OPINION
BRUGGINK, Judge.This is a suit for compensation for the alleged taking of riverbank property and a *575home by the United States Army Corps of Engineers (“Corps” or “Government”). The action is brought by Mr. William Eugene Owen, in his capacity as executor of the estate of his mother, Mrs. Caroline Payne, who is deceased. Plaintiffs claim is founded upon the fifth amendment to the Constitution of the United States. The specific basis for plaintiffs claim is that the Government “inversely condemned” the land and home of Mrs. Payne when it undertook certain improvements along the Tombigbee River in Greene County, Alabama, and that it has failed to pay just compensation for the taking. The court has jurisdiction pursuant to 28 U.S.C. § 1491(a)(1) (1982).1 After trial, the court concludes that plaintiff has not proven that he is entitled to compensation.
FACTUAL BACKGROUND 2
Mrs. Caroline Payne owned a parcel of land on the banks of the Tombigbee River in Greene County, Alabama. See Figure 1. [See Appendix.] The original tract was approximately 125 feet wide along the river, and 300 feet deep. It is lot 14 within a platted parcel known as the McAlpin Subdivision. The subdivision lies just below the outer bank of a curve made by the river. The opposite bank across from the site lies in Sumter County, Alabama. Mrs. Payne purchased her lot in July 1975 for $1500. At the time she purchased the property, it was subject to a flowage easement giving the Corps the right in perpetuity to occasionally flood the property above 76 feet mean sea level (“msl”) and permanently flood it below that level. In 1976, Mrs. Payne built a home on lot 14. (Referred to hereafter as “the Payne property”, or “the site”.) Figure 2 shows the McAlpin Subdivision. [See Appendix.]
The Tombigbee River flows from northeast Mississippi through west Alabama. It joins the Black Warrior River in the vicinity of Demopolis, Alabama, approximately 36 miles downriver from the Payne property. Further downstream, the Tombigbee joins the Alabama River before entering Mobile Bay. The river is over 400 miles in length. That portion of the river above Demopolis is known as the Upper Tombigbee.
From the very early days of settlement of the Upper Tombigbee region, there have been efforts by the federal government to improve the navigability of the river. During the wetter months of the year, the river was naturally navigable by large riverboat traffic approximately to Columbus, Mississippi, 85 miles upstream of the site. The natural channel of the Upper Tombigbee was not navigable by larger vessels during the summer months, however. Studies by the Corps were commissioned to evaluate the possibility of making the upper stretches of the river available to larger ships year-round, and Congress made numerous appropriations for dredging and snag removal to facilitate traffic. It is not *576clear from the record whether those efforts succeeded in making the Upper Tombigbee navigable for larger traffic year-round. It is clear that the Upper Tombigbee, including that portion adjacent to what later became the Payne property, was used for riverboat traffic during high water months until approximately the turn of the century. There was a small riverboat landing a few miles downstream of the site at Miller’s landing.
Two major Corps projects on the Upper Tombigbee are of significance to this action. The first was the completion in 1956 of the Demopolis lock and dam. Prior to that time the river had been free-flowing in the vicinity of the site. After completion of the Demopolis lock and dam, the river was impounded upstream of Demopolis beyond the area of the Payne property. The effect of impoundment was to raise the level of the river in the vicinity of the site by approximately five feet. Normal pool elevation above Demopolis became 73 feet msl. Nathaniel McClure, Chief of Planning of the Mobile District of the Corps, testified that the purpose of the 1956 impoundment was “to provide navigation on the Tombigbee Waterway.” Because this action of the Corps figured significantly in the testimony, it is worth pointing out that plaintiff does not and could not claim a taking by the Corps through that project, as this action was commenced more than six years after the lock and dam were completed. In addition, contemporaneously with completion of the Demopolis lock and dam, the Corps purchased, through condemnation proceedings, the flooding easements previously mentioned, which gave it the right to permanently “flood, overflow and submerge” below 76 feet msl at the subject property, and occasionally to flood the property above that level.
The second project was dredging and excavation work done in connection with the completion of the Tennessee-Tombigbee Waterway Project (“Tenn-Tom”). In order to shorten travel time between the Tennessee River and the Gulf of Mexico, Congress in 1946 3 authorized initial construction to link the Tennessee and Tombigbee Rivers and to improve the resulting shortcut to the Gulf. Actual construction did not begin until 1971, however. The project called for a canal to be built between the extreme upper reaches of the Tombigbee and the Tennessee River. The project also called for a new series of locks and dams as well as extensive dredging and some widening of the Tombigbee. The total rise to be accomplished in ten locks was 341 feet.
One component of the Tenn-Tom was the Demopolis Lake Navigation Channel, a stretch of the river which embraces the site in question. In 1976, design of that channel was approved, and in 1976 and 1977 the Corps conducted dredging operations in the vicinity of the site. For Corps construction planning purposes, this work was referred to as “Cut Area Seven.” As was the case along the entire length of the Tombigbee, the purpose of this particular work was to create a minimum channel 300 feet wide and 9 feet deep, capable of handling two-way barge traffic of 8-barge tows. In many places the natural channel was already that wide and deep.
Extensive surveys were made of the river bed to determine where dredging had to take place. It was determined that some deepening and widening was necessary in Cut Area Seven. Plans called for dredging to take place primarily along the west (Sumter County) bank of the river. No dredging was required on the east bank adjacent to the McAlpin Subdivision. In addition, it was determined that the channel would be further widened at the bend just upstream of the site. The river at that point takes a turn from a southeasterly direction to more of a southwesterly direction. As explained in expert testimony, this is due to the intermittent presence in the banks of material known as Selma chalk. This chalk is harder and more resistant than the alluvial soils generally present along the riverbank in this area. As the river encounters the chalk at a place *577referred to by the witnesses as the Boligee boat ramp, it curves to the south-southwest. Through dredging, the Corps excavated a portion of the inside of the curve formed at this point to make the turn less pronounced and easier to negotiate for barge traffic. No new interests in land were acquired by the Corps at the subject property in connection with this work.
The complaint alleges that the Corps modified the course of the Tombigbee by “dredging a large amount of the riverbed and the bank upstream and across from the Plaintiff’s property.” Plaintiff contends that the Corps’ actions had the effect of shifting the flow of the river somewhat. Instead of a strong flow of current impacting just below the Boligee boat ramp, plaintiff contends that the primary force of the current shifted downstream so that the velocity of the river increased along the bank at the Payne property. The asserted result of this change is that erosion significantly increased, undercutting the bank.
