Ideker Farms, Inc. v. United States

Opinion

In the United States Court of Federal Claims No. 14-183L (Filed: December 14, 2020)

) IDEKER FARMS, INC., et al., ) ) Plaintiffs, ) Taking; Fifth Amendment; Flooding; ) Arkansas Game & Fish; Flowage v. ) Easement; Just Compensation; Before- ) and-After Valuation; Consequential THE UNITED STATES, ) Damages; Interest Rate ) Defendant. ) )

R. Dan Boulware, St. Joseph, MO, for plaintiffs. Edwin H. Smith, Seth C. Wright, St. Joseph, MO, and Benjamin D. Brown, Washington, D.C., of counsel.

Brent Allen, Environment and Natural Resources Division, U.S. Department of Justice, Washington, D.C., with whom was Jean E. Williams, Deputy Assistant Attorney General, for defendant. Jacqueline C. Brown, Elizabeth McGurk, Brad Leneis, Terry Petrie, and Frank Singer, Washington, D.C., of counsel.

TRIAL OPINION

FIRESTONE, