No. 138-67 | Ct. Cl. | Jun 12, 1970
: This case was referred to Trial Commissioner C. Murray Bernhardt with directions to make findings of fact and recommendation for conclusions of law under the order of reference and Eule 134(h). The commissioner has done so in an opinion and report filed on September 17,1969. Exceptions to the commissioner’s opinion, findings and recommended conclusion of law were filed by defendant;
OPINION OB COMMISSIONER
Bernhardt, Commissioner: The issue met here is whether the defendant’s completion in 1960 of the Toronto Dam and Reservoir on the Verdigris River in Kansas created a backwater effect which has since subjected bordering farms owned or rented by the several plaintiffs to intermittent — but permanent and substantial — flooding, as they contend and the Government denies in this action under the Fifth Amendment for a constructive taking, or inverse condemnation as it is sometimes termed. The detailed facts in the report trailing this opinion narrowly but surely conclude that a com-pensable taking has occurred.
The plaintiffs own or rent on crop shares several farms flanking the river from 21.2 to 26.6 miles above the dam. The properties lie in the flood plain of the river and have historically been prone to occasional floods, particularly in the months between April and July when fledgling crops of wheat, alfalfa, corn, milo, oats and soybeans are most vulnerable to damage in their growth cycle. Prior to 1960 floods caused by heavy flows were less frequent and destructive than comparable flows since completion of the dam, even excluding from consideration the unprecedented series of 11 to 13 floods in 1961, which would have been just as injurious to plaintiffs in the absence of the dam. More pointedly, post-dam floods have lingered on the land much longer than before, a condition ruinous to most growing crops. Moreover, the altered flood pattern has impaired the land itself in places by scouring the soil, which condition ditching and diking has failed to prevent.
In preparation for the dam project the Corps of Engineers made abstruse studies which resulted in determinations to
Although there is no reason to doubt the technical ability of hydrologic engineers to forecast backwater effects of a proposed dam within a reasonable margin of error, to do so more data is required than apparently the Engineers had in this instance. Principally wanting was knowledge concerning the contributions of tributaries, adequate hydro-graphic studies of the terrain including elevations of river bottoms, banks and adjacent farmlands in the valley, and inflow measurements usefully near the properties in suit.
An analysis of statistics of post-dam floods compared with those antedating the dam (finding 21), while expectedly not in precise unison, yields the probability that the backwater effects of the dam were underestimated with the result that henceforth the plaintiffs may expect, intermittently but
That some plaintiffs claim a diminution in the value of their fee to properties rented out on crop shares to other plaintiffs claiming a permanent subjection of their tenant crop interests to intermittent but permanent damage does not prevent separate compensation for each of the interests thus fractioned (Cf. United States v. Twin City Power Co., 248 F.2d 108" court="4th Cir." date_filed="1957-09-18" href="https://app.midpage.ai/document/united-states-v-twin-city-power-company-243111?utm_source=webapp" opinion_id="243111">248 F. 2d 108 (CA 4, 1957)), but merely has relevance to the quantum of recovery, as does the nature and extent of the lease interests, which matters will be fit subjects of inquiry when the separated issue of damages arises for disposition.
1. The Toronto Dam and Reservoir were constructed by the United States primarily for flood control purposes on the Verdigris River in Woodson and Greenwood Counties, Kansas. The reservoir began operating as a detention basin in January 1959. The dam and reservoir became fully operational in March 1960 and on March 21, 1960, the water in the reservoir reached the top of the conservation pool. The United States operates and maintains the dam and reservoir.
2. The Toronto Dam is located at river mile 271.5. Its top is at elevation 946 feet, m.s.1.
3. Plaintiffs’ properties flank both sides of the Verdigris River between river mile 292.7 (21.2 miles above Toronto Dam) and river mile 298.1.
