REHNQUIST, C.J., delivered the opinion for a unanimous Court.

John B. Draper, Santa Fe, NM, for plaintiff.

David Willis Robbins, Denver, CO, for defendant.

Jeffrey P. Minear, Washington, DC, for the U.S.

Chief Justice REHNQUIST delivered the opinion of the Court.

This original action involves a dispute among Kansas, Colorado, and the United States over alleged violations of the Arkansas River Compact. The Special Master has filed a report (Report) detailing his findings and recommendations concerning the liability phase of the trial. Both Kansas and Colorado have filed exceptions to those findings and recommendations. We agree with the Special Master's disposition of the liability issues. Accordingly, we overrule the parties' exceptions.

* The Continental Divide in the United States begins at the Canadian border in the mountains of northwestern Montana. From there, it angles southeast through Montana and Wyoming until it enters Colorado. It then runs roughly due south through Colorado, following first the crest of the Front Range of the Rocky Mountains, and then shifting slightly west to follow the crest of the Sawatch Range. The Arkansas River rises on the east side of the Continental Divide, between Climax and Leadville, Colorado. Thence it flows south and east through Colorado, Kansas, Oklahoma, and Arkansas, emptying into the Mississippi River, which in turn flows into the Gulf of Mexico. As if to prove that the ridge which separates them is indeed the Continental Divide, a short distance away from the source of the Arkansas, the Colorado River rises and thence flows southwest through Colorado, Utah, and Arizona, and finally empties into the Gulf of Baja California.

The Arkansas River flows at a steep gradient from its source south to Canon City, Colorado, whence it turns east and enters the Royal Gorge. As it flows through the Royal Gorge, the Arkansas River is at some points half a mile below the summit of the bordering cliffs. The Arkansas River thence descends gradually through the high plains of eastern Colorado and western Kansas; its elevation at the Colorado-Kansas border is 3,350 feet. It then makes its great bend northward through Kansas, and from there flows southeasterly through northeastern Oklahoma and across Arkansas. The Arkansas River covers about 1,450 miles from its source in the Colorado Rockies to the point in southeastern Arkansas where it flows into the Mississippi River. It is the fourth longest river in the United States, and it drains in an area of 185,000 square miles.

The first Europeans to see the Arkansas River were members of the expedition of Francisco Coronado, in the course of their search for the fabled Southern Cities of Cibola. In 1541, they crossed the Arkansas River near what is now the Colorado-Kansas border. One year later, those in the expedition of Hernando DeSoto would see the Arkansas River 1,000 miles downstream at its mouth. The western borders of the Louisiana Purchase, acquired from France in 1803, included within them most, if not all, of the Arkansas River drainage basin. Zebulon Pike, in his expedition of 1805-1806, in the course of which he sighted the mountain peak named after him, traveled up the Arkansas River. John C. Freemont traversed the River in the other direction in his expedition of 1843-1844.

Today, as a result of the Kerr McClellan Project, the Arkansas River is navigable for ocean going vessels all the way from its mouth to Tulsa, Oklahoma. The Arkansas River is unique in that the pronunciation of its name changes from State to State. In Colorado, Oklahoma, and Arkansas, it is pronounced as is the name of the State of Arkansas, but in Kansas, it is pronounced Ar-KAN-sas.

The reach of the Arkansas River system at issue here is a fertile agricultural region that extends from Pueblo, Colorado, to Garden City, Kansas. This region has been developed in Colorado by 23 major canal companies and in Kansas by 6 canal companies, which divert the surface flows of the Arkansas River and distribute them to individual farmers. Report 35-38. Also relevant to this dispute, the United States has constructed three large water storage projects in the Arkansas River basin. Id., at 43-48. The John Martin Reservoir, located on the Arkansas River about 60 miles west of the Kansas border, was authorized by Congress in 1936, 49 Stat. 1570, and was completed in 1948. It is the largest of the federal reservoirs, and initially it had a storage capacity of about 700,000 acre-feet.1 Report 45. The Pueblo Reservoir, located on the Arkansas River about 150 miles west of the Kansas border, was authorized by Congress in 1962, and was substantially completed in 1975. Id., at 44. In 1977, the storage capacity of the Pueblo Reservoir was estimated to be about 357,000 acre-feet. Ibid. Finally, the Trinidad Reservoir, located on the Purgatoire River (a major tributary of the Arkansas River) was approved by Congress in 1958, and was completed in 1977. Id., at 43. The total capacity of the Trinidad Reservoir is about 114,000 acre-feet. Ibid.

