OPINION
BRUGGINK, Judge.This is a suit for compensation based upon the fifth amendment to the Constitution of the United States. Plaintiffs, owners of property interests in Baja California, Mexico, allege that the construction and operation of Hoover Dam and Glen Canyon Dam in the United States resulted in a substantially reduced flow of water in the Colorado River, causing the development of a sediment blockage below their properties. They allege that the river has either directly overflowed their land, or caused elevated groundwater conditions, resulting in the taking of a permanent flowage easement. After trial and a thorough examination of the record, the court concludes that plaintiff Jesus Mosqueda’s property was taken but that plaintiff Gordon Bailey’s was not.
I. PRIOR CASE HISTORY
Originally these consolidated cases involved thirty-six plaintiffs. The claims of all plaintiffs except for Mosqueda and Bailey, however, have been either voluntarily dismissed or dismissed for failure to answer interrogatories.
Prior to trial the court issued two rulings on dispositive motions filed by defendant. The first, a motion to dismiss or, in the alternative, for summary judgment, was granted in part, sub nom. Anderson v. United States, 7 Cl.Ct. 341, 346 (1985). The court held that under 28 U.S.C. § 1502 (1982) it lacked jurisdiction over plaintiffs’ claim based on the Treaty Respecting Utilization of Waters of the Colorado and Tia-juana Rivers and of the Rio Grande, Feb. 3, 1944, United States-Mexico, 59 Stat. 1219, T.S. No. 944 (“1944 Treaty”), but that it did have jurisdiction insofar as a taking *480was alleged.1 By its order of April 10, 1986, the court denied defendant’s second motion for summary judgment, finding that genuine issues of material fact existed. Trial was held September 14-25, 1987.
II. FACTUAL BACKGROUND
A. The River Delta Before 1972
The Colorado River originates primarily in Colorado, Utah, and Wyoming. It flows in a southerly direction through the United States, crosses the Northern International Boundary (“NIB”) just below Yuma, Arizona, proceeds southwesterly along the Mexican-American border for approximately twenty-two miles, crosses the Southern International Boundary (“SIB”), and finally flows through the Mexican State of Baja California to the Gulf of California. The entire river basin normally drains slightly more than fifteen million acre-feet of water per year. The properties in question lie in the lower Colorado River delta, an area that can be described generally as a large triangle at the base of the Colorado River. The delta is bordered on the north by the boundary between the United States and Mexico, on the west by the Cocopa Mountains, and on the east by the Sonoran Mesa. The triangle comes to a point at the north end of the Gulf of California. The delta region is characterized by a very gradual slope from Yuma, Arizona to the Gulf of California. The slope of the land decreases as the river approaches the Gulf.
The Hardy River bisects the Colorado River delta on roughly a north-south axis at the delta’s western edge. The Hardy is historically a river with considerably less water volume than the Colorado. The Hardy River originates within the delta near Volcano Lake and moves slowly southward until it joins the Colorado. Irrigation runoff constitutes a large proportion of the water that flows in the Hardy, and the river’s contact with an old volcanic bed gives it a brackish quality. The present junction of the Hardy and Colorado Rivers lies approximately fifty miles north of the Gulf of California, although a combination of natural and human-induced events have affected that location. Below the junction the combined river is known in Mexico as the Rio Hardy, but for purposes of distinguishing the rivers in this opinion, that stretch of river will be referred to as “the Colorado/Hardy.”
Figure 1 below is the court’s characterization of the Colorado River delta, based primarily on a map prepared in 1928 by Mr. Frank Adams, a consulting engineer with the American section of the International Water Commission. The Commission was created pursuant to congressional authorization in 1924 and 1927 to study the equitable use of the lower Rio Grande, lower Colorado, and Tiajuana River waters. The Adams map accompanied a 1930 report to Congress prepared by the Commission’s American section.2 The court has labeled several reference points, including the international boundaries, the Colorado River
*481Figure 1—The Colorado Delta, 1500-1900.
[[Image here]]
channel as it existed between 1500 and 1900, the Sonoran Mesa, the Cocopa Mountains, the Laguna Salada, the Hardy River, Montague Island, and the Gulf of California. The boundaries of the Laguna Salada Basin are based on a 1916 map of the delta prepared by E.C. La Rue, an hydraulic engineer with the United States Geological Survey.
The parties are in substantial agreement concerning the history of the Colorado River delta up until about 1930. Plaintiffs and defendant both rely heavily upon the 1937 Godfrey Sykes publication, The Colorado Delta, for this information. Sykes made numerous personal trips through the delta and his work is the major source of historical data on the river and development in the area.
In prehistoric times, the mouth of the Colorado was located in the vicinity of what is now Yuma, Arizona. What is now the delta was once the upper reach of the Gulf of California. Shifting of the earth’s crust resulted in a slight uplift in the center of what is now the delta, and caused the Colorado to flow in a westerly direction. For a period of time the river flowed in the direction of the Saltón Sea, a natural land depression in California. Eventually the river became redirected to the south, and for thousands of years it flowed over the entire region, spreading alluvial deposits and creating the present delta. The area *482of deposition includes the sites of plaintiffs’ properties.
Eventually the river developed a relatively stable path from what is now Yuma to the Gulf of California along the extreme eastern edge of the delta. This route is referred to as the “old navigable channel.” Records indicate that it existed in this approximate location from at least the mid-1500’s to 1900. It is probable that the old navigable channel was not entirely stable, as a comparison of maps drawn in 1858 and 1891 illustrates that the river meandered over an area several miles in width. During this time the channel was sufficiently deep to be navigable from the Gulf to Yuma.
Water has been taken out of the Colorado River for irrigation purposes since before 1900. At the turn of the century, water was being pumped out of the Colorado in the Yuma area through the Alamo Canal to the Imperial Valley in California. In 1901 a small cut was made in the west bank of the river to divert water directly into the Alamo Canal. This cut was made through the natural dike that builds up around any alluvial river, such as the Colorado. Natural diking may develop so that the river is actually “perched” on, and impounded by, its own silt deposit to such an extent that the river level may be higher than adjacent areas. The cut to the Alamo was made a few miles south of the NIB. Heavy water flows in 1905 eroded the diversion cut so that the river’s entire flow shifted into the Alamo Canal (see Figure 2 below). The water followed the natural gradient of the land and entered a channel known as the New River. The Colorado flowed through this channel in a northwesterly direction across California’s Imperial Valley until it emptied into the Saltón Sea. The river had no outlet to the ocean at this time.
Figure 2—The Colorado Delta, 1905-1907.
[[Image here]]
*483As a consequence of the river’s 1905 change in path, only minor flows, if any, existed in the old navigable channel for the next two years. In 1907 the diversion to the Alamo Canal was blocked and water again flowed in the old navigable channel. The river was unable to completely reestablish itself in that channel, however, due to the growth of vegetation in the old river bed.
The next significant change occurred in 1909, when the Colorado made a natural breakout to the west at a point close to the SIB, flowing in the Abejas River, a preexisting arroyo, and into Volcano Lake in the west-central part of the delta (see Figure 3 below). Levees were built on the northwest side of the river to prevent water from again flowing into the Imperial Valley. Most of the water, after overflowing the lake, eventually entered the Hardy River through numerous small channels, and was carried to the Gulf. During this time the river deposited silt in a broad area in and around the lake.
Figure 3—The Colorado Delta, 1909-1922.
[[Image here]]
To protect agricultural areas to the northwest from flooding and to prevent the possibility of another breakout to the Salton Sea, a four-mile-long diversion channel was excavated from the Abejas River to the Pescadero arroyo (see Figure 4 below). The Sykes book documents that the diversion cut, completed in 1922, began from a point approximately 7.5 miles west and one mile south of the point where the river crossed the SIB. As a result of this change, the point at which the Colorado waters joined the Hardy moved further to the southeast. Development of a distinct Pescadero channel was slower than expected because the arroyo was not large enough to contain the full volume of Colorado River waters. The portion of the drainage area in which there was no distinct channel was referred to by defendant’s witnesses as a subaerial fan, characterized by silt deposition due to generalized *484overflow. Between 1922 and 1929, the river deposited enormous quantities of silt, thus slightly raising the central delta area. To protect the developing Mexican agricultural interests northwest of the river from flooding, and to prevent another breakout to the north, dikes were constructed in stages to force the river south. An important northeast-to-southwest levee, completed in 1927 on the western bank of the river, was the fifteen-mile long Rodriguez levee.
Figure 4—The Colorado Delta, 1922-1929.
[[Image here]]
In 1927, a large canal, known as the Vacanora, was excavated between the Pes-cadero channel and the old navigable channel for irrigation purposes. Siphons were used to transport water over the natural dike from the Pescadero to the Vacanora. In 1929, when the siphoning proved inadequate, a break was made in the bank of the Pescadero so that water could flow directly into the Vacanora. Heavy water flows eroded this cut and the Colorado River again shifted to the east and soon adopted the Vacanora as its main channel (see Figure 5 below).
*485Figure 5—The Colorado Delta, 1929.
[[Image here]]
The Colorado River’s appropriation of the Vacanora Canal completely altered the course of the river according to Sykes. The south end of the approximately eighteen-mile-long canal initially formed the farthest reach of a stable and well-defined channel for the river. In shifting to the Vacanora, the Colorado abandoned the Pes-cadero basin that had developed earlier, and in which it seemed that a distinct outlet would eventually form to the Hardy. Pressure was now directed toward the meander zone of the old navigable channel, along the eastern side of the delta. High water and flooding consequently took place more towards the southeast portion of the delta.
By 1929, a diked railroad bed for the Mexicali & Gulf Railroad had been constructed across the delta in a generally northwest-to-southeast direction. At this time, the Vacanora channel lost its clarity of direction at approximately the point where it met the railroad bed. Defendant’s experts characterized the sheet overflow3 that occurred in that area as another su-baerial fan. Although sheet flows existed in the vicinity of the railroad line, according to Sykes, two distinct channels began developing below the railroad in 1930, one predominant to the other. He predicted (as of 1935) that either of these channels would provide the main outlet for the Colorado to the Gulf.
Aerial and satellite photographs introduced at trial detailed the progress of channel development subsequent to 1935. The aerial photographs depicted the lower delta for the years 1939, 1942, 1949, and 1962. On the 1939 photo, plaintiffs’ expert witness Wilbur Lockman 4 identified a relative*486ly defined channel extending from the Mexicali & Gulf Railroad almost directly south to the Hardy River, corresponding to one of the two channels described by Sykes. On the 1942 photo, however, it was impossible to identify a clear channel for more than a few miles below the railroad. Lock-man testified that the area southwest of where the channel lost definition was characterized by another subaerial fan, but that at this time the river was developing a new channel in a southwesterly direction. In Lockman’s opinion, the 1949 photograph showed that this channel had become stable and distinct, entering the Hardy via one of the previously abandoned channels of the Pescadero River. He identified the same channel on the 1962 aerial photograph.
Like Lockman, defendant’s expert Joseph Friedkin5 could not identify a defined channel on the 1942 photo continuous from the railroad southward to the Gulf. Unlike Lockman, Friedkin also testified that he could not identify a continuous main channel for the Colorado from the railroad to the Hardy River on the 1949 photograph. Instead, Friedkin testified that this and other photographs supported his conclusion that the development of the Colorado River since 1905 was characterized by successive subaerial fans. He described a process whereby with the passage of time, the clearly defined channel would extend further, establishing and then cutting through one subaerial fan after another, each successively closer to the ocean. His opinion was also based on personal experience navigating the river in 1949. Both Friedkin and Lockman testified, however, that by 1962 the Colorado maintained a single main channel to the Hardy. One factor permitting that development may have been that the Mexican Government had by that time channelized the river immediately above the Hardy juncture for a distance of approximately five miles. The development of the river channel between 1939 and 1972 is shown below in figures 6, 7, 8, and 9.
*487Figure 6—The Colorado Delta, 1939.
[[Image here]]
*488Figure 7—The Colorado Delta, 1942.
[[Image here]]
*489Figure 8—The Colorado Delta, 1949-1962.
[[Image here]]
*490Figure 9—The Colorado Delta, 1962-1972.
[[Image here]]
B. Development of Hoover and Glen Canyon Dams
Hoover Dam is located near Boulder City, Nevada approximately 225 miles above the NIB on the Colorado River. It was completed in 1935. Behind Hoover Dam, Lake Mead filled between 1935 and 1941. Lake Mead has a storage capacity of twenty-seven million acre-feet of water, which is approximately twice the average annual flow in the river basin. Flood control is one of the primary purposes of the dam. Other purposes include storage of water, river regulation, and delivery for consumptive uses.
