OPINION
Before RABINOWITZ, C. J., and CON-NOR, BOOCHEVER, BURKE and MATTHEWS, JJ. MATTHEWS, Justice.This is a land dispute. The land in question consists of approximately 117 acres added by accretion to the shoreline of the Gastineau Channel over the last fifty years. The appellants claim that Daniel Pederson, their predecessor in interest, homesteaded the shoreline to which the land accreted, and that the land belongs to them. The state claims that the federal government retained its interest in the shoreline and that the state now owns the accreted land by virtue of selection made under the Alaska Statehood Act.1 The lower court held in favor of the state.
The land in dispute formed part of the Mendenhall Valley Elimination — a large tract of land which was withdrawn from the Tongass National Forest by Presidential Proclamation in 1922 and opened to homesteading. The boundaries of the Menden-hall Valley Elimination had been defined by an unofficial survey conducted by a Forest Service surveyor in 1919. That survey established various boundary markers and corners, including corners along the Gasti-neau Channel which the superior court found to be meander corners. A meander corner establishes a point on a meander line. Meander lines are straight lines that approximate the sinuosities of the line of mean high tide. While the meander line remains fixed, the line of mean high tide may change over the years due to accretions or deletions. Patented land which extends to the line of mean high tide may be increased in size or diminished according to such accretions or deletions.2 At the time of the 1919 survey and thereafter, the land was subject to accretion caused by silt carried down from the Mendenhall Glacier and deposited at the mouth of the Menden-hall River. The land has also been gradually rising up as the glacier recedes, a process know as isostatic rebound.
In 1924 the General Land Office directed that an official survey of the Mendenhall Valley Elimination be prepared. This survey, U.S. Survey No. 1536, was made by Fred Dahlquist, a U.S. Cadastral Engineer. Dahlquist was instructed by the General Land Office that “the corners established by the Forest Service [in its 1919 survey] must be adopted for this survey whether or not the courses and distances of the connecting lines agree with those given.” Pursuant to these instructions, Dahlquist began his survey at corner number one of the 1919 survey and followed the prior survey all the way around the elimination. In his field notes Dahlquist consistently stated that he was proceeding on a true line until he reached corner ten, which meets the Gasti-neau Channel some distance easterly of the Pederson homestead. There he wrote that he was proceeding along a meander line. He continued to note that he was proceeding on the meander line until he reached corner fourteen. That corner is identical with Corner No. 5 of the Pederson home*270stead. His notes indicate that he proceeded “Thence ... on true line leaving meander line. . . . Over level, open grassy land ... to Cor. No. 1, the place of beginning.”
Dahlquist’s survey was approved and accepted by the General Land Office in 1926 and became the official survey of the Men-denhall Valley Elimination. A portion of the plat of this survey is attached as an appendix to this opinion.
In 1929 the appellants’ predecessor in interest, Daniel Pederson, filed a homestead application covering land in the southwest corner of the area surveyed by Dahlquist. The homesteading application described a tract of land “situate within the boundaries of U.S. Survey No. 1536 . containing approximately 160 acres.” The southern boundary of the property ran “beginning at Cor. No. 1, identical with Cor. No. 14, U.S. Survey 1536; thence ... to Cor. No. 2 identical with Cor. No. 1 said survey 1536. . . . ”3
In 1932, as required by law, Pederson’s homestead was surveyed. This survey, U.S. Survey No. 2136, followed along line 14-1 of No. 1536. The surveyor’s notes described the line as going “over level grassy meadows subject to overflow during extreme high tides.” This survey was approved in 1935. The plat of this survey is also appended to this opinion.
In 1937 the federal government issued Pederson a patent to the land he had homesteaded. The patent conveyed the land
embraced in H.E. Survey No. 2136, situate within the Mendenhall Elimination from the Tongass National Forest . containing one hundred fifty-nine acres and sixty-three hundredths of an acre, according to the Official Plat of the Survey of the said Land, on file in the General Land Office.
Pederson and his successors in interest continued to use the property as though it were shorefront property and have considered the approximately 117 acres that have accreted to the shore as their own. In 1966 the State of Alaska gave notice that it was selecting the accreted lands pursuant to the Alaska Statehood Act. The Bureau of Land Management tentatively approved the selection. This lawsuit followed.
Both parties agree that the issues in this case are controlled by federal law.4 When the federal government grants land via a patent, the patent is the highest evidence of title.5 Here, the patent incorporates by reference the plat of Survey No. 2136, made in 1932. The plat, including the surveyor’s field notes and descriptions, thus becomes a part of the patent and controls the extent of the lands conveyed.6
On its face, the 1932 survey seems clear, and indicates that line 5-6 (identical to line 14-1 of the 1924 survey) is a true line. By the time of the survey considerable accretion had already occurred in the area, and corner 6, (identical to corner 1 of the 1924 survey) was some distance in from the shoreline. The official plat of the 1932 survey does not show that the boundary of Pederson’s property was the Gastineau Channel. Rather, line 5-6 angles sharply away from the water’s edge. Corners 5 and 6 are not designated “meander corners.” Moreover the patent conveys to Pederson property amounting to 159.63 acres. This is the area of the property within the boundaries of the 1932 survey and does not include land visible on the plat extending beyond line 5-6.
*271On the face of the patent therefore, there does not seem to be any question that the land conveyed to Pederson did not include the shoreline.
The appellants seek to avoid this conclusion by referring to the surveys conducted in 1919 and 1924. Prior surveys and other extrinsic evidence are admissible if they are relevant to show the proper boundaries of a disputed tract of land.7 However, in this case reference to the prior surveys does not help appellants. According to the description contained in Pederson’s homestead entry, Pederson intended to stake a homestead that ran along line 14-1 of the 1924 survey. But line 14-1 does not appear to have been a meander line. The court below carefully reviewed the field notes, the plat and other evidence of the 1924 survey and concluded that although the surveyor intended to create a meander line from corner 10 through corner 14, the line from corner 14 to corner 1 was clearly not a meander line, it was a true line.