Twice before in this century, the States of Kansas and Colorado have litigated in this Court regarding their respective rights to the waters of the Arkansas River. See Kansas v. Colorado, 206 U.S. 46, 27 S.Ct. 655, 51 L.Ed. 956 (1907); Colorado v. Kansas, 320 U.S. 383, 64 S.Ct. 176, 88 L.Ed. 116 (1943). In the first suit, the Court denied Kansas' request to enjoin diversions of the Arkansas River by Colorado because the depletions alleged by Kansas were insufficient to warrant injunctive relief. Kansas v. Colorado, supra, 206 U.S., at 114-117, 27 S.Ct., at 674-676. In the second suit, Colorado sought to enjoin lower court litigation brought against Colorado water users, while Kansas sought an equitable apportionment of the Arkansas River. Colorado v. Kansas, supra, 320 U.S., at 388-389, 64 S.Ct., at 178-179. The Court granted Colorado an injunction, but concluded that Kansas was not entitled to an equitable apportionment. 320 U.S., at 400, 64 S.Ct., at 184. The Court suggested that the States resolve their differences by negotiation and agreement, pursuant to the Compact Clause of the Constitution. Id., at 392, 64 S.Ct., at 180. See U.S. Const., Art. I, § 10, cl. 3.

In 1949, after three years of negotiations, Kansas and Colorado approved, and Congress ratified, the Arkansas River Compact (Compact). See 63 Stat. 145; see also Report 5-6; App. to Report 1-17 (reprinting text of Compact). Article VIII of the Compact creates the Arkansas River Compact Administration (Administration) and vests it with the power and responsibility for administering the Compact. Id., at 11-15. The Administration is composed of a nonvoting presiding officer designated by the President of the United States, and three voting representatives from each State. Each State has one vote, and every decision, authorization, or other action by the Administration requires a unanimous vote. Id., at 12-13 (Article VIII-D).

The Compact's primary purposes are to "[s]ettle existing disputes and remove causes of future controversy . . . concerning the waters of the Arkansas River" and to "[e]quitably divide and apportion" the waters of the Arkansas River, "as well as the benefits arising from the construction, operation and maintenance by the United States of John Martin Reservoir." Id., at 1-2 (Articles I-A, I-B). Article IV-D, the provision of the Compact most relevant to this dispute, states:

"This Compact is not intended to impede or prevent future beneficial development of the Arkansas River basin in Colorado and Kansas by Federal or State agencies, by private enterprise, or by combinations thereof, which may involve construction of dams, reservoir, and other works for the purposes of water utilization and control, as well as the improved or prolonged functioning of existing works: Provided, that the waters of the Arkansas River . . . shall not be materially depleted in usable quantity or availability for use to the water users in Colorado and Kansas under this Compact by such future development or construction." Id., at 5 (emphasis added).

In 1983, Kansas conducted an independent investigation of possible violations of the Compact arising from the impact of increases in post-Compact well pumping in Colorado and the operation of two of the federal reservoirs. Report 9-10. In December 1985, Kansas brought this original action against the State of Colorado to resolve disputes arising under the Arkansas River Compact. Id., at 10. The Court granted Kansas leave to file its complaint, Kansas v. Colorado, 475 U.S. 1079, 106 S.Ct. 1454, 89 L.Ed.2d 712 (1986), and appointed Judge Wade H. McCree, Jr., to serve as Special Master, Kansas v. Colorado, 478 U.S. 1018, 106 S.Ct. 3330, 92 L.Ed.2d 736 (1986). Upon Judge McCree's death, the Court appointed Arthur L. Littleworth as Special Master. Kansas v. Colorado, 484 U.S. 910, 108 S.Ct. 254, 98 L.Ed.2d 212 (1987).

