OPINION
BRUGGINK, Judge.Plaintiffs in this action are the present or former owners of two parcels of land in Itawamba County, Mississippi. They contend that the activities of the United States, acting through the Army Corps of Engineers (the “Corps”), resulted in a taking of an easement on their property through flooding and sediment deposition. Jurisdiction attaches under the Tucker Act1 by virtue of the takings clause of the Fifth Amendment to the Constitution. The action is presently pending on remand from the Federal Circuit, which reversed an earlier grant of summary judgment in favor of the Government.2 The Claims Court had held that the facts alleged did not demonstrate that the flooding was inevitably recurring. The appellate court, reversed, finding that there were disputed issues of fact precluding summary judgment. The case was assigned on remand to the present judge. After trial and a visit to the property, the court concludes that a taking occurred, and that plaintiffs are entitled to recover.
FACT FINDINGS
The property in question consists of two parcels of bottom land along Twenty Mile Creek, close to its juncture with the East Fork of the Tombigbee River. Tract I consists of 483.5 acres. It lies about three miles river miles (“RM”) above the juncture. Tract II consists of 118 acres and is bordered by both Twenty Mile Creek and the East Fork of the Tombigbee. It lies about one and a half RM above the juncture. Plaintiffs Jimmy Crane, Johnny *449Crane, and Edwin Wilburn purchased both tracts in 1978.3 In June 1979, plaintiff Neal Turner bought both tracts from Wilburn and the Cranes.4
Twenty Mile Creek is a stream in northeastern Mississippi. It is about 30 miles in length and drains an area of approximately 174 square miles. The creek flows in roughly a southeasterly direction. It was once a meandering stream that flowed through the bottom land of the Tombigbee watershed between the East Fork and higher ground to the southwest. Because the natural stream bed had little carrying capacity it frequently topped its banks during heavy rains. For that reason, beginning as early as 1910, the creek was re-channelized to a somewhat different, and straighter, alignment slightly northeast of the old meander patterns. This early work did not drastically increase the carrying capacity of the creek, but apparently provided some flood protection, at least until the creek bed became silted or overgrown. The Federal Government became involved with the enactment of the Flood Control Acts of June 22,1936, and August 28,1937, which authorized clearing and rechanneliz-ing work in the lower reaches of Twenty Mile Creek and other tributaries of the Tombigbee.
In 1958, Congress authorized certain flood control projects on 22 streams entering the Tombigbee River, including Twenty Mile Creek. Flood Control Act, Pub.L. No. 85-500, 72 Stat. 307 (1958). Work on the Twenty Mile Creek portion was begun in December 1965, and completed in January 1967, when the completed project was turned over to the local sponsor, the Tom-bigbee River Valley Water Management District (the “Management District”). With respect to Twenty Mile Creek, the project was confined to the stretch from RM 0, which is the terminus of the creek at the point it joins the East Fork, to RM 11.1.
Work on the stretch of the creek between RM 0 and RM 9.1 involved channelizing the creek bed to a specified profile. Most of this work consisted of substantially widening and frequently deepening the channel. The particular shape of the profile depended on the stretch of the creek involved. The Corps’ design called for staging the elevation of the creek bed so that the channel was successively deeper as it approached the East Fork. From RM 9.1 to 11.7, work consisted solely of clearing and snagging along the natural channel.
The project was designed to accommodate a flood of 0.33 year recurrence frequency. To put it differently, it was not intended to handle floods which occur less frequently than once every four months. Presumably a flood that could be anticipated to occur only once a year, for example, would top the banks as designed.
In fact the execution of the work resulted in a very different profile than the one put on paper. The contractor who excavated the creek bed was instructed to place the dredged materials along both banks of the creek in a spoil pile of uniform height and shape. Although Corps witnesses at trial demurred at the suggestion that this spoil pile in fact constituted a levee, several internal Corps memos refer to the phenomenon as a levee, and the court’s observation confirms that assessment. The levee is virtually unbroken for miles on both sides of the creek, and appeared to be approximately 10 to 15 feet above the natural contour. It was deliberately left open at a few points where there were distinct natural drainage channels into the creek. One such point was at the south corner of Tract I, where the drainage ditch from the field fed into Twenty Mile Creek. There was a small bridge over the ditch to make driving the levee possible. The levee is driveable in reasonable weather, even at places where it drops down to the original level. The Management District maintains the channel and the levee to keep down vegetation.
*450The immediate effect of the project was to increase the water surface gradient of the creek between RM 11.7 and O. The creek was initially able to handle a substantially larger volume of water, and the gradient resulted in much faster discharge into the East Fork.
Soils in the Twenty Mile Creek bottom are highly erodible. Much of the bottom land was already in cultivation prior to the work in 1965-66, and considerably more was converted to row crop cultivation thereafter. Almost immediately after completion of the project, there were complaints from the upstream landowners of serious erosion problems. As will be discussed more fully below, increased velocity of the water flow after the project work added to the sediment carrying capacities of the creek, which in turn destabilized the stream bed, particularly upstream of RM 11.7. This was partly through a process known as headcutting, in which soil at the banks of a creek and at the margin of all its feeders sloughs into the water. In the succeeding years, there has been continuous collapse of the creek bank, and general erosion.
