Appeal by the state from a judgment adjudging that the state has taken a perpetual flowage easement upon the intervening property *402owner’s land and awarding damages for such taking.1
This case, involving an omitted property owner in condemnation proceedings, has been before this court in two prior appeals.2 In 1935 the state commenced condemnation proceedings to acquire by the power of eminent domain certain lands in Chippewa County for the purpose of creating and perpetually maintaining a reservoir for the storage and maintenance of an adequate supply of water in Lac qui Parle Lake and for the purpose of the regulation and control of floods. By the spring of 1937, the state had acquired title to the lands involved in its original proceedings and the building of dikes, dams, and control works had been finished. The final certificate of completion of the original proceedings was approved by the trial court subject, however, to the rights of the parties to any suit pending prior to the date of approval.
The case involved in this appeal bears the same title as the original proceedings but designates Charles Hacker’s administratrix, his wife, as the petitioner in intervention.3 The intervenor’s husband was a landowner who on November 7, 1942, filed a petition in intervention alleging that his lands located on the Minnesota River a few miles4 below the Lac qui Parle dam and water-control works, although not included in or acquired under the original condemnation proceedings, had been damaged through flooding and had, in fact, been taken as a result of the control and operation of the project.
After the second of the two prior appeals to this court (State, by Peterson, v. Anderson, 244 Minn. 581, 69 N. W. [2d] 688,5 — April 1, 1955), an award was made by commissioners and from their award *403both parties appealed to the district court. The district court then made an award of damages based upon the verdict of a jury and entered the judgment from which this appeal is taken.
The basic issue presented for review is whether the evidence sustains the finding that the state has taken a flowage easement upon intervenor’s land to an elevation of 942 project datum.6 If the evidence does sustain the finding, then an additional contention of the state to the effect that the imposition of a peipetual flowage easement up to an elevation of 942 constitutes an usurpation by the court of legislative power is without merit and may be disregarded.
In evaluating the evidence both as to the taking and the extent of the taking, it is helpful to set forth the major features of the entire water-control system constructed by the state. In order to control waters in the Minnesota Valley, the state has erected a series of control works on the Minnesota River. The portion of the river involved herein flows through the Minnesota Valley and junctions with the Chippewa River at Montevideo. The overall plan and effect of the control system will best be understood by beginning with the project farthest upstream and describing the intermediate units in order down to the confluence of the two rivers.
The first and uppermost project was constructed to divert the flow of the Whetstone River (except as to a small amount passing through a culvert) into Big Stone Lake. The Whetstone had formerly in a state of nature flowed into the Minnesota River IV2 miles southeast of the foot of Big Stone Lake. The project at Big Stone Lake consists of dikes and control works which impound and control the flow of water directly from Big Stone Lake into the Minnesota River. The quantity of water impounded in the lake has been increased over a state of nature by the waters of the Whetstone River. The 1937 construction of the dikes and control works makes it possible to raise the water in Big Stone Lake to a much higher level than in a state of nature.
Approximately 10 miles (as measured in a straight line) below the Big Stone Lake control works, the Minnesota River flows into Marsh *404Lake. In about 1937 the State of Minnesota constructed the Marsh Lake Dam which consists of a dike which holds the water to a higher level than it was in nature. This dike created a water-conservation project at Marsh Lake for there was no effective control works installed in the dike with which to regulate the level of the water retained. This was not the case at the Big Stone Lake control works.
After leaving the Marsh Lake dike on the easterly end of Marsh Lake, the Minnesota River flows to the westerly end of Lac qui Parle Lake. Approximately 1 mile below the southeast end of Lac qui Parle Lake, the state constructed a dam consisting of a core wall built into a highway and control works which allow the water level in the lake to be varied. The gates or bays of the control works may be closed to raise the water level 10 feet above that in a state of nature; the core wall itself, however, was constructed 7 feet higher. Since the flow over the control works, if the bays were closed, might not allow flood waters to flow rapidly enough to hold down the water level behind this dam, the water, according to one of the engineers, could rise to the 943 project datum level of the core wall or 17 feet above the level in a state of nature.
After leaving the Lac qui Parle control works, the Minnesota River flows past the Hacker farm and junctions with the Chippewa River approximately at Montevideo. The Hacker farm is located about four miles in a direct line below the Lac qui Parle dam.