Glen Canyon Dam is situated northeast of Hoover Dam, approximately 250 river miles upstream on the Colorado, near Page, Arizona. It was completed in 1963. Lake Powell, behind Glen Canyon, filled between 1963 and 1980. The lake has a storage capacity of twenty-five million acre-feet of water. Its primary purposes are to conserve water for the upper basin states, to regulate the delivery of water to the lower basin states, and to generate hydroelectricity.
C. Fluctuations in Water Flow
The parties presented evidence concerning the fluctuation of flow discharge rates at various points along the river for the years 1904 to the present. This information was derived from data provided by the International Boundary and Water Commission (“IBWC”). The court summarizes this data in Table 1 below, which lists the peak mean monthly flow rates in cubic feet per second (“cfs”) at three different locations.6 During the period 1904 to 1934, the *491peak monthly discharge at Yuma ranged from 13,000 to 130,000 cfs. The average peak monthly discharge for that thirty-one year period is 68,226 cfs.
During the time Lake Mead was being filled, flows into Mexico dropped to a range of 12,000 to 29,000 cfs as measured at Yuma. For the period from 1963, when Glen Canyon Dam closed, to 1980, when Lake Powell was filled, the range was between 2000 cfs and 12,000 cfs, with an average of 3833 cfs.7 As can be seen from these figures, one effect of impounding river water to fill the two lakes was to restrict dramatically the normal downstream flow. Consequently, the volume of water carried by the river into Mexico was greatly reduced during the times that the Colorado filled Lake Mead and Lake Powell.
Table 1—Peak mean monthly discharges in cfs for Colorado River, 1904-1985
Year YUMA NIB MORELOS
1904 44,000 — —
1905 78,000 — —
1906 84,000 — —
1907 97,000 — —
1908 42,000 — —
1909 104,000 — —
1910 57,000 — —
1911 63,000 — —
1912 108,000 — —
1913 47,000 — —
1914 111,000 — —
1915 47,000 — —
1916 59,000 — —
1917 93,000 — —
1918 62,000 — —
1919 35,000 — —
1920 130,000 — —
1921 111,000 — —
1922 98,000 — —
1923 85,000 — —
1924 54,000 — —
1925 42,000 — —
1926 60,000 — —
1927 58,000 — —
1928 64,000 — —
1929 79,000 — —
1930 50,000 — —
1931 21,000 — —
1932 62,000 — —
1933 57,000 — —
Table 1 — Peak mean monthly discharges in cfs for Colorado River, 1904-1985
Year YUMA NIB MÜRELOS
1934 13,000
1935 12,000
1936 7,000
1937 7,000
1938 7,000
1939 20,000
1940 10,000
1941 29,000
1942 28,000
1943 17,000
1944 17,000
1945 16,000
1946 12,000
1947 12,000
1948 15,000
1949 16,000
1950 12,000
1951 8,000 6,000 —
1952 17,000 17.500 —
1953 16,000 15.000 —
1954 7,000 7.000 6,000
1955 8,000 6.500 2,000
1956 4,000 4.500 1,500
1957 8,000 15.000 15,000
1958 10,000 13.500 10,000
1959 3,000 6.000 4,000
1960 2,000 4.500 4,000
1961 2,000 3.500 1,000
1962 2,000 3.000 2,000
1963 2,000 3.500 1,000
1964 2,000 3.000 500
1965 — 3.000 1,000
1966 — 3.500 500
1967 — 3.000 500
1968 — 3.000 500
1969 — 3.000 500
1970 — 3.000 500
1971 — 3.000 500
1972 — 3.000 1,000
1973 — 2.500 500
1974 — 2.500 500
1975 — 3.000 500
1976 — 3.500 1,500
1977 — 3.500 1,500
1978 — 3.500 1,000
1979 — 7.000 2,500
1980 — 12.000 7,500
1981 — 4.000 3,500
1982 — 3.500 2,000
1983 — 32,500 28,000
1984 — 25.000 22,000
1985 — 20.000 17,500
*492D. The Legal Regime
Because of the Colorado River’s tremendous importance as a lifeline in an otherwise arid region, a legal regime has developed to allocate the volume of the river among various competing uses. This regulation began in 1922 with the Colorado River Basin Compact (“1922 Compact”),8 entered into by several western and southwestern American states. It allocated the Colorado River’s anticipated annual discharge of fifteen million acre-feet within the United States among the upper and lower basin states. The 1922 Compact provided that the upper and lower basins were each entitled to 7.5 million acre-feet per year, with specific entitlements made for each state. In addition, Article 10 of the 1944 treaty obligated the United States to deliver 1.5 million acre-feet of water to Mexico each year. Consequently, 16.5 million acre-feet per year is allocated from a river that normally receives inflows of slightly more than fifteen million acre-feet. Not all water allocated is withdrawn, however. Under normal conditions, the consumptive uses of these waters in the United States and Mexico require between twelve and thirteen million acre-feet. Thus, overallocation of river waters is currently problematical only when the Colorado experiences flows below thirteen million acre-feet. Alden Briggs9 testified for defendant that overallocation is likely to become a greater problem upon completion of the Central Arizona Project, which is expected to result in utilization of an additional 1.5 million acre-feet of water.
The Bureau of Reclamation, part of the U.S. Department of the Interior, determines the timing of releases from Hoover Dam after receiving schedules of expected water requirements for the calendar year from all major users, including Mexico. These schedules list monthly requirements. During the year, major water users list expected daily requirements in weekly schedules sent to the Bureau. In addition to the regularly scheduled releases from Hoover Dam for consumptive uses, there are two types of regulatory releases. These are “flood control releases” and “additional releases,” previously known as anticipatory releases.
According to Briggs, flood control releases are made by the Bureau of Reclamation in accordance with regulations established by the Corps of Engineers under the authority of the Flood Control Act of 1944, ch. 665, § 7, 58 Stat. 887, 890-91 (codified as amended at 33 U.S.C. § 709 (1982)). The regulations require that on August 1 of a given year, there must be 1.5 million acre-feet of storage space available in Lake Mead. This amount of space must be increased during the succeeding months, so that on the following January 1 there is a total of 5.35 million acre-feet of unused storage capacity behind Hoover Dam to accommodate possible spring flooding. To meet these requirements, flood control releases are made by incremental steps between January 1 and July 1, and in equal monthly allotments between August 1 and January 1. See 33 C.F.R. § 208.11 (1987). The Bureau of Reclamation considers projected inflows and consumptive uses when making these releases.
Anticipatory, or additional, releases may also be made from Hoover Dam for purposes of flood control supplemental to the storage capacity-building releases. Briggs testified that they are authorized by the Colorado River Basin Project Act of 1968, Pub.L. No. 90-537, § 602, 82 Stat. 885, 900 (codified at 43 U.S.C. § 1552 (1982)). Anticipatory releases were made from Hoover Dam in 1979, 1980, 1981, and 1983, due to anticipated above-average inflows. These releases resulted in peak mean monthly flows that reached as high as 32,500 cfs at *493the NIB gauging station. For comparative purposes, during the period that Lake Powell was filled, 1963 to 1980, peak mean monthly flows at the NIB averaged under 4000 cfs.
The decision to make the anticipatory releases was made by the Secretary of the Interior after consultation with the governors of the seven basin states. After the Bureau of Reclamation received authorization to make these releases, the decision was relayed to the United States Commissioner to the IBWC. The U.S. Commissioner informed the Mexican Commissioner of the decision to make anticipatory releases, who in turn notified the Mexican Government. The Mexican Government was offered the opportunity to give its advice on the timing of the anticipatory releases, but the final decision for when to release water was made by the Bureau.
E. The Subject Properties
Plaintiffs Bailey and Mosqueda hold interests in real property in Baja California near the confluence of the Colorado and Hardy Rivers. Because the locations of the river channels have changed from time to time, and because flooding that began in 1979 has further altered the confluence area, the court will describe the locations of properties at issue in relation to the river delta as it existed in 1978, shortly before the flooding.
To aid in this description, the court refers to Figure 10 below. For illustrative purposes, the court has labeled the locations of the subject properties (indicated by M for Mosqueda and B for Bailey), the Colorado and Hardy Rivers, the Laguna Salada, and protective levees.10
Figure 10—Location of plaintiffs properties.
[[Image here]]
*494Plaintiff Bailey, an American citizen, owns a leasehold interest in a single lot in a small development called Campo Rio Hardy on the west bank of the Colorado/Hardy River. The lot is located on a stretch of the river below the 1978 confluence of the Colorado and Hardy Rivers, and above the earlier confluence between the Pescadero River and the Hardy. This property lies several miles south of an extension of the Rodriguez levee, and it is thus on the unprotected side of the levee.
Bailey first leased his lot in 1976, using it as a vacation spot for fishing, swimming, and boating. The cabin on that lot was built virtually entirely by Bailey, and by 1978 he had added a headwall and a boat ramp. In addition, Bailey kept several pieces of personal property on his lot, including a boat and household furniture.
Plaintiff Mosqueda, a Mexican citizen,11 owns several lots above the 1978 confluence of the Colorado and Hardy Rivers. Many of these lots are located adjacent to each other along the western bank of the Hardy River in an area known as Colonia Terenos Indios. Mosqueda also owns property across the Hardy on the eastern side of the Colorado River in an area known as Colonia Vicario. The Terenos Indios property, where his home and business are located, lies just north of the point at which the Hardy River crosses the Rodriguez levee,12 and is thus protected by the levee from the Colorado River. It is approximately five miles from the main channel of the Colorado. The Vicario property lies on the protected side of the Ockerson levee extension on the eastern side of the Colorado River.
In the early part of the century, much of the territory in the area around what is now Mosqueda’s land was populated by Mexican Indians. Testimony by Pascuela Dominguez, who has lived in the area since 1915, established that she and others utilized the area where Mosqueda now lives for cattle raising and gathering mesquite for firewood.13 This testimony was not challenged by defendant and the court accepts it as true. Although she stated that the land could be cultivated, she was unclear when farming began. A map developed in 1916 by E.C. La Rue for the U.S. Geological Survey shows that part of the lands comprising what is now Mosqueda’s property was considered irrigable, as was the area of what is now Bailey’s property. By 1928, a map developed by Frank Adams for the International Water Commission showed Mosqueda’s and Bailey’s land to be “undeveloped and unprotected delta lands of which at least 70% reported irrigable with river controlled.”
Mosqueda testified that prior to 1937, the land in the vicinity of his property was owned by the Colorado River Land Company, which was under a mandate from the Mexican Government to develop the area. When the company did not proceed, the Mexican Government appropriated the land, subdivided it, and sold it to Mexican citizens such as Mosqueda.
Mosqueda came to the area in 1942. In 1943, he bought his first parcel, lot 13 in Terenos Indios, and began farming there in 1944. Mosqueda’s farmland was initially irrigated through a canal system with water from the Colorado River.14 Mosqueda constructed a pond when he obtained his first parcel in order to store irrigation wa*495ter. To obtain drinking water, he dug wells in 1945. The water was fresh and safe for drinking. These wells were last used in 1968.15
In 1950, Mosqueda purchased the unnumbered lot where his cantina and park, described below, are located, although these enterprises were not developed until later. He purchased Terenos Indios lot 10 in 1955, lots 8 and 9 in 1956, lot 6 in the 1960’s, and lot 12 in 1972. These parcels, like lot 13, were used primarily for agricultural purposes. The cantina was opened in May 1971. Although the Terenos Indios properties are contiguous, Mosqueda does not assert a claim for compensation with respect to lots 6 and 12.16 Mosqueda bought his property in Vicario in 1950. He irrigated the land via a canal that was fed by water pumped directly from the Colorado River.
Mosqueda utilized his Terenos Indios property for farming, operating a cantina, and conducting a tourist business. The Vicario property was used strictly for agriculture. Mosqueda grew cotton, wheat, barley, milo, and rye grass. His testimony that the farm property was productive before the flooding was not challenged. The farm property for which Mosqueda seeks compensation totals 143.95 hectares.17
At his cantina, Mosqueda sold beer and bait to tourists visiting the area. He also rented “shades” where tourists could park their cars while they picnicked, boated, or fished. The shades were wooden frames covered by dried tules, a type of freshwater reed. Eight shades were located along the Hardy River and two shades in a park area that Mosqueda developed by planting hundreds of palm and eucalyptus trees between 1963 and 1973. Mosqueda also rented small residences along the Hardy River on an annual basis.
F. The Flooding
Experts for both sides testified that sometime between 1962 and 1972 an unusual sediment deposition began to form below the point where the Colorado and Hardy Rivers converge. There is no question that as this deposit grew it blocked the Colorado/Hardy channel and caused wides-cale floods in the lower delta in 1979-80 and in 1983-85. The blockage continues to cause a substantial backup of water affecting plaintiffs’ properties.
The experts agreed that the 1972 photo illustrates the first indications of ponding18 just south of Bailey’s lot. This was at a time when flows on the Colorado River in Mexico were very low. As shown in Table 1, supra, the eleven-year period beginning with the closing of Glen Canyon Dam in 1963 saw flows into Mexico no higher than 3500 cfs at the NIB. As ponding increased, a canal was dug in 1972 or 1973 south of Bailey’s property to direct water from the Colorado/Hardy into Laguna Sa-lada, a low-lying area to the west that is separated from the delta by the Cocopa Mountains.
In 1979, anticipatory releases were made from Hoover Dam. Water flow at the NIB increased somewhat, to 7000 cfs. Bailey’s property was completely inundated at this time. By November 1979, Mosqueda’s land would also have been flooded but for a sandbag wall he constructed along the Hardy River. Colorado River water backed up to the Rodriguez levee adjacent to Mosque-da’s property to such an extent that the water on the downstream (unprotected) side of the levee was at a higher elevation than on the upstream (protected) side. The pipes through which the river normally flowed were closed to avoid backup of the Hardy. Previously installed pumps were *496activated for the first time to force water from the protected side downstream over the Rodriguez levee.
Despite the diversion to Laguna Salada, flooding increased, as evidenced by 1979 and 1980 satellite photographs. The 1979 photographs indicate that no flows were entering the Gulf. Some water reached the Laguna Salada via the canal. The canal was improved in 1980, a year in which anticipatory releases due to wet weather resulted in flows of 12,000 cfs at the NIB. Although a satellite photograph taken that year shows more water in Laguna Salada, it also indicates increased flooding. By that time well over half of the unprotected, i.e., undiked, lower portions of the delta were flooded.
A 1981 satellite photo shows somewhat diminished standing water, corresponding to low flows in the Colorado that year. In 1983, a combination of greater water inflow in the Colorado basin and the completion of filling Lake Powell in 1980 led to a series of years of dramatically higher water, which has only partially abated. In 1983, 1984, and 1985 the NIB gauge showed flows of 32,500, 25,000, and 20,000 cfs respectively. The only figures presented with respect to 1986 show mean flows at the SIB, which is below the Morelos diversion, ranging from 7550 to 14,700 cfs. Flows decreased in 1987. Mean flows at the NIB were 17,300 cfs in January; by May flows were approximately 3300 cfs, with a corresponding decrease in flooding.
Two photographs taken in 1983 show extensive flooding. Water rose high enough to breach the tidal bar at three points. The last satellite photo, taken in 1985, shows that water had receded in the lowest reaches of the delta, but that the area around Bailey’s land was still completely flooded and that water was still backed up to Mos-queda’s property. Bailey testified that his land remained flooded until the summer of 1987.
Bailey’s house, headwall, and boat ramp suffered permanent structural damage. In addition, furniture was destroyed by the flooding and his boat was submerged by water, although Bailey acknowledged that the boat’s submergence may have been the result of vandalism.
Whereas Bailey’s property was flooded by direct overflow, Mosqueda’s property was primarily affected by a rising water table. The level of water in the Colorado River below the levee protecting Mosque-da’s property rose several feet higher than the level of the Hardy River on the protected side, preventing the Hardy from draining by gravity. This condition still existed at the time of trial. At some places on his property, the water table rose above the surface of the land. Elsewhere, the water table has remained less than one meter below the surface of the land, according to Lockman’s measurements in November 1986 and water table maps produced by the Mexican Ministry of Agricultural and Hydraulic Resources between 1982 and 1985.
III. DISCUSSION
The fifth amendment to the U.S. Constitution provides in part that private property shall not be taken for public use, without “just compensation.” It is well settled that flooding may result in a taking. Pumpelly v. Green Bay, 80 U.S. (13 Wall.) 166, 181, 20 L.Ed. 557 (1871); National By-Prods. v. United States, 186 Ct.Cl. 546, 575, 405 F.2d 1256, 1272-73 (1969); Berenholz v. United States, 1 Cl.Ct. 620, 626 (1982), aff'd, 723 F.2d 68 (Fed.Cir.1983). The applicable principles are well-settled. A taking may result from permanent flooding, Berenholz, 1 Cl.Ct. at 626, from intermittent, but inevitably recurring inundation, United States v. Cress, 243 U.S. 316, 328, 37 S.Ct. 380, 385, 61 L.Ed. 746 (1917); Barnes v. United States, 210 Ct.Cl. 467, 475, 538 F.2d 865, 870 (1976), or from elevated groundwater, United States v. Kansas City Life Ins. Co., 339 U.S. 799, 810, 70 S.Ct. 885, 891, 94 L.Ed. 1277 (1950). Flooding that can be characterized as a random occurrence, not inevitably recurring, does not amount to a taking of property. Wilfong v. United States, 202 Ct.Cl. 616, 622, 480 F.2d 1326, 1329 (1973). In addition, the flooding must result in substantial damage, Cress, 243 U.S. at 328, 37 S.Ct. at 385; Barnes, 210 Ct.Cl. at 475, 538 F.2d at 870, *497and must be the natural and probable consequence of Government activity, Bartz v. United States, 224 Ct.Cl. 583, 593, 633 F.2d 571, 577 (1980); Barnes, 210 Ct.Cl. at 477-78, 538 F.2d at 871. Finally, no taking occurs if the activity that caused the flooding actually benefited plaintiffs more than it has injured them. United States v. Spo-nenbarger, 308 U.S. 256, 60 S.Ct. 225, 84 L.Ed. 230 (1939).
Two previous cases have discussed whether a taking resulted from flooding caused by sediment blockage. Cotton Land Co. v. United States, 109 Ct.Cl. 816, 75 F.Supp. 232 (1948), also involved the Colorado River. In that case, the court found that impoundment of waters behind Parker Dam in Lake Havasu (below Hoover Dam) resulted in depositions of sand far upriver from the dam when swift flowing waters encountered the still water of the lake. The sand caused the river to back up and spread laterally over its banks, flooding the plaintiffs property. The court rejected the Government’s contention that causation was too attenuated to find liability. It held that the dam caused the sediment blockage and hence the flooding. See also United States v. Kansas City Life Ins. Co., 339 U.S. 799, 810, 70 S.Ct. 885, 891, 94 L.Ed. 1277 (1950) (although plaintiff’s land was not inundated, construction of dam resulted in elevated groundwater and insufficient drainage, constituting a taking).
In Barnes v. United States, 210 Ct.Cl. 467, 538 F.2d 865 (1976), plaintiffs were owners of land, or crops on land, situated adjacent to the confluence of the Niobrara and Missouri Rivers. For hundreds of years the swift flowing Niobrara River deposited sediment upon encountering the slow moving Missouri. This resulted in the formation of sediment accumulations in the Missouri River. In 1952, Fort Randall Dam was closed. It is located on the Missouri thirty-six miles above the confluence of the two rivers. Prior to the closing of the dam, seasonal high flows on the Missouri would scour out the sediment accumulations. After the dam closed, however, seasonal high flows were eliminated and the sediment continued to build up. By 1964, the sediment deposit was of tremendous proportions. The,lands did not experience serious flooding, however, until 1969 when excess waters were released from the dam.
After 1969, flooding continued on an intermittent basis. Even when the land was not inundated, its use was impaired by seepage and an elevated groundwater table. Lands along the Missouri River had been utilized for agricultural and recreational purposes before the flood. And although they had been subject to some minor flooding prior to 1952, this did not interfere with agricultural activity. Under those circumstances the Court of Claims found that a taking had occurred.
Applying these decisions to the facts here, the court concludes that to make a prima facie case, plaintiffs must show (1) that flooding resulted from sediment accumulation near the mouth of the Colorado River, (2) that the accumulation was due to low flows in the Colorado, (3) that the low flows were directly attributable to Hoover or Glen Canyon Dams, and (4) that the operation of these dams was within the exclusive control of the United States.
There is no real dispute as to whether a sediment accumulation exists and whether it has caused flooding. Defendant merely disagrees as to the location and cause of the blockage. Friedkin, for example, when asked what caused the flooding of plaintiff’s properties, testified that “from the evidence as I have studied it and reviewed it, the flooding and the high water table in the plaintiffs’ properties is due to the backwater ... from the sediment depositions, which were deposited by the river in the area downstream from the Rio Hardy, and it was the backwater from those, combined with higher-than-usual flows from the, the river, although not extremely high, which caused the flooding of the lands in that area.” Defendant’s other principal expert, Daryl Simons,19 was *498asked for his opinion as to the cause of the backwater effect at Mosqueda’s property: “Carrying forward from time of deposition [of sediment] to now, my opinion is that river-conveyed sediment has dropped out, causing blockage, which is causing backwater, which is aggravating the flooding problems on the Mosqueda property.”
Plaintiffs’ theory, in brief, is that low flows prevented the river from pushing successfully against tidal action, with the result that the ocean carried beach sediment upward into the mouth of the river. Defendant, on the other hand, argues that the blockage consists of river sediment that was dropped, in part due to low flows, before the river contacted the ocean. Hence, defendant locates the blockage slightly upriver from the area identified by plaintiffs. Before examining these theories in more detail, the court observes that from plaintiffs’ standpoint, the defendant’s explanation of the nature and location of the sediment is equally useful to establish liability. As long as the blockage is sufficiently attributable to operation of the dams, whether it was caused by beach sand or river sand does not matter. Because the question of ultimate causation is disputed by defendant, however, both theories will be considered.
In support of its theory, plaintiffs offered the testimony of Wilbur Lockman and the appendix to a 1982 Corps of Engineers report. Lockman’s explanation starts with the existence of a tidal sandbar across the entire boundary at which the delta meets the ocean. A crescent of elevated sand stretches from the Sonoran Mesa on the northeast to the Baja California peninsula on the southwest. This tidal bar, as Lockman described it, is a natural and inevitable result of contact between the ocean and shallow delta land. Historically, there has been a breach in this bar just north of Montague Island, where the Colorado/Hardy enters the Gulf. There is normally a tidal effect as far as twenty miles upriver from that point. The constant competition between the tide and the river has, until recently, maintained a scoured exit for the Colorado. Lockman testified that average flows approximating 20,000 cfs for one month during a year at the NIB would have been sufficient to maintain an open channel to the Gulf. This opinion was based on correlating aerial photographs that depicted the breaching of the tidal bar in 1983 with readings at the NIB gauge during the same period in time.
Lockman’s views are substantially supported by an appendix to the “1982 Review of Flood Control Regulation—Final Report” jointly produced by the Corps of Engineers and the Bureau of Reclamation (“1982 Report”). The report describes the blockage as follows:
A tidal sandbar has developed in the main channel of the Colorado River near its mouth, where it empties into the Gulf of California. The sandbar is about 14 miles upstream from Montague Island, about 78 river miles south of the Southerly International Boundary (pi. D-1).
According to a 1979 river profile prepared by the Ministry of Agriculture and Hydraulic Resources of Mexico, the tidal sandbar is about 18 miles long, with about 5 miles of its top above the water surface. At its highest point, the sand bar is about 6 feet above the normal river channel bottom. Plate D-4 shows a profile of the tidal bar. The sandbar has been there at least since 1972, when its presence was noted on a satellite photo.
The tidal sandbar blocks the mouth of the river, restricting free flow of water into the Gulf. The sandbar has acted as a dam, raising stages in the Colorado River a few miles upstream. The river, unable to follow its normal course, seeks paths to the Gulf on either side of the sandbar. Left to its own means, the river could, if high flows continue, reestablish a channel through the sandbar or claim as its own course one or both of the side routes. Mexico reports some flow through the sandbar but no evidence that the flow is cutting a new channel through the bar.
*499In this litigation, defendant takes the position that the conclusion in its 1982 Report is correct in part, but incorrect in two respects. With respect to location, Fried-kin testified that satellite photos taken in 1972 and 1979 demonstrate that the blockage developed further upstream than assumed by Lockman and the Corps report. He also testified that this sediment was not deposited by the ocean but rather by the river as the result of natural delta formation. According to Friedkin,. the sediment accumulation was simply the seventh in a series of “subaerial fans” formed by the river since the 1910’s. From defendant’s viewpoint, it was merely unfortunate that the sediment accumulation took place in a spot that caused flooding on plaintiffs’ lands. Although the 1982 Report was reviewed by the IBWC, Friedkin’s agency, and although Friedkin was present at meetings at which the entire report was discussed, he did not recall discussing the particular portions that the plaintiffs rely on. No objection was made by the U.S. section of the IBWC to the statements quoted above.
