admitted to the Union, the United States had a general, publicly stated policy of enclosing as inland waters areas between the mainland and closely-grouped fringing islands. If this general formula is applied to the Alaska’s Arctic Coast, the State argues, the waters of Stefansson Sound qualify as inland waters. Alaska maintains that this policy was in effect from the early 1900’s to 1971, when the United States published a set of charts strictly applying the arcs-of-circles method to Stefansson Sound. In Alaska’s view, relying solely on the Convention’s normal baseline approach to delimit the State’s submerged lands impermissibly contracts the State’s recognized territory from that which existed at the time of statehood. Since adopting the Convention’s definitions to give content to the Submerged Lands Act, we have never sustained a State’s claim to submerged lands based solely on an assertion that the United States had adhered to a certain general boundary delimitation practice at the time of statehood. In the Louisiana Boundary Case, we left open the possibility that Louisiana could claim ownership of certain submerged lands by demonstrating a “firm and continuing international policy” of enclosing waters between the mainland and island fringes as “inland waters.” 394 U. S., at 74, n. 97. Had that been the United States’ “consistent official international stance,” the Government “arguably could not abandon that stance solely to gain advantage in a lawsuit to the detriment of Louisiana.” Ibid. In that litigation, the State ultimately failed to demonstrate any firm and continuing international policy of enclosing waters behind island fringes as inland waters. See United States v. Louisiana, 420 U. S. 529, 529-530 (1975) (per curiam) (decree) (accepting Master’s recommendation that certain actions by the United States did not establish a general policy of applying straight baselines to near-fringing islands); Report of Special Master in United States v. Louisiana, O. T. 1974, No. 9 Orig., pp. 7-13. Alaska nevertheless claims that in the Alabama and Mississippi Boundary Case the Court identified a “firm and continuing” 10-mile rule for fringing islands. Alaska first contends that the Alabama and Mississippi Boundary Case precludes the Government from claiming that the waters of Stefansson Sound are not inland waters. The State then argues in the alternative that independent evidence supports its formulation of the rule. We address Alaska’s points in turn. B In the Alabama and Mississippi Boundary Case, the Court considered the States’ claim that the waters of Mississippi Sound constituted “historic” inland waters under Article 7(6) of the Convention. In discussing whether the States had shown that the United States had continuously asserted the inland water status of Mississippi Sound, the Court identified a general policy “of enclosing as inland waters those areas between the mainland and off-lying islands that were so closely grouped that no entrance exceeded 10 geographical miles.” 470 U. S., at 106. Alaska argues that the Government is estopped from questioning application of this general coastline delimitation practice to its Arctic Coast. Alaska recognizes the rule that the doctrine of nonmutual collateral estoppel is generally unavailable in litigation against the United States, see United States v. Mendoza, 464 U. S. 154, 160-163 (1984), but suggests that the policy considerations underlying this rule do not apply to cases arising under the Court’s original jurisdiction, where the Court acts as factfinder and the United States has an incentive to fully litigate all essential issues. We have not had occasion to consider application of nonmu-tual collateral estoppel in an original jurisdiction case, and we see no reason to develop an exception to Mendoza here. Even if the doctrine applied against the Government in an original jurisdiction case, it could only preclude relitigation of issues of fact or law necessary to a court’s judgment. Montana v. United States, 440 U. S. 147, 153 (1979); Mendoza, supra, at 158. A careful reading of the Alabama and Mississippi Boundary Case makes clear that the Court, did not attach controlling legal significance to any general delimitation formula. The Master in that case recited a series of statements and precedents following Mississippi’s admission to the Union supporting the view that the Federal Government had treated the waters of Mississippi Sound as inland waters. These statements included multiple references to a rule for closing gulfs, bays, and estuaries with mouths less than 10 miles wide as inland waters, Report of Special Master in Alabama and Mississippi Boundary Case, O. T. 1983, No. 9 Grig., pp. 40, 42, 48-49, 52, and to a rule for closing straits leading to inland waters, id., at 42, 49-50. In addition, the Master cited a 1961 letter from the Solicitor General to the Director of the United States Coast and Geodetic Survey concerning coastline delimitation principles for the Gulf of Mexico, proposing to treat “ ‘[wjaters enclosed between the mainland and offiying islands ... so closely grouped that no entrance exceeds ten miles’” as inland waters. Id., at 52. In excepting to the Master’s conclusion that the waters of Mississippi Sound qualified as historic inland waters, the United States argued that the “generalized ... formulations” recited by the Master could not support the States’ claim, without evidence of specific federal claims to inland waters status for Mississippi Sound. Exceptions of United States in Alabama and Mississippi Boundary Case, O. T. 1983, No. 9 Orig., pp. 32-33. The Court assumed that the United States’ position was correct, but concluded that the States had in fact identified “specific assertions of the status of [Mississippi] Sound as inland waters.” 