4. The determination to limit acquisition of property rights at river mile 292.7 was made by the Corps of Engineers after it concluded, as a result of studies, that the lands situated above that point that might be affected by inundation from the reservoir are within
5. In determining the upper limits of acquisition of property for the Toronto Dam and Reservoir Project the Corps of Engineers made studies of and analysed the possible or anticipated backwater effects of the proposed dam. A report dated April 1955 entitled “Backwater Effects of Toronto Dam” is referred to in several exhibits, but the report itself is not in the record. However, a chart was prepared which shows the results of the analysis. In preparing this chart, which is labeled “Toronto Reservoir Envelope Curve of Backwater Effects Profiles on Verdigris River” (in evidence as Defendant’s Exhibit No. 12), Mr. ~W. T. Moore, one of the defendant’s expert witnesses,
6. After determining the flood size to be used as a basis, the Corps of Engineers plotted a theoretical discharge of the flood through the reservoir in accordance with the proposed flood control regulation plan. The Engineers then selected for analysis three different size inflows entering the pool at varying levels of the pool, and plotted the backwater effects for each. They considered all areas contributing to the Toronto Beservoir Project in the determination of those flows, including Walnut Creek which enters the river from the west at mile 276.1 (4.6 miles above the dam), but lacked data as to the contributions of other minor tributaries above that, including Brazell Creek (at mile 283), Dry Creek (mile286.1), and West Creek (mile 291.3). There are no tributaries between the dam and the Coyville gauge. Nor were complete hydrologic surveys made of the Verdigris Kiver valley, particularly in the vicinity of the properties in suit, so except at a few isolated points elevation data as to banks and elsewhere in plaintiffs’ properties was not available.
7. (a) Using this engineering method it was determined that the water surface level of the inflow with the dam in -place would theoretically approach within one foot of the water surface level of an equivalent inflow without the dam at river mile 288 for a 105,000 c.f.s. inflow, at river mile 290 for a 45,000 c.f.s. inflow, and at river mile 294.8 for an 8,000 c.f.s. inflow. The three points at which each set of the “with and without” water surface profiles virtually merged (i.e., became asymptotic) were then connected in a curved line terminating upstream at river mile 298.6 where the 931 foot elevation of the stream bed conjoined the 931 foot elevation of the flood control pool, which would theoretically be the furthermost upstream point that backwater effects of the dam would reach and is referred to as the point of zero flow. This curve is shown in a yellow -hump on Defendant’s Exhibit 12
(b) According to the standard formula used by hydrologists the discharge of a stream is equal to its cross-sectional area times velocity. Velocity will change because of the change in slope or effective gradient. For example, gradient
8. The backwater envelope curve described in finding 7(a) served as a basis for determining the top of the flood control pool at 931 feet m.s.l., and for the acquisition of property in the construction phase of the project. Thus, the top of the easement taking line up to about river mile 287.4 was set at 936 feet, or five feet above the flat flood control pool area which was at 931 feet, in order to provide a five-foot freeboard for wind and wave action which, with the pool at flood control level, might damage surrounding property due to saturation and erosion, as well as to provide additional storage capacity in the event of a flood larger than a 50-year flood. Above river mile 287.4 the easement taking line follows the line of the envelope curve up to the latter’s apogee of about 939.5 feet at river mile 292.7, and thence curves down to 931 feet at river mile 298.6, the point of zero flow. Theoretically, and assuming the correctness of this backwater effect depiction, and assuming further that the banks of the river did not contain floodwaters, those parts of the plaintiffs’ properties lower than from 931 feet to 939.5 feet in elevation would be affected in certain parts by flood waters passing them in the river at a water surface elevation of between 931 and 939.5 feet, a theoretical maximum flooding differential of 8.5 feet.
(b) An undated dam design analysis states at page 2-b that above a point interpolated to be at river mile 287.5 land would be acquired to the elevation of the envelope curve plotted for the July 1951 flood (shown on Defendant’s Exhibit 12), and says nothing about cutting off the acquisition program at river mile 292.7, which raises the question as to whether this cutoff decision was not made at some subsequent time and represented a reversal of intentions on the part of the Corps of Engineers. Certainly the envelope curve portraying the July 1951 flood elevations shows backwater effects exceeding the 936-foot pool elevation at all points between river miles 287.5 and about 296.7, which stretch of the river is opposite the bulk of plaintiffs’ properties. A numbered series of successive design and real estate memorandums in evidence reflect an intention by the Corps of Engineers to acquire fee titles for properties below elevation
(c) An undated map in evidence as Plaintiffs’ Exhibit L reports that a five-year frequency flood produces a pool elevation of 927 feet which extends 300 feet horizontally beyond the conservation pool level of 901.5 feet, and a flood of 931 feet extends some unstated distance beyond that. While the 927-foot pool level is shown to extend northward to a termination point at river mile 292.7, the flood level of 931 feet is shown to terminate upriver just below river mile 298.
