The United States, the Gila River Indian Community (GRIC), and the San Carlos Apache Tribe brought suit against the Gila Valley Irrigation District (GVID), the state of Arizona, and other defendants, alleging that the various defendants were violating the terms of a 1935 Consent Decree which established the respective rights of the parties to the waters of the Gila River.

After trial, the district court entered judgment in favor of GVID on GRIC's claim that the Water Commissioner is violating the Decree by using the "stored water released" method of computing the amount of Gila River water available for additional apportionments. Judgment was entered in GRIC's favor on GRIC's claim that the Commissioner is violating the Gila Decree when calculating additional water apportionments by failing to make appropriate deductions for "losses for evaporation, seepage or otherwise," incurred while the water is in transit below the Coolidge Dam to the point of diversion at the Ashurst-Hayden Dam. Each appeals. We have jurisdiction under 28 U.S.C. § 1291, and we affirm the district court's judgment on both counts.

* By Acts of May 18, 1916, 39 Stat. 130, and June 7, 1924, 43 Stat. 475, Congress authorized the San Carlos Irrigation Project. The 1924 Act specifically authorized the construction of a dam, now known as the Coolidge Dam, on the Gila River. It is located on the San Carlos Indian Reservation and impounds the San Carlos Reservoir.

In 1925, the United States, acting on its own behalf and as trustee and guardian for the Pima and Apache Indians, brought an action in the United States District Court for the District of Arizona, seeking a determination of the relative rights and priorities of Indians and non-Indians to the waters of the Gila River. United States v. Gila Valley Irrigation Dist., Globe Equity No. 59. On June 29, 1935, the parties stipulated to the Gila Consent Decree.

In addition to setting out the rights of the parties to Gila River water, the Decree provided that the district court would appoint a Water Commissioner to carry out and enforce the Decree. Gila Decree Art. XII. The Gila Decree also provided for the district court's retaining jurisdiction to review the actions of the Water Commissioner and to enforce the Decree. Gila Decree Arts. XII, XIII.

The current litigation began in 1976, when the United States filed a petition to review the actions of the Water Commissioner. The petition alleged that the upper valley water users were pumping water out of wells in violation of the Gila Decree.1 The district court ordered the United States to give notice to individual upper valley defendants whose pumping allegedly was violating the Decree.

Thereafter, the action remained dormant until GRIC successfully moved to intervene as a plaintiff. On May 25, 1989, GRIC filed a nine-count Amended Complaint in Intervention as Plaintiff. Pursuant to a stipulation, the first four counts were bifurcated. After a trial on the merits, the district court filed its Findings of Fact and Conclusions of Law concerning these first four counts on April 3, 1990, and on August 14, 1990, entered judgment.

We "will not vacate findings of fact unless they are clearly erroneous. As long as findings are plausible in light of the record viewed in its entirety, a reviewing court may not reverse even if convinced it would have reached a different result." Wardley Int'l Bank, Inc. v. Nasipit Bay Vessel, 841 F.2d 259, 261 n. 1 (9th Cir.1988) (citation omitted).

Additionally,

[w]hen the district court's decision is based on an analysis of the contractual language and an application of the principles of contract interpretation, that decision is a matter of law and reviewable de novo. When the inquiry focuses on extrinsic evidence of related facts, however, the trial court's conclusions will not be reversed unless they are clearly erroneous.

Kern Oil & Ref. Co. v. Tenneco Oil Co., 840 F.2d 730, 736 (9th Cir.) (quoting Miller v. Safeco Title Ins. Co., 758 F.2d 364, 367 (9th Cir.1985)), cert. denied, 488 U.S. 948, 109 S. Ct. 378, 102 L. Ed. 2d 367 (1988).

Therefore, we will review the district court's interpretation of the Gila Decree de novo, but will defer to its factual findings on extrinsic evidence unless they are clearly erroneous. See id.; see also United States v. ITT Continental Baking Co., 420 U.S. 223, 233-37, 95 S. Ct. 926, 932-35, 43 L. Ed. 2d 148 (1975) (consent decree treated as contract when construing its meaning).