470 U. S., at 107; see id., at 108-110. In light of the Court’s assumption that specific assertions of dominion would be critical to the States’ historic title claim, we cannot conclude that any general delimitation policy identified in the Alabama and Mississippi Boundary Case is controlling here. The Court’s inquiry in the Alabama and Mississippi Boundary Case was not whether the States had demonstrated a “firm and continuing international policy” of enclosing waters between the mainland and island fringes as inland waters, sufficiently well defined to cover the waters of Mississippi Sound. Rather, the inquiry was whether the States had demonstrated that the Sound met the specific requirements for a historic inland waters claim under Article 7(6) of the Convention. In the context of that claim, the variation or imprecision in the United States’ general boundary delimitation principles might have been irrelevant because the State could point to specific federal assertions that Mississippi Sound consisted of inland waters. But variation and imprecision in general boundary de- * limitation principles become relevant where, as here, a State relies solely on such principles for its claim that certain waters were inland waters at statehood. The United States is therefore free to argue that any 10-mile rule is not sufficiently well defined to support Alaska’s claim that the waters of Stefansson Sound constitute inland waters. C Alaska argues that even if principles of collateral estoppel do not apply, the evidence before the Master established that the United States had a well-defined, “firm and continuing” 10-mile rule that would require treating certain areas along Alaska’s Arctic Coast as inland waters. The Master exhaustively cataloged documents and statements reflecting the United States’ views and practices on boundary delimitation, both in its international relations and in disputes with various States, between 1903 and 1971. The Master found that “the exact nature of the United States’ historic practice is a matter of some intricacy,” and concluded that any 10-mile rule was not sufficiently well defined to require treating the waters of Stefansson Sound as inland waters. Report 55. Alaska argues that the Master afforded “undue significance to minor variations in the way the United States expressed its otherwise consistent policy over time, ignoring the principle that minor uncertainties and even contradictions in a nation’s practice are legally insignificant.” Exceptions of State of Alaska 14 (Alaska Exceptions Brief). The relevant sources do not bear out Alaska’s claim. Of particular importance for our analysis is the position of the United States in its foreign relations between 1930 and 1949. In March 1930, the United States formally proposed certain principles for delimiting inland waters to the League of Nations Conference for the Codification of International Law. See 3 Acts of the Conference for the Codification of International Law, Territorial Waters 195-201 (1930) (Acts of the Conference). As the Geographer of the Department of State later observed, where the mainland and offshore islands are assigned individual 3-mile belts of territorial sea, there will remain “small pockets of the high sea deeply indenting territorial waters.” U. S. Exh. 85-223 (Boggs, Delimitation of the Territorial Sea, 24 Am. J. Int’l L. 541, 552 (1930)). Because such pockets would “constitute no useful portion of the high sea from the viewpoint of navigation,” ibid., the United States proposed that countries “assimilate” these small enclaves of high seas to the adjacent territorial sea where a single straight line of no more than four nautical miles in length would enclose an enclave, 3 Acts of the Conference 201. At the same Conference, the United States also proposed a rule for straits. Where a strait connected “two seas having the character of high seas,” the waters of the strait would be considered territorial waters of the coastal nation, as long as both entrances of the strait were less than six nautical miles wide. Id., at 200. Where a strait was “merely a channel of communication with an inland sea,” rules regarding closing of bays would apply. Id., at 201. Under those rules, waters shoreward of closing lines less than 10 nautical miles in length would be treated as “inland” waters. Id., at 198. The United States’ 1930 “assimilation” proposal is inconsistent with Alaska’s assertion that, since the early 1900’s, the United States had followed a firm and continuing 10-mile rule for fringing islands. If the United States’ policy had been to draw a baseline connecting islands no more than 10 miles apart, all waters between that line and the mainland would have been treated as “inland waters.” Under the 1930 formula, however, there were “small pockets of the high sea” between that line and the mainland, and those pockets would have been assimilated to territorial waters (that is, waters seaward of the coastline), not to inland waters (that is, waters enclosed by the coastline). Alaska now argues that the 1930 assimilation proposal “was at most one of the legally insignificant uncertainties or contradictions” rather than a change from a firm 10-mile rule. Alaska Exceptions Brief 25 (internal quotation marks omitted). Alaska took a different position before the Special Master, where