There is no question that in the period between 1978 and the present, 100 feet of lot 14 closest to the river has eroded. Other lots above and below the Payne lot were also affected, though none as severely as lot 14. (See Figure 3, which is a comparison of the 1974 survey with a 1990 post-erosion survey.) [See Appendix.] What was once solid ground on a bluff approximately 40 feet above the river edge has now collapsed and been washed away by the river. The Payne house, once about 40 or 50 feet back from the edge of the bluff, collapsed into the river sometime after November 1982. At the request of the plaintiff, and with the agreement of the Government, the court visited the site. All that remains of Mrs. Payne’s home are a few concrete slabs along the river edge and some pipes dangling out of the eroded bank.
The Government acknowledges that serious erosion took place after Mrs. Payne built her home. It contends, however, that this was a natural phenomenon, not caused by the Corps’ activities. The Government’s position is that the riverbank in the immediate vicinity of the Payne property has been eroding over thousands of years as the river channel slowly moves east at that point, and that it will continue to do so until it reaches resistant soil, such as Selma chalk. It contends that other phenomena have hastened that process. The immediate cause of the collapse, it contends, was a massive and virtually unprecedented flood along the Upper Tombigbee which took place in April, 1979. A secondary cause it points to is the prior disappearance of a protective sandbar from the toe of the Payne property, and the accompanying shift of the primary river channel to the east bank.
Plaintiffs witnesses
Plaintiff presented four witnesses who could testify to conditions at the Payne property before and after the 1976-1977 Corps dredging. Mr. Buddy Lavender moved to Boligee, Alabama, the closest community to the property, in 1939 at age 11. Mr. Lavender has fished the Tombig-bee near plaintiff’s property all his adult life. At one point, he was a commercial fisherman, using trot lines. Mr. Buster Payne, presently 68 years old, has lived in Sumter or Greene counties all his life. He is the brother-in-law of Mrs. Caroline Payne. His first recollections of the Payne property go further back than 1940. In 1975 he moved to a house in the McAlpin subdivision approximately 600 feet upriver from the Payne lot. The plaintiff, Mr. William Owen, moved to Boligee when he was six years old. He is presently 47. His recollection of the river goes back earlier than 1956. Mr. Walter Aman has lived in Boligee since 1929, and is familiar with the site from frequent fishing and boating trips.
These witnesses described the condition of the Tombigbee at the site during three distinct time periods: before the closure of the Demopolis Lock and Dam in 1956; before the completion of dredging in cut area seven in 1977; and after 1977. With respect to the pre-1956 period, Mr. Lavender described the river as approximately 90 to 100 feet wide. During low water, the river would be about 2 to 3 feet deep. There were few pleasure boats on the river at *578that time, but many people fished the river in skiffs. He and others also described a physical feature of the riverbed during this period which figured prominently in subsequent expert testimony. This was a large sandbar on the east (Greene County) side of the Tombigbee at and slightly downstream of the site. From aerial photographs and maps, it is obvious that the bar was over 1000 feet long and over 400 feet wide at its widest point. The bar was noted on navigation charts and referred to as the Epes Bar. See Figure 4. [See Appendix.]
The Epes Bar was of local interest for at least two reasons. First, it was a beach on which people from the area gathered for swimming, cookouts and boating. In the 1940’s a boat ramp had been cut through the bluff at a point immediately south of what later would become lot 14, and this gave access to the bar for cars, boats, and trailers. The bar was also important in that the Sumter County and Greene County road departments excavated sand and gravel there during the summer months. Mr. Lavender testified that a “lot” of sand and gravel was removed using front end loaders and a dragline. Excavation occurred all over the bar. Despite the quantities removed, he explained that the bar was always replenished during floods in the winter months so that it would reappear fully the following summer when the river level dropped.
During this period, the deepest part of the Tombigbee was always on the west side of the Epes Bar. The thread or deepest part of the river channel was referred to by experts as the “thalweg.” The thalweg of the river was, therefore, on the opposite side from what later was designated as lot 14. This is confirmed by early charts showing cross sections of the riverbed. The earliest chart in evidence is dated 1914. By chance, one of the cross sections taken over the years bisected the Epes Bar and the Payne property. This is also reflected on Figure 4, taken from a 1937-38 topographic map.
Mr. Lavender was very positive in his testimony that there was no serious bank erosion anywhere along this stretch of the river throughout the period before 1956. This is consistent with the testimony of the other local fact witnesses. Banks were heavily vegetated, and as the river rose and fell, they were left undisturbed, even if the river went over its banks. Some of the trees on the slopes were substantial.
The Demopolis Lock and Dam was completed in 1956. It was in connection with this work that the Corps purchased the permanent and periodic flooding easements affecting the site. After the Demopolis project, the river level rose permanently about five feet, flooding the Epes Bar. Mr. Lavender testified that he believed the bar was still present, though submerged, for some time after 1956. He drew this conclusion because his fish traps were resting on gravel rather than mud at that location. Mr. Aman testified that, around 1960, a dredging company from Mobile, the Rat-cliff Company, had a contract with the Corps that permitted it to mine gravel in the Upper Tombigbee. He recalls that Rat-cliff mined in the area of the Payne property during one summer.
After 1956 an island developed on the west side of the river opposite the Payne property. The island rose in what had once been the thalweg of the river during the existence of the Epes Bar. Over time this island grew, and eventually connected itself on the upstream end to the bank, forming a slough. It has, over time, become vegetated and firmly fixed. This is shown on Figure 1.
Mr. Owen testified that there had been no serious erosion prior to 1977. There may have been isolated small collapses of the bank, but not the extensive erosion which took substantial portions of lots 10 through 19 in the McAlpin Subdivision beginning in 1978. Mr. Owen testified that during the period 1978, about three feet of the bank eroded. The erosion accelerated in 1979 and continues to the present.
All of plaintiff’s fact witnesses agreed that after the dredging, the main force of the river current seemed to shift downstream approximately 200 yards. Whereas it had been hitting the east bank just below *579the Boligee boat ramp, it moved south to the area of the Payne property. The witnesses used several expressions to refer to what they saw. Mr. Payne referred to a change in the “pitch” of the river. Mr. Lavender said the “current” seemed to be hitting the Payne property and then “rico-chetftng]” toward the west bank. He also said that the current appeared “[w]ide and much more.” Mr. Owen said he saw more “swirling” in the river immediately adjacent to his mother’s property. Mr. Aman referred to a change in the “course” of the river.
Plaintiff’s only other substantive witness was Mr. Robert Murphy, a hydrologist. He has extensive experience in the study of surface and ground waters. He was employed for 30 years by the United States Geological Survey. Mr. Murphy has done work in the area of stream flow measurement and has used various gauging devices as well as aerial photography. He drew some distinction, however, as did other experts, between hydrology and hydraulics. The former, his area of expertise, involves measurement and quantification of water. The latter focuses on the physics or mechanics of water once it is on the ground or in channels. It deals with the forces and currents created by moving water. Prom the court’s standpoint, hydraulics is the more relevant focus in the case at bar, since plaintiff has framed the ultimate fact issue as whether flow in the Tombigbee was redirected as a result of the dredging.