10. The farming practice on the properties in question is to feed to cattle and hogs the grain raised on the land rather than to market the grain as a cash crop. The grain grown on
11. The damage done to a grain crop by reason of flooding varies according to such factors as the length of time the water stands on the cropland, the stage of development of the particular crop at the time of flooding, and the type of crop. Some grains (such as milo) withstand flooding better than others, but none of the grains on plaintiffs’ properties would withstand prolonged flooding.
12. (a) Eight plaintiff property-owners or tenants testified concerning the frequency of floods, and their consequences on crops and lands, prior to and following the completion of the dam in 1960. While there is disparity in their individual recollections of conditions, the collective result of their testimony as to flood frequency and drain-off characteristics is approximately as follows:
Except for the major flood in 1951 which covered all the properties in suit and took about 86 hours to drain back into the river, there were only two or three “nuisance” floods during the decade of the 1950’s, the chief one being in 1958. Generally speaking, that was in part an excessively dry decade except for 1951, as reflected in the rainfall and water statistics (findings 15-17). Following completion of the dam in 1960 there was a series of 11 to 13 floods in 1961, and there were one and sometimes two minor floods each year thereafter through 1967, except for 1966. Many of these post-1961 floods were characterized as “nuisance” floods, although the flood in June 1967 was considered to be somewhat worse than that. The testimony of these same witnesses as to how much longer it took the floodwaters to recede from their properties after 1960 than before varied from 24 to 48 hours longer, two or three days longer, two to four times as long, up to 50 hours, one to one and one-half days longer. Except for the 1951 flood which took about 36 hours to x’etum to the river, these witnesses testified that before completion of the dam in 1960 it would take from eight to 12 hours for floodwaters to recede from their properties.
(b) Plaintiffs’ testimony at trial as to flood frequency before and after completion of the dam in 1961 corresponded roughly with the contents of their responses to interroga
(c) In their answers to interrogatories in June 1966 in connection with Civil Action No. W-3204 then pending in the United States District Court for the District of Kansas (in evidence here as part of Plaintiffs’ Exhibit CC), most of the present plaintiffs described with some particularity the crops lost or damaged by floods in specified post-dam years, referring in their responses to verification consisting of conservation plan maps of the Department of Agriculture, which maps are not in the present record. In their testimony in the instant case all but one of the plaintiffs testified that his crop losses in 1961 were total (one said 50 percent) , but it is reasonable to conclude that similar crop losses would have been suffered in 1961 even without the dam, as happened in 1951 under comparable conditions except that the dam was not then in existence. In 1961 floods were so constant that the ground never dried out long enough to plant or produce crops successfully. Also, in their testimony in the instant case eight of the plaintiffs testified that their crop losses in floods from 1962 through 1967 varied with the
(d) Several of the plaintiffs installed drainage ditches and dikes, and filled lagoons, after the 1961 flood in an effort to improve drainage conditions and restore certain affected areas to farm use. Oil wells and equipment on certain of the properties have been raised on mounds about 15 feet high to protect them from floods. A short section of road on the east of the properties was raised about five feet after the 1961 floods to be made passable by school busses and milk trucks. Prior to 1960 certain portions of the plaintiffs’ properties were washed and scoured by occasional floods, but the scours were deepened considerably by the increased frequency of floods after 1960 and by the occasional backwater effect of the dam and reservoir in retarding the runoff after a flood. Whether these scoured areas are extensive in comparison to the total properties owned or occupied by the plaintiffs cannot be determined. It is expected that some of them are restorable to farm use after they dry out thoroughly. It would seem that the deposit of silt by floodwaters has been more injurious to crops than the scouring effect of the waters has been to the land, since whatever backwater effect exists causes the floodwaters to pond, which, in turn, would tend to promote silting as opposed to the scouring effect of a swift current. At page 3 of Eeal Estate Memorandum No. 2, dated April 4,1955, it is stated with reference to areas of the flood plain below plaintiffs’ properties (probably comparable to plaintiffs’ properties), that “The lowlands are subject to occasional flooding; with the exception of some crop losses, little damage is done, very little scouring or erosion was noted.”