Kansas advanced three principle claims, each involving an alleged Compact violation. See Report 58. First, Kansas alleged that increases in groundwater well pumping in Colorado in the years following adoption of the Compact have caused a significant decline in the Arkansas River's surface flow in violation of Article IV-D of the Compact. Second, Kansas claimed that Colorado's Winter Water Storage Program (WWSP)—a program whereby the Bureau of Reclamation of the Department of the Interior (Bureau of Reclamation) and Colorado use excess capacity at the Pueblo Reservoir to store a portion of the winter flow of the Arkansas River—violates the Compact. Third, Kansas claimed that Colorado's failure to abide by the Trinidad Reservoir Operating Principles (Operating Principles) constituted a violation of the Compact. Ibid.

The Special Master bifurcated the trial into a liability phase and a remedy phase. At the conclusion of the liability phase, the Special Master filed his Report, outlining his findings and recommendations. In his Report, the Special Master recommended, among other things, that the Court: (1) find that post-Compact well pumping in Colorado has "materially depleted" the "usable" flow at the Colorado-Kansas border (stateline) in violation of Article IV-D of the Compact, Report 336; (2) find that "Kansas has failed to prove that operation of the [WWSP] program has violated the [C]ompact," ibid.; and (3) "dismiss the Kansas claim arising from the operation of the Trinidad Reservoir," ibid.2

Both Kansas and Colorado have filed exceptions to the Special Master's Report. Kansas excepts to the Special Master's rejection of its (1) Trinidad Reservoir claim, see id., at 373-433; (2) WWSP claim, see id., at 306-335; and (3) preferred method for determining the usability of depletions of stateline flows, see id., at 291-305. Colorado excepts to the Special Master's determination that: (1) Kansas was not guilty of inexcusable delay in making its post-Compact well-pumping claim and that Colorado was not prejudiced by this delay, see id., at 147-170; (2) pre-Compact wells in Colorado are limited to pumping the highest amount pumped in the years during which the Compact was negotiated and that the highest amount of such pumping was 15,000 acre-feet per year, see id., at 182-200; (3) increases in usable stateline flows resulting from the operating plan for the John Martin Reservoir adopted by the Arkansas River Compact Administration in 1980 (1980 Operating Plan) were "separately bargained for" and, therefore, should not offset depletions caused by post-Compact well pumping in Colorado, see id., at 171-181; and (4) Kansas need only meet the "preponderance of the evidence" standard to prove a breach of Article IV-D of the Compact, see id., at 65-70.

We turn to the parties' exceptions.

In 1958, Congress authorized construction of the Trinidad Project, a dam and a reservoir system on the Purgatoire River slightly upstream from the city of Trinidad, Colorado. See id., at 382-388. Recognizing that Article IV-D of the Compact prohibited any development of the Arkansas River basin that resulted in a material depletion of usable river flow, the Bureau of Reclamation conducted studies regarding the future operation of the Trinidad Project. Id., at 388-390. The Bureau of Reclamation established Operating Principles whereby the Trinidad Project could be administered "without adverse effect on downstream water users and the inflow to John Martin Reservoir." Id., at 390 (internal quotation marks omitted). The Governor of Kansas reviewed the Bureau of Reclamation's proposed Operating Principles and indicated that if five additional conditions were accepted, then "Kansas would be in a position to approve the amended Operating Principles and to support completion of the project." Id., at 392-393. In June 1967, the Administration approved the Operating Principles as well as Kansas' five additional conditions. Id., at 395.

In 1979, Colorado began storage of water at the Trinidad Reservoir. Id., at 396. Kansas immediately complained that the Operating Principles were being violated. Id., at 397. In 1988, at the request of the Administration, the Bureau of Reclamation conducted a study of the Trinidad Reservoir. It concluded that two storage practices at the Trinidad Reservoir constituted a " 'departure from the intent of the operating principles.' " Ibid.