it argued that the United States “unequivocally embraced the ‘assimilation’ practice as the official United States position” between 1930 and 1949. Brief for Alaska on Island Fringes 54, 60-61; see Alaska Exh. 85-63 (Memorandum of United States in Response to Request of Special Master in United States v. California, O. T. 1949, No. 11 Orig., p. 19); Alaska Exh. 85-82 (Aide-Mémoire from the Department of State to the Government of Norway, Sept. 29, 1949, pp. 4-5). Alaska cannot explain why the United States would have pointed to the assimilation formula as its official position between 1930 and 1949 if a 10-mile rule for islands was in effect during that time. Nor does the United States’ proposal on straits demonstrate a policy of connecting near-fringing islands with straight baselines of less than 10 miles. If the mainland and offshore islands form the two coasts of a strait, under the United States’ proposal the strait would be treated as territorial waters (not inland waters) if it linked two areas of high seas. The distance between the fringing islands may have some bearing on whether those islands in fact form, the coast of a strait, but not on whether the waters they enclose are territorial or inland waters. In other words, under the 1930 proposal, the character of the waters to which a strait leads, not the distance between the islands forming one coast of the strait, determines the character of the strait itself. Rather than treating the mainland and a line connecting fringing islands as the two coasts of a strait, Alaska appears to view a passageway between two offshore islands, leading to the waters between the islands and the mainland, as a strait. With this geographic configuration in mind, Alaska argues that the proposal to apply a 10-mile bay-closing rule to a strait serving as a “channel of communication with an inland sea” is “fully consistent” with a 10-mile rule. Alaska Exceptions Brief 25. But even under this approach, a rule that straits leading to an inland sea are themselves inland waters is not equivalent to a simple 10-mile rule. Again, under the United States’ 1930 proposal, the character of the strait depends on the character of the waters to which it leads. A 10-mile bay-closing rule would apply only if the waters between the strait and the mainland were inland waters under some other principle. Under the simple 10-mile rule that Alaska advocates, the fact that the islands are less than 10 miles apart itself determines that the waters behind the islands are inland waters. In sum, although Alaska is correct that the United States’ position at the League of Nations Conference did not call for strict application of the arcs-of-circles method, ibid., neither the assimilation proposal nor the proposal for straits is fully consistent with a simple rule that islands less than 10 miles apart enclose inland waters. The discussion above leads to the conclusion that, if the United States had a 10-mile rule at Alaska’s statehood, that rule developed after 1949. Even if a rule developed within a decade of Alaska’s statehood could be considered a “firm and continuing” one, Alaska has not shown that any such rule would encompass the islands off its Arctic Coast. For the period between 1950 and Alaska’s statehood, Alaska focuses principally on the United States’ position in a series of disputes with States over ownership of submerged lands in the vicinity of near-fringing islands, rather than on positions taken in its international relations. First, in 1950, the State Department and the Justice Department proposed a boundary between Louisiana’s inland and territorial waters for use in the Louisiana Boundary Case. That boundary, known as the Chapman Line, followed certain barrier islands along Louisiana’s southeast coast, enclosing Chandeleur and Breton Sounds and Calliou Bay as inland waters. According to Alaska, the Chapman Line shows the use of a simple 10-mile rule. Second, in 1951, the Justice Department asked the State Department to outline the United States’ approach to demarcating inland and territorial waters, for purposes of submerged lands litigation between the United States and California. A letter from the Acting Secretary of State stated that an island “was to be surrounded by its own belt of territorial waters measured in the same manner as in the case of the mainland.” Alaska Exh. 85-94 (Letter from James E. Webb to J. Howard McGrath, Attorney General, Nov. 18, 1951, p. 3). The letter also drew upon the 1930 Hague proposals for straits, noting that the waters of a strait connecting high seas were never inland waters, but that bay-closing rules should apply to a strait serving as “a channel of communication to an inland sea.” Id., at 4. Third, in a submission to the Court in 1958, the United States commented that waters behind certain islands in Louisiana, Mississippi, and Alabama were inland waters. Brief for United States in Support of Motion for Judgment on Amended Complaint in United States v. Louisiana, O. T. 1958, No. 9 Orig., pp. 177, 254, 261. We agree with the Special Master that the United States did not exclusively employ a simple 10-mile rule in its disputes with the Gulf States and with California. The 1951 State Department letter in the California litigation merely echoed the United States’ proposal at the Hague Conference concerning straits leading to inland waters. As discussed supra, at 18-19, a rule applying 10-mile bay-closing principles to straits leading to inland waters would not always lead to the same result as a simple 10-mile rule. Under