Complaints over erosion led Congress in 1980 to fund modifications to Twenty Mile Creek. Pub.L. No. 96-304. This funding was for the purpose of constructing two “grade control” structures and to do related bank stabilization work. Subsequently, the Corps constructed a grade control structure at RM 11.7 and another at RM 19.9. The purpose of these structures is to break the velocity of the water flow and thereby reduce the headcutting effect during floods. In addition, several sections of the creek above 11.7 were protected by placing riprap along the banks or by planting willows.
The present lawsuit is not directly concerned with the upstream erosion and loss of farmland. Rather, it is concerned with a phenomenon that manifested itself somewhat later. It can best be summarized by setting forth plaintiffs’ theory of causation in one place since the court ultimately finds that it is correct. Plaintiffs contend that the erosion upstream of their property resulted in deposition of sand in the lower reaches of the creek and at the juncture of Twenty Mile Creek and East Fork. This deposition raised the general water surface level of the creek and, combined with the concentration of waters during floods which were now more elevated because of the levee, caused catastrophic “blow outs” of the levee onto plaintiffs’ lands.
The events relevant to the plaintiffs’ land are best drawn from the testimony of Jerry Wilburn and Neal Turner. Wilburn and his partners owned the land from 1978 through June 1979, and again from 1983 to the present. Turner owned the land in the interim period. Wilburn has lived all of his life (50 years) in Mantachie, Mississippi, a community a few miles from the site. He has been a farmer and, for a period, was sent by his neighbors to represent them in Jackson in the legislature. He has been familiar with the area of Tracts I and II for many years, and recalls ownership and use of the parcels back to the early 1960’s. Although Wilburn’s narrative is a bit disjointed, apparently both tracts have been farmed since at least the early 1960’s. Wilburn testified that prior to the Corps’ work in 1966, Twenty Mile Creek would periodically flood the entire bottom. This did not damage the farm land, however, since there was very little sand in the water. After the runoff, the land was useable.
Prior to 1963 or 1964, there were small fields on both tracts. In about 1963 or 1964 the fields were connected on Tract I so that virtually the entire parcel was put into cultivation. Wilburn also testified that Tract II was also more extensively cleared in the mid 1960’s, and that this followed previous cultivation of smaller fields on the tract.
Wilburn’s testimony is somewhat confusing on precisely when larger scale soybean farming began on Tracts I and II. He stated that one of the former owners, a man named Pierce, farmed the land from about 1964, first as a renter, then as owner. The first year he farmed, he achieved a spectacular harvest of 50 bushels of soybeans per acre. Thirty-five bushels is considered a high yield in that area. He also *451testified, however, that the first time he remembers a crop was sometime between 1966 and 1968. He presumably meant a soybean crop since he stated that the owners had planted smaller fields of corn, sunflowers and soybeans in the 1950’s.
According to Wilburn, Pierce consistently got good crops during the 14 years he farmed Tract I. This despite the fact that, upon the completion of the channel deepening, Pierce frequently had problems with sand being deposited at the south corner of Tract I during high flows in Twenty Mile Creek. Water came in through the gap in the levee and would “flood the[ ] fields, and throw[] all that sand out there.” When the water went down, Pierce would go in with a bulldozer and clean out the plug and level the sand. There was not so much sand that he could not spread it over his fields and keep them in production.
Despite these problems, Pierce was able to keep the drainage ditches on Tract I cleaned out and made a good crop in 1978. Wilburn and his colleagues decided to buy both tracts. As mentioned above, they promptly resold the properties in 1979 to Turner, who testified that both tracts appeared to have “good farming soil.” Almost immediately, Turner began to have the same problems with keeping sand-laden water from coming in at the gap in the levee. There were also minor breaks in the levee at both tracts in spring 1980. Turner was nevertheless able to make a crop of 20 to 25 bushels per acre in 1979 and a “pretty fair crop” in 1980 and 1981. In 1981, Turner experienced what is referred to as a “dry flood.” A dry flood occurs when there is very high water in the channel itself, with very little or no local rain. The high water is due to rapid runoff from upstream rain. The flood entered at the gap and silted his field, ruining a crop. He was able to replant, however, and got a decent crop. The Management District cleaned up the gap and sand in the vicinity.
Later in 1981 Turner decided to fill the gap to protect the upper tract. Consequently, he took out the bridge and hauled in enough dirt to complete the levee at that point. He redirected the drainage from his ditches on Tract I so that they exited onto the adjacent triangular parcel. From there the water flowed into Tract II and on into the East Fork. In most of 1982 Turner had no problems, and his crop that year was good.