At trial, Kansas argued that the Operating Principles were binding on the State of Colorado and that any departure from them constituted a violation of the Compact "regardless of injury." Id., at 408 (internal quotation marks omitted). Kansas, however, "offered no evidence, apart from the Bureau studies, to show that the actual operation of the Trinidad project caused it to receive less water than under historical, without-project conditions." Id., at 412. Instead, Kansas sought to quantify depletions by "comparing the flows into John Martin Reservoir 'as they would have occurred under the Operating Principles with the flows that occurred under actual operations.' " Id., at 409. The Special Master concluded that in order to prove a violation of the Compact, Kansas was required to demonstrate that "the Trinidad operations caused a material depletion within the meaning of Article IV-D." Id., at 431. The Special Master recommends that we dismiss Kansas' Trinidad claim because "Kansas has not established, and did not attempt to establish, such injury." Ibid.

Kansas argues that "[d]eparture from the Operating Principles is ipso facto a violation of the Compact, and it [is] entirely sufficient, for purposes of quantifying the effect of the violation, to compare the actual operation with simulated operation as it should have been under the Operating Principles." Kansas' Exceptions to Special Master's Report 12. But, it must be recalled, this is an original action to enforce the terms of the Compact.3 Article IV-D provides that the Compact is not intended to prevent future beneficial development of the Arkansas River basin—including dams and reservoirs —provided that the river flow shall not be materially depleted. The Compact thus permits the development of projects such as Pueblo Reservoir so long as their operation does not result in a material depletion of usable flow to Kansas users. For Kansas to prevail in its contention, it would have to show that the Operating Principles had the effect of amending the Compact by granting either party the right to sue the other for violation of the Operating Principles even though the violation resulted in no material depletion of usable flow at stateline. Although the Administration is empowered to "[p]rescribe procedures for the administration of th[e] Compact," App. to Report 11 (Article VIII-B(2)), it must do so "consistent with the provisions of th[e] Compact," ibid. (Article VIII-B(1)) (emphasis added); see also Report 416 ("[T]he Compact Administration was not delegated power to change the Compact"). The theory advocated by Kansas is inconsistent with Article IV-D, which allows for the development and operation of dams and reservoirs so long as there is no resultant material depletion of usable flows at stateline.

Thus Kansas, in order to establish a Compact violation based upon failure to obey the Operating Principles, was required to demonstrate that this failure resulted in a material depletion under Article IV-D. Kansas "has not established, and did not attempt to establish such injury." Id., at 431. We overrule Kansas' exception to the Special Master's dismissal of its Trinidad Reservoir claim.

In 1964, the Bureau of Reclamation and Colorado began planning a program to use excess capacity at Pueblo Reservoir in order to store a portion of the winter-time flow of the Arkansas River for beneficial use at other times. Under the WWSP, winter-time flow—much of which was used previously to flood uncultivated cropland—is instead stored at the Pueblo Reservoir. Kansas contends that the Special Master erred in finding that it had failed to prove that the WWSP had "materially depleted" usable stateline flows. We disagree.

In his Report, the Special Master concluded:

"Kansas has not proved that the WWSP has caused material Stateline depletions. Kansas' case has not been helped by its own contradictions in quantifying impacts to usable flow ranging during this trial from 255,000 acre-feet initially, to 44,000 to 40,000; nor by the fact that depletions are essentially eliminated if accretions are taken into account." Report 335.

The Special Master examined the computer models submitted by Kansas and Colorado and determined that "the depletions shown by the Kansas model are well within the model's range of error." Id., at 334-335. As a result, "[o]ne [could not] be sure whether impact or error [was] being shown." Id., at 335.

