This case comes before the court on plaintiffs’ exceptions to the recommended decision of Trial Judge C. Murray Bernhardt, filed August 10, 1979, pursuant to Rule 134(h), and on plaintiffs’ motion, filed April 21, 1980, for partial remand, having been submitted to the court on the briefs and oral argument of counsel. Upon consideration thereof, since the court agrees with the trial judge’s recommended decision, as hereinafter set forth**, it hereby denies plaintiffs’ motion for partial remand and affirms and adopts the decision as the basis for its judgment in this case. Accordingly, plaintiffs are not entitled to recover, and the petitions are dismissed.
OPINION OF TRIAL JUDGE
BERNHARDT, Trial Judge: Some 112 owners of farms riparian to the Iowa River above and below the Coralville Dam sue under 28 U.S.C. § 1491 for the Fifth Amendment taking by inverse condemnation of portions of their farms *585by means of recurring flooding alleged to be due to the construction and operation of the dam. It is concluded that they are not entitled to recover.
Coralville Dam, a component of the comprehensive flood control plan for the Mississippi River Basin, was placed in operation in September 1958. It is located at river mile 83.3 on the Iowa River, about 5 miles above Iowa City, Iowa. The dam controls an upstream drainage area of 3,084 square miles. The total drainage area of the Iowa River and its major tributary below the dam, the Cedar River, is about 12,640 square miles.
The plaintiffs fall into three groups. First, those upstream from the Coralville Reservoir, known as the "Mar-engo” plaintiffs. Second, those between the dam and the downstream confluence of the Cedar River tributary at about river mile 29 on the Iowa River, known as the "Lone Tree” plaintiffs. Third, those below the Cedar River confluence and river mile 6.2 above the mouth of the Iowa River, known as the "Wapello” plaintiffs.1 The flooding experience of each group varied according to location in relation to the dam. Their common complaint is that, since the construction of the dam and Reservoir, water lingers on their properties for longer periods and interferes with their farming procedures and production. Many complain of subsurface saturation (i.e., excessively high water tables) which kills crops and reduces yields, and of floods occurring later in the growing season than before which harm immature plants at times too late to replant.
Technical evidence establishes that flooding and subsurface saturation increase susceptibility of plants to diseases and parasites, lose soil nutrients through denitrification and leaching, create ponds which inhibit oxygen circulation in the soil and beneficial bacterial activity, delay farm operations and prevent uniform treatment, and lower soil temperatures.
The plaintiffs contend that the primary flood control purpose of the dam has been subverted to favor competing *586recreational and real estate interests of the land fringing the Coralville Reservoir by maintenance of high Reservoir levels designed to minimize unsightly marshy fringes at lower levels, thereby reducing the Reservoir’s storage capacity and requiring higher discharge rates from the dam on occasions of high inflows from heavy precipitation or snow melts. Plaintiffs also charge that the Corps’ regimen of high Reservoir levels — and consequently low storage capacity — is maintained to confer flood protection benefits on riparian owners on the parent Mississippi River into which the Iowa River discharges, to the plaintiffs’ detriment.
In regulating the discharges from the dam and controlling the levels of the Reservoir the Corps endeavors to maximize the efficiency of the one and optimize the storage capacity of the other by a constant juggling process. The regulatory scheme is based on a quantitative precipitation forecast prepared by the River Forecast Center. Rainfall predictions are added into streamflow data by use of a unit hydrograph to determine estimated Reservoir levels resulting from runoff and precipitation directly into the Reservoir. The Reservoir crew is kept constantly advised of imminent flood crests.
By these means the Corps has substantially enhanced the effective storage value of the Reservoir. For example, in 1973 the total river flow at Marengo (above the headwaters of the Reservoir) was 2,732,000 acre feet, or five times the 475,000 acre feet storage capacity of the Reservoir.