the former approach, the critical factor is where the strait leads, not the width of the strait. Alaska does not attempt to show that Stefansson Sound is a strait leading to inland waters. Nor does the 1950 Chapman Line reflect a “firm and continuing” policy of enclosing waters behind fringing islands as “inland waters.” The Chapman Line may be consistent with such a policy, but as the Master noted, no contemporaneous document explains the theory behind the Chapman Line in terms of a simple 10-mile rule. Report 85-88. Indeed, a 1950 draft memorandum from the State Department Geographer to the Justice Department opined that Chande-leur and Breton Sounds should be treated as inland waters not only because they were screened by a chain of islands that were less than 10 miles apart, but also because they were “not extensively traversed by foreign vessels” and because the islands covered “more than half the total arc of the territorial sea.” U. S. Exh. 85-400. These criteria go far beyond the simple 10-mile rule, and Alaska does not show how they would apply to Stefansson Sound. Finally, statements in the briefs filed by the United States in litigation with the Gulf States that certain waters behind offshore islands were inland waters do not explicitly rely on a 10-mile rule. Moreover, in our decision in United States v. Louisiana, 363 U. S. 1, 67, n. 108 (1960), we made clear that we did not take the Government’s concession that certain islands off Louisiana’s shore enclosed inland waters “to settle the location of the coastline of Louisiana or that of any other State.” These and other documents considered by the Master support his conclusion that Alaska has not identified a firm and continuing 10-mile rule that would clearly require treating the waters of Stefansson Sound as inland waters at the time of Alaska’s statehood. Indeed, we note that the result Alaska seeks would be in tension with the outcome of the Massachusetts Boundary Case, 475 U. S. 89 (1986), where, a year after deciding the Alabama and Mississippi Boundary Case, we concluded that the waters of Nantucket Sound are not inland waters. Following the Court’s decision in the Alabama and Mississippi Boundary Case, Massachusetts argued that a 10-mile rule would make the waters of Nantucket Sound inland waters. The Master in that case recognized that no entrance between the islands enclosing Nantucket Sound exceeded 10 miles, but nevertheless concluded that Massachusetts had not shown that the waters of Nantucket Sound were inland waters. Report of Special Master in Massachusetts Boundary Case, O. T. 1984, No. 35 Orig., pp. 69.2-70. We rejected the Commonwealth’s claim to inland waters status for Nantucket Sound, framed in its exception to the Master’s recommendation as an “ancient title” claim. Massachusetts Boundary Case, supra, at 105. If the case could have been resolved by reference to a simple 10-mile rule for all fringing islands, we need not have entertained such a claim. D In sum, we conclude that Alaska’s entitlement to submerged lands along its Arctic Coast must be determined by applying the Convention’s normal baseline principles. The Alabama and Mississippi Boundary Case does not foreclose this conclusion. The sources before the Master showed that, in its foreign relations, particularly in the period 1930 to 1949, the United States had advocated a rule under which objectionable pockets of high seas would be assimilated to a coastal nation’s territorial sea. Such a rule would have been inconsistent with the maintenance of a 10-mile rule for fringing islands. The United States also advocated a rule for treating the waters of a strait leading to an inland sea as inland waters, but that rule is not equivalent to Alaska’s simple 10-mile rule. Whether the waters of Stefansson Sound would be considered inland waters under the 1930 proposal for straits is unclear. Accordingly, we overrule Alaska’s first exception. 1 — 1 H-Í Alaska next excepts to the Master’s conclusion that a small gravel and ice formation in the Flaxman Island chain, known as Dinkum Sands, is not an island. Whether Dinkum Sands is an island affects Alaska’s ownership of offshore submerged lands in the feature’s vicinity. As discussed above, a State’s coastline provides the starting point for measuring its 3-mile Submerged Lands Act grant. See 43 U. S. C. §§ 1301(a)(2), 1311(a). Generally, the State’s coastline corresponds to a “baseline” from which, under the 1958 Convention, the United States measures its territorial sea for international purposes. Supra, at 8. Article 10(1) of the Convention defines an island as “a naturally-formed area of land, surrounded by water, which is above water at high-tide.” A line of ordinary low water along the coast of an island can serve as a baseline for measuring the territorial sea. See Arts. 10(2), 3. The Convention also permits a nation to claim a belt of territorial sea around certain features that are not above water at high tide, so long as they are located wholly or partly within the territorial sea bélt of the mainland or an island. Arts. 11(1)-(2). Again, for purposes of determining a State’s ownership rights under the Submerged Lands Act, we are concerned with a 3-mile belt of the territorial sea. See supra, at 8-9. Because Dinkum Sands is not within three miles of the nearest islands or the mainland, it does not meet the requirements of Article 11. Accordingly, Dinkum Sands has its own belt of territorial sea — and Alaska owns submerged lands beneath that belt — only if Dinkum Sands satisfies the requirements of Article 10(1). The issue here has been narrowed to whether Dinkum Sands is “above water at high-tide.” Dinkum Sands has frequently been submerged. Apart from daily shifts in the tide and seasonal shifts in sea level, the feature itself changes height. Report 275, 280-283, 309, n. 66. This phenomenon may be at least in part attributable to what the United States’ expert witness termed “ice collapse.” Dinkum Sands is formed by layers of ice and gravel mixed with ice. As the summer months approach, ice within Dinkum Sands melts and the feature slumps in elevation. 