We believe that the Special Master gave Kansas every reasonable opportunity to meet its burden of proving its WWSP claim. Kansas, however, failed to prove that operation of the WWSP program resulted in material depletions of usable flows in violation of Article IV-D. See ibid. Therefore, we overrule Kansas' exception to the Special Master's conclusion that Kansas had failed to prove its WWSP claim.

Article IV-D of the Compact permits future development and construction along the Arkansas River Basin provided that it does not materially deplete stateline flows "in usable quantity or availability." App. to Report 5 (Article IV-D) (emphasis added). In order to establish a violation of Article IV-D, Kansas was required to establish that development in Colorado resulted in material depletions of "usable" river flow. The Compact does not define the term "usable." Cf. Colorado v. Kansas, 320 U.S., at 396-397, 64 S.Ct., at 182 ("The critical matter is the amount of divertible flow at times when water is most needed for irrigation. Calculations of average annual flow, which include flood flows, are, therefore, not helpful in ascertaining the dependable supply of water usable for irrigation"). At trial, Kansas presented three methods for determining depletions of "usable" flow.

Kansas' first expert, Timothy J. Durbin, analyzed flow data for the period between 1951 and 1985 by plotting actual river diversions in Kansas against actual stateline flows. Report 293-294. Using these data, Durbin developed criteria to determine what river flows were usable. Durbin concluded that during the summer months, April through October, (1) 78% of the stateline flows were diverted; (2) flows greater than 40,000 acre- feet per month were not usable; and (3) flows greater than 140,000 acre-feet for the whole period were not usable. Id., at 293. With respect to the winter months, November through March, Durbin concluded that (1) 24% of the winter flow was diverted; (2) flows greater than 7,500 acre-feet per month were not usable; and (3) flows greater than 40,000 acre-feet for the whole period were not usable. Id., at 293-294.

After Colorado isolated errors in Durbin's analysis, Kansas presented a replacement case. Kansas' second group of experts, led by Stephen P. Larson, adopted the same methodology but revised certain exhibits and made minor corrections in data. As a result, Larson modified Durbin's coefficients, using 72% for the summer months and 25% for the winter months. Id., at 295.

Later, well after trial had begun, Kansas enlisted the aid of Brent Spronk, who proposed yet another method to quantify depletions of "usable" stateline flow. Id., at 300-305. Spronk attempted to determine the "percentage of days in each month when flows were being fully used in Kansas." Id., at 301. Instead of seasonal averages, the Spronk approach yielded coefficients that varied from month to month. Spronk then multiplied these monthly coefficients by the estimated depletions in flow predicted by Kansas' hydrological model. Id., at 301-302.

The Special Master concluded that "the Durbin approach, using Larson's coefficients, is the best of the several methods presented for determining usable flow" and that it provided "a reasonable way in which to determine depletions of usable flow." Id., at 305. We agree. Each of the three methods that Kansas proposed for calculating usable depletions required two steps: (1) a calculation of total depletions using the Kansas hydrological model, and (2) an application of "usability" criteria. See Brief for United States in Response to Exceptions of Kansas and Colorado 30. Each of the three methods proposed by Kansas was dependent on the Kansas hydrological model to estimate total depletions. The Spronk method required the Kansas hydrological model to predict accurately depletions for each and every month. Report 303. But as Durbin, Kansas' first expert, testified, Kansas' hydrological model was only a " 'good predictor' when 'looking at long periods of time.' " Id., at 303, n. 130 (quoting Durbin's testimony). Thus, the Spronk method required the Kansas hydrologic model to do something it was not designed to do, i.e., predict accurately depletions on a monthly basis. Id., at 303 ("The Spronk analysis assumes that the H-I model can accurately predict changes of Stateline flow on a monthly basis"). Because the Spronk method for determining "usable" river flows was less compatible with Kansas' hydrological model than the other methods proposed, we conclude that the Special Master properly rejected the Spronk method in favor of the Durbin approach, as modified by the Larson coefficients.