Article VIII of the Gila Decree provides that on January 1 of each year, the Water Commissioner shall apportion a quantity of water to the upper valleys based on the amount of water available for release from the Coolidge Dam:

[O]n the first day of January of each Calendar year, or as soon thereafter as there is water stored in the San Carlos Reservoir, which is available for release through the gates of the Coolidge Dam for conveyance down the channel of the Gila River and for diversion and use on the lands of the San Carlos Project for the irrigation thereof, then the Water Commissioner ... shall, to the extent and within the limitations hereinafter stated, apportion for the ensuing irrigation year to said defendants from the natural flow of the Gila River an amount of water equal to the above described available storage, and shall permit the diversion of said amount of water from said stream into the canals of said defendants for the irrigation of said upper valleys lands in disregard of the aforesaid prior rights of plaintiff used on lands below said reservoir....

Gila Decree Art. VIII(2).

The upper valley defendants also are entitled to additional apportionments of water based on additional amounts of water that are stored in the San Carlos Reservoir after January 1. Article VIII provides that the additional apportionments made to the upper valleys shall be equal to the amount of "accessions or newly available stored water supply" in the reservoir:

[I]f and when at any time or from time to time in any year, water shall flow into said reservoir after said date of first apportionment and shall be stored there and become added to the available stored water in said reservoir, the said commissioner shall make further and additional apportionments to said defendants of the natural flow of said stream as the same is available at the diversion points of said defendants, which said apportionments shall in turn correspond with and be equivalent in quantity to the amount of such accessions or newly available stored water supply....

Gila Decree Art. VIII(2).

The district court, in 1939, defined "accessions thereto" as "all waters flowing into said reservoir in excess of the amount discharged therefrom at the time of such inflow, allowance being made for loss through evaporation and seepage."

The Water Commissioner uses two different methods for calculating the amounts of the additional apportionments. When using Method 1, the Commissioner determines the amount of "accessions or newly available stored water supply" by measuring the rises in the height of the water in the reservoir.2 Essentially, the Commissioner determines the height of the water and uses capacity tables to calculate the amount of water stored in the reservoir. He then estimates the period of time the water will remain in the reservoir before all of it is released down-river for use by GRIC and other lower valley parties. The Commissioner next estimates the amount of evaporation which will occur over that period of time. The additional apportionment for that period is calculated as the amount of available stored water (determined by the height of water and capacity tables) minus the estimated evaporation losses for that period.

Method 1 has certain built-in sources of potential inaccuracy. First, the tables used to compute the volume of water in the reservoir are only accurate to the nearest 1,000 acre feet. Second, because the Commissioner is merely estimating both the length of time it will take the lower valley users to deplete all the stored water and the amount of evaporation losses which will occur over that length of time, total accuracy is impossible.

Method 2 is known as the "stored water released" method. Using this method, stored water released is calculated as the excess of the measured amount of water leaving the reservoir on a given day over the measured amount of water entering the reservoir on that same day. Water leaving the reservoir is measured by a gauging station located below the Coolidge Dam, and water entering the reservoir is measured by gauging stations located on two rivers which enter the reservoir. The Commissioner assumes that the excess amount of water released was previously stored in the reservoir, hence the name "stored water released."

When determining apportionments under Method 2, the Commissioner calculates the sum of stored water released for all the days since the prior apportionment, adds the amount of water stored in the reservoir on the day of the current apportionment, and deducts from that sum any previous apportionments made that year. The final figure is the amount of the additional apportionment.

Method 2, like Method 1, has two sources of potential inaccuracy. First, the gauges used to measure inflow and outflow are not particularly accurate. While the gauge used to measure outflow has a potential error of only plus/minus 5%, one inflow gauge has a potential error of more than plus/minus 15%, and the other gauge has a potential error of less than plus/minus 15%. Second, a certain amount of unmeasured inflow enters the reservoir, thus resulting in an improperly inflated figure for stored water released. These unmeasured inflows come from rain flowing into the reservoir, from underflow, and from tributaries and runoff from land below the inflow gauging stations and above Coolidge Dam.