Mr. Murphy expressed the opinion that a riverbank, such as the one below Mrs. Payne’s house, would tend to get saturated if in direct contact with water. Although not directly linked to Mr. Murphy’s subsequent conclusions, apparently plaintiff presented this fact to show that the bank was generally weakened by sustained contact with the river. Mr. Murphy was also of the view that the lower level of the bank of lot 14 is more saturated now than it was when the sandbar was present to shield the lot to some degree.
Mr. Murphy’s primary conclusion,4 with respect to the direct effect of the dredging on river currents, was that the widening of the inside of the curve above the McAlpin subdivision redirected the flow of the river toward lot 14. Mr. Murphy testified that he drew this conclusion after studying gauging data, storm data, annual peaks and flows, and river elevations. After considering his testimony, however, and examining these background materials, it is not at all clear how these sources lead him to that conclusion. The court is unable to understand how river gauges, rainfall data, or discharge volumes explain where the primary point of impact of the current is at any given place in the river. Mr. Murphy did no site-specific analysis. There was no empirical evidence concerning water direction or velocity.
When asked how he could tell the current was impacting in the vicinity of lot 14, Mr. Murphy responded that he could tell by observing the water. Like other fact witnesses, however, he did not explain his visual observations. The only witness who used any descriptive term in this connection was Mr. Owen, who referred to the presence of swirls in the river. The court is persuaded that the witnesses observed something about the river from which they reasonably concluded that water was moving faster at lot 14 than it had in the past. The lack of precision and the absence of any empirical analysis, however, makes it difficult for the court to weigh this testimony and compare it to contradictory expert testimony put on by defendant.
Defendant’s witnesses
Defendant presented five expert witnesses in an effort to establish its competing theories as to why lot 14 collapsed. Mr. Robert Earhardt is a meteorologist with the Corps. He is expert at analyzing clima-*580tologic data. He explained that an almost unprecedented flood occurred along the Upper Tombigbee during April 1979. The flood was associated with an extremely rare period of heavy rainfall between April 11 and 13 in the watershed to the northwest, although there had been several earlier heavy storms during the preceding six weeks. During April 11-13 a major storm line stalled for several days. Mr. Earhardt presented charts on which rainfall contour lines were superimposed on a map of the area. These contour lines, called isohyetal lines, demonstrate that during the three day period, up to 21.2 inches of rain fell in the headwaters of the Tombigbee. Using historical data, Mr. Earhardt stated that the “return period,” that is, the length of time before such an event could be expected to reoccur, is approximately 500 years. Mr. Earhardt referred to the rainfall upstream of the site as “spectacular”, and a profound and rare event. In the immediate vicinity of the Payne property, rainfall was not as heavy. There, approximately 10 inches fell in that same period. An 8 to 10 inch rainfall could be expected to occur once every 100 years in that area. The average annual rainfall in the area is 58 inches.
Mr. Harvey Blakeney is a geotechnical consultant. His expertise, developed in part after 28 years with the Corps, is in soils mechanics — the engineering properties of soil. He visited the site three times in connection with investigating plaintiff’s claim. He took soil boring samples and evaluated the character of the riverbank. He pointed out that there are occasional outcroppings of Demopolis chalk along the riverbank. The top of the Demopolis chalk at lot 14 is one or two feet above pool elevation. None is visible above that level at lot 14. The soil above the chalk ranges from sandy clay at the low end to silty sands and topsoil. A heavy outcropping of Selma chalk is present about 300 feet upstream from the Payne lot, on the same side of the river. Chalk is much more resistant to erosion and its presence accounts for the curve at that point.
Mr. Blakeney described the soils of lot 14 as relatively pervious and friable. They would erode easily. He also noted an unusual feature of the surface of lot 14. There is a large depression, in approximately the middle of what remains of the lot, which he testified was apparently caused by the 1979 flood.
Mr. Blakeney opined that a number of phenomena, in some combination, contributed to the collapse of the riverside of lot 14. He noted first that the disappearance of the sandbar prior to the dredging, with the consequent shift of the thalweg to the east side of the river, exposed the toe of lot-14 to erosion. This is consistent with the observations of Mr. Murphy. Mr. Blake-ney also observed that the soil at the surface of the lot was weakened by the removal of small trees, bushes and vines, in preparation for construction of the Payne home. This had the effect of eliminating a root mass holding the surface together at that location. These factors were then acted upon by the April 1979 flood. He stated that the drawdown after flood peak was very fast, and that the bank, already weakened from beneath, and consisting of naturally friable soil, was unable to support its own weight due to saturation. He stated that the dredging of the channel in the vicinity had no effect on this process, since it did not cut into the toe of the slope.
Mr. Blakeney made his second trip to the site at a time when the river level was elevated. He observed eddies or whirlpools all over the surface of the river, but no particular turbulence adjacent to lot 14.
Mr. Kenneth Underwood is a hydraulics engineer with the Corps. Like Mr. Murphy, he drew a distinction between hydrology and river hydraulics. Mr. Underwood began by giving a synopsis of the immensity of the April flood in terms of river volume. At the height of the flood, on April 18, 1979, the river was discharging 314,000 cubic feet of water per second (“cfs”) at Demopolis. This was far above the previous historic high flow of 250,000 cfs. (Data at Demopolis has been collected since 1929.) At Demopolis, the river crested at 24 feet above flood stage. The highest river elevation reached at the site was 111 feet msl. Normal pool elevation of the *581Demopolis channel is 73 feet msl. This meant that the Payne home was at least six feet deep in water. Up river from the site, at Gainesville, the previous 100 year high flow was 201,000 cfs. During the April flood, the flow reached 257,000 cfs. In Underwood’s opinion, the flood of April 1979 was one that would only reoccur every 300 to 400 years.
Mr. Underwood gave expert testimony on a number of key points. He used aerial photographs to point out previous erosion of the riverbank, and to demonstrate the eastward migration of the river channel at this location. With respect to the latter point, the court could observe from photographs the progression of parallel crescent shapes indicative of deposition of soil from earlier locations of the channel. This was confirmed also by Mr. Phillip Lamoreaux, whose testimony will be discussed below. Mr. Lamoreaux indicated, however, that the geologic and hydrologic phenomena thus observed were very gradual, on the order of “tens of thousands of years.”
Mr. Underwood also testified that he could observe evidence of bank erosion at the site of the Payne property or in the immediate vicinity from aerial photographs. Using paired photographs viewed through a stereoscope, Mr. Underwood stated that irregular small white scallops along the riverbank on the east side of the river were definite evidence of significant erosion pri- or to 1976. Only in the 1937 photographs was Mr. Underwood able to detect erosion directly at the Payne site. He explained that when the bank collapses due to undercutting from the river it tends to fracture downward in a crescent shape, which leaves a scar visible in an aerial photograph. The court viewed one set of slides through the stereoscope and was able to observe the shapes referred to by Mr. Underwood. It would have to accept his judgment, however, as to what caused the phenomenon. Mr. Murphy, plaintiff’s expert, examined a set of photographs and stated that there was no apparent damage. He conceded, however, that although he has used a stereoscope, he is not trained in stereo viewing.