With respect to location, a close review of the 1982 Report, the testimony of defendant’s experts, and the aerial photographs persuades the court that the conflict between the parties is more apparent than real, and, in any event, not fatal to plaintiffs’ case. First, the court notes that the Corps (and Lockman) described the accumulation as beginning fourteen miles above Montague Island and stretching upstream a distance of eighteen miles. Fried-kin testified that the deposit initially began twenty-five miles above Montague Island. This corresponds approximately to the middle of the accumulation point described by plaintiffs and the Corps. Friedkin testified that as of 1972, the accumulation extended ten to twelve miles upstream. He went on to state that sediment eventually accumulated nearly as far upstream as Mosque-da’s property. In describing the primary blockage on the 1972 aerial photograph, however, he circled an area that includes nearly all of the site described by plaintiff. In terms of location of the initial blockage, therefore, the parties are largely in agreement. Plaintiffs’ locus goes slightly farther downstream, and defendant’s slightly farther upstream, but there is substantial overlap. The more extensive upstream sedimentation later caused by the flooding itself is not inconsistent with Lockman’s theory as to why the flooding began.
The second point of disagreement concerns whether the blockage was the result of precipitation out of the river or sand carried in by the ocean. Once again, the court notes that the answer does not affect plaintiffs’ case. There is no real question that the blockage began and was allowed to develop because of low flows in the Colorado River. Simons testified that the amount of sediment a river can carry is in large measure a function of velocity. Velocity increases, in part, as a function of the volume of water in the river. Lower, and hence slower, flows decrease the amount of sediment that the river can carry. The result is that the sediment begins dropping out of the water earlier when flows decrease.
Defendant’s experts admitted that the principles of hydrology underlying Lock-man’s analysis are correct. Simons concurred that low flows could cause the river to be blocked by a tidal bar: “The tidal action is relatively constant, and with very low flows, you can block the channel. We have evidence of that.” He was asked whether a river normally maintains a break in a tidal bar upon entering the ocean. He responded, “Yes. Except during periods of low flow, there can be total blockage.”
The court finds that the most likely scenario, given the indisputable presence of the sediment accumulation, is that both parties are correct. While the upstream blockage is probably the result of river sediment precipitating out during periods of low volume and velocity, the initial reason for that deposition was probably the tidal blockage further downstream. As Friedkin and others indicated, the tidal effect reached as far as twenty miles from the mouth of the river. This is within the area where plaintiffs contended that deposition occurred. Both parties agree that the river can be blocked by the ocean dur*500ing periods of very low flow. When velocity in the Colorado River approached zero due to low flow and tidal resistance, river sediment would precipitate out. Ocean sand would also accumulate because the river would lack volume to scour the sand out to sea. Once any blockage developed, and so long as flows remained low for a period of time, this point of equilibrium between ocean and river would result in an ever-increasing deposition upriver as more and more sediment-bearing water was slowed or actually stopped. This in turn would yield more deposition and more backup.
The effect of water volume in creating the blockage is illustrated by the following excerpt from plaintiffs’ cross-examination of Friedkin:
Q: Okay, then during the period from ’62 to ’72, waters were reaching the junction of the Colorado and Hardy and were carrying sediment ... downstream in the Hardy/Colorado toward the Gulf, right?
A: Yes.
Q: And is it your finding then that it was those sediments that between 1962 and 1978 formed this plug that caused Mr. Mosqueda’s problem?
A: They initiated the formation of the plug.
Q: What was it that stopped those sediments from being transported all the way down, ... to the estuary or to the Gulf?
A: Well, because ... they picked up sediment there, the grades were flattening as it moved down, and they had picked up sediments in this clean, expedited, straight channel and when they hit the flatter slopes again down below, they began deposition, ... it was overloaded by the time it reached there again and deposited sediment.
Q: Do you think that if those velocities had been a little bit faster, maybe quite a bit faster, that would have helped them to move on downstream?
A: Sure, sure.
Q:. Is that yes?
A: Yes, uh-huh.
The correlation between volume and blockage was conceded in the appendix to the 1982 Report: “In the last 30 years, the high degree of reservoir control has permitted few large flows to flush the river channel. As a result, sediment has built up and vegetation has encroached, reducing the main channel capacity.”
When asked what level of flows would maintain a channel to the Gulf, Friedkin responded that flows ranging from 5000 to 15,000 cfs at the NIB for a “sufficient length of time” would be required. While this figure is less than the 20,000 cfs Lock-man estimated would be necessary, a review of historical flows in the river for the period after Lake Powell began filling in 1963 and until the first flood releases were made in 1979 shows that the peak mean monthly discharge levels at the Yuma and NIB gauges never exceeded 3500 cfs. Such an extended period of low flows in the Colorado was unprecedented. In contrast, the preceding twenty-one year period from 1942 to 1963 shows that the peak mean monthly flows measured at the Yuma and NIB gauges failed to reach 5000 cfs only five times. Assuming that Friedkin's estimate is correct that flows between 5000 and 15,000 cfs would maintain a channel, it is inescapable that flows in the Colorado River during the period Lake Powell was filling were, by defendant’s own testimony, inadequate for the river to scour a channel below its confluence with the Hardy in the years immediately preceding the flooding.
The court finds that low flows in the Colorado River caused the sediment accumulation. The court rejects defendant’s theory that the flooding is attributable to the natural formation of a subaerial fan in the vicinity of plaintiffs’ properties. The flooded condition of the delta near the Hardy-Colorado juncture, as depicted on the satellite photographs, does not conform to the descriptions provided by defendant’s *501experts of previous subaerial fans in several aspects.
First, each of the earlier fans that Fried-kin identified was characterized by sheet overflows around and over the river’s banks, resulting in fan-shaped deposition. In each case the river did not back up as it has in the present situation. It merely became diffuse for a time before reassembling at a lower point. Second, in each of the six earlier fans, the river inevitably cut through the fan and reestablished a main channel. Here, by contrast, the opposite has occurred. An examination of the 1972 and 1985 photos reveals that the deposition has in fact extended further upstream. Third, the deposition has not been the result of sheet overflow. Here, the accumulation developed because the river lacked the punch to penetrate to the ocean. Fried-kin himself testified that it was during a period of low flow in the 1960’s that the river did not develop a subaerial fan, but instead maintained a single channel, due to lack of volume. Fourth, although the gradient of the land gradually slopes downward from Mosqueda’s property to the Gulf, the satellite photos taken between 1972 and 1985 show that water backed up for several miles upstream. If, in fact, natural formation of a subaerial fan was occurring, it would be expected that the Colorado/Hardy would shift to a more favorable gradient but still find its way to the ocean. A comparison of the 1972 and 1985 photos fails to show development of an alternate channel following the high flows. The court finds that the present flooding cannot be explained as a natural process of the river delta.
Plaintiffs must next prove that the low flows were in fact attributable to the Hoover and Glen Canyon Dams. In support of this contention, plaintiffs presented the records of flow discharge rates on the Colorado River dating from 1904. This evidence is summarized in Table 1, supra. It is not disputed that the functioning of the Hoover and Glen Canyon Dams greatly reduced the flows on the lower Colorado River between 1935 and 1980. River regulation by Hoover Dam beginning in 1935 decreased the Colorado’s scouring ability by dramatically reducing the level of seasonal peak flows. Flows were especially low after the closing of the Glen Canyon Dam in 1963, as documented by the IBWC through the use of river gauges at the Northern and Southern International Boundaries. The consecutive sixteen-year period of flows below 4000 cfs is unprecedented for this century.
Defendant does not contest the fact that the flows on the lower Colorado were sharply reduced after the closing of the dams. However, it attributes the low flows in the vicinity of the plaintiffs’ property primarily to upstream diversions, particularly at Morelos Dam, where water is diverted to irrigate the Mexicali Valley. Plaintiffs in turn do not contest the fact that water is diverted at Morelos Dam, but argue that a much greater volume of water was retained behind the dams than was diverted for Mexican agriculture.
The relevant inquiry is whether flows sufficient to maintain a channel to the Gulf would exist given the upstream diversions but in the absence of the dams. Briggs from the Bureau of Reclamation testified that the Colorado River normally has a flow of fifteen million acre-feet per year. Present use of the Colorado River is between twelve and thirteen million acre-feet, including the 1.5 million acre-feet obligated to Mexico under the 1944 Treaty. Assuming average flows and average consumption, then normally between two and three million acre-feet per year would flow past the Morelos gauging station, which is 0.7 miles below the Morelos diversion. Actual records maintained by the IBWC show that the average yearly flow at the Morelos gauge from 1954 through 1962 was 1,333,-054 acre-feet, ranging from a low of 176,-626 acre-feet in 1961 to a high of 3,957,780 acre-feet in 1958. However, the IBWC records show that for the years 1963 through 1978, the highest yearly flow was 580,656 acre-feet occurring in 1978. In other years during that period, flows were much less, exceeding 200,000 acre-feet only in 1973, 1974, 1975, and 1976.
It is undisputed that flows at the More-los gauging station, as evidenced by the *502IBWC records, were consistently very low during the period immediately before flooding first appeared on the aerial photographs, namely between 1962 and 1972. There are three possible explanations for decreased flows. One possibility is that the Colorado River received diminished inflows during this period. Defendant did not make this contention, and certainly presented no evidence to that effect. The only evidence on this issue showed that the peak mean monthly inflows above Lake Powell for the years 1969 through 1980 were generally between 44,000 and 74,000 cfs. Exceptional years were 1976 and 1977, when the peak mean monthly inflows were 32,000 and 6000 cfs, respectively. Although not directly translatable into total yearly inflows, these figures indicate that inflows were normal. Briggs stated that seventy percent of the Colorado River’s inflows occurred above Lake Powell. The court finds that the blockage was not the result of diminished inflows.
A second possibility is that low flows at the Morelos station were the result of excess diversions in Mexico or elsewhere. The IBWC records do not support this possibility, however. The amount of water annually diverted at Morelos between 1963 and 1978 remained fairly constant between 1.2 million and 1.4 million acre-feet,20 as measured by the gauge just inside the intake canal. The average amount diverted was even more during the period between 1950 and 1962 when it is undisputed that no blockage or flooding occurred. Furthermore, Briggs testified that the major diversions in the United States were already operating in 1963. This leaves one explanation for the reduced flows after 1963, namely the filling of Lake Powell behind Glen Canyon. The court finds that water in excess of normal consumptive uses in both countries was retained behind Glen Canyon Dam, rather than diverted at More-los. In addition, regulation by Hoover Dam reduced peak flows that otherwise would have helped to maintain a scoured channel on a seasonal basis. The court thus finds that the low flows causing the blockage were attributable to the filling of Lake Powell and river regulation by Hoover Dam.
Plaintiffs have also established that the operation of the dams was within the exclusive control of the United States. When Lake Powell filled, Glen Canyon Dam retained all water in excess of the 7.5 million acre-feet allocated to the lower basin states and the 0.75 million acre-feet share of water allocated to Mexico from the upper basin states. With respect to the excess releases that actually caused the flooding, Briggs testified that decisions of when and how to release waters for flood control purposes rested with the Corps of Engineers and the Bureau of Reclamation. Downstream users and the IBWC were notified of plans to make anticipatory releases and were invited to comment, but the final decision to release waters was with the U.S. Government. Furthermore, Briggs testified that flood control releases to ensure reserve capacity behind the dams is nondiscretionary, as the Bureau of Reclamation must carry out the regulations set by the Corps of Engineers. Consequently, the court concludes that the construction of the dams, the filling of the lakes, and the flood control releases beginning in 1979 were under the exclusive control of the United States. Plaintiffs have thus proved that the flooding of their properties was the natural and probable consequence of actions of the United States.
A. Inevitable Recurrence
At trial, plaintiffs made the requisite showing that the flooding was permanent or inevitably recurring. Cress, 243 U.S. at 328, 37 S.Ct. at 385; Barnes, 210 Ct.Cl. at 475, 538 F.2d at 870. Plaintiffs’ properties have been subject to surface or subterranean flooding conditions for more than seven years. Mosqueda’s property has been subject to high water since November 1979. The parties stipulated that Bailey’s property has been completely flooded on two separate occasions for extended periods since 1979. In addition, the parties have stipulated that it is not feasi*503ble to remove the sandbar or dredge a channel through it. Although the river reestablished a channel across or below the lower portions of the sediment deposition, this has not eliminated the persistent backwater effect abutting Mosqueda’s property further upstream.
Expert and lay witness testimony for the Government confirmed the extensive and recurring nature of the flooding in the vicinity of plaintiffs’ properties. Simons testified that without channelization, flooding would remain a problem for property in the area of the Colorado-Hardy confluence. Friedkin testified that the entire area surrounding the juncture of the two rivers, including the properties of Bailey and Mos-queda, will be subject to flooding so long as the blockage remains. In addition, Briggs indicated the likelihood of future flood control releases. He testified that excess runoff years occur periodically and that flows are highly variable. The court finds that the flooding experienced between 1979 and 1986 will likely persist in Mosqueda’s case or, with respect to Bailey, reoccur.
In addition, evidence in the form of water table maps and salinity testing demonstrates the permanent nature of the damage to Mosqueda’s property by previous flooding. Maps prepared by the Mexican Ministry of Agricultural and Hydraulic Resources show that the water table at Mos-queda’s property remained within one meter of the surface in 1982, 1983, and 1985. This high water table has substantially increased the salinity of Mosqueda’s farmlands. Mosqueda testified that in most areas where water had receded from the surface and he had attempted to replant, crops would not grow. The damage to Mosque-da’s property was confirmed by the testimony of Government expert witnesses Bums Sabey and Robert Ohmart.21 Both testified that the high water table had forced subsurface salts to the surface. Even as water slightly receded, salts remained. Sabey testified that these salts could be leached out only by the infusion of tremendous amounts of fresh water, and even then the leaching would only be effective if there is sufficient drainage, which does not exist due to the elevated water table. The present high groundwater level will remain so long as the waters of the Colorado/Hardy River on the downstream side of the levee remain higher than the level of the Hardy on the upstream side. This differential has persisted since the first flooding in 1979 and has shown no sign of subsiding.
Moreover, the present water table on Mosqueda’s property is below the surface only because water is pumped from the protected (Hardy) side of the levee to the unprotected Colorado/Hardy side. The water table on the unprotected downstream side is several feet higher than on the protected side. In addition, as defendant’s witnesses testified, the poor permeability of the soil in the area adds to the difficulty in leaching the salts out of the top layer of soil.
Based on the above evidence, the court finds that Bailey’s property is subject to inevitably recurring inundation, and that Mosqueda’s land is subject to inevitably recurring subsurface flooding that has rendered portions of his land unsuitable for agriculture.
B. Benefits Analysis
Even though there exists a direct connection between defendant’s actions and a taking of plaintiffs’ property, no liability attaches if the dams involved conferred benefits on these particular plaintiffs greater than the damage they inflicted. See Sponenbarger, 308 U.S. at 266-67, 60 S.Ct. at 229 (holding that the far-reaching benefits from a flood control program outweighed the possibility of future floods). The burden of proof is on defendant to demonstrate that the federal project benefitted the property in excess of the injury it caused. Willis v. United States, 50 F.Supp. 99, 102 (S.D.W.Va.1943), *504modified and aff'd, 141 F.2d 314 (4th Cir.1944). The benefits analysis can be considered in the context of whether there was any taking at all, Sponenbarger, 308 U.S. at 266-67, 60 S.Ct. at 229, or whether the value taken was originally provided by the project. See United States v. Miller, 317 U.S. 369, 376, 63 S.Ct. 276, 281, 87 L.Ed. 336 (1943). When the Government denies that a taking occurs, “only benefits inuring specifically to the condemnee, rather than to the community at large, are relevant to an analysis of the government’s liability under the Fifth Amendment.” City of Van Buren, Ark. v. United States, 697 F.2d 1058, 1062 (Fed.Cir.1983).
Defendant alleges that without Hoover and Glen Canyon Dams, Mosqueda would have been unable to irrigate his land, and that neither his nor Bailey’s properties could have been developed without the flood protection that the dams afforded. With respect to these assertions, the court finds that the benefits doctrine applies differently as between Bailey and Mosqueda. Takings cases are particularly fact dependent. Yuba Goldfields, Inc. v. United States, 723 F.2d 884, 889 (Fed.Cir.1983). Whereas Mosqueda’s property lies on the protected sides of the Rodriguez or Ocker-son levees, Bailey’s property lies on the unprotected side, a mere ten feet from the edge of the Colorado/Hardy River. The history of land development in the lower Colorado delta demonstrates the importance of protection by the levee system.
The flow discharge records for the Colorado River show that in the period 1906-1935 (pre-Hoover Dam), the peak mean monthly flows at the Yuma gauging station were often in excess of 50,000 cfs. On occasion they exceeded 100,000 cfs. The record is clear, based on the testimony of experts for both parties and the Sykes book, that between 1905 and 1935 the Colorado River was unable to maintain a continuous channel due to a combination of natural events and human interventions. During the 1929-1935 period, the experts for both parties disagreed on the amount of sheet flow from the river, but it is undisputed that some occurred. Friedkin testified that since the development of the federal reservoirs, much greater control exists over water coming from the United States into Mexico. Because the volume of peak flows has been reduced, so has the number of times that the river would flood due to cyclical high flows. The degree of fluctuation in the Colorado River thus has been substantially reduced since the construction of Hoover Dam. The relevant inquiry is whether, absent the Hoover and Glen Canyon Dams, what use and benefit plaintiffs would have had of their properties.
The 1928 Adams map that serves as a basis for Figure 1, supra, shows that the delta north of the Mexicali & Gulf Railroad was substantially developed with irrigated farming prior to the construction of the Hoover Dam. Lands already being irrigated at that time were protected by a system of levees on the west bank of the Colorado River. Because Hoover Dam was not completed until 1935, the levee system in place at that time clearly provided sufficient flood protection for agricultural development. The land where Mosqueda’s and Bailey’s properties now lie was not at that time protected by levees, but was described by the 1928 map as “at least 70% irrigable with the river controlled.”
The Sykes book confirms the progressive development of more and more lands in the lower delta as the levee and canal system was continuously expanded and strengthened. As noted above, some levees were in place before the construction of Hoover Dam, notably the Rodriguez levee. The original fifteen-mile levee was built in 1927 off the west bank of the Colorado River to prevent the river from overflowing into the Abejas Channel. Over the years this levee was extended in a southwesterly direction until it reached the Mosqueda property in 1950-52. Friedkin testified that the levee extensions were made in an effort by the Mexican Government to force the river to maintain a predictable path and thereby protect agricultural developments. He further testified that the extensive investment in irrigation would prompt the Mexican Government, in the event of a break, to attempt to maintain the river’s alignment within the existing levee system.
*505Although Friedkin also stated that it would be easier to maintain the levees with the protection of the dams, he did not state that without the dams it would be impossible. Given the fact that the river was leveed before the dams were constructed, the court concludes that the benefit to Mos-queda from the dams was attenuated and minor. The real beneficiary, as Friedkin indicated, was the Mexican Government, which was able to expend less money and effort to maintain the levees. While there is evidence that extremely high flows broke the Pescadero levees in 1926, the court concludes that this menace would not be sufficient to prevent development of otherwise irrigable lands protected by levees. First, there is the indisputable fact that farming continued to expand within the protected areas despite the occasional floods. Second, no evidence exists to establish that flooding actually occurred in the vicinity of Mosqueda’s land. Although Friedkin guessed that water might have been present on the land as a result of the 1941 flood, there was no direct evidence of that. In fact, Mosqueda and Dominguez testified that there was no flooding as far back as 1925. This is consistent with the location of Mosqueda’s land, which, unlike the areas flooded in 1926, is not adjacent to the Colorado River, but is in fact five miles away. Finally, the type of flooding that took place earlier resulted from bank overflow or redirection of the river due to unusually high flows. That type of flooding abates when normal flows return. The flooding presently affecting the plaintiffs’ properties is due to a ponding effect and shows no signs of abating.
Similarly, the extensive irrigation of the Mexicali Valley in 1928 demonstrates that Hoover Dam was not necessary for such utilization of the Colorado River. This suggests that Mosqueda would also have been able to irrigate his farm without river regulation provided by Hoover Dam. The court thus finds that defendant has not proven that the dams provided special benefit to Mosqueda. Consequently, Mosqueda may recover for the taking of his property.
Bailey’s leasehold property is situated on the Colorado/Hardy River on the unprotected side of the Rodriguez levee. The court takes the current alignment of the levees as a given. Although at some point in the past it may have been possible to build the levee to cross the Hardy River at a location sufficiently downstream to protect Bailey’s property, in fact it was not done. The court finds this distinction between Mosqueda’s and Bailey’s properties to be critical. Without either Hoover Dam or the levees, the Bailey property would likely have been subject to flooding in periodic high flow years, particularly given the location of his property below the confluence of the two rivers. Sykes’ documentation demonstrates that the Colorado overflowed its banks and caused substantial flooding several times prior to construction of the Hoover Dam. In 1891, the river flooded severely above Yuma, largely destroying the town. In 1903, flows reaching 73,000 cfs at Yuma caused the river to overflow the banks of the old navigable channel, flooding the drainage area of the Abejas River. Floods also occurred after the river abandoned the old navigable channel. The Abejas channel flooded in 1916, a year when the peak mean monthly flow was 59,000 cfs. The Pescadero channel flooded to such an extent in 1926 when the peak mean monthly flow was 60,000 cfs that it destroyed a partially completed levee for a length of two miles. Similarly, the Vacanora Canal was unable to contain the river’s flow in 1929, 1930, and 1932, when the peak mean monthly flows registered 79,000, 50,000, and 62,000 cfs, respectively. Sykes reported that the 1932 flood included the flooding of the Colorado/Hardy channel, with water extending into Laguna Salada for a distance of eight or nine miles.
Beginning in 1904, the first year for which the court has data, peak mean monthly flows at the Yuma gauge exceeded 50,000 cfs twenty-two times before Hoover Dam was completed in 1935. Flows exceeded 100,000 cfs five times in that period. By comparison, during 1932, when Sykes commented on flooding in the Colorado/Hardy area, flows were 62,000 cfs.
*506The small group of cabins comprising Campo Rio Hardy, where Bailey’s property is located, did not begin developing until the late 1950’s. At that time Hoover Dam was closed and the level of peak flows was dramatically reduced. After 1935, the peak mean monthly flow never again exceeded 33,000 cfs.22 Bailey’s lot is located immediately adjacent to the water. His cabin begins ten feet from the bank of the river. At the time he first leased the lot in 1976, flows had been at historic low levels at the NIB for several years, most immediately due to the completion of Glen Canyon Dam. Peak mean monthly flows had been in the 2500-3500 cfs range during this period.
This review of historical data, contrasted with the location of Bailey’s property, persuades the court that without Hoover and Glen Canyon Dams, frequent and larges-cale overflow would have prevented development of Bailey’s lot. Because those uses and improvements for which he seeks compensation were made possible by the dams, the court concludes that the flooding of Bailey’s property did not result in a com-pensable taking.
C. Date of Taking
In Dickinson v. United States, 331 U.S. 745, 749, 67 S.Ct. 1382, 1385, 91 L.Ed. 1789 (1947), the Supreme Court held that a properly owner need not undertake piecemeal or premature litigation when the taking results from an ongoing event. A claim for a taking by flooding arises when the permanent nature of the flooding or its inevitable recurrence can fairly be perceived. This determination is in the nature of a jury verdict. Barnes, 210 Ct.Cl. at 480, 538 F.2d at 873. Consequently, in the present case the court must determine when the flooding condition became stabilized so as to make Mosqueda reasonably aware of its permanent nature.
Mosqueda’s flooding began in November 1979. The extent of the problem did not become apparent for several months, however. Mosqueda could not have been aware of the extent to which his land was affected until it became clear that the backwater did not result from possible seasonal flooding, and until it became obvious that subterranean flooding had ruined the productivity of the soil. In 1979 he planted barley and wheat on several lots that did not have standing water. Within five or six weeks after germination, the crops had died due to scalding from too much salt. The full extent of the flooding did not become apparent until the following year, however. In the late fall of 1980, Mosqueda attempted to plant reduced crops of barley and wheat in the areas where water had receded from the surface. Again the crops died. This second crop failure provided Mosqueda with clear evidence that the agricultural productivity of his soil was ruined. By that same time Mosqueda was confronted with the fact that the differential between the Hardy and the Colorado Rivers at his property was persistent. The court finds that the taking therefore occurred on January 1, 1981.
D. Damages
The Supreme Court has stated that “the destruction of privately owned land is ‘a taking’ to the extent of the destruction caused.” United States v. Kansas City Life Ins. Co., 339 U.S. 799, 809, 70 S.Ct. 885, 891, 94 L.Ed. 1277 (1950). Just compensation requires payment of the full monetary value of the interest taken. Almota Farmers Elevator & Warehouse Co. v. United States, 409 U.S. 470, 473, 93 S.Ct. 791, 794, 35 L.Ed.2d 1 (1973).