The first really serious flooding occurred in December 1982. A major blowout of the levee at two places resulted in deposition of two to three feet of sand on Tract I and plugging of all the drainage ditches. The main blowout that inundated Tract I came from a point not adjacent to Turner’s property, but upstream. There was flooding and sand deposition on a large portion of Tract I, which still had a crop of mature soybeans on it. There was also a breach at Tract II, directly onto Turner’s land. The deposition at Tract II obviously came from Twenty Mile Creek, as opposed to the East Fork, because flotsam from a large grade control structure upstream beached itself on the tract.
According to Wilburn, the same type flooding and sand deposition occurred in 1983, after which the condition of the fields was “horrible,” particularly Tract II. It was virtually covered in sand to varying degrees. Where the deposition was minor, there was standing water. Eventually willow trees began to grow at the south end of the tract. Wilburn stated that the problems in 1982 and 1983 were far worse than what Pierce had faced in the 60’s and 70’s. Turner testified that he tried to plant a crop in 1983 on Tract I by spreading out the sand, but was unsuccessful. As he put it, Tract I looked like a beach. He was faced with hundreds of acres of sand. Portions of Tract II have, according to Turner, constantly been under water since then. During this period, Wilburn also observed that at the juncture of the East Fork with Twenty Mile Creek water from the latter was coming in such volume and velocity that the East Fork flowed upstream for a period.
By 1983, Turner was no longer able to farm the property, and was therefore unable to make the mortgage payments. Wilburn and his partners took the parcels back and the mortgage and note were cancelled. *452Wilburn tried to repair the damage to the lower tract by having the sand leveled, the growth cut, and the ditches cleaned out. He paid $42,000 to have the work done. This succeeded in draining the lower tract into the East Fork. He also smoothed and disked the upper tract and then rented both parcels to Douglas Kitchen. Kitchen was able to get a crop in 1985 on a total of 380 acres. Wilburn characterized it as “pretty good,” yielding about 17 or 18 bushels per acre.5 One reason production was not higher is that beans would sprout and grow quickly in the sand, but, because the soil had little water retention capacity, the plants would burn. 1985 was a dry year, and Kitchen had no trouble getting equipment onto the fields. The following year Kitchen was unable to plant for reasons unrelated to the lawsuit.
In 1987 Kitchen decided to cut back on his farming activity. Wilburn and his partners elected to participate in a land set-aside program involving planting of timber. They paid a timber management company to plant pine trees. The pines have not prospered. Today about 10 percent survive, and from observation of some of them, the court judges they have not taken to the location. An experiment with planting oaks also failed.
More recently there have continued to be “blowouts” of the levee and deposition of sand on both tracts. Larry Dillard, formerly employed with the Management District, stated that there was a break in the levee in early 1990, which left the upper field standing in water. The court observed the land in May, 1991. The lower tract appeared to be a virtual lake, with occasional mounds of sand covered with willow growth. The upper tract did not appear to have any standing water, but is mostly covered with a layer of sand.
Aerial photographs track part of the events referred to by the witnesses, as well as earlier conditions of the property. The earliest aerial photograph was taken in 1937. Tract I is not visible. Tract II is completely forested. A triangular piece between Tract I and Tract II is completely cleared. At that time virtually the entire bottom on the southwest side of the channel was under cultivation, at least so far as visible in the photograph. In addition, a few parcels southeast of Tract II between the channel and East Fork were cleared. An undated photograph taken between 1937 and 1951 shows that about half of Tract I was cleared and cultivated. A picture taken in 1958 shows the same conditions. Perhaps half of Tract I was cleared for cultivation. None of Tract II appears to have been cleared. There are no aerial photographs from which it can be more precisely determined when Tract II was first cleared. A topographic map initially prepared in 1965, however, by using green shading to indicate uncleared areas, reflects that all of Tract II was uncultivated and only a quarter of Tract I. The first photograph which reflects the more extensive clearing of Tract I and the first major clearing of Tract II was taken in 1969, after the channelization. A picture taken in 1974 only shows Tract II. That parcel has been completely cleared and appears to be under cultivation. There is a clearly noticeable stream flowing through the middle of the tract, picking up drainage from the triangular parcel to the north, which in turn drains Tract I. A different picture taken in 1974 shows a closeup of the gap in the levee at the south corner of Tract I. The channel is open and splays of sand emanating from the drainage ditch are clearly visible. A photograph of both tracts taken in March 1979 shows ponding on Tract II, but Turner testified that this water drained.
A photograph taken November 1980 shows that drainage from Tract I is directed toward the gap in the levee at the south corner of the lot. Tract II is still being cultivated, but it is apparent that the track of the drainage ditch is much widened so that there is ponding on parts of the southeast end of the parcel. A photograph taken in March 1985 shows two developments. *453Tract I reflects a very large splay of sand emanating from a breach in the levee at a point above plaintiffs’ property. The gap in the levee at the south corner of Tract I has been closed and drainage appears to be going through well-defined ditches exiting to the east-southeast. The center of Tract II has a large splay of sand and substantial portions of the tract appear to be in standing water. More recent photographs, taken on the ground, show extensive ponding and sand deposition, as well as some recent breaches of the levee on both tracts.