The district court found that "[t]he Water Commissioner's methods of computing apportionments, both the initial apportionment and supplemental apportionments, are in compliance with the Decree, and are the most practical methods of computing apportionments." The court also concluded as a matter of law that, except for the requirement of deducting transit losses, discussed infra Part IV, "[t]he Water Commissioner's method of supplemental apportionments based on stored water released is otherwise appropriate under the Decree."

GRIC argues that the Commissioner's use of Method 2--stored water released--violates the Decree. It puts forth two arguments, neither of which is persuasive. GRIC first argues that Method 2 does not comply with the Decree as a matter of law. GRIC notes that, under Article VIII(2) of the Decree, additional apportionments are to equal the reservoir's "accessions or newly available stored water supply," and that "accessions" were defined by the district court in 1939 as "all waters flowing into said reservoir in excess of the amount discharged therefrom at the time of such inflow." GRIC argues that Method 2 is improper because, instead of subtracting reservoir outflow from reservoir inflow to determine "accessions or newly available stored water," this method merely attempts to determine measured depletion from the reservoir's stored water supply. In other words, Method 2 allegedly does not comply with the language of the Decree because it measures depletions, not accessions, to the water supply.

We conclude that Method 2 does not violate the Decree as a matter of law. The Decree merely requires that supplemental apportionments be based on accessions or newly available stored water in the reservoir. Even though the district court specifically defined "accessions" as "all waters flowing into the reservoir in excess of the amount discharged therefrom at the time of such inflow," neither the Decree nor the district court's instructions mandated that any particular method be used to measure the accessions. Method 2 does not attempt to redefine "accessions" or to alter the manner in which supplemental apportionments are made; it simply is one method by which the Commissioner attempts to measure the actual amount of accessions to the reservoir. Even though Method 2 measures "stored water released" or "depletions," the purpose of such measurements is to calculate the actual accessions; thus, Method 2 cannot fairly be said to violate the Decree as a matter of law.

GRIC's second argument is that the Commissioner's use of Method 2 is "clearly erroneous, contrary to law, and prejudicial to Indians' immemorial right." GRIC argues first, that Method 2 is so highly inaccurate that the district court erred in finding that this method was an appropriate means of calculating supplemental apportionments, and second, that the Decree must be "construed most strongly in favor of the Indians," and that Method 2 cannot be used because it results in an interpretation of the Decree that does not provide appropriate protection to GRIC, which is composed of Indians.

This boils down to an attack on the accuracy of Method 2 as follows: A large quantity of unmeasured water flows into the reservoir. This water is unmeasured because the gauges are inaccurate, and because some water enters the river below the inflow gauges but above Coolidge Dam. When all reservoir inflows--both measured and unmeasured--are passed through the dam and not retained in the reservoir, those inflows are natural flow waters, not stored waters. If, however, the unmeasured reservoir inflows that pass through the dam as natural flow waters are measured as reservoir outflows, then those waters will improperly be counted as "stored water released" under Method 2. Thus, the existence of unmeasured inflow, which then is measured as outflow, results in the over-calculation of stored water released.

GRIC's position is supported by some evidence in the record. For example, all the parties admit that the gauges are inaccurate and that at least a certain amount of unmeasured water flows into the reservoir. Even if supported by the record, however, the inaccuracies cited by the GRIC are insufficient to warrant reversal. The district court found that both Method 1 and Method 2 are "in compliance with the Decree, and are the most practical methods of computing apportionments." Implicit in this finding is the district court's factual determination that Method 2 is a sufficiently accurate method of calculating the reservoir's "accessions or newly available stored water supply."

