Opinion of the court filed by Circuit Judge DYK. Dissenting opinion filed by Circuit Judge NEWMAN.
DYK, Circuit Judge.
The Arkansas Game and Fish Commission (“the Commission”) filed a physical takings claim against the United States in the Court of Federal Claims (“Claims Court”), alleging that the United States had taken its property without just compensation. The Commission claimed that temporary deviations by the Army Corps of Engineers (“the Corps”) from an operating plan for Clear-water Dam during the years 1993 to 2000 caused increased flooding in the Commission’s Dave Donaldson Black River Wildlife Management Area (“Management Area”). This flooding, in turn, caused excessive timber mortality in the Management Area. The Claims Court concluded that the United States had taken a temporary flowage easement over the Commission’s property and awarded a total of $5,778,757.90 in damages. Ark. Game & Fish Comm’n v. United States, 87 Fed.Cl. 594, 617, 647 (2009). Because we conclude that the Corps’ deviations did not constitute a taking, we reverse.
Background
The Commission owns the Management Area, which is located in northeast Arkansas and consists of 23,000 acres on both banks of the Black River. The Commission operates the Management Area as a wildlife and hunting preserve and harvests timber, thereafter reforesting the harvested areas. The Corps completed construction of the Clearwater Dam in 1948. The dam is located in southeast Missouri approximately 115 miles upstream of the Commission’s Management Area. The reservoir created by the dam is called Clear-water Lake. The primary purpose of the dam was to provide flood protection.
A
Whenever the Corps constructs a dam, it adopts a water control plan reflected in a Water Control Manual. These water control plans detail release rates, safety features, and other operating instructions. They are required by Corps regulation. U.S. Army Corps of Engineers, Eng’r Reg. No. 1110-2-240, at 2 (Oct. 8, 1982) [hereinafter EM 1110-2-240]. The plans are developed by regional district commanders “in concert with all basin interests which are or could be impacted by or have an influence on project regulation” and then submitted to the Corps for approval. Id. at 3, 4.
This case concerns release rates from the dam established by the plan for Clear-water Dam and deviations from the planned rates. The purpose of regulating release rates is to control the flow of the Black River in order to reduce the adverse effects of flooding in downstream areas. The dam cannot be operated in a manner that completely eliminates flooding because water must be released from the [1368]*1368dam, and the released water will, to some extent, cause flooding. If the release rates are lower, the height of the flooding is decreased but the period of flooding is increased. If the release rates are higher, the height of the flooding is increased but the period of flooding is decreased. Apparently, agricultural interests favored a lower release rate even though this would lead to longer periods of flooding, while the Commission and those located near Clearwater Lake preferred a higher release rate to return the lake to its normal level more quickly. As will be seen, the claim here is that the Commission’s property was damaged by the temporary adoption of a lower release rate during the growing season, prolonging the release period and damaging trees.
In order to understand the parties’ respective positions, it is necessary to describe the background of the 1953 plan and the deviations from that plan that are challenged here. The plan for Clearwater Dam was developed over a period of several years and was finally adopted as part of the Clearwater Lake Water Control Manual in 1953. After the Corps completed construction of Clearwater Dam in 1948, it experimented with release levels for five years. Releases were measured by the maximum height of the water at the Poplar Bluff Gauge in the Black River, which is downstream of the dam but upstream of the Management Area. The first maximum levels at Poplar Bluff during this early period were 12 feet during agricultural season and 14 feet during non-growing season. The Corps concluded that “operating experience” showed these high releases negatively affected too many downstream areas. J.A. 9865. Therefore, in 1950, the Corps reduced the maximum release levels to 10.5 feet during growing season and 11.5 feet during non-growing season. After three years during which “no consistent problems [were] encountered,” the Corps approved the first Clear-water Lake Water Control Manual in 1953.1 J.A. 9865. Under the Manual’s “normal regulation,” releases were regulated so that the water height at Poplar Bluff did not exceed 10.5 feet during growing season and 11.5 feet during non-growing season. The maximum release levels at Poplar Bluff allowed for the quick release of water during the growing season, so flooding occurred in short-term waves rather than over extended periods. The Manual’s normal regulation somewhat mimicked the natural flooding patterns in the region. During the 1993-2000 period, the Manual’s “normal regulation” releases were the same as under the original 1953 plan.