Using cross sections for 1914, 1937 and 1971, as well as aerial photographs, Mr. Underwood demonstrated that the thalweg of the river has moved from the west to the east side of the channel at the site. This shift took place at least by 1971, significantly prior to the dredging complained of in this action. This is shown by a comparison of Figures 1 and 4.
Finally, Mr. Underwood testified that, in his opinion, the widening of the inside of the curve above the site could not have redirected the flow toward lot 14. If anything, according to Mr. Underwood, making the curve wider would decrease the overall downstream velocity. As a matter of hydraulic principles, increasing the area available to the river for negotiating the turn would have the effect of slowing it down.
Mr. Phillip Lamoreaux, a consulting geologist and hydrologist, also testified for the Government. He is expert in the area of geomorphology, which includes the study of the effects of erosion. He is the principal in P.E. Lamoreaux and Associates, a geology and hydrology consulting firm with a wide practice. Mr. Lamoreaux held the position of State Geologist in Alabama from 1961 to 1976. His firm was employed by the Government to conduct a study of the hydrology and geomorphology of the site. A team was assembled that included a hydrologist, a hydrogeologist, an expert in photo geology, and a technician to place test wells. The team dug auger holes to evaluate the character of the soil down to the top of bedrock, and placed peizometers in wells to measure the water table. It also evaluated existing surveys, cross sections, aerial photographs, and other data.
Mr. Lamoreaux testified that the auger holes showed that lot 14 is part of a broad flood plain composed of alluvial materials deposited by present and previous movements of the Tombigbee River over underlying chalk. Most of lot 14 is composed of sedimentary materials all the way down to the river level. It is 49 feet from the top of the bluff to the beginning of the chalk layer at the edge of the lot. He noted, however, that there is a buried bluff of chalk beneath lot 14, so that the chalk level *582is considerably lower on the river half of the lot than on the other end. Lamoreaux explained that this reflected an old meander scar of the river.
Mr. Lamoreaux concurred in Mr. Underwood’s assessment that the thalweg of the river made a significant shift between 1937 and 1971. He testified that the progress of the river to the southeast was inevitable over a long period of time; that eventually it would move to the limits of the chalk escarpment.
It was Mr. Lamoreaux’s opinion that the rapid erosion of lot 14 was due in part to the actions of the April flood. Relying on Corps data on river elevations, he concluded that the river elevation in the vicinity of the site dropped rapidly in the days immediately following the flood peak around April 16. The alluvial soil of lot 14 and immediately adjacent lots was completely saturated from days of exposure to rain and flooding. So long as the river was at flood stage, however, the water in the river gave some support to the bank of lot 14. It was his surmise that as the river level dropped, the soil of the bluff collapsed due to the presence of water in the soil and the natural tendency of the groundwater to exit along the face of the riverbank. When combined with the disappearance of the Epes Bar and the shift of the thalweg, the bank was no longer able to support itself.
Defendant’s last witness was Mr. Carroll Winter, Supervisory Civil Engineer with the Corps in Little Rock, Arkansas. Mr. Winter has been with the Corps for 26 years, including an assignment to the Mobile Division from 1971 to 1979, when he worked on geotechnical aspects of the Tenn-Tom. After the April flood, Mr. Winter toured the Tombigbee by boat in the vicinity of the Payne property, checking for damage. He observed bank failures similar to those at lot 14 in several locations hundreds of yards above and below the site.
DISCUSSION
It has long been settled that the Commerce Clause gives the federal govemment extensive powers to control navigation in the nation’s waterways. Kaiser Aetna v. United States, 444 U.S. 164, 173, 100 S.Ct. 383, 389, 62 L.Ed.2d 332 (1979); Gibbons v. Ogden, 22 U.S. (9 Wheat) 1, 6 L.Ed. 23 (1824). In Gilman v. Philadelphia, 70 U.S. (3 Wall.) 713, 724-25, 18 L.Ed. 96 (1866), the Court wrote that “Commerce includes navigation. The power to regulate commerce comprehends the control for that purpose, and to the extent necessary, of all the navigable waters of the United States which are accessible from a State other than those in which they lie____ This necessarily includes the power to keep them open and free from any obstruction to their navigation____” The power to regulate navigation confers upon the federal government a dominant servitude which extends to the entire stream and stream bed below the ordinary high-water mark. Federal Power Com. v. Niagara Mohawk Power Corp., 347 U.S. 239, 249, 74 S.Ct. 487, 493, 98 L.Ed. 666 (1954).
The navigational servitude does not extend beyond the ordinary highwater mark. Consequently, when fast lands are taken, compensation must be paid. United States v. Rands, 389 U.S. 121, 123, 88 S.Ct. 265, 267, 19 L.Ed.2d 329 (1967). Drawing a distinction between fast lands and lands subject to the navigational servitude can be problematic, however, as this case demonstrates. As the Federal Circuit explained in this action, there must be horizontal limits to the extent of a riverbed, or else imposition of the navigational servitude up to the ordinary high-water mark theoretically could extend indefinitely. It found that limit to be “the land beneath and within the navigable stream’s high water mark.” Owen, 851 F.2d at 1410. Because the land affected in the case at bar is outside the riverbed thus defined, the court explained that the Corps could not rely on the navigational servitude as a complete defense. It held that the permanent washing away of fast land as a result of government construction to alter5 the flow of the river can constitute a taking.
*583There is no question that the Payne home and much of the land eroded was horizontally beyond and vertically above both the navigational easement and the permanent flowage easement.6 Consequently, if plaintiff has shown that the damage to the home and lot were the consequence of the dredging that occurred in 1976 and 1977, he may recover.7
Plaintiff bears the burden of proving that the actions of the Corps were the direct and proximate cause of damage to the Payne property. Sanguinetti v. United States, 264 U.S. 146, 149-50, 44 S.Ct. 264, 265, 68 L.Ed. 608 (1924); Owen, 851 F.2d at 1418; M.R.K. Corp. v. United States, 15 Cl.Ct. 538, 545 (1988); Baskett v. United States, 8 Cl.Ct. 201, 210 (1985), cert. denied, 478 U.S. 1006, 106 S.Ct. 3300, 92 L.Ed.2d 714 (1986). To paraphrase the test of Tri-State Materials Corp. v. United States, 213 Ct.Cl. 1, 8, 550 F.2d 1, 5 (1977), the issue is whether plaintiff has demonstrated that, but for the dredging in 1976 and 1977, the property would have been undamaged. See also Sanguinetti, 264 U.S. at 150, 44 S.Ct. at 265 (damages must be the “direct and necessary result of the structure”); Loesch v. United States, 227 Ct.Cl. 34, 47, 645 F.2d 905, 914, cert. *584denied, 454 U.S. 1099, 102 S.Ct. 672, 70 L.Ed.2d 640 (1981) (erosion due to floods, which would have occurred absent the dams).