Mosqueda seeks compensation for several types of damage. The most significant is the loss of value to his farmland. In addition, he seeks compensation for farming losses, the taking of trees and permanent improvements, business losses, and the costs of mitigating the flooding. He makes the following claims:
Depreciation in Farmland: $142,258.59
Loss of value of trees: 87,847.00
Farming loss, 1979-80 47,498.00
1980- 81 22,758.00
1981- 82 13,473.00
River residences: 27,600.00
*507Parking income at river: $ 24,354.00
Parking income at park: 22,050.00
Sandbags and cleanup: 17,247.00
Shades: 9,653.00
Total: $414,738.59
Mosqueda’s testimony concerning the amount of these losses was not meaningfully challenged. The question for the court is whether the proof offered was adequate and whether the type of damages sought are recoverable in a takings case.
1. Diminution in Value of Farmland
Mosqueda claims that the value of 143.95 hectares of agricultural land was substantially diminished by the flooding. Defendant does not dispute Mosqueda’s testimony that this land was agriculturally productive prior to the flooding. In addition, defendant does not dispute the accuracy of the production record for Mosqueda’s lands for the years 1979-80, 1980-81, and 1981-82 showing farming losses suffered on lots in Colonia Terenos Indios and Colonia Vicario. No crops were planted on the flooded lands after 1981 because Mosqueda concluded from his recent experience that the lands were no longer suitable for their most productive use, farming.
When less than an entire interest is taken, just compensation is measured by the difference between fair market value of the property immediately before and after the taking. United States v. Miller, 317 U.S. 369, 376, 63 S.Ct. 276, 281, 87 L.Ed. 336 (1943); Georgia-Pac. Corp. v. United States, 226 Ct.Cl. 95, 107, 640 F.2d 328, 336 (1980); Cloverport Sand & Gravel Co. v. United States, 6 Cl.Ct. 178, 188 (1984). Although the determination of just compensation requires a nonrigid approach, see Georgia-Pac. Corp., 226 Ct.Cl. at 106-07, 640 F.2d at 336-37, the preferred method for determining fair market value is the use of comparable sales. Cloverport Sand & Gravel Corp., 6 Cl.Ct. at 180; see United States v. 179.26 Acres of Land, 644 F.2d 367, 371 (10th Cir.1981). This method requires evidence of sales of properties reasonably similar to the land in question. Cloverport Sand & Gravel Corp., 6 Cl.Ct. at 189. A past or present owner may also offer an opinion as to the value of his land, so long as it is based on “substantial data.” United States v. Sowards, 370 F.2d 87, 92 (10th Cir.1966); Cloverport Sand & Gravel Corp., 6 Cl.Ct. at 189-90.
With respect to his land before the taking, Mosqueda testified that he estimated a value between $1300 and $1400 per hectare. To support this, he offered information concerning two sales of nearby farmland that were not affected by the flooding. Mosqueda testified that these sales reflected the value of his land before the flooding despite the fact that they took place in 1987.23 One piece of farmland sold for $1298.70 per hectare; the other for $1329.80. He also based his assessment on general knowledge of the value of productive farmland, information from the farmers association to which he belongs, and conversations with land brokers. After averaging the two sales, Mosqueda based the pre-flood value at $1314.25 per hectare. Applied to the 143.95 hectares that were flooded, this yields a value of $189,186.29.
Mosqueda’s estimate of the post-flood value of his land was based in part on his view that the land had only marginal value because it would not sustain crops. He was able to find only one comparable sale of land affected by flooding. The sale occurred in 1986 at a price of $326 per hectare. The lot was nearby and had also been farmed. Based on a decline in value of $988.25 per hectare, Mosqueda claims $142,258.59 as the decrease in value of his land.
Mosqueda’s testimony was not challenged on cross-examination, nor did defendant present its own evidence of land values. The court finds that plaintiff Mos-queda is entitled to $142,258.59 as part of his just compensation award for subterranean flooding of his agricultural property. This property consists of lots 8, 9, 10, and *50813 in Colonia Terenos Indios, and lots 9A and 9C in Colonia Vicario.
2. Crop Losses
Mosqueda asserts that he suffered crop losses between 1979 and 1982. Damages may be claimed for mature crops destroyed at the time of the taking. Barnes, 210 Ct.Cl. at 482, 538 F.2d at 874; Henry v. United States, 8 Cl.Ct. 389, 393 (1985). Mosqueda testified that the wheat and barley he planted in 1979 grew only to a height of six inches before dying one-and-a-half months after germination. Nor is there any evidence that subsequent plantings matured. Consequently, the court finds that Mosqueda may not recover for his asserted crop losses.
3. Loss of Trees
It is undisputed that Mosqueda lost hundreds of mature trees. The court has not found any precedent squarely holding that trees that are not being grown as a crop can be a separate item of compensation. Presumably, damage to trees would normally be incorporated indirectly as part of the pre- and post-taking land valuation. Here, the trees in question were not standing on land for which Mosqueda seeks to show a reduction in fair market value. That appears to bring this case within Cooper v. United States, 827 F.2d 762 (Fed.Cir.1987), where the appellate court allowed recovery for trees, independent of the land, due to destruction by standing water. What is not clear from Cooper is whether the trees were being cultivated as a crop. The trial court’s reference to Barnes suggests that they were. See Cooper v. United States, 11 Cl.Ct. 471, 475, rev’d, 827 F.2d 762 (Fed.Cir.1987). Nevertheless, the appellate court’s decision does not appear to turn on that distinction: “It is undisputed that timber was destroyed. Numerous cases hold that the destruction of a property interest is a compensable taking within the meaning of the fifth amendment.” Cooper, 827 F.2d at 763 (citations omitted). Moreover, the Supreme Court has held that “for all the Government takes, it must pay.” United States v. Dickinson, 331 U.S. at 750, 67 S.Ct. at 1385; see First English Evangelical Lutheran Church v. County of Los Angeles, — U.S. -, 107 S.Ct. 2378, 96 L.Ed.2d 250 (1987).
That the trees in the case at bar had an ascertainable economic value is indisputable. The trees for which Mosqueda claims compensation were not native to the area. He purchased them, planted them, and maintained them. Mosqueda testified they were important to his business use of the property, and their value can be expressed in terms of purchase price and maintenance. The court concludes that Mosqueda is entitled to recover for the value of trees as an element of compensation.
With respect to 210 palm trees that died as a result of the flooding, Mosqueda asserts damages of $69,300. Mosqueda calculated the value of each tree by multiplying the average height of his trees (twenty-two feet) by $15.00 per foot, the replacement cost obtained from a local nursery, to obtain a figure of $330 per tree. Mosque-da also lost several hundred eucalyptus and other non-native trees. These were mature trees, with heights of thirty to forty-five feet. For these trees Mosqueda asserts a value of $18,547. This figure was calculated by adding the cost of the trees ($1067), installation costs ($200), and cost of maintenance ($17,280) based on wages actually paid over an eighteen-year period.
The Supreme Court has instructed that the essential question is, “What has the owner lost?” United States v. Fuller, 409 U.S. 488, 490-91, 93 S.Ct. 801, 803, 35 L.Ed.2d 16 (1973). Indemnity can be measured in different ways, depending on circumstances. United States v. Miller, 317 U.S. 369, 373-74, 63 S.Ct. 276, 279-80, 87 L.Ed. 336 (1943). Defendant did not provide any basis for questioning Mosqueda’s analysis or evidence. Having concluded that the trees can be a separate item of compensation, the court finds that Mosqueda’s cost-based proof as to value is sufficient. The court finds that the value of the trees lost was $87,847.00.
4. Damage to Shades and River Residences
With respect to the shades, Mosqueda failed to establish a taking. Al*509though it is not controverted that the shades suffered some damage, Mosqueda testified that only one of ten fell down as a result of the flood, and it was repaired. According to Mosqueda, the others are “a menace” and will have to be repaired “eventually.” By his own testimony, however, his shade rental business along the Hardy River has returned to the same level as before the flooding. The shades are apparently still serving their intended purpose. The court cannot award compensation for future repairs, and Mosqueda provided no evidence as to diminution in value.
Similar shortcomings in proof apply to the claim for damage to the river residences. Mosqueda claims $6000 for what it would cost to repair doors, a floor, walls, a foundation, and to repair or replace pipes, wiring, etc. None of these repairs have actually taken place. The court finds that Mosqueda’s method of proof on this aspect of the claim is too speculative and imprecise.
5. Loss of Business Income
Mosqueda claims damages for losses to his parking business at the river between 1980 and 1983, to his parking business at the park between 1980 and 1987, and from river residences that he would have leased between 1980 and 1987. Case law in the area of fifth amendment takings has developed in such a way that compensation cannot be awarded for loss of business income. Bothwell v. United States, 254 U.S. 231, 233, 41 S.Ct. 74, 75, 65 L.Ed. 238 (1920); Georgia-Pac. Corp., 226 Ct.Cl. at 147, 640 F.2d at 361; A.G. Davis Ice Co. v. United States, 362 F.2d 934, 936 (1st Cir.1966). Although compensation can be obtained for the fair market rental value of property to be used for a particular business when there is a temporary taking, see Kimball Laundry Co. v. United States, 338 U.S. 1, 7, 69 S.Ct. 1434, 1438-39, 93 L.Ed. 1765 (1949); Yuba Natural Resources, Inc. v. United States, 821 F.2d 638, 641 (Fed.Cir.1987), Mosqueda provided no evidence as to what he could have obtained to rent out the land as business property. Although the gross rental income from the residences is some indication, it does not reflect what a willing lessee would have paid to lease the property for purposes of renting the residences. Mosqueda’s figures were not discounted for overhead and related costs. The same defect is even more apparent with respect to loss of income from parking. The court concludes that Mosqueda cannot recover those amounts.
6. Mitigation Costs
Finally, Mosqueda claims costs incurred in an attempt to mitigate the effect of the flooding. These amounts are seen in the law as consequential and not as a natural and direct result of the flooding. See Manigault v. Springs, 199 U.S. 473, 484-85, 26 S.Ct. 127, 132, 50 L.Ed. 274 (1905). Mosqueda’s claim for sandbag and clean-up costs therefore is not compensable.
IV. CONCLUSION
The court concludes that plaintiff Mosqueda, but not Bailey, is entitled to recover compensation. With respect to Mosqueda, the court finds that defendant’s actions in building and operating Hoover and Glen Canyon Dams has resulted in the taking of a subterranean flowage easement in portions of Mosqueda’s land, and a direct taking of trees. The flowage easement exists on lots 8, 9, 10, and 13 in Colonia Terenos Indios, and lots 9A and 9C in Colonia Vicario. By paying the compensation ordered herein, the Government is entitled to maintain and operate Hoover and Glen Canyon Dams in such a way that groundwater due to backup from the Colorado River may persist regularly to within two feet of the surface of the lots listed above. The court finds the taking has decreased Mosqueda’s property in value by $142,-258.59. Mosqueda also has proven the destruction of trees worth $87,847.00. He is thus entitled to $230,105.59 in compensation, plus interest from the date of taking (January 1, 1981) to the date of payment by defendant. Foster v. United States, 3 Ct.Cl. 738 (1983), aff'd, 746 F.2d 1491 (Fed.Cir.1984), cert. denied 471 U.S. 1053, 105 S.Ct. 2112, 85 L.Ed.2d 478 (1985). Entry of *510final judgment will await determination of fees and expenses.
Mosqueda is entitled to recover reasonable attorneys’ fees and expenses under the Uniform Relocation Assistance and Real Property Acquisition Policies Act of 1970, 42 U.S.C. § 4654(c) (1982). King v. United States, 205 Ct.Cl. 512, 504 F.2d 1138 (1974); Houser v. United States, 12 Cl.Ct. 454, 475 (1987). Plaintiff has not furnished the court with materials to support a specific claim. Consequently, plaintiff Mosqueda is allowed to and including March 30, 1988 to make an appropriate submission. The response and reply periods thereafter will be controlled by RUSCC 81(e).
It is so ORDERED.
ORDER ON RECONSIDERATION
On March 4, 1988, the court issued an opinion making findings and conclusions with respect to the two remaining plaintiffs in this action. The court concluded that plaintiff Bailey was not entitled to recover for a taking, but that plaintiff Mosqueda was. Entry of judgment was delayed pending determination of attorneys’ fees. On March 22, 1988, defendant attempted to file an untimely motion for amendment and/or reconsideration of judgment. Without explanation, the motion was submitted four days out of time.
The motion was made under RUSCC 59(a)(1). Because no judgment has been entered, it should have been made pursuant to RUSCC 83.2(f). Unlike RUSCC 59, extensions to file under RUSCC 83.2(f) are not subject to the limitation of RUSCC 6(b). See Brown v. United States, 5 Cl.Ct. 1, 17, aff'd, 741 F.2d 1374 (Fed.Cir.1984). The court has therefore granted leave for the motion to be filed in order to explain more fully why it is being denied, and to express the court’s concern with the way the matters raised in the motion have been handled.