The Clearwater Lake Manual allowed for deviations from the “normal regulation” releases for (1) emergencies, (2) “unplanned minor deviations,” such as for construction or maintenance, and (3) “planned deviations” requested for agricultural, recreational, and other purposes. J.A. 9907-OS. The deviations in question here fell into the latter category. According to the Manual, the Corps was “occasionally requested to deviate from normal regulation.” Id. at 9907. Planned deviations had to be approved by the Corps’ Southwestern Division, which was required to consider “flood potential, lake and watershed conditions, possible alternative measures, benefits to be expected, and probable [1369]*1369effects on other authorized and useful purposes.” Id. As described in the Manual, these requests were for specific activities that required deviations only for limited periods of time, such as the harvesting of crops, canoe races, and fish spawning. Therefore, the approved deviations were by their nature temporary.
The temporary deviations here began in 1993, forty years after the adoption of the Water Control Manual.2 In 1993, the Corps approved a “planned deviation” from the Water Control Manual’s approved releases for a three month period from September 29 to December 15, 1993, lowering the maximum release level to six feet at the request of agricultural interests that desired slower releases to “allow[] farmers more time to harvest their crops.” J.A. 8237-38. No permanent change was made to the Water Control Manual at that time. But in the same year, the Corps fostered creation of the White River Ad Hoc Work Group (“White River Group”) to “recommend minor changes to the approved regulating plan[s]” for dams across the White River Basin, which included Clearwater Dam. Id. at 8242. In other words, the White River Group was to propose permanent changes to approved plans, including the Clearwater Lake Water Control Manual.
The Corps’ regulations provided that water control plans “will be revised as necessary to conform with changing requirements resulting from [new] developments.” ER 1110-2-240 at 2. The regulations also required that “plans will be subject to continuing and progressive study by personnel in field offices of the Corps.” Id. Substantive revisions “requirefd] public involvement and public meetings,” and the Corps was required to provide a report to the public at least 30 days before the meeting “explaining] the recommended ... change ... explaining the basis for the recommendation ... [and providing] a description of its impacts.” Id. at 4a. Moreover, in making revisions, the Corps was also required to “comply with existing Federal regulations,” such as the National Environmental Policy Act (“NEPA”). Id. at 2, A-l.
The White River Group included private recreational, agricultural, navigational, and hydropower interests, as well as state and federal agencies such as the Commission. The Commission objected to deviations from the approved water releases in the 1953 plan that would lower the release rates because such deviations would prolong the release period. The lower maximum release rates meant that water would evacuate from Clearwater Dam more slowly, causing consistent downstream flooding in the Management Area during the tree growing season. (The tree growing season lasts, on average, from April 4 to October 11 each year, and the “critical months” are June, July, and August.) Higher maximum release rates, on the other hand, would result in short-term waves of flooding that would quickly recede. The Commission specifically complained that the lake was 530 feet deep on April 15, 1994, and that, at the lower maximum rates, it took sixty days of constant releases for the lake to reach its targeted summertime level, flooding the Management Area for most of that two-month period.
In 1994, the White River Group, unable to recommend permanent revisions to the release plan, proposed a year “interim operating plan” that called for temporary [1370]*1370deviations during an eight month period. J.A. 8244. The Corps approved “the proposed interim operating plan” as a deviation from April 1994 through April 1995, allowing deviations from the normal release rates from April through November. Id. at 8241. The plan set the maximum target level at 11.5, instead of 10.5 feet, for the first two weeks in April and then set the maximum level at 8 feet for the next month and 6 feet from mid-May through November. In February 1995, when the White River Group again proved unable to propose a final plan, the Corps approved an extension of the “interim operating plan ... to continue as a deviation” for another 12 months through April 1996, allowing deviations from April to November. Id. at 8251. The Corps noted that it would “monitor the effectiveness of the interim operating plan” during that period. Id.