The evidence adduced at trial, consisting in large measure of government documents, establishes plaintiffs’ basic theory of causation beyond any doubt. The Corps observed and made note in a series of investigative reports of the fundamental elements later relied on in plaintiffs’ causation theory. A trip report in 1975, for example, focused on the backwater effect below the juncture of Twenty Mile Creek and the East Fork. The author writes that “the implementation of the Twentymile Creek channel improvement project has magnified the amount of silt and sand that is coming from that watershed.” An undated “Interim Report on the Tombigbee (East Fork),” states,
Sedimentation problems have increased along Twenty Mile Creek since the 1965 work. Due to changed flow conditions, head cutting and bank erosion along Twenty Mile Creek, large quantities of silt and sand are being transported to the mouth of the creek and into the Tombig-bee River and Flood plain____ On other occasions larger floods cause levee failure and large quantities of material are deposited in the flood plain of Twenty Mile Creek.
Elsewhere, that same report recites, “A second major factor in recent localized problems near the Mouth of Twenty Mile, is the channelization and construction of levees along Twenty Mile Creek in the mid 1960's.” A study completed in February 1979, for example, documents the degradation of the stream bed in the upper reaches of the creek, and the opposite effect downstream: “sediments are being deposited, raising the bottom of the channel and reducing the cross-sectional area ... thereby raising the water surface above that predicted for post-construction conditions.” The study also notes the increased velocities after channelization. In a report dated March 1990, the writers observe that this process of degradation and aggradation “occurred along much of the channel between 1980 and 1989, continuing earlier trends.”
In a design report issued in May 1983 titled “Bank Stabilization Project Twenty-Mile Creek Interim Measures,” the authors recite,
Shortly after the flood control project was completed, bank caving and erosion problems became evident____ Excavation below RM 9.1 increased channel width and lowered the water surface level, which increased upstream flow velocity and initiated headcutting. During higher flows, whirlpools formed below the narrower natural channel, and eroded the adjacent banks____ In addition, adjacent croplands were being flooded in the downstream reaches of Twenty Mile Creek because eroded sediments had deposited in the stream bed, increasing the flow stages.
In a memorandum dated December 30, 1985, Benton Odom, Chief of the Corps’ Coastal Engineering and Hydraulic Design Section, recites,
[Ejrosion and overbank silt deposits occurred along Twenty Mile Creek ... before the 1966 flood control project. The rate of erosion and overbank deposits seems to have increased during a period from the early 1970’s to 1983. Factors that have probably contributed to this are: (1) highly erodible soils, (2) poor soil conservation practices, (3) periods of unusually heavy rainfall, and (4) increased velocities due to the 1966 flood control project in the lower reach of Twenty Mile Creek.
(Emphasis supplied.)
The 1983 report notes that between 1967 and 1978, the upper reaches of the channelized portions of the creek were degrading rapidly while the lower portions, in plain*454tiffs’ vicinity, were aggrading. Consequently, river stages for the same volume of water were decreasing at the upper end of the creek and increasing adjacent to plaintiffs’ land. The report documented an increase in velocity of water throughout the channelized creek. It also noted the “tremendous loss of land as the channel has widened.” Id. at 2.24. The latter is a reference to upstream portions of the creek. Much of this soil was deposited at the juncture of East Pork and Twenty Mile Creek, in the bed of the lower portions of the creek, or as will be discussed, on adjoining tracts of land.
In “Technical Report No. 15, Bank and Channel Erosion Problems at Twenty Mile Creek, Mississippi,” dated September 1989, the authors recite that “it is apparent that no allowance was made in the design for the prevention of the severe erosion and deposition problems which have occurred.” It goes on to explain:
13. Stream Conditions. Construction of the project through the excavation of a wider, and in instances deeper, channel with the removal of brush, trees, and snags has contributed to the alteration of flow conditions in the stream. Under the original construction work, the thalweg [lowest thread of the stream] was lowered from the mouth to a point 9.1 miles upstream by channel excavation. (Plate 1). The project also involved clearing and snagging along the natural channel for 2.6 additional miles upstream from the excavated section. The combination of these changes lowered the water-surface elevation immediately following completion of the project in the 11.7-mile reach____ Under project conditions, the increased hydraulic gradient of the modified channel resulted in a steeper water-surface slope with increased water veloc-ities____ The increased velocities and increased slope added to the sediment-carrying capabilities of the stream. Because of this changed condition, soil was eroded by the flow from the banks and bottom of the stream. This unstable condition was extended upstream about 10 miles, primarily by the process termed “head cutting.”
15. Basic problem. The three basic problems identified ... are as follows:
a. Erosion____
b. Bank sloughing____
c. Sediment deposition. This problem is confined to the lower 5.5 miles of the project and is caused by the high level of transported sediments entering the low-velocity stream area near the mouth of the creek.