The evidence presented by the appellants is not persuasive to the contrary. It consists primarily of an analysis of statements taken out of context from the 1924 survey field notes. The appellants argue, for example that because the field notes describe the course travelled between corners 14 and 1 as “over level, open grassy lands . subject to inundation by the high tides of Gastineau Channel” the surveyor stayed on the meander line. They reason that since he was on a meander line when he reached corner 14 and remained “level” as he traveled to corner 1, he never left the meander line. However the surveyor clearly stated that from corner 14 he was continuing to corner number 1 “on true line leaving meander line.” This is explained by the appellants “as indicating that the meander line of the channel was not identical at that point with the mean high tide line and that while he was continuing along the mean high tide line, he was no longer following the shoreline.” This ignores the fact that at each corner of the survey, the surveyor meticulously entered whether he was following a true line or a meander line. Line 13-14 was a meander line, clearly marked as such in the field notes. In contrast, the entry for line 14-1 can have only one meaning — that the surveyor was back on a true line. This inference is further bolstered by the insert to the plat of the 1924 survey which lists the course and distance of all meander lines in the survey. The line between corners 14 and 1 is not listed as being a meander line.
The appellants reliance on even earlier surveys is also misplaced. Corner 14 of the 1924 survey corresponds — as per the surveyor’s instructions — with corner number 14 of the 1919 survey. This corner in turn corresponded to corner number 4 of an earlier survey conducted in 1914. Corner 4 in the 1914 survey was designated a meander corner. From this the appellants conclude that corner number 14 of the 1924 survey was a meander corner. Similarly, corner 1 of the 1924 survey began at the mean high tide line at the same place that corner 1 of the 1919 survey began. However, regardless what the corners represented in either the 1914 or the 1919 survey, they were not designated meander corners in the 1924 survey.8 Since the 1924 survey was the first official survey of the area, we deem it to be controlling of the question. Until rights are given to private parties the government can change the boundaries of *272surveyed lands as it pleases.9 It appears that the government did so in this case.
The appellants have advanced one more item of evidence that should be discussed. This is Shore Space Restoration Order No. 274, a document which was issued in 1935. The statute to which the order refers prohibits any homestead from extending more than 160 rods along a navigable body of water. The order waives that requirement for “that tract of land situated on the Gastineau Channel and described as U.S. Survey No. 2136. . . . ” From the existence of the order, and the language therein, the appellants argue that the Pederson homestead must have been located on the shoreline, or the order itself would have been unnecessary.
Although this evidence is persuasive, we conclude as did the trial court, that without more, it is not sufficient to outweigh the evidence in favor of the state’s position in the case.
We reach the holding in this case 'with some reluctance, for the appellants’ situation is compelling. Pederson and the appellants obviously believed they owned down to the shoreline. They have used the property as shorefront property and have exercised the incidents of ownership over the accreted land for the last fifty years. Unfortunately, the evidence seems clear that both the 1932 and the 1924 surveys did not convey to a meander line, and the surveys must control.
AFFIRMED.
*273
*274
. 72 Stat. 339 (July 7, 1958) (as amended).
. See Hawkins v. Alaska Freight Lines, Inc., 410 P.2d 992, 994 (Alaska 1966).
. Corner number 1 in the application became corner number 5 in the later 1932 survey.
. See Hughes v. Washington, 389 U.S. 290, 292, 88 S.Ct. 438, 19 L.Ed.2d 530 (1970); State v. Pankratz, 538 P.2d 984, 988 (Alaska 1975).
. United States v. Stone, 69 U.S. (2 Wall.) 525, 535, 17 L.Ed. 765 (1864); Walton v. United States, 415 F.2d 121, 123 (10th Cir. 1969).
. Jefferis v. East Omaha Land Co., 134 U.S. 178, 194-95, 10 S.Ct. 518, 33 L.Ed. 872 (1890); Cragin v. Powell, 128 U.S. 691, 696, 9 S.Ct. 203, 32 L.Ed. 566 (1888); United States v. Reimann, 504 F.2d 135, 140 (10th Cir. 1974); Walton v. United States, 415 F.2d 121, 123 (10th Cir. 1969); Spar Consol Mining and Devel Co. v. Miller, 568 P.2d 1159, 1161-62 (Colo.1977).
. See Graham v. Gill, 223 U.S. 643, 645, 32 S.Ct. 396, 56 L.Ed. 586 (1912); French Glenn Live Stock Co. v. Springer, 185 U.S. 47, 54, 22 S.Ct. 563, 46 L.Ed. 800 (1902). The appellants also suggest that, rather than being merely evidence of the proper interpretation of the 1932 survey the 1924 survey controls the boundary question entirely. This is based on the premise that Pederson’s right to the land was fixed by the survey in effect when he entered the homestead in 1929. We need not decide the issue, since it is our view that the 1924 survey did not create a meander line in any event. Even if the appellants’ theory was correct, they would lose anyway.
. It is agreed by the parties that the line between corners thirteen and fourteen is a meander line, and that corner fourteen is thus a meander corner with respect to that line.
. Cox v. Hart, 260 U.S. 427, 436, 43 S.Ct. 154, 67 L.Ed. 332 (1922); United States v. Reimann, 504 F.2d 135, 138 (10th Cir. 1974); Warner Valley Stock Co. v. Calderwood, 36 Or. 228, 59 P. 115, 116 (1899).