In its motion to reconsider, defendant argues: (1) that the court failed to resolve questions of law involving the application of Articles 17 and 20 of the Treaty Respecting Utilization of Waters of the Colorado and Tiajuana Rivers and of the Rio Grande, Feb. 3, 1944, United States-Mexico, 59 Stat. 1219, T.S. No. 994 (“the Treaty”), to plaintiffs’ claim, including jurisdictional, choice of law, and remedy issues; (2) that the court impermissibly intruded on the foreign relations of the United States and Mexico; and (3) that the bases for many of the court’s findings of fact are unclear.
In its order of July 2, 1987, the court, in denying a motion for enlargement of time to file a third motion for summary judgment, set out portions of procedural chronology that are relevant here:
This action was filed on January 25, 1984. Defendant has filed two previous motions for summary judgment. This court has issued two orders which in part granted those motions and permitted the case to proceed. Discovery was to close on December 31, 1986. The action has been scheduled for trial four times. On March 4, 1987, defendant’s counsel wrote the court that “Defendant will not file [an] additional dispositive motion in this case.”
At the most recent conference with the court on May 20, 1987, defendant indicated it would be unprepared to proceed with the pretrial scheduled for May 29, 1987, and that it planned to file a third motion for summary judgment. At that time the court told the parties that the case would proceed to trial but that if defendant filed a motion for summary judgment, it would be considered. By order of June 9, 1987, defendant was allowed until June 26, 1987 to file its motion. Defendant’s current motion for an extension comes on the last day for filing. It is filed 40 days after the conference at which defendant was told the court planned to proceed to trial. It comes four months after February 27, 1987, the deadline set by the previously assigned judge for filing dispositive motions during the status conference held January 8, 1987. It comes six months after the close of discovery. It comes seven months after the order of November 24, 1986, highlighting the areas potentially remaining for summary judg*511ment, and three and a half years after the lawsuit commenced. (Footnote omitted.)
In addition to the facts recited in that order, not mentioned were the following:
(1) In the order of January 25, 1985 (as amended May 31, 1985), Anderson v. United States, 7 Cl.Ct. 341, the court (Nettesheim, J.) denied in part without prejudice defendant’s motion to dismiss, or alternatively for summary judgment, which raised the precise legal grounds now advanced, and stated:
[B]ecause there is no evidence as to whether a forum exists or has been established by Mexico, the record at present is insufficient under Rule 56(c) to suspend these actions while plaintiffs pursue their remedies under the laws of Mexico.
Id. at 343 (footnote omitted).
Defendant may renew its motion on this [a different issue] or any other grounds not herein decided.
Defendant shall renew its motion for summary judgment by February 28, 1985.
Id. at 346 (emphasis added).
(2) The order of February 22, 1985 recites: If summary judgment proceedings are not renewed by July 1, 1985, the case will be set for trial.
(3) During the status conference of May 31, 1985, the court stated:
It also appeared to me, and I’m basing these comments solely on what was adduced in the Summary Judgment proceedings that Defendant had very substantial arguments which at that point had not been supported by affidavit dealing with whether the operation of the dam, meaning the discharges, was a product of an arrangement with a foreign country and in furtherance of foreign policy which would provide a separate basis for denying relief.
Now, both of these specific points were not developed by way of affidavit and if they are supported by way of affidavit by July 1, 1985 when Defendant must renew its Motion for Summary Judgment, it would seem to me that it might be very difficult to refute.
Transcript at 13.
(4) The following dialogue took place at the conference of September 13, 1985:
THE COURT: The date I notified you that the case would be set for trial if the summary judgment proceedings were not renewed by July 1, 1985 was on February 22 [1986].
If, as to the remaining plaintiffs, the government wishes to file a motion for summary judgment, it of course may do so, but it’s got to make sure that that motion is on file, responded to, and argued if it wants argument before trial.
Transcript at 10.
THE COURT: ... I am going to urge Miss Young, if there’s any basis for summary disposition, to get it in. I will give any motion you file prompt consideration.
MS. YOUNG: Very well, Your Honor.
THE COURT: And if you really think you can dispose of this case, do it; if you don’t, let’s try it.
Id. at 13.
MS. YOUNG: Well, Your Honor, there is a question floating around with regard to the treaty, whether or not the plaintiffs belong in a claims court and—
THE COURT: And that’s what we were trying to flesh out.
Id. at 13-14.
THE COURT: I think that you do have some arguments that might well at least flesh out some fundamental issues. If you don’t want to file it, fine. We can deal with these, you know, in the trial context.
MS. YOUNG: It’s not that I didn’t want to file it, your Honor. Given your statement which clarifies your problem with the administrative decision, I can certainly correct that and will speedily do so.
Id. at 16.
(5) At the status conference of November 22, 1985, the court stated:
*512I do know what the Government’s showings are because I have had your motion for summary judgment. It was my view, as I hinted broadly in my opinion, if you could have backed up some of those assertions with affidavits, this case was appropriate for summary disposition. It was my view that under the precedents of the Federal Circuit, that absent any evidentiary showing on those points, summary judgment could not be granted. But, if you could support the assertions you made with affidavits, they were very serious assertions. And, carried significant legal authority behind them. So, I think that you would achieve substantial economies to both your clients to exhaust summary judgment insofar as you are relying on what I refer to as overriding legal issues.
Transcript at 8-9.
(6) The Government’s second motion for summary judgment was filed January 21, 1986 only with respect to the issues of whether plaintiffs could establish the causation elements necessary to find a taking, whether United States dams benefitted plaintiffs, and whether plaintiffs’ injuries were consequential. The legal issues previously raised were not reasserted. The following occurred at oral argument on April 10, 1986:
THE COURT: The problem we have is this. I’ll be glad to hear argument, but for a variety of reasons, I think we have genuine issues of material fact that preclude summary judgment at this time.
Transcript at 9.
At any rate, I will respect the parties’ decisions as to whether or not we should proceed with summary judgment or should proceed with trial. I am perfectly willing to do both. I told Defendant we’d consider another summary judgment motion if it was filed on time, and it was. And I will—if Defendant wishes—and I’ll give her time to think about it—consider a supplementation to the motion.
Defendant, tell me this, Ms. Young. You did not move for summary judgment on the specific ground that I had mentioned in the last order that the actions of the United States in administering the discharge of water from the Lower Colorado River in New Mexico was essentially a sovereign function. There’s a lot of information in the affidavits about that.
So, I don’t want you to move for summary judgment on a ground that isn’t going to hold water, to make a bad pun. So, analyze very thoroughly if you want to proceed with that, but that’s where I think that issue is.
Id. at 13-15 (emphasis added).
(7) In the order of April 10, 1986 (Nettesh-eim, J.), the following appears:
It was suggested at 7 Cl.Ct. 345 that defendant may have a good argument that the sovereign’s conduct of foreign relations cannot be examined judicially. However, defendant has not moved for summary judgment on this ground. Perhaps it cannot, because the decision to close the Hoover and Glen Canyon Dams, which is the root of plaintiffs’ causation theory, may have been made by the United States alone.
(8) Defendant had not refiled a motion for summary judgment by the status conference of November 7, 1986:
THE COURT: [E]ven though Defendant had moved twice for summary judgment and been rejected twice, ... there are some very significant problems with this case that lead me to believe that it might be disposed of on a summary basis.
I am still concerned with the fact that unlike in most taking cases that we appear to be dealing with a situation here where the acts and the releases specifically were undertaken pursuant to an international agreement between two sovereigns.
I am not so sure that a court can examine into such acts, especially with respect to any involvement of the Government of Mexico. However, I have yet had elucidated enough that would give me any basis for summary judgment on that ground, because nobody has submitted the affidavit saying incontrovertedly that these were sover*513eign acts undertaken in concurrence with Mexican government officials, and what the implications allegedly of these acts are.
Transcript at 18-19 (emphasis added).
It is whether the decisions to make releases are made in the exercise of international relations by this government with another foreign government. And that point by affidavit really has not been brought out. I mean that was the ground on which I suggested in I believe the first opinion on the motion for summary judgment that it is something that we would look at later or that you could raise again.
I certainly did not deny it with any prejudice at all. Go back over that particular order, and I think that you will see that the grains were there, but it was not quite drawn together.
Id. at 23-24.
But I do want this issue of the international implications completely decided by pre-trial. I mean it is not something that should rear its head after trial. You either have got enough to make an argument or you do not. And if you do not, do not make it.
Id. at 27 (emphasis added).
(9) The order of January 8, 1987 (Nettesh-eim, J.) provides:
3. The pretrial conference shall be held at 1:00 p.m. on Friday, May 29, 1987. ... Any motion for summary judgment filed by defendant shall be fully briefed and argued on that date.
(10) Another deadline for filing a motion for summary judgment was set for February 27, 1987. On March 4, 1987, defendant’s counsel wrote the court:
Please be advised that the defendant will not file additional dispositive motion in this case. Even though the deadline set by you fell last Friday, I felt obligated to confirm my decision with this letter.
Based on that representation, at the status conference of March 18,1987, the court stated:
We are at the point that we agree summary disposition is inappropriate, and we are also at the point where we have been engaged in pretrial proceedings for far too long.
Transcript at 2.
Despite the background recited above, on May 20, 1987, nine days before a then-scheduled pretrial conference, defendant was given the opportunity to address what it referred to as lately discovered jurisdictional defects. A deadline of June 26, 1987 was set, despite the nearness of trial. Defendant failed to meet that deadline and the case proceeded to trial.
The legal issues now raised in defendant’s motion for reconsideration were either not briefed, argued, or tried before the court, or have been previously addressed. None was raised in defendant’s post-trial briefing. It is thus remarkable that in light of this procedural history and after a two-week trial, the defendant should make the following statement in its motion:
[T]he court has not ruled upon the government’s defenses based upon Articles 17 and 20 of the 1944 Treaty. Defendant contends that the provisions of the 1944 Treaty preclude litigation in the Claims Court. (Emphasis added.)
If the court applies the provisions of Articles 17 and 20 of the 1944 Treaty, it must find that the treaty provides for the resolution of claims arising under it. Since there is a specific mechanism within the treaty to which plaintiffs never resorted, the court has no need to assume jurisdiction of Mosqueda’s claims.
The substantive issue1 of whether Articles 17 and 20 bar recovery is precisely the one addressed in the order of January *51425, 1985. Defendant’s motion was denied in part because it required further factual development. The court specifically stated that the issue could be renewed in a later motion for summary judgment. It has not been the subject of a motion since May 1985, however.
Consequently, the motion for reconsideration is not only four days late in terms of RUSCC 83.2(f), it comes four years after the complaint and after thousands of hours of investment of time by the parties, the court, the witnesses, counsel, and court staff. For that reason alone, it is denied. See Prestex, Inc. v. United States, 4 Cl.Ct. 317, 318, aff'd, 746 F.2d 1489 (Fed.Cir.1984) (“As an initial matter, post-opinion motions to vacate or reconsider are not favored. This is especially true where a party has had a fair opportunity to argue [or] litigate the point in issue.”) (citing General Elec. Co. v. United States, 189 Ct.Cl. 116, 117-18, 416 F.2d 1320, 1321 (1969)).
Because defendant refers to some of the grounds for its motion as jurisdictional, however, the court will nonetheless explain why it will not change its ruling.
Treaty Arguments
Defendant argues that Articles 17 and 20 of the Treaty deprive this court of jurisdiction over Mosqueda’s claim. In effect, defendant’s position is that the Treaty precludes all judicial review. There is some question as to whether judicial review of constitutional claims can be totally precluded. See Cardenas v. Smith, 733 F.2d 909, 918-19 (D.C.Cir.1984). It is certain, however, that “judicial review is precluded only where there is clear and convincing evidence of an intent to restrict access to the courts.” 733 F.2d at 919.
Defendant first argues that Article 17 of the Treaty prevents any individual from bringing suit for the type of flooding that occurred here. Article 17 reads in pertinent part:
The use of the channels of the international rivers for the discharge of flood or other excess waters shall be free and not subject to limitation by either country, and neither country shall have any claim against the other in respect of any damage caused by such use.
Each Government declares its intention to operate its storage dams in such manner, consistent with the normal operations of hydraulic systems, as to avoid, as far as feasible, material damage in the territory of,the other. (Emphasis added.)
Article 17 does not bar Mosqueda’s claim. First, the article addresses “the discharge of flood or other excess waters.” It does not pertain to the cause of the flooding as found by this court, namely the development of sand and sediment blockage due, not to the discharge of water, but to the retention of water by Hoover and Glen Canyon Dams. See March 4 opinion at 497-503. Second, even if the flooding did depend on the release of excess waters, the article prohibits one country from asserting a claim against another. It does not speak to suits by a citizen of either country for flooding due to the release of Colorado River waters. Consequently, Article 17 does not provide a jurisdictional or substantive defense against the claim alleged in this case.