The final project in the water-control system is near Watson on the Chippewa River, upstream from the confluence of the Chippewa and Minnesota Rivers. A dike and dam were constructed across the Chippewa near Watson. As a part of the Watson project, a diversion channel called the Watson Sag was constructed to carry about two-thirds of the water -from the Chippewa River into Lac qui Parle Lake. The only water which did not flow down the Watson Sag into the Lac qui Parle Lake had to flow either through a culvert (4x4 feet) or over the top of an overflow spillway. The culvert was constructed so as to allow water to flow down the Chippewa River when the river was at low levels. The overflow spillway has an elevation of 944 project datum. When the water level falls below 944 project datum, all the water (with the exception of the flow through the 4x4 culvert) passes *405through the Watson Sag into Lac qui Parle Lake. It is significant that the water which enters Lac qui Parle Lake from the Watson Sag would never in a state of nature flow past the Hacker farm as it must once it has entered the lake above the Hacker farm.
At the trial, the owner of the property, Charles Hacker, testified to the flooding of his property for six consecutive years commencing with 1942. At the time of the trial he was 61 years old and had lived on his farm all his life. He testified that prior to 1942 the only flood he could remember which prevented cropping on his lowlands was the flood of 1919. The first flood he could remember was 1908 but this flood, as well as other floods prior to 1942 (except the one of 1919), did not stay on the lowlands during the cropping season. This testimony from a man who had spent his entire 61 years on the property supplies an evidentiary base to sustain a finding that, with only one exception, it was not until 1942 that his property was flooded during the cropping season. This witness said that in 1942 approximately 240 acres of pasture and cultivated fields were flooded. In 1943, 225 to 240 acres of bottom lands were flooded and stayed under water practically all summer. In 1944 and 1945, Hacker testified that the flooding conditions were such that the lowlands were not cropped at all, with the exception of six acres of sweet corn in 1945. Further flooding occurred in 1946 and 1947.
Conflicting testimony at the trial presented opposing explanations for the six consecutive annual floods on the Hacker farm beginning with the year 1942. The intervenor’s witness was M. E. Chamberlain, a civil engineer, who has resided in the area around the control project from 1876 to the time he testified. During this time, Chamberlain had conducted surveys of parts of the control project and he had inspected the entire area many times. This witness expressed the opinion that the 1942 and the succeeding annual floods on the Hacker farm were caused by increased water flowing past the property during the cropping season and that the water-control project of the state had caused this increased water flow. Chamberlain testified that the Watson Sag carried waters of the Chippewa River into the project, increasing the watershed area of Lac qui Parle Lake by 1,850 square miles. Sidney A. Frellsen, an engineering expert appearing for the state, con*406firmed this estimate by testifying that the watershed of the lake was 4.000 square miles prior to the diversion of the Chippewa River and 6.000 square miles after the diversion.
Chamberlain further testified that the Minnesota River in a state of nature, prior to the control project, had an early runoff of spring floods and that the water level of the river opposite the Hacker farm was then low enough to allow the property to drain prior to the cropping season. After the control project was constructed, however, Chamberlain said that the increased impoundment of water at Big Stone Lake, the retention of waters in Marsh Lake by a dike which prevented the normal discharge of its waters downstream, together with the increase in the water volume of Lac qui Parle Lake by the diversion into it of waters through the Watson Sag, had produced the maintenance of a high water elevation in the valley which in turn materially raised the level of the water in the river below the Lac qui Parle dam.
Taking the evidence as a whole, we can only conclude that it reasonably supports the trial court’s finding of a taking. The testimony of parties who had lived in the valley for many years established that prior to 1942 the Hacker farm was cropped for every year except 1919. The argument that in 1942 and 1943 floods were the result of extraordinary and unprecedented rainfall is disproved by the fact that in those years the Minnesota River was not flooding over the top of the Lac qui Parle dike, but was up to an elevation of 934, with “just a little water spilling over the spillway,” as testified to by Hacker.
The conclusion that the flood of 1942 was not the result of unprecedented rainfall is clearly supported by the testimony of the state’s engineer, Frellsen. On cross-examination, he said:
“Q. You weren’t able or didn’t retain in the reservoir during the year 1942 these waters, and let them out at the capacity of the river channel, did you?
“A. Too much water.
“Q. You let that out far in excess of river channel capacity, didn’t you?
“A. Well—
“Q. What was the answer?
“A. It could just as well be ‘yes’-
*407“Q. Is it ‘yes’?
“A. Yes.”
Later in the same cross-examination, FreUsen testified that the watershed of Lac qui Parle Lake had been increased 50 per cent by the Watson Sag. In view of this testimony, the trier of fact could reasonably conclude that the cause of excess water below the dam was attributable to the control works and not to extraordinary or unprecedented rains.