7 Tr. 986-987, 8 Tr. 1060-1062 (July 23, 1984). Alaska and the United States agree that “high-tide” under Article 10(1) should be defined as “mean high water,” an average measure of high water over a 19-year period. Cf. United States v. California, 382 U. S. 448, 449-450 (1966) (per curiam) (entering decree defining an island as “above the level of mean high water” and defining mean high water as “the average elevation of all the high tides occurring over a period of 18.6 years”); Borax Consol., Ltd. v. Los Angeles, 296 U. S. 10, 26-27 (1935) (approving definition of “mean high tide line” based on “average height of all the high waters ... over a considerable period of time,” at least 18.6 years). They disagree over how frequently a feature of variable elevation such as Dinkum Sands must be above mean high water to qualify as an island. Based on the drafting history of Article 10, the Master concluded that an island must “generally,” “normally,” or “usually” be above mean high water. Report 302. Applying this standard, the Master reviewed historical hydrographic and cartographic evidence and the results of a joint monitoring project conducted by the parties in 1981 and 1982. He concluded that Dinkum Sands is frequently below mean high water and therefore is not an island. Id., at 310. Alaska excepts to this conclusion on three grounds. First, Alaska challenges the legal conclusion that Article 10(1) requires an island to be above mean high water at least “generally,” “normally,” or “usually.” Second, Alaska disputes the Master’s factual finding that Dinkum Sands is frequently below mean high water. Finally, Alaska argues that Dinkum Sands should be treated as an island when it is in fact above mean high water. We And no error in the Master’s conclusion. A In the proceedings before the Master, the United States took the position that an island must be “permanently” above mean high water, Brief for United States on Dinkum Sands 17-29, while Alaska argued that Article 10 permits a feature “to slump on occasion below” mean high water but still qualify as an island, Brief for Alaska on Dinkum Sands 64. The Master essentially rejected the United States’ position in favor of a somewhat more lenient standard, under which an island must “generally,” “normally,” or “usually” be above mean high water. Although Alaska now objects to this standard, Alaska Exceptions Brief 44-45, 51, it sets forth no clear alternative. Alaska’s observation that “an island that is occasionally submerged is no less an island,” id., at 45, is not inconsistent with the Master’s approach. If Alaska is now implicitly claiming that a feature need appear only episodically above mean high water to qualify as an island, its position is without merit. Because Article 10(1) does not specify how frequently a feature must be above mean high water to qualify as an island, we must look to the Convention’s drafting history for guidance. See Louisiana Boundary Case, 394 U. S., at 42-47. In urging that the Master’s interpretation of Article 10(1) is inconsistent with the development of that provision, Alaska focuses on the fact that earlier drafts specified that an island must be “permanently above high-water mark.” Report 297 (citing J. Francois, Report on the Régime of the Territorial Sea, [1952] 2 Y. B. Int’l L. Comm’n 25, 36, U. N. Doc. A/CN.4/53 (in French; translation from Alaska Exh. 84A-21, p. 41)); see Alaska Exceptions Brief 50. The eventual deletion of the modifier “permanently,” in Alaska’s view, suggests that Article 10(1) contains no implicit modifier at all, such as “generally,” “normally,” or “usually.” Alaska’s reading of Article 10(l)’s drafting history is selective. In fact, the drafting history supports a standard at least as stringent as that adopted by the Master. The provision was first introduced at the League of Nations Conference for the Codification of International Law, held at The Hague in 1930. A preparatory committee offered the following as a basis for discussion: “In order that an island may have its own territorial waters, it is necessary that it should be permanently above the level of high tide.” 2 Conference for the Codification of International Law, Bases of Discussion, Territorial Waters 54 (1929). A subcommittee revised the definition but retained the element of permanence: “An island is an area of land, surrounded by water, which is permanently above high-water mark.” 3 Acts of the Conference 219. When the International Law Commission of the United Nations revived the work of the Conference in 1951, a special rapporteur reintroduced the subcommittee’s definition. Report 297. In 1954, the British delegate proposed adding the modifier “in normal circumstances,” so that an island’s status would not be questioned because it was temporarily submerged at high tide in an “exceptional cas[e].” See Summary Records of the 260th Meeting, [1954] 1 Y. B. Int’l L. Comm’n 92. The Commission adopted that proposal, id., at 94, and in its final report defined an island as “an area