Defendant next argues that Article 20 ousts this court of jurisdiction. That article provides in pertinent part:
Each Government shall assume responsibility for and shall adjust exclusively in accordance with its own laws all claims arising within its territory in connection with the construction, operation or maintenance of the whole or part of the works herein agreed upon, or of any of the works which may, in the execution of this Treaty, be agreed upon in the future. (Emphasis added.)
In defendant’s view, Mosqueda cannot sue in this court until he demonstrates that he cannot obtain adequate relief in the Mexican courts. Defendant relies on a statement in the order of January 25, 1985 addressing defendant’s first dispositive motion. The court stated: “Although plaintiffs’ complaint, as construed by this court, may state a cause of action apart from the Treaty, article 20 provides that all claims *515arising in Mexico in connection with the operation of the dams involved in this litigation must be adjudicated according to the laws of Mexico.” Anderson v. United States, 7 Cl.Ct. at 343. The succeeding language, however, suggests that the court was applying that analysis to the American, not Mexican, nationals. Moreover, it closed that analysis by stating that defendant had not put before the court “evidence as to whether a forum exists or has been established by Mexico.” Anderson, 7 Cl.Ct. at 343. More importantly, however, Article 20 does not, for two reasons, apply.
First, by its terms, Article 20 relates to claims connected with projects agreed upon in the Treaty, or future works agreed upon in execution of the Treaty. It does not address claims caused by other projects. It could not have referred to the construction of Hoover Dam, which was built before the 1944 Treaty, and does not, by its terms, include Glen Canyon Dam. The Colorado River projects agreed upon in the Treaty are described in Article 12. These works include: a diversion structure below the Northern International Boundary on the Colorado (i.e., what became the Morelos Diversion); Davis Storage Dam and Reservoir; canals and other facilities for conveying water to Mexico; and gauging stations.
It is clear that neither dam in question in this litigation was constructed in furtherance of the Treaty. Hoover Dam was built pursuant to the authority of the Boulder Canyon Project Act, ch. 42, § 1, 45 Stat 1057, 1057 (1928) (codified at 43 U.S.C. § 617 (1982)). It closed in 1935, nine years before the Treaty was signed. Glen Canyon Dam was built pursuant to the authority of the Colorado River Storage Project Act, ch. 203, § 1, 70 Stat. 105, 106 (1957) (codified as amended at 43 U.S.C. § 620 (1982)). That latter act does not indicate that the authorized projects were in furtherance of the Treaty. Because Article 20 does not contemplate legal claims relating to the operation of Hoover and Glen Canyon Dams, it does not preclude or limit this court’s jurisdiction of Mosqueda’s claim.2
The second reason why the court views Article 20 as inapplicable is that it is far from clear that the remedy alluded to there permits a Mexican citizen to sue the United States in a Mexican tribunal. The more natural reading is that each country adjudicates claims by its own citizens against itself arising out of the operation of the joint project.
Defendant also appears to assert that even in Article 20 does not deprive this court of jurisdiction, it at least requires the court to apply Mexican law to Mosqueda’s claim. As discussed above, Article 20 does not apply to Mosqueda’s claim. Even if it did, the question of whether a taking occurs is a matter of United States federal law. See Johnson v. United States, 202 Ct.Cl. 405, 418, 479 F.2d 1383, 1390 (1973). Although property interests are defined by reference to local law, 202 Ct.Cl. at 418, 479 F.2d at 1390, defendant did not raise any dispute concerning whether Mosqueda had a property interest cognizable in Mexico. Based on Mosqueda’s testimony, the court found that he did.
Intrusion Into Foreign Affairs
Defendant argues that the court’s findings of fact and conclusions of law somehow intrude upon the conduct of foreign relations between the United States and Mexico. In its brief in support of its first dispositive motion, defendant argued that summary judgment should be granted because “plaintiffs fail to state a claim upon which relief can be granted by this court because the United States is without authority to take their properly by eminent domain.” In the course of developing this argument, defendant stated in its brief that “Mexico is also a sovereign state, engaged in full diplomatic relations with the United States and subject to no control or exercise of jurisdiction by the United States within its borders. For the United States to condemn property in Mexico is a practical impossibility without offending the sovereignty of Mexico.” Defendant’s motion to dismiss for failure to state a claim was denied without prejudice. With respect to defendant’s foreign relations argument, the court stated that “defendant’s support for this factual proposition consists only of letters *516from agency counsel, which, however well reasoned, are not substitutes for an affidavit or other evidentiary material.” Anderson, 7 Cl.Ct. at 345. Defendant did not renew its motion for summary judgment on this ground, nor did it submit affidavits concerning the United States’ conduct of foreign relations with Mexico, nor did it raise the foreign relations issue in its pretrial briefing.
Defendant did make recitations concerning interference with foreign affairs, however, in its June 26, 1987 motion for an extension of time to file a dispositive motion. In denying that motion, the court stated: “While it may well be that there will develop during pre-trial and trial serious defects in plaintiff's case of a factual or legal nature, the court is persuaded that given the previous history of the case, the only way to expose and frame those questions is to proceed to trial.” Order of July 2, 1987.
Despite this observation by the court, defendant failed to develop this issue. It was certainly not tried to the court and no mention was made in defendant’s post-trial brief. The motion gives no reason, either legal, or based upon the factual record, why the ruling would impermissibly intrude on relations between the two countries, and none are apparent to the court.
It is difficult for the court to understand what other issues defendant could consider unresolved. Defendant in its pretrial briefing did raise for the first time the issue of whether 28 U.S.C. § 2502 (1982), limiting an alien’s right to sue in the Claims Court, barred plaintiff Mosqueda’s claim because he is a foreign national. That argument was rejected by the court in its September 9, 1987 order, in part because defendant later conceded that there was no bar. Also, defendant never reasserted any alleged difficulty with the case due to the fact that the land was outside the United States, despite repeated invitations to do so. See March 4 opinion at 480 n. 1.
Additionally, while defendant did, in its pretrial brief, claim that Mexican law created a servitude of drainage, this assumes Article 20 of the Treaty applies to this case. As stated above, the issue of whether a taking occurs is controlled by United States federal law. In any event, it is far from clear that Mexican law, assuming it applies, bars recovery. While there is a provision in the Mexican Federal Civil Code that states, “Lower estates are subject to receive the waters which naturally ... fall from the higher estates....” Codigo Civil para el Distrito Federal [C.C.D.F.] art. 1071, the next article provides that when that occurs, the downstream owners are entitled to indemnity, C.C.D.F. art 1072.
To the extent defendant raises in its current motion an issue with respect to whether the acts leading to the taking were within United States’ discretion and control, that issue was resolved in the March opinion. It was found there that construction and operation of the dams were within United States control. The only effect of the Treaty was that it called for a minimum delivery of water. It was found, however, that retention of waters in excess of those mínimums due to the filing of Lakes Mead and Powell caused the blockage.
Finally, the order of January 25, 1985 disallowed, on defendant’s motion, any possible claim that plaintiffs were entitled to relief directly under the Treaty, Anderson, 7 Cl.Ct. at 343. Under these circumstances, the court is at a loss to understand how defendant can take the position, either that the court has not ruled on issues raised by defendant, or that other matters were preserved and tried to the court.
Findings of Fact
The court has reviewed its findings of fact and concludes that they were proper.
CONCLUSION
For the reasons expressed above, the motion for reconsideration is DENIED.
. Defendant also argued in its first dispositive motion that the United States lacks authority to condemn property in Mexico and consequently, that this court cannot find that there has been an "inverse” condemnation. This court held that resolution of the issue required further factual inquiry and denied this aspect of defendant’s motion without prejudice, Anderson, 7 Cl.Ct at 345. Defendant has not reasserted this basis for its motion. But see Porter v. United States, 204 Ct.Cl. 355, 496 F.2d 583 (1974) (just compensation clause can be applied outside United States when it is the United States that takes the property); Seery v. United States, 130 Ct.Cl. 481, 127 F.Supp. 601 (1955) (plaintiff, an American citizen, stated a cause of action for taking of real and personal property in Austria); Turney v. United States, 126 Ct.Cl. 202, 115 F.Supp. 457 (1953) (United States found liable for taking of personal property in Philippines belonging to alien corporation); Juda v. United States, 6 Cl.Ct. 441 (1984) (Marshall Islanders could invoke just compensation clause).
. Report of the American Section of the International Water Commission, H.R. Doc. No. 329, 71st Cong., 2d Sess. (1930).
. Sheet overflow is the spread of water extending for some distance over the river's banks.
. Lockman, a civil engineer, was qualified as an expert in the areas of hydrology, fluid mechan-*486tes, and interpretation of aerial and satellite photography.
. Friedkin, presently a consulting engineer, has extensive experience with the lower Colorado delta from his previous employment with the U.S. Government. He testified at trial both as a lay witness and as an expert witness. As an expert witness, he was qualified in the areas of hydraulics, soil mechanics, alluvial rivers, and interpretation of aerial and satellite photography.
. Mean monthly flow equals an amount halfway between the extreme low and extreme high flows measured daily during the course of a month. The peak mean monthly flow is determined by identifying the month having the *491highest mean monthly flow over the course of the year.
. The flow measurements for 1963 and 1964 were taken at the Yuma gauge. The measurements in subsequent years were taken at the NIB gauge.
. Signed November 22, 1922, the Compact grew out of “An Act to permit a compact or agreement between the States of Arizona, California, Colorado, Nevada, New Mexico, Utah, and Wyoming respecting the disposition and apportionment of the waters of the Colorado River, and for other purposes,” ch. 72, 42 Stat. 171 (1921). The Compact itself was approved by the Boulder Canyon Project Act, ch. 42, § 13, 45 Stat. 1057, 1064 (1928) (codified at 43 U.S.C. § 617l (1982)).
. Briggs, a hydraulic engineer with the Bureau of Reclamation, is Chief of the Water Scheduling Branch for the Lower Colorado Region.
. Numerous other levees exist above Mosque-da’s property. Only those directly relevant to this litigation are reflected.
. In its order dated September 9, 1987, the court concluded that 28 U.S.C. § 2502 (1982), limiting an alien’s right to sue in the Claims Court, did not bar Mosqueda’s suit in the present case. This order was based in part on defendant’s acknowledgement that § 2502 did not deprive the court of jurisdiction over Mos-queda’s claim.
. Additions to the Rodriguez levee extended to Mosqueda's property by 1952. The court refers to both the extended portion and the original fifteen-mile-long structure as "the Rodriguez levee.”
. Dominguez was born in 1915. Although she could not identify precisely the years in which these activities were performed, the court finds it reasonable to assume her memory is accurate for events occurring since 1925.
. Mosqueda did not know the exact location where water was diverted from the Colorado River when he first came to the property. Since 1950, water has been diverted at Morelos Dam, 1.1 miles below the NIB. See infra p. 33.
. Mosqueda has used bottled drinking water since 1968.
. Mosqueda testified that he does not seek compensation for lot 6 because it is at a higher elevation than the others and has remained productive. He does not seek compensation for lot 12 because the Mexican Government has already taken an easement in this property, allowing it to remove dirt in order to build up the levee that crosses the Hardy River.
. One hectare is approximately equivalent to 2.47 acres.
. Ponding refers to the existence of still waters spread laterally over the river's banks because of channel blockage.
. Simons, a consulting engineer, was qualified as an expert in the areas of applied geomorphol*498ogy, river mechanics, hydraulic structures, erosion and sedimentation, irrigation and drainage, and the interpretation of aerial and satellite photography.
. Some of this water returns to the Colorado and Hardy Rivers as irrigation runoff.
. Sabey, an agronomist at Colorado State University, was qualified as an expert in the area of soil sciences. Ohmart, an ecologist at Arizona State University, was qualified as an expert in the areas of native vegetation, related soils and water analysis, tree dating, and the interpretation of aerial and satellite photographs.
. The final data for the Yuma gauge are for 1964. More recent data reflect the NIB gauge.
. While the court considers that it has authority to adjust post-taking sales figures for the effect of time, defendant did not challenge Mos-queda's assertions or provide a means to discount Mosqueda’s figures. Consequently, any adjustment by the court would be speculative.
. In addition, the motion notes that the court erred in referring to earlier dismissals of most plaintiffs as either voluntary or for the failure to answer interrogatories. Defendant states that this is incorrect because eight of the 36 plaintiffs were dismissed on the ground that their failure to raise taking claims in then-pending district court litigation precluded raising those claims in the Claims Court. Anderson v. United States, 7 Cl.Ct. 341, 345 (1985). Defendant is correct.
. To that extent, the dicta in the order of January 25, 1985 to the contrary is vacated.