Plaintiff’s theory of the case
In evaluating the evidence presented, it becomes obvious that plaintiff is relying in some measure on a post hoc ergo propter hoc (after an event therefore because of the event) analysis. Plaintiffs witnesses testified that, prior to the dredging there was no erosion, but after the dredging there was significant erosion. Consequently, the conclusion offered is that the erosion was caused by the dredging. That type of analysis, however, can easily lead to wrong inferences if there are other elements of causation at work. A number of decisions of this court and its predecessor court have rejected a post hoc ergo propter hoc approach to proving causation in circumstances similar to the ones at bar. Loesch, 227 Ct.Cl. at 45, 645 F.2d at 914 (allegation of erosion caused by construction of locks and dam on Ohio River); Yazel v. United States, 118 Ct.Cl. 59, 72, 93 F.Supp. 1000, 1003 (1950) (collapse of riverbank after construction of levee); Baskett, 8 Cl.Ct. at 210-11 (erosion on river bank allegedly caused by high-lift dam); Rhoads v. United States, 6 Cl.Ct. 278, 280 (1984) (testimony that there was no erosion prior to construction of dam found not persuasive).
Loesch and Baskett present circumstances remarkably similar to the present case. There the courts were presented with factual testimony that prior to construction of locks and dams on the Ohio River there had been no erosion. The parallel to Loesch includes the fact that the expert for the plaintiffs in that case was not an expert in river mechanics or hydraulics. He had been engaged by the plaintiffs 3 or 4 weeks prior to trial. He presented the conclusion that the locks and dams in question were the sole causative factor, without supporting empirical data or study, and without considering other possible factors. The court noted that “opinion evidence is only as good as the facts upon which it is based,” Loesch, 227 Ct.Cl. at 46, 645 F.2d at 915, and found the opinion not helpful. In both cases, the courts found that erosion on rivers is an extremely complex matter in terms of causation, and that greater weight should therefore be given to the testimony of experts. Loesch, 227 Ct.Cl. at 45; Bask-ett, 8 Cl.Ct. at 212.
Mr. Murphy first observed the site “6 or 8 weeks” prior to trial. He did not conduct any tests and did not present any empirical evidence. In fairness to Mr. Murphy, the court notes that, unlike the expert in Loesch, Mr. Murphy did examine hydrologic and climactic data in reaching his conclusion. As discussed above, however, he did not explain in what way his examination of that data aided his conclusion. There was no attempt to link data concerning amounts of rainfall or river stages with the hydraulic properties of the river, and the connection is not obvious. The court is left with the bare assertion that widening the curve and dredging the existing channel caused a change in the way the river current hit the Payne lot. None of plaintiff’s evidence is helpful, therefore, in understanding the processes involved. In this respect it is noteworthy that when Mr. Murphy was asked “[w]hat physical data specifically do you rely upon to assess the changes in the momentum that you’ve described ... [,] changing from upstream of lot 14 to the vicinity of lot 14 ...,” he responded, “[vjisual observation. This would be visual observation of the flow of the current and then standing down at the vicinity of the plaintiff’s property looking back toward the curve. It was obvious that the current was coming to that point.”
The Government’s theory of the case
In examining defendant’s presentation, it becomes obvious that the Government, too, focused very little attention on an examination of the hydraulic effects of the dredging. In an almost incidental fashion, Mr. Underwood offered the opinion that widening the curve, and dredging in other places downstream of the bend, would not have the effect of increasing the volume or velocity of the current in the area of lot 14. He explained that widening and deepening the channel would tend to slow the river by *585giving it a greater catchment. That explanation sounds plausible, however, and stands unrebutted.8
The Government’s primary approach to the evidence was to establish other possible causes of the erosion, unrelated to the dredging. A number of interrelated factors emerge from the evidence: the general trend of the river channel to the east, the disappearance of the Epes Bar on the east side of the channel, the appearance of a sandbar and subsequent narrowing of the channel on the west side, the related movement of the thalweg to the east bank, erosion caused by the flood of April, 1979, and collapse of the riverbank caused by the rapid decline of waters after the April flood. One possibility is that these factors operated in some combination with each other.
The court has the benefit of numerous aerial photographs that include the site. There is one photograph for each of the following years: 1937, 1941, 1950, 1955, 1959, 1965, 1968, 1979, 1981, 1982, 1985. In addition, there are maps of the area showing the river channel in 1936-37,1954, and 1974. Also in evidence are the results of cross section surveys done at the site by the Corps in 1914, 1937, and 1971. These exhibits are relatively large scale and, when viewed sequentially, provide a picture of what occurred during the period between 1914 and 1985. The most obvious change in the river channel which emerges from these exhibits is the dramatic narrowing of the channel directly adjacent to the McAlpin subdivision. As confirmed by plaintiff’s witnesses, during the 1930’s and 1940’s there was a very large sandbar immediately adjacent to the bank at what later became lots 13 through 25 of the McAlpin subdivision. The map from 1936-37 reflects a sandbar over 1000 feet long and approximately 400 feet wide at its widest point. Witnesses testified that the sandbar was attached to the east bank. One could walk or drive from the bank directly onto the bar, unless the river was at high stage.
Cross sections of the river for this same period of time show that the deepest part of the channel was on the west, or Sumter County, side of the river. Photographs from 1950 and 1955 show the river at high stage. The bar is not visible, although Mr. Payne testified that it was present during this time. Another characteristic of the channel at the McAlpin subdivision also becomes apparent from these photographs, as well as the photograph taken in 1959, and the map from 1954. The channel was much wider at the site than at any nearby location. It is approximately three times as wide as the river channel at Boligee boat ramp.
In the photographs taken in 1950 and 1959, apparently during high water, there is evidence visible of a small island in what used to be the thalweg of the river. The photograph taken in 1965 and the map drawn in 1974 both show that by those dates the island had developed into a very large sandbar on the west side of the channel opposite the site. The sandbar is approximately the same length, though somewhat narrower than the previous Epes Bar, which had disappeared by this time. A cross section survey done in 1971 also shows that the channel has shifted from the west to the east side of the channel. At this time there is no longer a sandbar abutting the toe of lot 14. Later photographs show that what began as a sandbar on the west side eventually became vegetated and substantially fixed to the bank. The channel alignment map, prepared in 1976, as well as aerial photographs after *586that date, show that the channel at the McAlpin subdivision is no longer wider than at other points along the river in the vicinity.