of land, surrounded by water, which in normal circumstances is permanently above high-water mark,” Report of the International Law Commission to the General Assembly, Art. 10, U. N. Gen. Ass. Off. Rec., 11th Sess., Supp. No. 9, U. N. Doc. A/3159, p. 16 (1956). In 1957, an internal State Department memorandum evaluating the Commission’s work suggested that the words “permanently” and “in normal circumstances” appeared to be inconsistent and could both be omitted, because “current international law does not purport to solve such minor problems” as how to treat formations that would be submerged at unusually high states of high tide. Alaska Exh. 84A-21 (Memorandum from Benjamin H. Read, Islands, Drying Rocks and Drying Shoals, Sept. 1957, p. 11). The United States presented that position at the 1958 United Nations Conference on the Law of the Sea, arguing that “there is no established state practice regarding the effect of subnormal or abnormal or seasonal tidal action on the status of islands.” 3 United Nations Conference on the Law of the Sea, Official Records: First Committee (Territorial Sea and Contiguous Zone), Summary Records of Meetings and Annexes, U. N. Doc. A/CONF.13/C1./L.112, p. 242 (1958). The Conference adopted the United States’ recommendation, and excised the words “permanently” and “in normal circumstances” from the definition of an island. As the Master recognized, in including the phrase “in normal circumstances,” the Convention’s drafters had sought to accommodate abnormal events that would cause temporary inundation of a feature otherwise qualifying as an island. Report 300. The United States’ view that the international definition of an island need not address abnormal or seasonal tidal activity ultimately prevailed. But the change from the Commission’s draft to the final language of the Convention did not signal an intent to cover features that are only sometimes or occasionally above high tide. In fact, the problem of abnormal or seasonal tidal activity that the 1954 amendment addressed is fully solved by the United States’ practice of construing “high tide” to mean “mean high water.” Averaging high waters over a 19-year period accounts for periodic variations attributable to astronomic forces; nonperiodic, meteorological variations can be assumed to balance out over this length of time. See 2 A. Shalowitz, Shore and Sea Boundaries 58-59 (1964). Accordingly, even if a feature would be submerged at the highest monthly tides during a particular season or in unusual weather, the feature might still be above “mean high water” and therefore qualify as an island. What Alaska seeks is insular status not for a feature that is submerged at abnormally high states of tide, but for a feature that rises above and falls below mean high water— a tidal datum that has already accounted for the tidal abnormalities about which the drafters of Article 10(1) were concerned. Even if Article 10(1)’s drafting history could support insular status for a feature that slumps below mean high water because of an abnormal change in elevation, it does not support insular status for a feature that exhibits a pattern of slumping below mean high water because of seasonal changes in elevation. Alaska nevertheless contends that there is support for according island status to features more “ephemera[l]” than Dinkum Sands. See Alaska Exceptions Brief 45-50. The authorities Alaska cites all predate the Convention and are therefore unhelpful in construing Article 10(1). Alaska also relies on an analogy to the “mudlumps” of the Mississippi delta, features whose status under the Convention has never been determined. See Report of Special Master in United States v. Louisiana, O. T. 1974, No. 9 Orig., p. 4 (filed July 31, 1974) (concluding that Louisiana’s Submerged Lands Act grant could be measured from two mudlumps, but not deciding whether the mudlumps were islands under Article 10(1) or low-tide elevations under Article 11(1)); United States v. Louisiana, 420 U. S. 529 (1975) (overruling exceptions). In sum, the Convention’s drafting history suggests that, to qualify as an island, a feature must be above high water except in abnormal circumstances. Alaska identifies no basis for according insular status to a feature that is frequently below mean high water. B In disputing the Master’s factual conclusion that Dinkum Sands is “frequently below mean high water,” Report 39, Alaska relies on three cartographic sources. First, two nautical charts produced following a 1949-1950 survey of the Beaufort Sea by a United States Coast and Geodetic Survey party depict Dinkum Sands as an island, consistent with a survey note describing a “new gravel bar baring about three feet” at mean high water. Alaska Exh. 84A-203 (U. S. Coast and Geodetic Survey, Descriptive Report to Accompany Hy-drographic Survey H-7761, p. 3); see Alaska Exh. 84A-202 (U. S. Coast and Geodetic Survey, Addendum to Descriptive Report to Accompany Hydrographic Survey H-7760, p. 4). Second, in 1971, an ad hoc interagency group known as the Baseline Committee, charged with delimiting the United States’ coastline, produced baseline charts treating Dinkum Sands as an island. Third, a 1979 map developed for a joint federal-state oil and gas lease sale in the Prudhoe Bay area assigned ownership of a 3-mile belt of territorial sea around Dinkum Sands to Alaska. As Alaska appears to acknowledge, see Alaska Exceptions Brief 53, the 1971 baseline chart and the 1979 leasing map were based on the 1949-1950 survey rather than independent observations. In 1956, the United