In February 1996, the White River Group formed a subcommittee of solely Black River interests (“the Black River Group”), including the Commission, to recommend a plan for Clearwater Dam. The Black River Group also could not reach consensus on a proposed permanent deviation plan for Clearwater Dam. Instead, the Corps approved a new “Interim Operating Plan” proposed by the Black River Group for Clearwater Dam for another 12 months through April 15, 1997, with deviations occurring April through November. The Corps said it would “continue to monitor the effectiveness” of both the White River and Black River “interim operating plans.” Id. at 8254. The new Black River Group interim plan differed from the prior interim plan recommended by the White River Group. It set the release rate at 6 feet in June and at 5 feet from July through November. This interim plan lapsed in April 1997, and no new interim plan was immediately adopted, leaving the 1953 Manual release rates in place. The Corps did, however, approve a temporary planned deviation in accordance with the Manual from June 3, 1997, to July 5, 1997, to prevent possible flooding and another temporary planned deviation from June 11, 1998, to November 30, 1998, in response to a request from agricultural interests.
The White River Group disbanded in 1997 when it recommended a final plan for the rest of the White River Basin, but the Black River Group continued to work on a plan for Clearwater Dam. The Black River Group finally recommended a proposed plan on September 15, 1998, and that plan was approved as a temporary deviation for 13 months from December 1, 1998, to December 31, 1999, with deviations occurring during the entire 13 month period. This proposed plan also differed from the interim plans approved in prior years, setting the maximum level at 4 feet from mid-May through November but increasing the maximum level if the lake behind Clear-water Dam filled to a certain volume.
In 1999, the Corps began the process of adopting a revised permanent release plan for Clearwater Dam. In keeping with federal regulations, the Corps prepared an Environmental Assessment (“EA”). An EA is a brief report, without detailed descriptions or data, indicating possible environmental consequences that can help determine whether a more extensive Environmental Impact Statement (“EIS”) is necessary pursuant to NEPA. An EIS is necessary where there is a possibility of significant environmental impacts. In comments made on the draft EA, the Commission objected to the proposed changes to the Water Control Manual. The Corps agreed that the proposed revision would require an EIS under NEPA.
Meanwhile, the Corps approved the continuation of the 1999 deviations for 11 [1371]*1371months through December 1, 2000, as “a temporary Water Control Plan.” J.A. 8294. In May 2000 and March 2001, the Corps and Commission together conducted tests to determine the environmental impact on the Management Area when certain levels of water were released from Clearwater Dam. In doing so, the Corps confirmed that tree roots would be flooded under the proposed plan, which could potentially damage or destroy the trees. Therefore, the Corps declined to further pursue a permanent revision to the 1953 Water Control Manual and returned to the releases set out in the original Manual.
In sum, after approving a planned deviation in 1993, the Corps approved three different interim deviation plans with different release rates during the period from 1994-2000. During some portions of that period no interim plan was in place. When temporary plans were in place, for all but two years (1999 and 2000), the release rates deviated from the 1953 plan during only part of the year, usually April through November. During this entire 1994-2000 period, efforts were made to propose a permanent amendment to the 1953 plan, but no permanent revision to the plan was ever adopted. The effort was finally abandoned in 2001, and the unamended 1953 plan continued to govern. The chart below summarizes the details of these temporary deviation plans.3
Plan Date of Plan Deviations (and Date) Normal Regulation
1993 Deviation Request Under Original Plan 9/29/93-12/15/93 6 ft 10.5 ft or 11.5 ft4 (9/29/93-12/15/93)
1994 White River Group Interim Plan 4/1/94-4/15/95 11.5 ft 10.5 ft (4/1/94-4/14/94)_
8 ft 10.5 ft (4/15/94-5/14/94)_
6 ft (5/15/94-11/30/94) 10.5ft or 11.5 ft
11.5 ft 10.5 ft (4/1/95-4/15/95)_
White River Group Interim Plan Extension 4/15/95-4/15/96 (4/15/95-5/14/95 10.5 ft
6 ft (5/15/95-11/30/95) 10.5 ft or 11.5 ft
11.5 ft 10.5 ft (4/1/96-4/15/96)_
1996 Black River Group Interim Plan 4/15/96-4/15/97 11.5 ft 10.5 ft (4/15/96-5/20/96)_