A review of the record indicates that the district court's findings are at least plausible, and that those findings therefore cannot be deemed clearly erroneous. See Wardley Int'l Bank, 841 F.2d at 261 n. 1. Method 2 may not be perfect, but Method 1 also is less than perfectly accurate. GRIC points out that because the gauges are so inaccurate, the potential error in Method 2 is significantly greater than any potential error in Method 1. While this point is well taken, it is insufficient to render the district court's findings clearly erroneous. Even though Method 1 may be better than Method 2, and even though Method 2 measurements potentially could be quite inaccurate, the evidence is insufficient to demonstrate that Method 2 actually is so untrustworthy as to render clearly erroneous the district court's finding that Method 2 is acceptable. Additionally, GRIC conceded at oral argument that it has no right under the Decree to an accurate method of measurement. Thus, even if Method 2 is not accurate, that in and of itself does not mandate that the Commissioner exclusively use Method 1.

GRIC next argues that the district court's ruling violates the principle that "agreements relating to Indian rights [are] to be construed most strongly in favor of the Indians." United States v. Ahtanum Irrigation Dist., 236 F.2d 321, 340 (9th Cir.1956), cert. denied, 352 U.S. 988, 77 S. Ct. 386, 1 L. Ed. 2d 367 (1957). The Ahtanum court also noted that in water rights cases involving Indians, the agreements "must be construed as reserving to the Indians, who previously owned substantially all of the waters, everything not clearly shown to have been granted [to non-Indians]." Id. at 341. Thus, claims GRIC, the district court should have ruled that Method 2 was improper, because by doing so the court would properly have been construing the Decree "most strongly" in favor of GRIC.

To bolster this claim, GRIC also points to our earlier decision involving an interpretation of the Gila Decree. In United States v. Gila Valley Irrigation District, 454 F.2d 219, 222 (9th Cir.1972), we noted that the Commissioner could not permit water, other than apportioned water, "to be diverted as against the water rights of plaintiffs except in strict accordance with the priority schedule set forth in Article V of the Decree." That holding, GRIC argues, demonstrates that the Decree is to be narrowly construed, and that we should not permit the use of Method 2 because to do so would allow the upper valley defendants to receive non-apportioned water in violation of the Decree.

This argument is not persuasive because the district court's judgment does not run afoul of the rule that "agreements relating to Indian rights [are] to be construed most strongly in favor of the Indians." Ahtanum, 236 F.2d at 340. Indeed, on this one issue, the district court never construed the terms of the Decree in any manner. Rather, the court simply made a factual determination as to the practicality of a method of measuring reservoir accessions. While Ahtanum and Gila Valley may require that a court interpret the Decree in a manner favorable to GRIC, neither case requires that findings of fact (such as the finding that Method 2 was a practical way of measuring reservoir accessions) be slanted in favor of GRIC.3IV

As we have already discussed, the Decree requires that the Commissioner make apportionments of Gila River water to the upper valley defendants based upon the quantity of water stored in the reservoir.4 The Decree further states that

in calculating apportionments of the stored water supply the Water Commissioner shall make appropriate deductions for losses for evaporation, seepage or otherwise that may be suffered between the time of the apportionment and that of the diversion of a corresponding quantity of water from the stream.

Gila Decree Art. VIII(2).

The Commissioner, when calculating the quantity of water to be apportioned, does make deductions for evaporation losses,5 but such losses are only measured from the time of apportionment until the time the Commissioner estimates that the water would be released from Coolidge Dam. The district court specifically found that "[w]hen calculating apportionments, the Water Commissioner has never made deductions for seepage, or for transit losses below Coolidge Dam to the point of diversion at Ashurst-Hayden Dam."

The failure to deduct the quantity of water lost in transit during the two-day journey from the Coolidge Dam to the Ashurst-Hayden Dam results in a significantly higher quantity of water being apportioned to the upper valleys. Although the testimony varied, somewhere between 8% to 31% of the total water released could be lost in transit.