In response to complaints from downstream farmers the Army Corps of Engineers (hereafter "Corps”) inaugurated Plan 8 in 1963 and early 1964 for regulation of the dam’s operation, but it did not resolve the complaints. Under Plan 8 the Reservoir elevation starts the year at 680 msl2 and gradually (starting February 1) lowers to 670 msl by February 15, where it remains through June 15. The spring conservation pool level of 670 msl is exclusively for flood control, i.e., to increase Reservoir capacity to accommodate spring floods which customarily occur in the spring months. From June 15 to September 15 the Reservoir elevation is *587maintained at 680 msl. The summer conservation level of 680 msl benefits recreational uses of the Reservoir and insures against drought conditions. To some extent the operation of the dam, originally designed for the exclusive purpose of flood control, has been modified as stated to also benefit competing but subservient recreational and real estate interests.
Although plaintiffs contend that the dam is operated to benefit riparian owners on the Mississippi River to plaintiffs’ detriment, in reality the dam is operated to modify the effects of each flood in sequence as it occurs. No preference is shown to non-Iowa River interests, but on the other hand neither are Iowa River interests favored.
The Corps concedes that the dam lacks ideal storage capacity, but has shown that a larger dam was neither economically nor physically feasible. The size of the dam was limited by the geography of its location, and even if a larger dam could have been built the consequent greater storage capacity would have caused a higher pool elevation which would have been detrimental to the Amana Villages upstream from the head of the Reservoir.
Plaintiffs contend also that sedimentation, which reduces the storage capacity of the Reservoir, is one of the causes of their problem. Increased sedimentation, according to upstream plaintiffs, has extended the backwater effect of the Reservoir, while downstream plaintiffs allege that the reduced capacity of the Reservoir due to sedimentation has forced the Corps to increase dam discharges, thereby elevating the stream level. The contention is refuted by engineering data. Although the most recent 1975 survey shows that since 1959 approximately 20 percent of the summer conservation pool (680 msl) has been filled with sediment (capacity reduced from 50,000 acre feet to 40,300 acre feet), and that the storage capacity below 670 msl has been reduced by approximately 50 percent (from 17,000 acre feet to 10,000 acre feet), sedimentation has had a negligible impact on the flood control capacity of the Reservoir. This is explained by the V configuration of the Reservoir, causing the bulk of its storage capacity to exist at higher elevations. The sedimentation rate has exceeded expectations because the inflow has been greater than the *58830-year average, but at the end of 100 years of operation it is estimated that the Reservoir will retain 90 percent of its flood control pool at 712 msl. The sedimentation that has occurred thus far is insufficient to have produced the flooding and saturation effects of which the plaintiffs complain.
The Marengo plaintiffs, located at river mile 138 approximately 55 river miles upstream from the dam, have failed to show that sedimentation at present levels can cause the backwater effect they claim. Field surveys made in 1960, 1965 and 1969 show that the furthest upstream any backwater influence from Coralville Reservoir has occurred is between river miles 120 and 124.
All three groups of plaintiffs argue that the level of the ground water table has been raised by greater streamflows which, although of lower elevation, nonetheless remain high for an extended period of time. However, plaintiffs have failed to show that the ground water basin, in general, is responsive to streamflow. There is a relationship between rainfall and ground water level, but the response of ground water level to rainfall is very slow compared to the response of stream level to rainfall. Rain water percolates into the ground and recharges the ground water basin, which acts as a huge underground reservoir. The level of the ground water basin does not synchronize with changes in stream level. An exception to this general rule is with respect to the land immediately adjacent to the streambanks and extending back from it as much as 200 to 300 feet. At trial plaintiffs introduced evidence on this "bank storage” theory to the effect that water will move out of the streambed and into adjacent land when the level of the water in the stream is higher than the water table of the adjacent land. For example, on November 14, 1969 (an excessively wet year), the stream level was higher than the adjacent ground water level and there existed a bank storage condition for about 200-300 feet back from the river. However the theory is of minor significance in the present situation since most of the time the Iowa River is an effluent stream carrying ground water out of the basin rather than an influent stream carrying water into the river banks and recharging the ground water table. The bank storage phenomenon *589when it occurs is normally confined to relatively narrow strips of land parallel to the river.