6 ft 10.5 ft (6/1/96-6/30/96)
[1372]*13725 ft 10.5 ft (7/1/96-11/30/96)_
1998 Deviation Request Under Original Plan 6/11/98-11/30/98 5 ft 10.5 ft or 11.5 ft _(6/11/98-11/30/98)_
1998 Black River Group Proposed Plan 12/1/98-12/31/99 11.5 ft 10.5 ft or 11.5 ft (12/1/98-5/14/99)_
4 ft 10.5 ft or 11.5 ft (5/15/99-11/30/99)_
Black River Group Proposed Plan Extension 1/1/00-12/1/00 11.5 ft (1/1/00-5/14/00) 10.5 ft or 11.5 ft
4 ft 10.5 ft or 11.5 ft (5/15/00-11/30/00)_
B
In 2005, the Commission brought suit against the United States under the Tucker Act, 28 U.S.C. § 1491, claiming that the temporary release rate deviations during the 1993-2000 period constituted a taking of a flowage easement entitling the Commission to compensation. The Commission alleged that the deviations caused repeated increased flooding and damaged and destroyed timber in the Management Area. The United States denied that a taking had occurred. The United States ai’gued that any increased flooding was only temporary and constituted, if anything, a tort rather than a taking. The United States also argued that the damage was not substantial enough to constitute a taking and that the effects in any event were not predictable, again defeating a takings claim.
A hearing was held before the Claims Court from December 1, 2008, to December 12, 2008, in which eighteen witnesses testified. In addition to addressing the nature of the deviations from the 1953 plan (described above), the hearing addressed two other issues: the substantiality of the flooding and the predictability of the alleged damage.
With respect to the substantiality of the flooding, the Commission’s testimony showed that the Management Area flooded regularly during the 1993-2000 period, including during the tree growing season. The presence of standing water or saturated soil during tree growing season can weaken the roots of the multiple species of oak trees located in the Management Area, which can also render the oaks more susceptible to drought conditions.
The flooding was not consistent from year to year. The parties’ experts disagreed as to the amount of increased flooding. Two experts for the Commission, Drs. Heitmeyer and Overton, testified about the extent of the increased flooding. Both used the height of the Corning water gauge along the Black River (downstream of the Poplar Bluff gauge) as a proxy for when flooding occurred, observing that flooding occurred when the Corning gauge reading was over 5 feet. Dr. Heitmeyer reported that the gauge exceeded 5 feet (the level at which flooding occurred) during the tree growing season about 29 more days per year from 1993-1999 than from 1949-1992. Dr. Overton compared the gauge readings from 1981-1992 to the readings from 1993-1999. He explained that the Corning gauge reached a level where substantial flooding occurred (between 8 and 10.5 feet) an average of 24 more days per year during the tree growing season from 1993-1999 than from [1373]*13731981-1992. Dr. Overton’s data also indicated that flooding occurred in the Management Area an average of 8.5 more days per year from 1993-1999 during the “critical” tree growing period of June to August.
During 1999 and 2000, the region suffered a moderate drought. Therefore, according to both the United States and the Commission, the deviations during 1999 and 2000 did not have much practical effect because the water flow was low and did not cause flooding. However, the Commission witnesses testified that drought conditions increased tree mortality in the Management Area. Trees weakened by six years of additional flooding during the tree growing season could not survive the drought.
The final question addressed at the hearing was the predictability of the results following the government’s actions. The Claims Court found that the Corps was unaware in 1993 that deviations would cause additional flooding in the Management Area. Indeed, one Corps official testified that the Corps believed the downstream effect of the deviations would diminish before the water reached Arkansas. However, the Commission sent repeated letters to the Corps claiming that the ongoing deviations were causing additional flooding in the Management Area and also raised its concerns in the White River Group and Black River Group meetings. Still, in its 1999 draft EA, the Corps would have found “no significant impact” based on the proposed revisions to the plan. The Corps only changed its mind about the effects of the deviations after conducting the May 2000 and March 2001 test releases with the Commission which confirmed the possible adverse impacts on the Management Area and led the Corps to return to the release rates in the original 1953 plan.