The flooding complained of by plaintiffs is further explained in a final report titled “Tombigbee River (East Fork) Study,” dated January 1988. The writers point out that the effect of channelizing is to increase flow efficiency and quicken removal of flood runoff. The sediment carrying capacity of channelized streams, as well as the generation of sediment, is also amplified. Unfortunately, as the report points out, the East Fork itself has not been channelized, and thus it has a relatively slow, meandering flow and conseqtiently little sediment carrying capacity. The result is predictable. At the confluences of the East Fork with the easterly flowing channelized streams, large deposits of sand have built up.
The 1988 report also points out that the spoil piles erected along Twenty Mile Creek (which it refers to as levees) have the beneficial effect of holding higher flood flows within the banks of the creek. It goes on to explain the negative effect of that phenomenon, however:
Conversely, these levees have been breached in the past during large floods resulting in large amounts of sediment to be deposited over a relatively small area. This phenomen[on] of levee breaching has been observed to be aggravated by landowners on the lower reaches of these tributaries. On the lower end of Twenty Mile Creek, landowners have blocked gaps in the levees which were left during construction for interior drainage____ [W]hen the levee is breached by either the means of overtopping or erosion through increased channel velocities, the result is a sudden, catastrophic amount *455of sediments consisting mainly of fine sand to be deposited in a fan shaped delta____
The 1988 report concludes that excessive surface water in the lower reaches of Twenty Mile Creek is related in part to “[ijncrease in upland erosion,” “[ijncreased sediment accumulation in the floodplains of the tributaries,” and the “general lessening of stream gradients."
The reference in the 1988 Corps report to actions by landowners blocking gaps in the levee is probably a reference to actions of Turner. As indicated above, he plugged the drainage outlet at the southwest comer of Tract I after a series of smaller sand splays. The plain thrust of the 1988 report, as well as the import of the trial testimony of Douglas Otto, a Corps Engineer, was that such gaps in the levee were useful, since they created de facto sedimentation ponds on adjacent tracts of land. I.e., the force of the flood was dissipated when the rising waters could find natural exit points in the levee. The good news for downstream landowners after the levees were constructed was thus that they would be protected from even higher flood levels than the project design anticipated. The bad news for plaintiffs was that the gap in the levee on their property would be a safety valve for absorbing excess water and sediment. The import of the report, and the Government’s position at trial, was that it was not sporting for plaintiffs to try to protect themselves from bearing the brunt of the upstream erosion.
DISCUSSION
Liability
The law applicable to the facts above is not contested. Plaintiffs can recover for a constitutional taking if they have shown that their land is subject, because of government action, to permanent or inevitably recurring floods. Turner v. United States, 901 F.2d 1093, 1095 (Fed. Cir.1990); Hartwig v. United States, 202 Ct.Cl. 801, 809, 485 F.2d 615, 620 (1973). Such facts establish the taking of a flooding easement.
The court has no difficulty finding that there is a causal connection between the Corps’ activities and the damage to plaintiffs’ land. That causal connection is virtually conceded in Corps documents. Channelization led to higher velocities and flood stages, which in turn added to the sediment carrying capability of the water and greatly accelerated erosion in the upper reaches of Twenty Mile Creek. This soil was deposited at the lower reaches of Twenty Mile Creek and immediately below its juncture with the East Fork. This deposition raised the river stage along the channel adjacent to plaintiffs’ land.6 Over time, the process had occurred to such an extent that the tremendous flows captured in the upper reaches of the creek could no longer be contained in the now shallower lower reaches. This was exacerbated by the damming effect of the sediment deposits at the East Fork.7
Although this process may have been made worse by increased cultivation upstream and by the naturally erodible soils, there is no suggestion that the damage would have occurred without the channelization in 1965 and 1966. The condition of the soils of course was a given at the time the Corps did its work.
*456Despite the seeming abundance of evidence that the source of plaintiffs’ difficulties was Twenty Mile Creek, the Government offered a different causation theory at trial, primarily through the testimony of Douglas Otto. Based on a topographic map, and a few assumptions, Otto attempted to demonstrate that the drainage ditch on Tract I served a basin of nearly 12,000 acres. Otto concluded that the flooding on both tracts was the result of Turner’s closing off the gap in the levee at Tract I and trapping flow from the drainage ditch, and funneling it in a new direction, namely onto Tract II. This theory was fanciful.
To place 12,000 acres of drainage in perspective, this is nearly 19 square miles. Twenty Mile Creek, which, by the court’s own observation appears at least as substantial, if not larger, than the East Fork of the Tombigbee River, part of a major river system, drains 174 square miles, only nine times as much as that purportedly drained by the ditch through Tract I. Although the site visit did not involve walking all of Tract I, the entire lower portion of the tract is visible from the adjoining, raised road embankment. And the point at which the main stem of the drainage ditch crosses under that embankment was visible. Despite the heavy rains contemporaneous with the site visit, there was no clearly obvious water flow through the property. Otto’s theory thus is apparently inconsistent with observable physical conditions.