Several of defendant’s witnesses confirmed the shift of the thalweg and narrowing of the channel. The significance of these developments was best explained by Messrs. Blakeney and Lamoreaux. So long as it existed, the Epes Bar formed a protective barrier to the riverbank at the site, by directing the primary flow of the Tombig-bee to the other side of the channel. When that barrier was gone, the McAlpin subdivision was directly exposed to the primary force of the current. He went on to testify that this factor weakened the bottom of the slope, and was subsequently exacerbated by other factors: removal of the protective root mass from the top of lot 14, the inherently unstable nature of sedimentary soil, and the April 1979 flood.
By at least 1974, therefore, the current of the river during normal and low periods had shifted. Whereas it had been shunted to the west bank by the Epes Bar in the past, after 1974, the overall channel had narrowed significantly, and the thalweg of the river was now directed along lot 14. It is noteworthy that these developments all predate the Corps activities complained of. When dredging took place in 1976 and 1977, no work was done to widen or deepen the channel adjacent to lot 14. The channel on the east bank was already deep enough.
On this critical issue, the court notes that Mr. Lavender’s explanation of the shift of the current draws more attention to the development of the island and narrowing of the channel than it does to the upstream dredging. When asked what difference he noticed in the river current after dredging, he stated:
[After the river was raised] we had an island out in front there that we played over it. We’d go there and build temporary structures on there, you know, to camp during the summer and to have parties. That closed off that river on the west side. It stopped up both ends of that river and raised a sharp bank there instead of ... a little sloping island out in the river. On the east side it moved that curve to the east side of the river after the dredging.
The impression left is that the development of a substantial island on the west bank is what later directed the current to the east side.
The court notes also that Mr. Murphy’s opinion that the dredging redirected the main current against lot 14 is also linked to the disappearance of the sandbar. In response to a request to explain why he concluded that “the slope of the river has been increased and thereby the velocity has been increased,” he testified that:
[P]reconstruction before dredging the river had sandbars in it. It had curves that were reasonably narrow in width. In the case of the Payne property the water came down around the curve and then was directed over against the opposite bank with a fairly sharp curve to go around the sandbar and gravel bar in question.
During construction the regions of the channel were somewhat straight and certainly the curves were enlarged.9
He clearly assumed in drawing his conclusions, that the remains of the Epes Bar were dredged by the Corps in 1976' and 1977: “The Corps has since dredged out the location where the bar was.” He later explained, however, that he had not seen the 1971 Corps cross sections of the channel for the lot 14 area, and did not realize that the cross section survey showed that the thalweg had already moved by 1971. When asked what basis he had for his assumption that the Epes Bar was still in existence just prior to the 1976 and 1977 *587dredging, he stated that he had “no basis.” During cross-examination he conceded that the thalweg had shifted by at least 1976 and displaced the Epes Bar.
Mr. Murphy’s conclusion that the dredging caused a redirection of the current is therefore subject to question. His opinion presumed that it was the 1976 and 1977 dredging that eliminated the sandbar protecting the east bank of the river at lot 14. In fact the channel had shifted by that time. When combined with his observations that the continuous exposure of the toe of the slope to water caused saturation and weakening, his conclusions tend to support defendant’s theory of the case.
It is not clear what caused the shift of the thalweg of the river and the disappearance of the Epes Bar. The Government presented evidence that there had been extensive gravel mining of the Epes Bar prior to 1956. That mining appears to have been of minimal effect since the bar was replenished every year. There was additional mining by a private contractor, Ratcliff, however, during one period in approximately 1960. Mr. Aman testified that the company was on the river in the vicinity of Boligee for approximately three months. The company used pumps to suction up sand and gravel onto barges, which were then shipped to Mobile. He testified that Ratcliff was getting the gravel “as far down to the bottom as they could.” In sum, it is not clear whether the disappearance of the bar was due to mining or some natural processes. What is clear is that by 1971 the bar was gone and the thalweg had shifted from the west side to the east. At the time of the 1979 flood, therefore, the main channel of the river was directly adjacent to lot 14.
The evidence on the question of whether there was erosion prior to 1978 is equivocal. Most of plaintiff’s fact witnesses testified that there had been no erosion prior to 1977. Mr. Underwood of the Corps explained, however, that certain distinctive shapes on stereo-paired photographs of the area prior to 1977 could only be bank collapse or erosion. While Mr. Murphy disputed this characterization, he, unlike Mr. Underwood, was not trained in evaluating stereo-paired photographs.10 Mr. Payne also stated that one of the landowners immediately upstream of Mrs. Payne had experienced erosion of about 8 to 10 feet of bank prior to the dredging. The court finds that there had been some instances of erosion in the general vicinity prior to 1977, but that they were not as significant as what occurred at the McAlpin subdivision after 1977. Even if it is assumed, however, that there had been no significant erosion prior to the dredging in question, for rea*588sons already discussed, there is still insufficient evidence from which to conclude that erosion from 1978 forward was caused by the Corps. Moreover, the court notes that two witnesses testified that there was erosion in several places along the river after the April 1979 flood. Mr. Winter stated that comparable erosion took place several hundred yards above and below plaintiffs property. Mr. Lavender also testified that there had been erosion “in different places up and down the river.” This phenomenon substantially weakens plaintiff’s claim that the dredging in and below the bend at the McAlpin subdivision caused the damage, but it is consistent with defendant’s argument that the bank collapse was caused by the 1979 flood.
After examining the evidence supporting the various explanations of the collapse, it is apparent to the court that while absolute certainty is not possible, certain conclusions can be drawn. Initially, the court concludes that plaintiff has not established that the 1976 and 1977 dredging was a factor in the erosion. There is simply no credible demonstration of a link between the Corps’ actions and the damage to the property. Plaintiff’s expert testimony was conclusory and premised on critically incorrect assumptions. While the fact witnesses were credible, their testimony was too imprecise and superficial to form a basis for liability. While the fact witnesses were no doubt sincere in their conclusions that the flow only changed after 1978, this testimony is overcome by the incontrovertible evidence that significant alterations had taken place in the channel and were firmly established by at least 1971. The most important of these changes were the shift of the thalweg, coincident with the disappearance of the Epes Bar and narrowing of the channel. Long before the Corps work at issue, the hydraulic characteristics of the river had changed in a way that experts for both parties agreed would affect the banks. In conjunction with the massive flood which occurred in April 1979, and the generally saturated and unstable soil conditions, these changes in the river channel caused erosion to the plaintiff’s property.
CONCLUSION
Plaintiff has not met his burden of proving that the dredging was a cause of the erosion. Defendant established other forces and developments that independently are sufficient to explain the damage. It is not possible or necessary to resolve which factor predominated.11 Given the absence in plaintiff’s presentation of an explanation for why the dredging would *589redirect the main force of the current,12 and an absence of any persuasive evidence linking the dredging with a change in the way the river affected lot 14, the court finds that defendant has offered more persuasive counter explanations as to what caused the damage to lot 14. The Clerk is directed to dismiss the complaint. No COsts.