This conclusion is further supported by the analysis of infrared aerial photographs supplied by the government. Infrared aerial photography and "remote sensing” technologies were used at trial in assessing relationships of streamflows and the agricultural uses of plaintiffs’ farms bordering the river. Developed during World War II, infrared photography not only enables reliable detection of dead or decaying vegetation at early stages because of the early loss of infrared reflection, but has an enhanced ability to register ground moisture. Totally saturated soil appears black in the film. As soil dries out to 10-20 percent moisture content it becomes progressively lighter on the infrared film.
The accuracy of infrared photography was verified by actual onsite saturation tests correlating the May 11, 1976, infrared photography with observed data. Samples taken from the darkest areas on the imagery showed saturation of 38 percent. Samples taken in light-tone areas showed 25 percent saturation. The 25 percent area supported travel by farm vehicles and appeared capable of cultivation. While recognizing the limitations of infrared photography, the government’s exhibits are convincing evidence that the bank storage effect is physically limited to narrow strips of land on the river banks.
Normal annual rainfall over the Iowa and Cedar River Basin averages from around 30 inches over the headwater region to about 35 inches at the mouth of the river. During the April to September growing season Iowa normally receives about 70 percent of its total annual precipitation. Rainfall peaks in the month of June at a normal amount of 5 inches. Since construction of the dam rainfall of from one to four inches above annual norms was experienced in 1960, 1969, 1970, 1971 and 1974. In 1959, 1961, 1965, 1972 and 1973 annual rainfall averaged 5 inches or more above normal. Thus, in 10 of 16 post-dam years from 1959 through 1974 the annual rainfall was either above normal or sharply above normal.
High precipitation levels resulted in high runoff and high stream levels. Eighty of the 120 months between 1967 and *5901976 had above normal streamflows. Thirty-one consecutive months from July 1972 to January 1975 had above normal streamflows. Twenty-six consecutive months from July 1970 to August 1974 were in the "excessive flow” or upper quartile of streamflows. July 1969 had the highest monthly flow in all 888 months of recorded flows on the Cedar River.
The government prepared an elaborate "benefits analysis” by means described in Appendix A following the findings. Each of the years from 1959 to 1974 was analyzed for the potential of raising a crop, using two hypotheses: one, regulated flows with Coralville Dam in position, and two, unregulated flows without Coralville Dam. The benefits analysis reconstructed the crops saved, replants saved, yield reduction, and overtopping of levees and gravity drains allowed. In all but one area, wherever the regulatory effect of the dam was more than negligible crop benefits resulted from the dam. In essentially low water years which resulted in good crop years, such as 1963, 1964, 1967 and 1968, as well as the wetter years 1959, 1961 and 1971,3 crop results would have been the same with or without the dam in place. Plaintiffs’ complaints of flooding are confined to essentially high water years.
The benefits analysis for the more troublesome wet years 1960, 1962, 1965, 1966, 1969, 1970, 1972, 1973 and 19744 demonstrates direct and clear benefits to the Lone Tree plaintiffs since the operation of the Coralville Dam. (See finding 58). Only a small measure of benefit or detriment was provided the Wapello plaintiffs by the dam during the wet years listed. This is because the flow at Wapello, where the river widens substantially and has a much larger capacity, is the product chiefly of unregulated tributaries, *591especially the Cedar River. In the Wapello area the dam controls only about one-fourth of the total Iowa-Cedar River watershed embracing over 12,000 square miles above Wapello. The plaintiffs in the Wapello area are beyond meaningful control by the Coralville Dam. The problem of ungauged and unregulated tributaries contributing significant amounts to the streamflow of the Iowa River is also relevant to a lesser extent to the Lone Tree plaintiffs situated between the Coralville Dam and the confluence of the Iowa and Cedar Rivers.