It is also based on the assumption that the northeast bank of the levee intercepts what would otherwise be numerous other lower points in the upstream topography, and thus funnels drainage onto plaintiffs’ land. Otto testified that he has walked much of the levee and does not believe it has other outlets above plaintiffs’ land, at least, presumably, within the watershed of the drainage ditch. The topographic map in question indicates that the levee edge of the drainage area at issue is at least seven miles long. The court finds that it is simply incredible, without better support, that there are not in fact other points of access to the creek, or other exits from the drainage area, other than across Tract I.
Perhaps most telling, the physical evidence makes it clear beyond contradiction that the source of the sand was Twenty Mile Creek. It is a fact that there were breaches in the levee. Witnesses observed water coming through these breaches onto the fields, not the converse. On the ground photographs show that the levee eroded on the field side and sand is splayed out away from the breach in a fan. These tell-tale fans are patent in aerial photographs.
If Otto were correct, then Pierce should not have had sand deposition problems during the early 1970’s, although he did, and Turner should not have had similar problems before he plugged the gap, but he did. Finally, there are the numerous references in the Corps’ own documents to breaches in the levee resulting in deposition from the creek onto adjoining lands. In light of this overwhelming evidence, the suggestion that Turner caused the mass deposition of sand on his property is not remotely credible.8
From the evidence discussed above, the court also finds that the flooding of plaintiffs’ land is inevitably recurring. The dynamics of the erosion-sedimentation-flooding cycle were initially manifested on the property in question during Pierce’s ownership. It has continued ever since. The levee has been breached on numerous occasions, spreading over at least 10 years. Both tracts are presently still affected by events that reached a catastrophic level in 1983. In its present condition Tract I is not useable for row crop cultivation. Tract II is not suitable for any sort of cultivation. The court concludes that both tracts have been permanently damaged, and are sub*457ject to a continuing easement of flooding from Twenty Mile Creek.
The very evidence that aids in finding that the flooding is inevitably recurring also tends to suggest that the problem arose more than six years prior to the commencement of this action. The complaint was filed February 22, 1988. If the dynamics of the taking had stabilized prior to February 22, 1982, and if Turner should have known the extent of the taking at that point, then plaintiffs would be barred. See 28 U.S.C. § 2501 (1988). As the Supreme Court stated in United States v. Dickinson, 331 U.S. 745, 67 S.Ct. 1382, 91 L.Ed. 1789 (1947), a property owner need not undertake piecemeal or premature litigation if the taking results from an ongoing event. But he is obligated to proceed when “the consequences of inundation have so manifested themselves that a final account may be struck.” Id. at 749, 67 S.Ct. at 1385.
Up until December 1982, Pierce and Turner had been able to cope with the periodic small scale flooding. The land never recovered from the catastrophic events commencing in December 1982, however. In that flood, a mature crop was destroyed, and the sand intrusion reached an extent and a depth which made it impossible to restore the land to productivity. The court finds that the taking was no earlier than December 1982, and thus the suit is timely. In order to fix a specific date from which to assess interest, the court will use the end of the 1983 growing season. By December 1, 1983, Turner had given up any hope of making the land productive for row crops and had allowed Wilburn and his partners to repossess. By that date, damages had become sufficiently apparent and stabilized to permit suit. Accordingly, the court will use December 1, 1983 as the date of taking.
One final issue as to liability remains. The Flood Control Act of 1958, through which the work at issue was authorized, required that “local interests” agree to certain conditions, including, for example, provision of necessary lands and easements. The local interests were represented by the Management District. In a resolution dated August 23, 1963, the district agreed to “Maintain all works after completion in accordance with regulations prescribed by the Secretary of the Army.” From this, the Government concludes that the Management District, and not the Federal Government, is liable for the loss in value of plaintiffs’ lands. There are two defects in this analysis.
The first is a factual one. The Government’s defense did not include an attempt to interpose lack of some unspecified maintenance as a supervening cause which would exculpate it from liability. It is plain from the discussion above that failure was built into the design. In the absence of evidence that, but for the commissions or omissions of the Management District, the damage would not have occurred, there is ample basis to attribute liability to the Federal Government. The fact that the Management District may independently have contributed to the damage, is, of course, not material. Only the United States can be the defendant here. The second deficiency in this argument is that the resolution does not purport to be a transfer and acceptance of liability for the construction itself. Since the damage was not linked to maintenance responsibilities assumed by the Management District, the resolution is of no consequence in this action.
Damages
Three experts provided evidence as to the financial impact of the easement imposed by the project. The opinion of real estate appraiser P.O. Beard, one of plaintiffs’ two experts, came into evidence only in the form of his appraisal. He was unable to testify, and the parties agreed that his report was admissible. There was very little difference in approach between Beard’s methodology and that of Fred Spencer, another expert appraiser for plaintiffs, who did testify, and James Davis, who testified on behalf of the Government and offered his own report. All three appraisers gave opinions as to the value before the flooding damage and after the damage. This is an appropriate way to *458establish the impact of a taking. See United States v. Miller, 317 U.S. 369, 376, 63 S.Ct. 276, 281, 87 L.Ed. 336 (1943); Georgia-Pacific 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).