The Special Master concluded that Kansas was not guilty of inexcusable delay in making its well-pumping claim, and that Colorado had not been prejudiced by Kansas' failure to press its claim earlier. Id., at 170. Colorado has excepted to this determination. Colorado argues that the equitable doctrine of laches should bar Kansas' claim for relief. See Colorado's Exceptions to Special Master's Report (Colorado's Exceptions) 24-64. We overrule Colorado's exception.

The defense of laches "requires proof of (1) lack of diligence by the party against whom the defense is asserted, and (2) prejudice to the party asserting the defense." Costello v. United States, 365 U.S. 265, 282, 81 S.Ct. 534, 543, 5 L.Ed.2d 551 (1961); see also Black's Law Dictionary 875 (6th ed. 1990) (" 'Doctrine of laches,' is based upon maxim that equity aids the vigilant and not those who slumber on their rights. It is defined as neglect to assert a right or claim which, taken together with lapse of time and other circumstances causing prejudice to the adverse party, operates as bar in court of equity"). This Court has yet to decide whether the doctrine of laches applies in a case involving the enforcement of an interstate compact. Cf. Illinois v. Kentucky, 500 U.S. 380, 388, 111 S.Ct. 1877, 1883, 114 L.Ed.2d 420 (1991) (in the context of an interstate boundary dispute, "the laches defense is generally inapplicable against a State"); Block v. North Dakota, 461 U.S. 273, 294, 103 S.Ct. 1811, 1823, 75 L.Ed.2d 840 (1983) (O'Connor, J., dissenting) ("The common law has long accepted the principle 'nullum tempus occurrit regi'—neither laches nor statutes of limitations will bar the sovereign"); Colorado v. Kansas, 320 U.S., at 394, 64 S.Ct., at 181 (in the context of a suit seeking an equitable apportionment of river flows, facts demonstrating a delay in filing a complaint "might well preclude the award of the relief [requested]. But, in any event, they gravely add to the burden [the plaintiff] would otherwise bear"). We need not, however, foreclose the applicability of laches in such cases, because we conclude that Colorado has failed to prove an element necessary to the recognition of that defense. See Costello, supra, 365 U.S., at 282, 81 S.Ct., at 543.

Colorado argues that Kansas knew or should have known by 1956, or at the latest, before 1968, that both the number of post-Compact wells and the amount of post-Compact pumping in Colorado had increased substantially. Colorado's Exceptions 37, 39. Colorado argues that by 1956 Kansas had sufficient information about increased well pumping in Colorado and its potential impact on usable stateline flows to call for an investigation to determine if a Compact violation existed. Id., at 46.

The Special Master concluded that prior to 1984, Kansas had made no formal complaint to the Administration regarding post-Compact well pumping in Colorado. Report 155-156. Nevertheless, the Special Master concluded that Colorado's evidence did not "deal with the issue of impact on usable flow at the Stateline," id., at 161, and did "not demonstrate that [the Kansas officials] were aware of the number of wells, the extent of Colorado's pumping, or the impact or even potential impact of pumping on usable Stateline flows," id., at 164. The Special Master explained the difficulty of assessing the impact of increases in post-Compact well pumping on usable stateline flows because of changing conditions during the 1970's and early 1980's:

"The 1970s were generally dry years and some reduction in flow was to have been expected. Pueblo Dam came on line in 1976 and began to regulate native flows. Transmountain imports increased, which to some extent provided an offset to pumping. The 1980 Operating Plan was placed into effect, which Colorado alleges offset the impacts of increased pumping downstream from John Martin Reservoir. The Winter Water Storage Program was instituted. Moreover, there was no quantitative or specific entitlement against which depletions to usable flow could be judged. Nor were there any agreed upon criteria for establishing what flows were usable." Id., at 162-163.

As late as 1985, Colorado officials refused to permit an investigation by the Administration of well development in Colorado because they claimed that the evidence produced by Kansas did not " 'suggest that well development in Colorado has had an impact on usable stateline flows.' " Id., at 163 (quoting memorandum of Mr. J. William McDonald, chief of the Colorado delegation to the Administration). In light of the vague and conflicting evidence available to Kansas, we conclude that Colorado has failed to demonstrate lack of diligence, i.e., inexcusable delay, on the part of Kansas.