After acknowledging that the Commissioner has never made deductions for transit losses, the district court concluded, as a matter of law, that "[t]he Decree requires the Water Commissioner, when making apportionments, to make deductions for transit losses below Coolidge Dam to the point of diversion at Ashurst-Hayden Dam, and for seepage." The court also directed the Commissioner "to propose a method of calculating seepage and transit losses below Coolidge Dam and to the point of diversion of Ashurst-Hayden Dam."6

GVID first argues that the district court's ruling is contrary to the express language of the Decree. Although GVID points to several different clauses in support, we conclude that the district court correctly interpreted the terms of the Decree.

First, GVID argues that the Decree requires the Commissioner to take deductions from the "stored water supply," but that after the water is released from the reservoir it is no longer "stored." Thus, transit losses should not be deducted because such losses simply are not losses to the "stored water supply." The language of the Decree, however, is not reasonably susceptible to the construction GVID urges. The Decree states that when

there is water stored in the San Carlos Reservoir, which is available for release through the gates of the Coolidge Dam for conveyance down the channel of the Gila River and for diversion and use on the lands of the San Carlos Project ..., then the Water Commissioner ... shall ... apportion for the ensuing irrigation year to said defendants from the natural flow of the Gila River an amount of water equal to the above described available storage....

Gila Decree Art. VIII(2) (emphasis added).

This provision explicitly provides that the "water stored" which is to be used in calculating the apportionments to the upper valley defendants is that water in the reservoir which is available both for release through the Coolidge Dam and "for diversion and use on the lands of the San Carlos Project." Water which is released from the dam but which is lost in transit will not be available for diversion and use on the lands of the San Carlos Project. Therefore, in order for the apportionment to the upper valleys to be equal to the amount of stored water available for use and diversion on the lands of the lower valleys (as is expressly required by the Decree), the "appropriate deductions for losses for evaporation, seepage or otherwise" must be made up to the point at which the lower valley users actually divert the water for use, at the Ashurst-Hayden Dam. The deduction of transit losses thus complies with the unambiguous mandate of the Decree to have the upper valley apportionment be equivalent to that quantity of stored water which actually is available for diversion in the lower valleys.

GVID also points to the testimony of an expert witness, Gookin, to support its contention that water in transit is not stored water. This argument misses the mark because in calculating an apportionment, the Commissioner must look to the quantity of stored water which becomes available for diversion and use on the San Carlos Project. Therefore, even if "stored water" is defined as the water kept behind the Coolidge Dam, a deduction for transit losses is still required in order to determine that quantity of stored water which is actually available for use in the lower valley.7

The next argument made by GVID relies on the statement in the Decree that

the Water Commissioner shall make appropriate deductions for losses for evaporation, seepage or otherwise that may be suffered between the time of the apportionment and that of the diversion of a corresponding quantity of water from the stream.

Gila Decree Art. VIII(2) (emphasis added).

Focusing on the word "corresponding," GVID claims that because the upper valley defendants do not divert water from the reservoir, but rather divert a quantity of water from the Gila River above the reservoir that "corresponds" to the quantity of water stored in the reservoir, the Decree must require deductions for losses that occur between the time of apportionment and the time the upper valley defendants divert a corresponding quantity of water from the river upstream. Therefore, continues GVID, if a corresponding quantity is diverted from the river above the reservoir before the water stored in the reservoir is released, the water stored in the reservoir will not have suffered any transit losses. As a result, concludes GVID, deductions for transit losses are inappropriate.

The Decree, however, does not read this way. When read in light of the portion of Article VIII which refers to stored water available for release and diversion on the lands of the San Carlos Project, it is apparent that the phrase "the diversion of a corresponding quantity of water from the stream" refers to the lower valley users ' diversion of water in a quantity that corresponds to the final apportionment made to the upper valley users. The Decree specifically states that the apportionments of water to the upper valley must be equivalent in quantity to the amount of water actually available for diversion and use on the San Carlos Project. The only way to achieve the mandated equality is to make the requisite deductions for losses occurring between the time of apportionment and the time that the lower valley users actually divert the water for their own use. Thus, the term "corresponding quantity of water" must refer to the lower valley users' diversion of water if the Decree's basic rule of equivalency is to be respected.