The parties also disputed the issue of causation and the amount of damage caused by the flooding. The Commission’s expert testified that half of the damaged trees would die within five years and the living damaged trees were worth half of their original value. The United States argued that this testimony was inadmissible because so-called cruise maps on which the testimony was based had been destroyed and that the expert evidence in any event used unreliable guesses and was inadmissible under the Federal Rules of Evidence and Daubert v. Merrell Dow Pharmaceuticals., Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). The Commission also sought regeneration costs to restore areas suffering from invasions of new wetland vegetation to their pre-deviation condition. The United States contended that the Commission failed to prove the United States’ actions resulted in the devaluation from invasive growth and, even if it did, the damages were only consequential.
C
The Claims Court concluded that the deviations from the 1953 Water Control Manual were “interim deviations” that were “approved at various times from 1993 to 2000.” 87 Fed.Cl. at 603. The court also found that:
“Certainly no permanent flowage easement in the Management Area was taken by the flooding attributable to the Corps’ deviations from the operating Plan for Clearwater Dam.... [A] temporary flowage easement is a necessary foundation for the Commission’s takings claim, as has always been evident from the Commission’s pleadings and proofs.”
[1374]*1374Id. at 617 (emphasis in original); see also id. at 619-20 (finding appropriation was “temporary rather than permanent”). The court concluded that a takings claim could be based on such a temporary activity. Id. at 618.
The Claims Court relied on the observations by Drs. Heitmeyer and Overton and determined that the deviations caused a substantial enough increase in flooding to constitute the taking of a flowage easement. It also held that the flooding was the predictable result of government action because with any “reasonable investigation of the effects of the deviations ... it would have been able to predict” that the deviations would cause flooding and damage timber in the Management Area. 87 Fed. Cl. at 623. The Claims Court also concluded that the increased tree growing season flooding caused increased timber mortality in the Management Area, and therefore awarded damages based on the loss of timber and regeneration costs for some areas that had been inundated by new wetland vegetation. The Claims Court awarded $5.5 million in damages for dead or declining timber, relying on the Commission expert’s testimony about the value of the trees.5 It also awarded $176,428.34 in damages for regeneration costs, but it granted damages only for areas classified as having suffered “severe” rather than “heavy” or “moderate” effects.
The United States appealed, contending that no taking had occurred, and that if it had, the damages were overstated. The Commission cross-appealed, contending that the Claims Court should have awarded additional damages for regeneration. We have jurisdiction pursuant to 28 U.S.C. § 1295(a)(3).
Discussion
Determining whether a taking has occurred is a “question of law based on factual underpinnings,” and as such, we review the Claims Court’s legal analysis and conclusion de novo and its factual findings for clear error. See Ridge Line, Inc. v. United States, 346 F.3d 1346, 1352 (Fed. Cir.2003).
I
In general, if particular government action would constitute a taking when permanently continued, temporary action of the same nature may lead to a temporary takings claim. See First English Evangelical Lutheran Church of Glendale v. Cnty. of Los Angeles, 482 U.S. 304, 328, 107 S.Ct. 2378, 96 L.Ed.2d 250 (1987). However, cases involving flooding and flow-age easements are different. Both Supreme Court precedent and our own precedent dictate that we must distinguish between a tort and a taking. An injury that is only “in its nature indirect and consequential,” i.e. a tort, cannot be a taking. Sanguinetti v. United States, 264 U.S. 146, 150, 44 S.Ct. 264, 68 L.Ed. 608 (1924). The Supreme Court has long recognized that to be considered a taking overflows must “constitute an actual, permanent invasion of the land, amounting to an appropriation of and not merely an injury to the property.” Id. at 149, 44 S.Ct. 264 (emphases added). The Court has stated that an invasion is permanent when there is a “permanent condition of continual overflow” or “a permanent liability to intermittent but inevitably recurring overflows.” United States v. Cress, 243 [1375]*1375U.S. 316, 328, 37 S.Ct. 380, 61 L.Ed. 746 (1917); see also United States v. Kansas City Life Ins. Co., 339 U.S. 799, 810 n. 8, 70 S.Ct. 885, 94 L.Ed. 1277 (1950) (quoting Cress and finding a taking where the plaintiffs land was “permanently invaded by the percolation of ... waters”).