Accordingly, we overrule Colorado's exception to the Special Master's conclusion that the defense of laches should not bar Kansas' well-pumping claim.

The Compact prohibits "future beneficial development of the Arkansas River basin" that "materially deplete[s]" the usable flows of the Arkansas River. App. to Report 5 (Article IV-D) (emphasis added). Because some wells in Colorado were in existence prior to the Compact, both parties agree that a certain amount of post-Compact well pumping is allowable under the Compact. Report 182. Kansas and Colorado, however, dispute the extent of this allowance. The Special Master determined that the "highest annual amount shown to have been pumped during the negotiations, namely 15,000 acre-feet, should be allowed under the [C]ompact." Id., at 200. Colorado makes both a legal and a factual challenge to this determination. Colorado's Exceptions 66-73, 73-84.

Colorado argues as a legal matter that the Compact does not limit the pumping by pre-Compact wells to the highest amount actually pumped in pre-Compact years; rather, Colorado claims that the limit on its pre-Compact pumping is the maximum amount that Colorado law permitted or the maximum amount of pumping possible using wells existing prior to the Compact. Id., at 69-70. In support of its position, Colorado argues that the Special Master failed to consider the subsequent practice of the parties, i.e., Kansas' failure to object to replacement of centrifugal pumps with turbine pumps or increased pumping by pre-Compact wells, and that Article VI-A(2) of the Compact supports its position.4

We conclude that the clear language of Article IV-D refutes Colorado's legal challenge. Article IV-D permits "future beneficial development of the Arkansas River basin . . . which may involve construction of dams, reservoir, and other works for the purposes of water utilization and control, as well as the improved or prolonged functioning of existing works: Provided, that the waters of the Arkansas River . . . shall not be depleted in usable quantity or availability. . . ." App. to Report 5 (emphasis added). Regardless of subsequent practice by the parties, improved and increased pumping by existing wells clearly falls within Article IV-D's prohibition against "improved or prolonged functioning of existing works," if such action results in "materia[l] deplet[ions] in usable" river flows. Ibid.; see Texas v. New Mexico, 462 U.S. 554, 564, 103 S.Ct. 2558, 2565, 77 L.Ed.2d 1 (1983) ("[U]nless the compact to which Congress has consented is somehow unconstitutional, no court may order relief inconsistent with its express terms"). Article VI-A(2) of the Compact, which begins with the phrase, "Except as otherwise provided," App. to Report 10, must be read in conjunction with and as limited by Article IV-D. We agree with the Special Master that "new wells, the replacement of centrifugal with turbine pumps, and increased pumping from [pre-Compact] wells all come within [Article IV-D]." Report 194.

Second, Colorado argues as a factual matter that the Special Master unreasonably relied upon faulty reports by the United States Geological Survey (USGS) and the Colorado Legislature to conclude that the greatest amount of annual pre-Compact pumping in Colorado was 15,000 acre-feet. Colorado's Exceptions 73-74. The Special Master concluded:

"There is no precise answer to the amount of [pre-Compact] pumping. . . . That amount must simply remain as an estimate of water use that affected the general allocation of water between the states when the [C]ompact was being negotiated. Two responsible reports, one published by the USGS and one prepared by the Colorado legislature, reached similar conclusions as to the amounts of Colorado pumping during the 1940s. . . . They have since been used by the Colorado State Engineer. I have relied on these reports and recommend that the highest annual amount shown to have been pumped during the negotiations, namely 15,000 acre-feet, should be allowed under the [C]ompact." Report 199-200.

Although the ultimate responsibility for deciding what are correct findings of fact remains with the Court, Colorado v. New Mexico, 467 U.S. 310, 317, 104 S.Ct. 2433, 2438, 81 L.Ed.2d 247 (1984), in this instance, we are in full agreement with the Special Master. Accordingly, we overrule Colorado's exception.