The state’s other expert witness, Adolph F. Meyer, expressed the opinion that the flow of water past the Hacker farm was less during the 1942 flood than it would have been in a state of nature. This' argument made from public records of the U. S. Geologic Survey conflicted with Chamberlain’s observation that an earlier runoff resulted in dry cropland during the cropping season. The testimony of Hacker as to the preceding periods in the history of the river support Chamberlain’s conclusion. John Wilkinson, a resident of the county for 32 years, and a neighbor of Hacker’s from 1915 to 1936, corroborated Hacker’s testimony in all material respects. The direct observations of both Hacker and Wilkinson discredit the state’s argument based upon a statistical approach since the chart used by Meyer indicated that the Hacker farm was flooded in earlier years, a proposition manifestly erroneous in the light of the testimony by eyewitnesses. When Meyer was confronted with the statement that crops were raised on the Hacker farm in periods when, according to his charts and calculations of the cubic-feet-per-second flow past the Hacker property, the land should have been flooded, he failed to offer a satisfactory explanation. The finder of fact could reasonably conclude therefore that Meyer’s statistical calculations were inaccurate and entitled to little evidentiary weight.
Meyer, the state’s expert witness, also testified that floods of varying degrees recur in 25-, 50-, and 75-year cycles. He also indicated that he foresaw recurring flood levels. In view of Meyer’s cyclic theory, the fact that the Hacker farm had flooded only once earlier (in 1919) as seriously as it flooded in 1942, 1943, 1944, 1945, 1946, and 1947, is significant since these six consecutive annual floods do not fit any cyclic pattern. Reasonably, the six consecutive floods can be explained only on the basis that something had altered the normal course of nature and that the altering factor was the state’s water-control project. *408In fact the evidence would not sustain a finding that the six years of flooding resulted from unusual rainfalls. Furthermore, there is no evidence that the river channel opposite and below the Hacker farm had undergone any change in the capacity of its flow. Eyewitnesses effectively discredited any contention that a sandbar had developed in the channel during the few preceding years, or even during preceding decades.
The state contends that a taking of the Hacker farm by flooding is a physical impossibility. The claim is that the riverbed at the Hacker farm is 6V2 feet below the level of the river at the dam. The highest recorded tailwater at the dam was 936.3 during the period of flooding, so that by reducing that figure 6.5 the water level at the Hacker farm is only 930. Therefore, expert-witness Meyer concludes the property could under no circumstances be taken to an altitude of 942, and the state argues that to conclude otherwise is to find that water can run uphill. The statement in the record cited by the state does not mention “riverbeds,” but states that the water elevation at the Hacker farm is 6V2 feet below the elevation of the tailwater at the dam. The argument may be disposed of by noting that when the tailwater is at an altitude of 928, the river has flooded the Hacker farm. The water which then floods the Hacker farm comes onto the property overland having first crept over the riverbank about 2 miles upstream. After the water has filled the channel and jumped the bank, it can in no sense be said to be running uphill. As the channel fails to carry the water downstream, it should be expected that the water will in effect back up to the level of the tailwater below the dam.
The evidence also supports a finding of a taking upon to 942 project datum. Mr. Chamberlain established the level of the water on the Hacker farm from the testimony of Charles Hacker. Chamberlain found that the water on the Hacker farm had gone to an altitude of 937 and allowed an additional five feet when he said:
“A. I should say 942, allowance for seepage and some way of access [sic], and the possibility of slightly higher flood.
“Q. Is that the usual computation * * *?
“A. What I have been using * * * is about four feet above the highest known flood.”
*409Although Chamberlain had earlier estimated the altitude of the flood water as 935, he revised this estimate to 937 for the year 1942. The additional 5-foot allowance was then made to fix the flowage rights which he considered as taken to an altitude of 942. This 5-foot allowance was based on a combination of factors. In an earlier estimate, Chamberlain had allowed 3 feet for seepage and because the water “may go higher.” The subsequent increase in his estimate to 942, while related to 1942 flood data, took into account the information that “1943 was a little higher, and” that “the flood of 1919 came within 75 feet of his [Hacker’s] bam, which is considerably higher.” The estimate of Chamberlain clearly allowed for seepage, floods which had gone higher than the 1942 flood, and wave action. The finding of a taking to 942 project datum is reasonably sustained by Chamberlain’s estimates and by the evidence as a whole.
Since the evidence, taken in the light most favorable to the trial court’s findings, sustains a taking of intervenor’s land to an elevation of 942 project datum, the judgment of the trial court must be, and is, affirmed.
Affirmed.
Upon Appeal from Clerk’s Taxation of Costs.
On March 21, 1958, the following opinion was filed:
Appellant, State of Minnesota, appeals from the clerk’s taxation of costs.