*590APPENDIX
[[Image here]]
FIGURE 1
1968 Topographic map
*591[[Image here]]
PXGüR® 2
McAlP^-n Subaivision
*592SURVEY MAP OF LOT NO. 14. MCALPINE SUBDIVISION,COMPARING SURVEYS OF 1974 AND 1990.
[[Image here]]
FIGURE 3
Comparison of 1974 and 1990 surveys showing area of erosion
*593[[Image here]]
FIGURE 4
1936-37 TopograPhic map showing location of Epes Bar
. Prior to the filing of this action, the decedent brought an action in the United States District Court for the Northern District of Alabama. The complaint asserted claims under both the Federal Tort Claims Act ("FTCA”), 28 U.S.C. § 1346(b) (1982), and under the little Tucker Act, 28 U.S.C. § 1346(a)(2) (1982). The district court’s dismissal of the FTCA claim was affirmed on the grounds that the Government was insulated from liability for its negligence, if any, by the discretionary function exception. Payne v. United States, 730 F.2d 1434 (11th Cir.1984). The Tucker Act claim for inverse condemnation was in excess of $10,000 and hence beyond the jurisdiction of the district court. Id.
. This claim was initially dismissed. In reliance on Pitman v. United States, 198 Ct.Cl. 82, 457 F.2d 975 (1972), and Ballam v. United States, 806 F.2d 1017 (Fed.Cir.1986), cert. denied, 481 U.S. 1014, 107 S.Ct. 1889, 95 L.Ed.2d 496 (1987), the court held that there was no set of facts upon which plaintiff could recover for government-caused erosion outside the bed of a navigable stream. In reversing and remanding, the Federal Circuit overruled those portions of Pit-man and Ballam upon which the dismissal was based. Owen v. United States, 851 F.2d 1404 (Fed.Cir.1988). The court specifically held that a cause of action can be framed on the assertion that Corps dredging and excavation caused the erosion of lands beyond the bed of a navigable stream. The court remanded for a determination of whether the claim could be factually established. The court preserved for resolution the question of whether the Tombigbee was a navigable stream at the time the actions complained of occurred. The court concludes below that the Tombigbee was navigable prior to the actions complained of. See supra note 7. More importantly, though, the court’s finding that plaintiff failed to prove its contentions renders the issue of navigability nondispositive.
. Rivers and Harbors Act of 1910, 33 U.S.C. § 421 (1982). See generally Environmental Defense Fund, Inc. v. Alexander, 614 F.2d 474 (5th Cir.), cert. denied, 449 U.S. 919, 101 S.Ct. 316, 66 L.Ed.2d 146 (1980).
. Mr. Murphy was also asked questions concerning the effects of stockpiling dredged materials along the banks of the river, and of shortening the river by eliminating Rattlesnake Bend, 25 miles downstream of the site. These issues were not raised in pretrial materials and are considered beyond the scope of trial. In any event, there was no evidence specifically linking those phenomena with a change in the way the current impacted the east bank of the river at the McAlpin subdivision.
. The court distinguished Bedford v. United States, 192 U.S. 217, 24 S.Ct. 238, 48 L.Ed. 414 *583(1904). There the Supreme Court held that property owners downstream of a newly constructed revetment could not recover from the United States for the loss due to erosion of substantial amounts of fast land. In Bedford, the Court found it significant that the construction was intended to preserve the natural course of the Mississippi River. In addition, the effects of the revetment, as against damage already resulting from an impending change in the river course, was conjectural. The Federal Circuit held that, on the facts alleged, Bedford was not a bar here, since it had to be assumed that the dredging altered rather than preserved the flow of the Tombigbee. The parties have not directed evidence or argument at this question. After trial, the court concludes that a different result should not obtain. While dredging arguably defines more clearly an existing channel, there was no contention that the construction at issue here was intended to protect against erosion or flooding. The Government was not attempting, as it was in Bedford, to arrest an errant river. The dredging was intended to facilitate commercial use of the river.
. The parties did not attempt to define with precision the relationship between the navigational servitude and the Corps easements. The difference in imposition on the Payne property between the two dominant estates is not important, however. The damaged property was above 76 feet msl, and the occasional right to flood beyond that level would not insulate the Corps from liability for a permanent taking.
. The Federal Circuit held that "a successful demonstration by the [plaintiff] that the river was nonnavigable prior to construction would be a valid basis for upholding the [plaintiffs] general claim that the government's navigational servitude would not foreclose recovery of compensation by the government.” Owen, 851 F.2d at 1417. In view of the court’s primary holding that the possibility of liability existed despite the navigational servitude, it is not clear how a finding of nonnavigability would further aid plaintiff. In any event, plaintiff has not demonstrated that the river was nonnavigable prior to 1977, and defendant has demonstrated the contrary.
Defendant correctly points out that the Supreme Court has held that a stream is not non-navigable merely because artificial aids must be employed to make it navigable. United States v. Appalachian Electric Power Co., 311 U.S. 377, 407, 61 S.Ct. 291, 299, 85 L.Ed. 243 (1940). Navigability has to be assessed in light of the purposes for which it is being asserted. Kaiser Aetna, 444 U.S. at 170-71, 100 S.Ct. at 387-88. Here, the assertion is made in connection with a massive project to facilitate the movement of commercial traffic in a river that was first utilized for that purpose over a century ago. There is a direct connection, therefore, between the assertion of navigability and the acts alleged to constitute a taking.
To be designated as a navigable stream, it is not necessary that the stream be navigable in fact throughout all seasons of the year. Economy Light & Power Co. v. United States, 256 U.S. 113, 123, 41 S.Ct. 409, 413, 65 L.Ed. 847 (1921). Nor does disuse of the stream alter its natural character for navigability. Id. at 124, 41 S.Ct. at 413. The river had been utilized even in its unimproved condition. The Tombigbee was utilized by large commercial riverboats until at least the turn of the century, and there has been continuous Corps activity on the river in terms of snag removal, bridge relocation, sandbar removal, and later on, construction of locks and dams for navigational purposes. Mr. McClure testified that one of the purposes of impounding the river in 1956 was to provide for navigation. Although no specific evidence was presented concerning the nature of river traffic after completion of the Demopolis Lock and Dam in 1956, the court will take judicial notice of the fact that locks are used to facilitate river navigation. There was evidence that the completion of the dam created a lake at least five feet deeper than the natural riverflow. The court concludes that the Tombigbee was navigable prior to the improvements complained of.