States Coast and Geodetic Survey resumed charting Dinkum Sands as a low-tide elevation, based on observations of a Navy vessel made the prior year. It is undisputed that one of the members of the Baseline Committee persuaded the Committee to treat Dinkum Sands as an island based solely on his personal observation of Dinkum Sands as a member of the 1949-1950 survey party. See Alaska Exh. 84A-207 (Department of State, Memorandum to Members of the Baseline Committee, Minutes of Oct. 10,1979, Meeting, p. 2) (noting that the Committee “has used Dinkum Sands as a basepoint for determining the breadth of the territorial sea . . . because early surveys showed Dinkum Sands to be above high water and Admiral Nygren had personally observed it above high water”). The 1979 leasing map relied on the 1971 baseline chart in assigning Dinkum Sands its own 3-mile belt of territorial sea. The question, then, is whether the 1949-1950 survey party’s conclusion that Dinkum Sands is three feet above mean high water, taken together with visual observations of Dinkum Sands above water, undermines the Master’s factual finding that Dinkum Sands is “frequently below mean high water.” Report 309. It does not. Alaska emphasizes that Dinkum Sands has been observed “many times . . . above water” and only “occasionally . . . submerged.” Alaska Exceptions Brief 44. But visual observations of Dinkum Sands are not dispositive; the question is not whether Dinkum Sands is above or below high tide on any given day, but where the feature lies in relation to mean high water. To address precisely this problem, the parties jointly commissioned a $2.5 million study to calculate mean high water in the feature’s vicinity and to determine the feature’s elevation in relation to that datum. First, using a year of tidal readings, the National Ocean Survey computed a mean high-water datum at Dinkum Sands and calculated an error band to account for the fact that the level would ordinarily be based on 19 years of readings. Second, an engineering firm measured Dinkum Sands’ highest points in March, June, and August 1981. Comparing the feature’s highest elevation measurements to the mean high-water level, the Master found that Dinkum Sands was not above mean high water at any time it was surveyed. The two highest points of the survey were within the error band for the mean high-water level, but the Master found this fact to be of little weight because the measurements were likely taken from piles of gravel disturbed by the March measurements, rather than from Dinkum Sands’ true highest points. Alaska continued to measure Dinkum Sands in relation to mean high water in 1982 and 1983. The feature was found to be above mean high water on a visit in July 1982. By September, the feature had fallen in elevation, possibly by more than a foot, see Report 281-282, placing it below the mean high-water datum. Between May and July 1983, the feature was observed above water several times, although its elevation in relation to mean high water was not known. Based on two helicopter observations of the feature and estimates of sea level in relation to mean high water, the Master concluded that Dinkum Sands could have been above high water by a matter of inches in September 1983. The Master found that the feature was “consistently” below mean high water in 1981 and below mean high water by September — the end of the open water season — in both 1981 and 1982. Id., at 309. Relying largely on the 1981-1983 data, the Master concluded that Dinkum Sands is not an island. Alaska makes no mention of the 1981 joint monitoring project. The Master discussed the State’s methodological objections to the results at length, see id., at 255-269, and we see no reason to revisit the Master’s conclusion that those objections are unpersuasive. Alaska does not explain why the Master should have relied on a single August 1949 measurement of Dinkum Sands’ elevation in relation to mean high water rather than on the exhaustive survey expressly designed to determine Dinkum Sands’ status under Article 10(1) of the Convention. In contending that Dinkum Sands has been above mean high water except on a “handful of occasions,” Alaska recognizes that Dinkum Sands slumps in elevation during the open water season between late July and September. Alaska Exceptions Brief 54. Alaska suggests that natural processes build up Dinkum Sands “just... prior to the autumn freeze-up,” and that the feature then remains above mean high water for 9 to 10 months of the year. See ibid. There is no basis in the record, however, for concluding that Dinkum Sands is above mean high water during the winter months. During the winter, the area is completely covered by pack ice. The sole measurement of the feature’s elevation during the winter was that taken in March 1981, and it was then below mean high water. Report 286. But even if the record demonstrated that the feature remained above mean high water until “ice collapse” caused it to slump, that would not compel a ruling in Alaska’s favor. Although Article 10(1)’s drafting history may suggest that a feature submerged at abnormally high tides does not lose its insular status, it does not support the broader conclusion that a feature with a seasonal loss in elevation, bringing it below mean high water, qualifies as an island. See supra, at 27. In sum, we find no error in the Master’s conclusion that Dinkum Sands is frequently below mean high water and therefore does not meet the standard for an island. C Alaska finally urges a compromise resolution, under which Dinkum Sands