In Cress, the Court found a taking where the erection of a lock and dam on the Cumberland River “subjected] [the plaintiffs land] to frequent overflows of water.” Cress, 243 U.S. at 318, 37 S.Ct. 380. The Court explained that these intermittent overflows showed “that this [was] not a case of temporary flooding or consequential injury,” where takings liability would be denied, “but [instead] a permanent condition, resulting from the erection of the lock and dam.” Id. at 327, 37 S.Ct. 380 (emphasis added). It further stated that there was “no difference of kind ... between a permanent condition of continual overflow [which had previously been found to constitute a taking] ... and a permanent liability to intermittent but inevitably recurring overflows.” Id. at 328, 37 S.Ct. 380 (emphasis added).
In United States v. Dickinson, 331 U.S. 745, 746-47, 67 S.Ct. 1382, 91 L.Ed. 1789 (1947), the Court also found a taking based on a permanent condition. In Dickinson, the United States constructed a dam, and the “water above the dam was ... impounded to create a deeper channel,” thereby permanently raising the river level, “permanently flood[ing]” some of the adjacent land, and subjecting more of it to intermittent overflows. Id. at 746-47, 67 S.Ct. 1382. Even though Dickinson later “reclaimed most of his land which the Government originally took by flooding,” the Court found that the reclamation did not “change[ ] the fact that the land was taken [in the first place] ... and an obligation to pay for it then arose.” Id. at 751, 67 S.Ct. 1382. The nature of the government’s action remained permanent, even though the reclamation had mitigated some of the effects. Again, the Court distinguished between invasions that were permanent or temporary in character. Summarizing these cases, in Loretto v. Teleprompter Manhattan CATV Corp., 458 U.S. 419, 428, 102 S.Ct. 3164, 73 L.Ed.2d 868 (1982), the Supreme Court noted that it has “consistently distinguished between flooding cases involving a permanent physical occupation, on the one hand, and cases involving a more temporary invasion ... that causes consequential damages within, on the other.”
Numerous cases from our predecessor court have similarly held that inherently temporary conditions cannot result in the taking of a flowage easement. See Barnes v. United States, 538 F.2d 865, 870 (Ct.Cl. 1976); Nat’l By-Products, Inc. v. United States, 405 F.2d 1256, 1273 (Ct.Cl.1969); Fromme v. United States, 412 F.2d 1192, 1197 (Ct.Cl.1969). “The plaintiff must establish that flooding will ‘inevitably recur,’ in the phrasing of the Cress case.” National By-Products, 405 F.2d at 1273. “[G]overnment-induced flooding not proved to be inevitably recurring occupies the category of mere consequential injury, or tort.” Barnes, 538 F.2d at 870.
The Commission argues that under our decision in Ridge Line a permanent invasion is not required and that appropriation of a “temporary flowage easement” is sufficient. Commission’s Br. 39. This argument is without support. A panel decision of this court cannot, of course, overturn Supreme Court precedent or our precedent. In any event, we did not purport to do so in Ridge Line. The case involved a permanent condition — runoff caused by [1376]*1376the construction of a postal facility.6 As such, we had no reason to address whether an inherently temporary condition could be considered a taking. The holding in Ridge Line is fully consistent with the Cress case’s emphasis on the “inevitably recurring” nature of intermittent flooding.
In Ridge Line we noted that the Claims Court had erred in “confin[ing] its analysis of liability to whether the government’s actions constituted a permanent and exclusive occupation.” 346 F.3d at 1352, 1354 (internal quotation omitted). We concluded that “permanent destruction or exclusive occupation by government runoff is not always required for a successful taking! ]•” Id. at 1354. In other words, we explained, the “occupation” need not be exclusive and the destruction need not be “permanent.” Id. at 1352. “[I]ntermittent flooding of private land can constitute a taking of an easement.” Id. at 1354. Thus, we confirmed that intermittent but inevitably recurring flooding could constitute a taking and that continuous inundation was not required. Id. at 1357. But, citing Barnes, we noted that “government-induced flooding not proved to be inevitably recurring occupies the category of mere consequential injury, or tort.” Id. at 1355 (quoting Barnes, 538 F.2d at 870). The condition leading to the “intermittent, but inevitably recurring” flooding, id. at 1357, must be permanent. Otherwise, it could not be “inevitably recurring.”