P. 3: “* * * The Verdigris Diver floodplain near Quincy is 1% miles wide and in the upper reaches narrows to 14 to !4 mile. Southeast of Madison it is as much as 2 miles wide in places. * *
P.4: “* * * Flooding and loss of land to the river as a result of scouring and stream channel changes are the chief hazards associated with these [alluvial] soils.” *****
P. 6; “* * * Flooding on the mainstem is more frequently caused by a succession of storms than by single intense storms. The latter type, however, has caused extremely high stages at Madison and Quincy.”
P. 7: “During the wettest period (1940-52) there was a total of 30 major and minor floods or an average of 2(4 floods per year. During the years 1922 to 1957 major floods occurred on the average of one every 1.78 per year. Minor floods through years 1940 to 1952 averaged 1.58 per year.”
[Note. See table of major and minor floods from 1922 through 1957 on P. 7 of Defendant’s Exhibit No. 1.]
P. 8: “It is stated that the flood of July 12, 1951, had a peak discharge of 130,000 c.f.s. at the Coyville gauge, the greatest on record, and ‘about 1% chance of being equalled or exceeded in any one year’.”
$ ‡ ‡ $
P. io: “* * * The 1957 flood was within four to six feet of the 1951 flood crest at Madison, and above it at several points. While there are no statistics on the May 1957 flood, most farmers feel that the land damage equalled or exceeded that of 1951. Because of the time of year, May, the soil was under cultivation and land scouring was extremely severe in the upper reaches of the watershed, and in the lower reaches extreme sedimentation occurred.”
‡ 3;
P. 14: “Periods of zero flow have been experienced during prolonged drouths. Some of the most severe pe*564 riods were: * * * June to September 1953, 1954, 1955, and from July 1956 to March 1957.”
(b) Further evidence to the same purport appears in a preliminary appraisal of water problems in the Walnut-Verdigris unit of Kansas, published by the Kansas Water Resources Board in June 1960, which states:
The relatively flat flood plains are subject to frequent overflows * * * most of the over-bank flows in the unit occur between April and July, at a time when heavy damages will be sustained by the growing crops.
(c) In addition, a map dated in 1956, in evidence as Plaintiffs’ Exhibit E, reflects that the Santa Fe Railroad line, lying in the eastern flood plain of the Verdigris River and a substantial distance back from the river, was under water in 1926.
(d) An April 1939 map of the Verdigris Watershed (Plaintiffs’ Exhibit C) indicates by a dashed line the approximate overflow limits of the river as then known, which overflow area contains most of the properties in issue here.
14. A detailed study of the daily water surface elevations of the reservoir for the years 1961 through 1965, taken from the monthly reservoir regulation charts maintained by the Corps of Engineers, suggests some inconsistency between the known facts of post-dam floods of the properties in issue and the reliability or accuracy of the envelope curve of backwater effects in evidence as Defendant’s Exhibit 12, upon which the Government’s land acquisition procedures were based. By establishing for each of the post-1961 floods the contemporaneous maximum pool elevation, and then superimposing on such elevations the envelope curve of backwater effects of a 50-year, five-day flood, it would appear that theoretically none of the post-1961 floods could possibly have produced a backwater surface high enough to flood plaintiffs’ properties, assuming as stated by defendant’s experts that the reservoir at levels such as those which then prevailed receives and dissipates the inflow swiftly enough to avoid backwater-induced floods. The maximum water surface elevations of the
Fall 1962 -918.68 feet for September, and maximum of 908.12 feet in balance of fall 1962.
196S -905.90 feet, occurring in March.