Plaintiffs contend that the construction and operation of the dam generated inevitably recurring floods and constituted a "taking” of plaintiffs’ properties by the government. Although plaintiffs concede that government operation made skillful use of the inadequate Reservoir capacity, they contend that in order to alleviate flooding in different downstream reaches of the Iowa River, or even the Mississippi River, it has been necessary at times for the Corps to create river stages in the Iowa River reaches downstream from the dam which are at times unavoidably detrimental to farming operations in the growing season. Plaintiffs further contend that the result of the prolonged elevated river stages has been the blockage of natural drainage into the stream, thereby causing subsurface saturation of adjacent lands. According to plaintiffs’ argument this saturation has been and will continue to be a source of damage to plaintiffs’ lands and will predictably worsen as sedimentation progresses in the storage pool.
Defendant denies any taking, contending that plaintiffs failed to prove that water problems experienced by all classes of plaintiffs were due to, or the consequences of governmental action, i.e., attributable to the dam. Defendant maintains that water problems experienced by downstream plaintiffs were not attributable to the dam. It asserts that water problems experienced by downstream plaintiffs occurred despite and not because of the dam, and that the net benefits conferred by the dam to these plaintiffs were substantially in excess of any minor damage attributable to the dam.
Plaintiffs say that Iowa State law applies, since real property complaints are inherently a local matter. Under *592Iowa’s constitution benefits are excluded from consideration when part of a tract is taken, the measure of damages being the difference in value of the tract before and after the taking, irrespective of any benefit. 3 Nichols on Eminent Domain 8.6211[16]. Iowa law is not applicable. In Kohl v. United States, 91 U.S. 367, 374 (1876), the court stated: "if the United States have the power [of eminent domain] it must be complete in itself. It can neither be enlarged nor diminished by a state. Nor can any state prescribe the manner in which it [the power] must be exercised.” In United States v. 19.86 Acres of Land in East St. Louis, 141 F. 2d 344 (1944), while interpreting a portion of United States v. Miller, 317 U.S. 369 (1943) dealing with the proper measure of compensation in condemnation proceedings, the Seventh Circuit stated that "the forms and methods of procedure afforded by the law of the state do not affect questions of substantive right.” The court in Johnson v. United States, 202 Ct. Cl. 405, 418, 479 F. 2d 1383, 1390 (1973), held that the issue of what constitutes a "taking” is a "federal question” governed entirely by federal law, but that the meaning of "property” as used by the Fifth Amendment will normally obtain its content by reference to state law. See also United States Ex Rel T.V.A. v. Powelson, 319 U.S. 266, 279 (1943). The character of the property involved is not at issue here; the issue is whether there has been a Fifth Amendment taking of plaintiffs properties. Clearly the issue is governed by federal law, and not the law of the State of Iowa.
We have noted earlier the area rainfall statistics, a rainfall preponderance during the April-September growing season, and the abnormal prevalence of rainfall during 10 of the 16 post-dam years. In those excessively wet years the Reservoir’s high levels were often accompanied by high downstream water levels of both the Iowa River and its regulated and unregulated tributaries. Faced with the need to reduce the high levels of the Reservoir because of actual or impending heavy rainfalls in the area the Corps had no alternative to increasing the rate of discharge from the dam, despite temporary consequences to downstream farmers.
*593Excessive precipitation was the root cause of the flooding experienced by plaintiffs in the wet years of which they complain. The government’s manipulation of releases from the dam played only a secondary role. The United States is not liable for flood damages unless directly attributable to governmental action. Indirect or consequential damages are not compensable. In several flooding cases plaintiffs have failed to recover under their Fifth Amendment claim because they failed to prove the element of inevitably recurring floods. Fromme v. United States, 188 Ct. Cl. 1112, 1118, 412 F.2d 1192, 1196 (1969); National By-Products, Inc. v. United States, 186 Ct. Cl. 546, 575-78, 405 F. 2d 1256, 1272-74 (1969); North Counties Hydro-Electric Co. v. United States, 108 Ct. Cl. 470, 70 F. Supp. 900 (1947); North Counties Hydro-Electric Co. v. United States, 138 Ct. Cl. 380, 151 F. Supp. 322, cert. denied, 355 U.S. 882 (1957); B. Amusement Co. v. United States, 148 Ct. Cl. 337, 180 F. Supp. 386 (1960). Here the damage was not the natural consequence of government action. Therefore we need not speculate over the likelihood of recurrence of the physical conditions in the Iowa River Basin causing the damage experienced by plaintiffs during the years in issue.