Ridge Line also clarified that, in distinguishing between a tort and a taking, courts must additionally consider 1) whether “the government’s actions were sufficiently substantial to justify a takings remedy” and 2) whether the harm suffered by the plaintiff was “the predictable result of the government’s action.” 346 F.3d at 1355, 1356. The Ridge Line court relied on Sanguinetti where the Supreme Court found no taking in that case because it was unclear whether any substantial additional flooding actually occurred and whether the overflow was “the direct result of the [canal]” and thus would be reasonably anticipated by the government. 264 U.S. at 149, 44 S.Ct. 264. Similarly, in John Horstmann Co. v. United States, 257 U.S. 138, 146, 42 S.Ct. 58, 66 L.Ed. 171 (1921), the Supreme Court found no taking where flooding was caused by the movement of underground percolating waters “which no human knowledge could even predict.”
The parties in this case vigorously dispute whether the extent and frequency of flooding satisfied the substantiality requirement and whether it was predictable. However, we need not decide whether the flooding on the Management Area was “sufficiently substantial to justify a takings remedy” or “the predictable result of the government’s action,” Ridge Line, 346 F.3d at 1355, 1356, because the deviations were by their very nature temporary and, therefore, cannot be “inevitably recurring” or constitute the taking of a flowage easement. For this reason, we also need not address the parties’ dispute as to the calculation of damages.
II
Most government-induced flooding cases involve overflows caused by permanent structures or improvements, such as dams, canals, or levees. See, e.g., Dickinson, 331 U.S. at 746-47, 67 S.Ct. 1382 (involving the [1377]*1377construction of a dam and the resulting rise in the river level); Sanguinetty 264 U.S. at 147, 44 S.Ct. 264 (involving a canal insufficient to carry away flood waters); Cress, 243 U.S. at 327, 37 S.Ct. 380 (involving erection of a lock and dam); Ridge Line, 346 F.3d at 1351 (involving runoff caused by construction of Post Office facility); Nat’l By-Products, 405 F.2d at 1259-61 (involving a levee built on one side of a creek without a corresponding levee on the other side). Permanent conditions often, but not always, yield inevitably recurring flooding. Compare Cress, 243 U.S. at 327, 37 S.Ct. 380 (finding inevitably recurring flooding), with National By-Products, 405 F.2d at 1274-75 (finding no taking because plaintiffs did not prove floods caused by levees would inevitably recur).
As with structural cases, in determining whether a governmental decision to release water from a dam can result in a taking, we must distinguish between action which is by its nature temporary and that which is permanent. But in distinguishing between temporary and permanent action, we do not focus on a structure and its consequence. Rather we must focus on whether the government flood control policy was a permanent or temporary policy. Releases that are ad hoc or temporary cannot, by their very nature, be inevitably recurring.
The Commission’s entire theory is contrary to governing law. The Commission does not contend that the deviations from the 1953 plan and the resulting flooding were inevitably recurring. Rather, the Commission contends that temporary deviations are sufficient. In its brief, it contended “that [the fact that] the Corps eventually stopped [the] deviations only renderfed] the taking temporary,” Commission’s Br. 41, and therefore claims “the [United States] appropriated a temporary flowage easement for which it must pay,” id. at 39 (emphasis added). At oral argument, the Commission similarly argued that it “do[es] not think the law requires [the deviations] to be permanent or that the government intended ... to make [them] permanent.” Oral arg. at 18:17-18:25, available at http://www.cafc. uscourts.gov/. It argued again that “the fact that ... the United States abandoned its easement only makes [the taking] temporary.” Id. at 20:33-20:41.