1964
October -902.09 feet
November -920.15 feet
1965
May -901.66 feet
June -922.96 feet
Bearing in mind the advice of Government witnesses that the upper limit of backwater effects according to the plotted envelope curve marks a point of zero flow, and that at the elevation of 913 feet the upper end of the reservoir intersects the stream bed at approximately river mile 292.7 (the lower limit of plaintiffs’ properties), if, as we are advised, there is negligible backwater effect upstream from the point of zero flow, it is difficult to understand how the floods of 1963, October 1964, and May 1965 could have occurred when the maximum pool elevations were far below 913 feet, as reflected in the above schedule. Moreover, with maximum pool elevations at 918.68 feet (fall 1962), 920.15 feet (November 1964), and 922.96 feet (June 1965), the point of zero flow in each case would be somewhere midway between the upper and lower limits of plaintiffs’ properties, so unless the Government’s theory is in error there could have been no backwater flood effects for the northern half of plaintiffs’ properties on those occasions. Yet, they were flooded. There is, in short, a rough synchronization between post-1960 flooding episodes and contemporaneous pool levels approaching the southern limits of plaintiffs’ properties, which leads one to suspect that when the upper limit of the reservoir approaches or exceeds the lower limits of plaintiffs’ properties, heavy inflows lose their velocity upon reaching the reservoir head and thence tend to spread out over the river banks and onto plaintiffs’ properties. This condition is promoted by the inferred fact that the channel capacity of the river as it passes plaintiffs’ properties is substantially less than in the lower reaches down to the dam.
15. Historically the two most severe floods on the properties in suit occurred in 1926 and 1951. Between 1951 and the
Number of Tear Flows Tear Number of Flows
1950 _ 2 1961 13
1951_ 7 1962 4
1952 _ 1 1963 1
1953-1956 _ 0 1964 2
1957 _ 4 1965 6
1958 _ 3 1966 0
1959 _ 5 1967 3
1960 _ 1
16. The long-term average flow of the Verdigris River at the damsite, both before and since the completion of the dam, is approximately 300,000 acre feet per year. The years 1953 through 1956 were a period of extended drought. In 1953 the annual flow past the present damsite was about 9,000 acre feet, or 3 percent of the average long-term flow. In 1954,1955 and 1956 the flows were between 20,000 to 30,000 acre feet, or less than 10 percent of the long-term average. In 1961 more than 1,000,000 acre feet of water flowed into the reservoir, which is more than three times the average annual runoff based on a long-term average.
17. The average annual rainfall over the Verdigris River Basin which drains into the Toronto Reservoir is approximately 36 inches. In 1961 it was approximately 55^ inches. During 1961 above-normal rainfall occurred over the basin in the months of February, March, May, July, August, September, October and November. In September 1961 from 9.9 to 12.2 inches
18. All of the floods which have occurred in the area involved since the Toronto Dam and Reservoir began operation in 1960 were analysed by the Corps of Engineers. The September 1961 flood had the maximum discharge during this period and produced the highest pool elevation (928.4
19. The Corps of Engineers analysed the September 1961 flood and prepared an envelope curve of backwater effects of the dam and reservoir during that flood, which presentation is in evidence as Defendant’s Exhibit No. 4. The backwater effects of the dam and reservoir in the September 1961 flood above river mile 292.7, as theoretically plotted, are all below the easement-taking line plotted from the theoretical 50-year flood as described in finding 8, supra. The maximum elevation of the reservoir in the September 1961 flood was 928.4 feet, but the backwater effect of the dam and reservoir theoretically produced water surface levels above river mile 292.7 as high as 937.1 feet, an excess of 8.7 feet over the maximum reservoir level. Whether or not this exceeded the elevations of the river banks facing the plaintiffs’ properties cannot be precisely determined since the elevations of the river banks were not completely surveyed, as stated in finding 6, supra. However, it is concluded from the testimony and other evidence in the record that the river banks were topped in many places in the September 1961 flood and major portions of the plaintiffs’ croplands were flooded, as well as sloughs and low-lying areas of marginal value. Scoured areas were further scoured.
20. The basic soil in plaintiffs’ properties is a type of clay that has a very high consistency and is erosion resistant. There is considerable humus in the topsoil and the tilling of such soil makes it more susceptible to erosion by a flood of sufficient magnitude. However, the backwater effects of the dam and reservoir serve to slow down the flow of water entering the upstream limits of the reservoir, so that downstream from that point erosion is decreased rather than increased. Wetness or drynes9 of the soil on plaintiffs’ properties does not appreciably affect floods or flood runoffs.