For factual reasons heretofore detailed the operation of the dam and Reservoir had no influence in producing the conditions of which the plaintiffs complain in the Marengo area upstream from the Reservoir, and little if any influence on the conditions complained of by plaintiffs in the Wapello area. As to the plaintiffs in the Lone Tree area, occasionally the lower areas of their farms bordering the river became too wet to farm. However the great majority of these conditions occurred in excessively wet years and would in all likelihood have happened without the existence of the upstream dam. The remaining instances, where the operation of the dam caused sustained high levels of the stream which kept the lower areas of the adjacent farms too damp to farm, and which might properly be attributed primarily to the dam rather than to meteorological events, were heavily countervailed by the benefits to the farmlands as a whole, whether to rescue them from the damaging effects of floods or to save them from the consequences of drought conditions. The benefits analysis study provided by *594the government, and which was not effectively challenged by the plaintiffs, demonstrated these benefits quite conclusively.
It was held in United States v. Sponenbarger, 308 U.S. 256, 266-67 (1939), that "if governmental activities inflict slight damage upon land in one respect and actually confer great benefits when measured in the whole, to compensate the landowner further would be to grant him a special bounty.” This principle was applied in Ark-Mo Farms, Inc. v. United States, 209 Ct. Cl. 116, 530 F. 2d 1384 (1976), where the government’s unchallenged hydrological data showed that the river control project had in fact decreased peaks, duration and frequency of high level floods at plaintiffs farm. Citing Sponenbarger, the court in Ark-Mo Farms held that it was a case of at most "little injury in comparison with far greater benefits conferred.” To similar effect see the recent opinion in Accardi v. United States, 220 Ct. Cl. 347, 599 F. 2d 423 (1979). The Fifth Amendment does not make the government an insurer against all damages from floods which may be incidental to projects conferring major benefits far outweighing detriments.
Plaintiffs are not entitled to recover.
CONCLUSION OF LAW
Upon the trial judge’s findings and foregoing opinion which are adopted by the court, the court concludes as a matter of law that plaintiffs are not entitled to recover and, therefore, plaintiffs’ petitions are dismissed.
Although the court adopted the trial judge’s separate findings of fact and appendix (A), which are set forth in his report, they are not printed herein since such facts as are necessary to the decision are contained in his opinion.
There is an immaterial discrepancy in the defendant’s facts of record classifying four plaintiffs as being within the Lone Tree category of plaintiffs even though their properties lie along the Iowa River slightly below the Cedar River confluence. The discrepancy does not alter the results.
Meaning: 680 feet above mean sea level.
The level of precipitation during 1959 was strongly above average; however, the heaviest rainfall occurred during the end of March through the beginning of April. This allowed sufficient time for the soil to dry and permit a normal planting. 1961 also recorded strongly above normal rainfall but here again the rains missed the planting season. In 1961 March, July and September were among the wetter months while April, May and most of June were rather dry. Therefore, under both regulated and unregulated modes the crop yields were similar.
Although 1962 and 1966 did not have above normal rainfall in total, the rainfall that did occur was primarily during the growing season. In 1962 the end of May and the beginning of June, as well as the entire month of July, were wet. In 1966 the entire spring was quite wet, with the month of May experiencing the heaviest rainfall. The timing of the rainfalls is quite important. Where the growing season experiences unusually heavy and frequent storms the difference between the benefits gained from the regulated versus the unregulated condition is substantial.