The Commission’s concession that the government action was temporary in nature is fully consistent with the established facts. The undisputed facts are clear that the governmental action was designed to be temporary and that the Corps never approved a permanent change in the preexisting flow rates. As discussed above, all of the deviations from 1993 to 2000 were approved only as temporary or interim deviations. The multiple interim plans differed. Even where deviations were the same in consecutive years, such as in 1994 and 1995, the Corps had to approve an extension of the interim deviation plan for the second year.
The Commission itself similarly referred to the deviations as temporary or interim. During oral argument, the Commission noted that when deviations first began they were temporary and made in response to particular requests by specific interests and that the permanent revisions to the plan were never approved. Moreover, in an internal memorandum summarizing the first meeting of the Black River Group in 1996, a Commission official described the 1994-1995 deviations as “a temporary interim plan.” J.A. 9145. In a later memorandum, a second Commission official acknowledged that “[d]uring the [1378]*1378years[ ] 95, 96, and 97[,] the so-called interim plan was tried.” Id. at 2434. Martin Blaney, the Commission’s Statewide Habitat Coordinator, testified at trial that an “interim plan” had been used from 1998-2000. Id. at 1811. Similarly, Robert Zachary, the Wildlife Supervisor for the Commission, testified that the Corps’ “interim plan of operation” caused tree growing season flooding in 1996. Id. at 1687.
Therefore, the numerous deviation plans were inherently temporary, and the Corps, despite considering permanent revisions in 1999, never approved any of the deviations as a permanent policy. In order to make any permanent change to the release rates, Corps regulations required compliance with federal regulations. The Corps did not even begin the EA or EIS necessary to comply with NEPA until 1999, and it never completed the necessary steps to implement a permanent revision of the Water Control Plan. As such, the plaintiffs have not met their burden to prove that the increased flooding would be “inevitably recurring” because the deviations were explicitly temporary.
Binding decisions of the Court of Claims, our predecessor court, in Barnes and Fromme found no taking under similar circumstances. In Barnes, the Fort Randall Dam was constructed in 1952, and the government began to release water “to evacuate the excess water accumulation caused by rains and melting snows” in 1969. 538 F.2d at 868. The releases caused intermittent flooding from 1969-1973 and again beginning in 1975. Id. at 868-69, 872. Noting that “the flooding [was] of a type which will be inevitably recurring,” id. at 872, the court determined that a taking had occurred but held that it did not occur until “the permanent character of intermittent flooding could fairly be perceived” in 1973, id. at 873. Consequently, it did not allow the plaintiffs to recover for crop damage sustained from 1969-1973 because it was not obvious that the releases and the flooding would be permanent until 1973. Id.
In Fromme, the Corps constructed a channel and levee near the plaintiffs property. 412 F.2d at 1194. During construction of this project, from 1965-1967, the defendant maintained a “temporary spoil bank,” which caused the plaintiffs land to flood “for a substantially longer time than would have been the case if the spoil bank and partially completed levee had not been in existence.” Id. at 1195. The court held that no taking occurred during the 1965-1967 period because the spoil bank only “represented a temporary situation,” and the channel would not lead to the “inevitably recurring floodings which the Supreme Court [had] stressed ... in Cress.” Id. at 1196-97. Hence, both Barnes and From-me indicate that flooding must be a permanent or inevitably recurring condition, rather than an inherently temporary situation, to constitute the taking of a flowage easement.7
Because the deviations from the 1953 plan were only temporary, they cannot [1379]*1379constitute a taking. The actions at most created tort liability. We recognize that in other contexts the distinction between a temporary and permanent release plan may be difficult to define. The government cannot, of course, avoid takings liability by characterizing inevitably recurring events as merely a series of temporary decisions. Here, however, the Corps’ regulatory scheme has itself clearly distinguished between permanent and temporary release rates. The deviations in question were plainly temporary and the Corps eventually reverted to the permanent plan. Under such circumstances, the releases cannot be characterized as inevitably recurring.
Ill
For the foregoing reasons, we reverse the Claims Court’s decision that the United States had taken a flowage easement on the Commission’s land without just compensation.
REVERSED.
Costs
No costs.