It is well settled that the state cannot be taxed costs upon appeal in this case unless a statute so provides. This is so because the state, when acting in its sovereign capacity, cannot be taxed costs and disbursements except as otherwise provided by law. State ex rel. Smiley v. Holm, 186 Minn. 331, 243 N. W. 133. Moreover, M. S. A. 607.01, providing for allowance of costs and disbursements to the prevailing party upon an appeal, is not expressly applicable to the state, and so under § 645.27 may not be applied to the state. Section 645.27 provides:
“The state is not bound by the passage of a law unless named therein, or unless the words of the act are so plain, clear, and unmistakable as to leave no doubt as to the intention of the legislature.”
*410The sole question for decision is the applicability of § 117.20, which provides:
“Subd. 4.7 At any time within 40 days from the date of the filing of the report, any party to the proceeding may appeal from any award of damages embraced in the report, or from any omission to award damages, by filing with the clerk a notice of such appeal; which shall specify the particular award or failure to award appealed from, the nature and amount of the claim, the land to which it relates, and the grounds of the appeal; and upon appeal the prevailing party shall recover costs and disbursements.” (Italics supplied.)
If this provision is applicable to this appeal, the state recognizes that it must pay the costs on the, appeal. The argument advanced by the state is that the statute applies only to an appeal from an award of damages or from an omission to award damages, and that the appeal made by the state in this case was upon the validity of the trial court’s finding of a taking of property to a certain elevation.
The state seeks support for its position in State, by Peterson, v. Bentley, 224 Minn. 244, 28 N. W. (2d) 179, 770, wherein we denied taxation of costs under the statute. In that case we denied costs since the appeal therein was not from any award or failure to award damages but from the trial court’s orders denying motions by the state either to amend orders granting the right to intervene or for a new trial. In the instant case, however, involving an appeal from a judgment, the award of costs is controlled by our determination in State, by Peterson, v. Bentley, 231 Minn. 531, 546, 45 N. W. (2d) 185, 194. In that case, upon appeal from a judgment awarding damages, we awarded costs under § 117.20, subd. 4 (then § 117.20[2]), without regard to the particular grounds for the attack upon the award.
In the instant case the appeal was from a judgment which not only adjudicated that the state had taken a flowage easement up to a certain elevation but also, as an integral part thereof, awarded damages to the landowner for such taking. The very basis of the award was the finding of the taking. As a matter of commonsense, the state would *411have no objection to acquiring an easement to almost any elevation if it were not subject to an award of damages in proportion to the extent of such easement. The very purpose of the state’s appeal herein was to reduce the award of damages by showing that the evidence did not justify a finding of a taking of a flowage easement to an elevation of 942 project datum. Where, upon an appeal from a judgment which adjudicates both the taking of a specified flowage easement and the amount of damages to be awarded therefor, the state challenges the extent of the taking as a ground for undermining the award of damages, costs are to be awarded to the prevailing party pursuant to the obvious meaning of § 117.20, subd. 4. It would indeed be a travesty upon the unmistakable meaning of § 117.20, subd. 4, to hold that an appellant who limits his attack to one of the grounds for an award of damages is not attacking the award itself.8
The clerk’s taxation of costs is affirmed.
Affirmed.
This case arises under the old intervention procedure which has now been supplanted by mandamus as the proper remedy to be followed by an omitted property owner. See, State, by Peterson, v. Anderson, 220 Minn. 139, 19 N. W. (2d) 70; State, by Peterson, v. Anderson, 239 Minn. 144, 146, 147, 58 N. W. (2d) 257, 259.
See, State, by Peterson, v. Anderson, 239 Minn. 144, 58 N. W. (2d) 257, and Id. 244 Minn. 581, 69 N. W. (2d) 688.
See footnote 1.
Estimated as about 8 river-channel miles or about 4 miles as the crow flies.
See footnote 1.
The record in this case does not include the record of the trial on the question of damages after April 1, 1955, but does include the record in the two prior appeals.
Prior to amendment by L. 1957, c. 728, this subd. 4 appeared as § 117.20(2) and period of appeal was limited to 30 days.
It should be noted that if the determination of the trial court were embodied in a judgment merely mandamusing the state to condemn lands omitted in the original proceedings, the review of that judgment in this court could be without taxation of costs since a judgment so mandamusing the state, although a final adjudication upon the issue of whether land has been taken, would not at that stage involve an award of damages. See, State, by Peterson, v. Anderson, 239 Minn. 144, 58 N. W. (2d) 257; State, by Peterson, v. Anderson, 220 Minn. 139, 19 N. W. (2d) 70.