Beard concluded that the highest and best use of both tracts of land prior to the taking was for row crop agriculture, principally soybeans, with some fringe areas around the periphery of the property suitable for production of timber. Based on an examination of four comparable sales, he concluded that the unaffected value of the land was $725 per acre. This compared with a value of $100 per acre after the flooding, which he derived from comparable sales of timberland. He concluded that the property was reverting to timberland, principally of a type of low value hardwood. Damages were thus $625 per acre, for a total of $375,600.
Fred Spencer also testified for plaintiffs. He did not prepare a report, but offered a summary sheet of his figuring. He based his conclusions of the before and after value on three comparable sales, two of which were also used by defendant’s appraiser. Spencer particularized the two tracts somewhat more than did Beard. The upper tract contained 457 acres of cultivatable land, and 26.3 acres of woodland. Based on his three comparable sales, he calculated a before value of $850 per acre based on use as row crop land and $250 per acre for the woodland. The after value he placed at $150 per acre for the 483 acre upper tract and $100 per acre for the lower tract. His assessment was that the lower tract would not grow anything of value. The upper tract he assigned more value to because of the possibility that it might sustain timber regrowth, although he discounted that likelihood.
James Davis testified as an expert appraiser for defendant. The effective date of his appraisal was December, 1982. He broke down Tract I into three portions. 65 acres are upland not affected by the flooding and siltation (plaintiffs do not contest this point). 347.1 are croplands, and 71.4 acres are woods, ditches, etc. He broke the lower tract into two parcels as well, 102.6 acres of cropland and 15.4 acres of woods, etc. He agreed with plaintiffs’ appraisers that the highest and best use of the cropland was primarily for production of soybeans. Although he also testified that portions of the lower tract never should have been cleared because it was too subject to inundation due to its location. In terms of fixing a value at those usages, he relied on nine comparable sales from the area. Based on those sales and an upward adjustment for inflation, he arrived at a figure of $775 per acre for 514.7 acres of cropland on both tracts and $275 for 86.8 acres of timberland, for a total before value of $422,800. The court notes that Wilburn and his partners purchased the land in 1978 for $260,750 (this did not include the 65 acres of upland, which Wilburn already owned) and sold the entire parcel to Turner the following year for approximately $380,-000.
Davis candidly observed, presumably with respect to Tract I, that “[tjhere has been a terrific amount of siltation, which I think has resulted in a soil on the property that is not capable of producing crops in its present condition.” The after value of the upper cropland, other than the 65 acres of uplands, was reduced in his view to $200 per acre, the value he assigns to cutover timberland with no new growth. In his estimation it was compatible with growth of softer hardwoods, such as sycamores and cottonwood, which have commercial value as pulpwood. As to Tract II, Davis believed that there had been no damage to the 15.4 acres of woodlands, and only minimal damage to the remaining 102.6 acres of cropland. He discounted the present value of the cleared portions of Tract II by 25 percent and arrived at an after-taking figure of $580 per acre. The total after value was thus $203,200. The difference between the before value ($422,800) and the after value yields a total diminution in value of $219,600, according to Davis.
The court notes that there is very little difference between the appraisers with respect to the before-value of cropland. Beard used the figure $725 per acre, Spenc*459er used $850 per acre, and Davis used $775 per acre. Since Beard and Davis represent a substantially larger number of comparable sales, and since Beard was plaintiffs’ witness, the court will use Beard’s figure. That figure requires some adjustment however, as to Tract I. Spencer testified that he assigned timberland values to 26.3 acres. Davis did the same thing with respect to 71.4 acres, although he apparently included ditches and roads in that figure. The court will use Spencer’s figure, instead of Beard’s, because ditches and roads should be considered part of the productive cropland. Accordingly, 26.3 acres of Tract I will be valued as timberland, and 457.2 acres as cropland. As to the value of woodlands, Spencer testified that the before value of woodlands should be $250 per acre. Davis used a figure of $275 per acre. Beard did not identify any land as standing timber so his evaluation is not available. The court once again uses Spencer’s figure, as the highest amount requested by plaintiffs.
Accordingly, the before value of Tract I is calculated as follows:
26.3 acres of timberland @ $250 per acre $ 6,575
457.2 acres of cropland @ $725 per acre $331,470
Total $338,045
As to after value, Davis’ testimony that Tract I is compatible with growth of softer hardwoods went unrebutted, and the court accepts his figure of $200 per acre. Accordingly, the after value of Tract I is as follows:
65 acres of upland, unaffected @ $725 per acre $ 47,125
418.5 acres of timberland @ $200 per acre $ 83,700
Total $130,825
The difference between the before and after values of Tract I is $207,220. This represents the total diminution in value attributable to the imposition of the easement.