would be deemed an island when above mean high water. Alaska attempts to find support for its position in this Court’s recognition in prior cases of the concept of an “ambulatory coast line.” Alaska Exceptions Brief 55. In adopting the 1958 Convention to aid interpretation of the Submerged Lands Act, we recognized that the Convention treats a nation’s coastline as its modern, ambulatory coastline. See United States v. Louisiana, 394 U. S. 1, 5 (1969) (Texas Boundary Case); Louisiana Boundary Case, 394 U. S., at 32-34. Shifts in a low-water line along the shore, we acknowledged, could lead to a shift in the baseline for measuring a maritime zone for international purposes. In turn, the State’s entitlement to submerged lands beneath the territorial sea would change. An island may very well have its own ambulatory coastline. What Alaska seeks here, however, is not an entitlement to submerged lands seaward of a gradually accreting or eroding shore. Rather, Alaska’s ownership of submerged lands around Dinkum Sands would appear and disappear periodically, depending upon whether the feature was above or below mean high water. Not only does Article 10(1) of the Convention not support such a reading, but Alaska’s position makes a sensible application of other provisions of the Convention impossible. The Convention separately categorizes features that are below mean high water, but above water at low tide. See Art. 11. In addition, under Articles 10(2) and 3, an island’s belt of territorial sea is measured from the line of low water. As Dinkum Sands’ elevation shifts and the feature slumps toward the mean high-water datum, below the mean high-water datum, and possibly below the low-water datum, the baseline for measuring the surrounding maritime zone would shift and then disappear. Quite apart from the fact that Alaska’s proposal would lead to costly and time-consuming monitoring efforts, we agree with the Master that Alaska has identified no precedent for treating as an island a feature that oscillates above and below mean water. IV Alaska’s third exception concerns the ownership of submerged lands within the National Petroleum Reserve-Alaska (Reserve), a 23-million acre federal reservation in the northwestern part of the State. The Reserve’s seaward boundary runs along the Arctic Ocean from Icy Cape at the west to the mouth of the Colville River at the east. When this litigation began, Alaska and the United States disputed the location of the Reserve’s boundary, focusing in particular on whether the boundary followed the sinuosities of the coast or instead cut across certain inlets, bays, and river estuaries. Alaska initially conceded federal ownership of submerged lands within that boundary. In light of this Court’s decision in Montana v. United States, 450 U. S. 544 (1981), and with the consent of the United States, the Special Master granted Alaska relief from its concession, and Alaska claimed ownership of submerged lands beneath certain coastal features within the Reserve’s boundaries. Order of Special Master in United States v. Alaska, O. T. 1983, No. 84 Orig. (Jan. 4, 1984). A separate proceeding concerning ownership of submerged lands beneath inland navigable waters is pending in Federal District Court, Alaska v. United States, Nos. A83-343, A84-435, A86-191 (D. Alaska), and has been stayed until resolution of the present case, see Report 347, n. 4. The parties no longer dispute the location of the Reserve’s boundary. Accordingly, we consider only the Master’s recommendation concerning the ownership of submerged lands beneath certain coastal features within that boundary. The Master concluded that the United States retained ownership of the submerged lands in question at Alaska’s statehood. That conclusion rested principally on three premises: first, that the United States can prevent lands beneath navigable waters from passing to a State upon admission to the Union by reserving those lands in federal ownership (as opposed to conveying them to a third party); second, that Congress had authorized the President to reserve submerged lands with a 1910 statute known as the Pickett Act; and third, that the 1923 Executive Order creating the Reserve reflected a clear intent to reserve all submerged lands within the boundaries of the Reserve and to defeat the State’s title to the submerged lands in question. Alaska excepts to the Master’s conclusion on several grounds, arguing that the Government did not show a sufficiently clear intent to reserve submerged lands or to defeat state title and that the 1923 Executive Order was promulgated without proper authority. We discuss some background principles and then consider these arguments in turn. A The Property Clause, Art. IV, § 3, cl. 2, provides that “Congress shall have Power to dispose of and make all needful Rules and Regulations respecting the Territory or other Property belonging to the United States.” In Shively v. Bowlby, 152 U. S. 1, 48 (1894), the Court concluded that this power extended to granting submerged lands to private parties, and thereby defeating a future State’s equal footing title, “to carry out . . . public purposes appropriate to the objects for which the United States hold the Territory.” We agree with the Special Master that Congress can also reserve submerged lands under federal control for an appropriate public purpose, and thus resolve a question left open in Utah Div. of State Lands, 482 U. S., at 201, in the United States’ favor. As drawn by the Master, the boundary of the Reserve encompasses both those lands that would ordinaril