GVID also argues that its proposed reading does not result in any inequity because the upper valley defendants, unlike the lower valley users, are limited to a total consumptive use from the river of 120,000 acre feet. Because this consumptive use is measured by gauging stations on the river and not by the actual use of water on farms, the upper valley farmers suffer losses from such uncontrollable sources as evaporation and seepage, but this lost water is still counted toward their consumptive use limitation. Why, GVID asks, are the lower valley users entitled to deduct transit losses while the upper valley users remain uncompensated for similar losses? The answer is that the Decree explicitly protects the lower valley users by providing that losses occurring between the time water is released from the dam until the time the water is available for diversion in the lower valley are to be deducted when making apportionments. The Decree provides no analogous protection for upper valley water users. We decline to depart from the unambiguous language of the Decree just because adherence to that language may favor one party over another to some degree.

The GVID's next argument is that, while the Decree specifically mentions evaporation and seepage losses, it does not mention transit losses. The word "otherwise," argues the GVID, should not be interpreted to refer to transit losses because such losses are quite significant. GVID claims that the drafters of the Decree would not have intended a mere catch-all phrase such as "otherwise" to encompass such a significant loss. Additionally, portions of Articles V and VI specifically deal with transit-type losses, and the drafters could have mentioned transit losses in Article VIII also. The fact that they did not do so, argues GVID, again demonstrates that the drafters had no intention of allowing deductions for such losses.

This argument is not convincing. The phrase "evaporation, seepage or otherwise" refers to physical events that deplete the quantity of available water. The word "otherwise" refers to other unenumerated physical events, such as runoff or leakage perhaps, which could deplete the water supply. "Transit losses," on the other hand, merely is a phrase used to identify the place at which such losses due to physical events can occur. Because the phrase "evaporation, seepage or otherwise" does not enumerate or limit the locations at which such physical losses can occur, the failure to enumerate "transit losses" does not preclude the court from deducting losses incurred while the water is in transit from the Coolidge Dam to the Ashurst-Hayden Dam.

Actually, the relevant phrase in the Decree for determining the locations at which losses may be deducted is "that may be suffered between the time of the apportionment and that of the diversion of a corresponding quantity of water from the stream." In accordance with this phrase, any water loss due to physical events such as evaporation that occurs between the Coolidge Dam and the Ashurst-Hayden Dam (where the water is actually diverted for use in the lower valleys) should be deducted by the Commissioner.

GVID also contends that the failure of any party from 1936 to 1984 to object to the Commissioner's failure to deduct transit losses demonstrates that the original intention of the parties was that transit losses should not be deducted. GVID relies on Blinderman Construction Co. v. United States, 695 F.2d 552, 558 (Fed.Cir.1982), for the proposition that when construing the Decree, the contemporaneous construction by the parties to it must be given great weight.

While it is true that rules of construction, such as the rule in Blinderman, may be appropriate in interpreting a consent decree, see ITT Continental Baking Co., 420 U.S. at 238, 95 S.Ct. at 935, we will only apply such rules if the terms of the decree are ambiguous. See, e.g., ITT Continental, 420 U.S. at 238 n. 11, 95 S. Ct. at 935 n. 11 ("evidence of events surrounding [a consent decree's] negotiation and tending to explain ambiguous terms would be admissible in evidence") (quotation omitted and emphasis added); Shakey's Inc. v. Covalt, 704 F.2d 426, 434 (9th Cir.1983) ("Any ambiguity is resolved by resort to common cannons of interpretation."); Pavlik v. Consolidation Coal Co., 456 F.2d 378, 380 (6th Cir.1972) ("When clear contract language itself reveals the intent of the parties, there is no need to turn to rules of construction.") The fact that the parties disagree as to the Decree's meaning does not in and of itself establish that the Decree is ambiguous. International Union of Bricklayers v. Martin Jaska, Inc., 752 F.2d 1401, 1406 (9th Cir.1985). Rather, the terms of the Decree unambiguously support the district court's interpretation. As a result, the Blinderman rule of construction is inapplicable.

AFFIRMED.