Tract II is somewhat more problematic. The aerial photographs show no clearing on this tract in 1963. The entire tract is cleared in 1969. The work at issue on the creek took place in 1966. The Corps documents strongly suggest that a significant amount of Twenty Mile Creek bottom land was put into crop production immediately after the channelization. It is noteworthy that a comparison between the topographic data and the aerial photographs shows that virtually all the clearing in the lower portion of the triangle formed between the east Twenty Mile Creek levee and the East Fork of the Tombigbee River is land with a higher elevation than Tract II. The land to the east, which Davis described as containing one of the best stands of hardwoods he had ever seen is at approximately the same elevation and proximity to the East Fork. It does not show clearing through any of the aerial photograph series.
The logical inference is that, if Tract II was ever suited to row crop agriculture, it was due to the initial protective effects, or anticipated effects, of the channelization. Davis testified that those portions of Tract II now covered with silt or subject to standing water should never have been converted to row crops. He thinks they should have been left as timberland. The court agrees. If the value plaintiffs allege was taken was the very value conferred by the project alleged to have caused the damage, plaintiffs cannot recover. See United States v. Sponenbarger, 308 U.S. 256, 266-67, 60 S.Ct. 225, 229, 84 L.Ed. 230 (1939). The court thus finds that the highest and best use of Tract II was timberland, at $250 per acre. The total before value was thus $29,500.
As to the after value of Tract II, Davis testified that only a portion of Tract II was affected by sedimentation and repeated flooding. He therefore discounted the cropland portions of Tract II by 25 percent. The court disagrees. Davis saw the property on two occasions only, both in the month of November, 1990. Tract II was not as wet then as it has been on numerous other occasions, including the time the court viewed it. An aerial photograph taken in October 1988 also shows that sand appears to be deposited on approximately half the tract. Photographs taken in 1985 *460clearly show that well over half the tract is either under sand or water. More recent on the ground photographs, as well as the court’s inspection, suggest that the parcel is virtually completely affected. Defendant did not put on evidence that the land could be productive, even at 75 percent of former productivity. There seems to be no basis for finding that less than all 118 acres are affected. The easement is thus imposed on the entire parcel.
Beard’s report and Spencer’s testimony indicate a post-taking value of $100 per acre, for a total of $11,800. The court is not inclined to question this figure. Portions of the land may ultimately be suitable for softer hardwood growth, or as a hunting wetland, but those values are highly speculative at this point. The total diminution in value as to Tract II was thus $17,-700.
CONCLUSION
The court concludes that plaintiffs have demonstrated that authorized actions of the Government resulted in the taking of an easement on their property for periodic inundation through flooding. The imposition of that easement resulted in a diminution in value of $224,920. Plaintiffs are entitled to recover that amount, plus interest from the date of taking to the date of payment, along with attorneys’ fees and expenses under the Uniform Relocation Assistance and Real Property Acquisition Policies Act of 1970, 42 U.S.C. § 4654(c) (1988). The parties are directed to confer on the amount of the judgment and to whom it is to be paid, the wording of the easement to be executed, and the amount of interest, attorneys’ fees, and expenses. On or before August 5, 1991, the parties are to make their report, including whether a deed conferring an easement has been executed. Entry of judgment will be deferred.
. 28 U.S.C. § 1491(a)(1) (1988).
. Turner v. United States, 901 F.2d 1093 (Fed.Cir.1990).
. What will be referred to as Tract I throughout this opinion includes 80 upland acres purchased by Wilburn and his partners sometime prior to the purchase of the balance of Tract I.
. The total purchase price was $380,000. The terms were $50,000 down, and $11,000 per year plus interest for 30 years.
. Roy Doss, a retired County Extension agent, familiar with soils and crops in the area, testified that 17 or 18 bushels is a marginal yield. Average soybean yield for all soils is 22 bushels per acre. Yield in the Twenty Mile bottom tends to be much higher.
. In a report prepared for the Corps by a private firm sometime before September 1981, it is noted that even by that date, "the effective slope [of the designed channel] is now reduced to about half of the original slope, by a combination of upstream head-cutting and downstream deposition. The excessively high velocities induced by the as-constructed slope could be expected to result in severe erosion of fine-grained materials."
. Corps documents reflect that as early as April 1968, the problem of deposition at the mouth of Twenty Mile Creek was being discussed. In a memo dated April 1, 1968 from the Washington Corps of Engineers headquarters to the South Atlantic Division Engineer, concern was expressed that sediment coming from the recently channelized tributaries of the East Fork would overwhelm the limited sediment carrying capacities of that stream. It is noteworthy that the memorandum suggests the use of sediment catchment basins, which could be periodically maintained.
. Otto also suggested the somewhat inconsistent criticism of Turner’s gap closing activities that, if the gap were open, then Twenty Mile Creek would have a natural safety valve during floods. This idea is the same one discussed in the Corps report of 1988 in which the author expresses some pique at downstream landowners protecting themselves by closing drainage channels into the creek. In any event, an easement would be imposed although it might be more akin to a complete taking if sedimentation were deliberately funneled onto the property.