In April 1980, the Administration adopted a resolution concerning the method for operating John Martin Reservoir (1980 Operating Plan). Report 47. The 1980 Operating Plan divides the water conserved in John Martin Reservoir into separate accounts. Kansas is allocated 40% of the conservation storage, with the remaining 60% being divided in specified percentages among the nine canal companies in Colorado Water District 67. Id., at 173. The Special Master concluded that the 1980 Operating Plan for the John Martin Reservoir was "separately bargained for" and therefore should not offset depletions caused by post-Compact well pumping in Colorado. Id., at 180-181. Colorado takes exception to this ruling.

Colorado argues that increases in usable stateline flows resulting from the 1980 Operating Plan should offset depletions to usable stateline flows. Colorado's Exceptions 85. Colorado maintains that the Administration adopted the 1980 Operating Plan "for more efficient utilization of water under its control because of changes in the regime of the Arkansas River," id., at 91, "including [post-Compact] well pumping in Colorado and Kansas," ibid.; see also App. to Report 107 (Resolution Concerning an Operating Plan for John Martin Reservoir) ("WHEREAS, the Arkansas River Compact Administration . . . recognizes that, because of changes in the regime of the Arkansas River, the present operation of the conservation features of John Martin Reservoir does not result in the most efficient utilization possible of the water under its control"). We disagree.

As Colorado acknowledges, the resolution adopting the 1980 Operating Plan "does not state that [post-Compact] well pumping in Colorado or Kansas was a cause of changes in the regime of the Arkansas River." Colorado's Exceptions 88. In fact, Colorado argues in a separate part of its brief that "Kansas had made no complaint about well pumping in Colorado to the Compact Administration . . . before 1984." Id., at 32. The 1980 Operating Plan expressly reserves the parties' rights under the Compact, stating that "[a]doption of this resolution does not prejudice the ability of Kansas or of any Colorado ditch to object or to otherwise represent its interest in present or future cases or controversies before the Administration or in a court of competent jurisdiction." App. to Report 116. The Special Master concluded:

"The 1980 Operating Plan provided benefits to both Kansas and Colorado which were separately bargained for. There is no evidence to support the claim that benefits to Kansas were in settlement of its well claims. Colorado received ample consideration under the agreement for the 1980 plan without a waiver of Kansas' well claims. The benefits received by Kansas under the plan should not be offset against compact violations, and should not be a bar to any of the Kansas claims in this case." Report 180-181.

We agree with the Special Master's resolution of Colorado's claim. Accordingly, we overrule Colorado's exception.

Finally, Colorado argues that Kansas is required to prove its well-pumping claim by clear and convincing evidence. Colorado's Exceptions 91. The Special Master, relying upon Nebraska v. Wyoming, 507 U.S. ----, 113 S.Ct. 1689, 123 L.Ed.2d 317 (1993), concluded that the proper burden of proof for enforcing an interstate compact is the preponderance of the evidence standard. Report 70. The Special Master noted that the Nebraska Court had drawn a distinction between actions seeking to "modify" a judicial decree and actions seeking to "enforce" a judicial decree. See Nebraska, supra, at ----, 113 S.Ct., at 1695. ("[W]e find merit in [the] contention that, to the extent that Nebraska seeks modification of the decree rather than enforcement, a higher standard of proof applies"). The Special Master concluded that an action seeking to enforce an interstate compact stood on the same footing as an action enforcing a judicial decree, and therefore was subject to the "preponderance of the evidence" standard. Report 70.

We need not, however, resolve this issue. The Special Master concluded that "regardless of which burden of proof applies" he had "no difficulty in concluding that [post-Compact] pumping in Colorado ha[d] caused material depletions of the usable Stateline flows of the Arkansas River, in violation of the Arkansas River Compact." Id., at 263. We agree with this determination, and thus overrule Colorado's exception.

For these reasons, we overrule the exceptions filed by the States of Kansas and Colorado. We remand the case to the Special Master for determination of the unresolved issues in a manner not inconsistent with this opinion.

It is so ordered.