21. From the foregoing facts it is concluded that whenever the Verdigris Fiver discharges as much as 19,000 c.f.s. at the dam, the Toronto Dam and Reservoir serve to increase the normal backwater effect of the river in the reaches opposite the plaintiffs’ properties so as to top the banks and flood major portions of the croplands, as well as sloughs and
Ultimate FindiNG
22. It is concluded that the Toronto Dam and Beservoir are an appreciable factor in producing a backwater effect on the Verdigris Kiver opposite the plaintiffs’ properties which permanently subjects such properties to (a) substantial intermittent flooding beyond the normal frequency and severity of flooding which existed before the dam was in being, and (b) a longer period for such floods to recede from the properties. These conditions reduce the value of the
CONCLUSION on Law
Upon the foregoing findings, of fact, which are made a part of the judgment herein, the court concludes as a matter of law that the plaintiffs are entitled to judgment in an amount to be determined by further proceedings under Eule 131(c).
All elevation references -will be to “mean sea level”.
Description of the individual properties and interests therein can await treatment of the damage issue after final disposition of the liability issue which was separated for trial pursuant to Rule 47(c) (1964 ed.). [Since September 1, 1969, Rule 131 (e).] These interests, which as to some plaintiffs are fee interests and as to others are crop-sharing tenant interests, are described more particularly in the plaintiffs’ answers to interrogatories filed in May and June 1966 ini Civil Actions W-3204 and 3205 in the U.S. District Court for the District of Kansas, in evidence here as part of Plaintiffs’ Exhibit CC. The answers to the interrogatories establish that plaintiffs, J. C. King and Lodena King, McConnaughey, and Marvin Winter disclaim a talcing of land and restrict their claims of talcing to personal property consisting of crops. Exhibit 1 to the petition describes in detail the real and personal property interests as to each plaintiff.
While the word “within” is used in official reports, it is clear that what is meant is “outside" or “beyond” the limits of backwater effects.
Mr. Moore at the time of trial was Assistant Chief of the Engineering Division, Tulsa District, Corps of Engineers.
According to defendant’s answer to interrogatory 8 (in evidence Rere as part of Plaintiffs’ Exhibit CC), in an earlier proceeding in the U.S. District Court for the District of Kansas, Civil Action No. W — 3205, dated April 27, 1966, the stream bed profile of the river was plotted from low points developed from a survey of channel sections. Plaintiffs’ Exhibit 4 reflects that there were only three such survey points resulting from a 1959 stream bed survey in the entire distance of the properties in suit (5.4 miles, from mile 292.7 to mile 298.1), so the depiction in Plaintiffs’ Exhibit No. 4 of the stream bed profile is far from accurate except at three connected points in a space of 5.4 miles. It is presumed that the bank elevations in that reach of the river were also measured at the same three widely separated points.
The dimensional distortion which appears in the envelope curve of backwater effects in Defendant’s Exhibit 12 is caused by the use of two different scales in the chart, i.e., the horizontal scale is about 2% Inches to the mile while the vertical scale is 2% inches to each 10 feet.
Paradoxically, a map attached to the end of Plaintiffs’ Exhibit BB reflects high -water marks of approximately 960 feet at river mile 298 and 947 feet at river mile 292, at which locations, respectively, the stream bed elevations are shown by interpolation as 930 feet and 907 feet. Nothing else in the record remotely reports high water marks at such high levels, except indirectly the fact (reported in finding 2, supra) that the designed maximum pool elevation was 940.6 feet.
The Blight variance reflects different reporting sources.
This statement was based on the witness’ interpretation of discharge data rather than upon personal observation of the occasions.
The pre-1960 flows of over 19,000 c.f.s. averaged 30,150 c.f.s. and coincided with monthly total rainfalls averaging 8.37 inches per month, while the post-1960 flows of over 19,000 c.f.s. averaged 37,900 c.f.s. and coincided with monthly total rainfalls averaging 7.46 inches per month.