. Plaintiff introduced a supplemental general design memorandum used by the Corps in planning the Demopolis Lake Navigation Channel for the purpose of showing that the Corps anticipated sloughing of the banks along the river. As Corps employees explained, and as the court concludes from the context of the highlighted portions of the memorandum, the concern was with sloughing along the banks immediately adjacent to dredging. The erosion in question could not have been caused in that way, since there was no dredging along the bank in front of lot 14. Nor does the Corps letter of June 25, 1979 aid plaintiffs case. While the Chief of the Real Estate Division did state that it was "conceivable” that the dredging in question caused "some alteration in the flow characteristics of the river," this was in the context of a general denial of liability on other grounds.
. There was a suggestion that dredging the entire length of the Upper Tombigbee added to the velocity of the river. There was also testimony that the channel length was shortened by approximately 12 miles by eliminating Rattlesnake Bend. The implications of these circumstances were never fully explained, and were not accompanied with any empirical data. Mr. Murphy did not know the drop in elevation between the upper and lower ends of Rattlesnake Bend, and did not have any data with respect to the effect of eliminating the bend on the velocity of the current. In any event the pretrial materials did not raise these matters.
. Plaintiff objected to Mr. Underwood’s testifying from stereo-paired photographs, on the ground that they had not been furnished in time for plaintiffs counsel or expert to examine them prior to trial. Plaintiff has maintained that objection after trial. The stereoscopic photographs were admittedly not identified during pretrial proceedings by the Government as exhibits, because the Government did not intend to introduce them. They were put into evidence at the court’s direction in order to better preserve the testimony of Mr. Underwood and Mr. Murphy. For several reasons the court declines to reconsider its trial rulings permitting testimony from the stereo-paired photographs. First, Mr. Underwood's testimony regarding predredg-ing erosion is not determinative of the result in any way. The court accepts the fact witnesses’ statements to mean that there was relatively little major erosion prior to the 1970’s. In any event, in the absence of Mr. Underwood’s testimony there would still be insufficient evidence to warrant a finding that the erosion after 1978 was caused by defendant. Second, at the time plaintiff’s exhibits 12A and 16A (stereo-paired photographs) were offered, plaintiff did not object. Third, there was no formal discovery, and there was no failure on the part of the Government to comply with informal or court-ordered discovery. Fourth, plaintiff’s counsel and expert witness were furnished ample opportunity outside the courtroom during the week-long trial to have access to the photographs. Finally, a remarkable lack of diligence and cooperation displayed by lead counsel for plaintiff made pretrial preparation and discovery extremely difficult in this case. (This does not reflect on other counsel for plaintiff.) There were instances in which the court’s orders were not fully complied with. Some of the problems encountered in pretrial are captured in the numerous orders and conferences necessitated by plaintiff’s counsel. In reviewing counsels’ representations and sworn statements concerning events surrounding discovery, the court is inclined to accept government’s counsels’ representations on disputed points.
. Accepting in principle that a dirt bank may shear when it is over saturated, other factors persuade the court that this phenomenon is not the primary explanation for the collapse of lot 14. The first is that, as Mr. Murphy explained in his rebuttal testimony, the drawdown rate of the river in the vicinity of the site was approximately two feet per day in the first five days after the flood, a rate that he felt was not unusual. The drop was more substantial between April 21 and 25, when the river fell approximately 23 feet. The total drop between midnight on April 16 and midnight on April 27 was 38 feet, an average of approximately 3‘/2 feet per day.
The second reason has to do with the disappearance of trees on lot 14. Mr. Owens testified on rebuttal that before the flood there were substantial trees, up to 24 inches in diameter, on the top of the bank. Some of these trees had already started disappearing prior to April, 1979. He came back to the home four or five days after the flood peak and the remaining trees were gone. This is confirmed by aerial photographs taken during the flood. Even at a time when water was still over the top of the bank, no trees are visible at the edge. If the erosion was primarily due to collapse as the river drew down, the trees would have been visible.
In this latter regard, it is noteworthy that Mr. Lamoreaux testified that, in floods of this magnitude, vegetation is not a guarantee against erosion or bank collapse. Apparently significant erosion was taking place during the flood itself, despite vegetation on the bank. It is also uncontroverted that the erosion and bank collapse process continued after the April flood subsided. This suggests that the initial flood damage took place during the flood itself, but that shearing may have occurred later, when the decline of the river level accelerated.
The final reason the court finds that most of the damage to lot 14 occurred from erosion rather than shearing is that the fact witnesses consistently described a continuous erosion process in which the top of the bank, held together by trees, was undercut from below prior to collapse.
. Mr. Charles Ming did not appear on plaintiffs original (December 29, 1988) pretrial list of witnesses. Plaintiff was directed to resubmit a new witness list by order of November 14, 1989, and did so on November 28, 1989. Mr. Ming did not appear on that list. The time for concluding discovery of expert witnesses closed on March 2, 1990. On March 26, two days prior to the pretrial conference, plaintiff moved to amend the witness list to add Mr. Ming and one other expert witness. Amendment was disallowed as untimely and prejudicial to defendant. At trial, plaintiff again sought to put on the testimony of Mr. Ming. Although the court sustained its earlier ruling, it permitted plaintiff to make an offer of Mr. Ming’s testimony. While the court sees no need to amend its rulings, it makes the following observations. First, Mr. Ming’s testimony was not rejected on the basis of a lack of qualification. On the contrary, he is highly trained and experienced in hydrology and river hydraulics. He did not, however, do any specific flow calculations, tests or studies or present any data in connection with the river adjacent to the Payne property. Second, Mr. Ming was not contacted by plaintiffs counsel until six weeks prior to trial. The Claims Court phase of this litigation began on February 4, 1985, ample time for plaintiff to have secured expert witnesses. Third, Mr. Ming was asked a critical question concerning whether the dredging across from the McAlpin Subdivision increased the flow against lot 14. His answer undermines plaintiffs announced theory of the case: “That minor cut doesn’t, but coupled with everything else it does, yes.” Fourth, Mr. Ming’s testimony did not support plaintiffs theory of the case (that dredging in the vicinity of Cut Area Seven caused a redirection of the flow of the current to the bank of lot 14). Instead, Mr. Ming offered the opinion that the "velocities in the channel have increased and so has the discharge" all along the Upper Tombigbee due to the generalized clearing, dredging, and shortening associated with the Tenn-Tom project. As part of his explanation, Mr. Ming relied on data concerning the placement of spoils piles of dredged materials, apparently on the theory that they impede flow on the upper banks of the river. Although the connection to erosion was never made directly, presumably the increased velocity and discharge caused erosion, not just to lot 14, but all along the river. This theory of the cause of the damage to the Payne property is not within a fair reading of the issues raised in the pretrial materials. The parties were warned that issues of relevance at trial would be controlled by the pretrial submissions. Orders of July 5, 1985 and March 29, 1990.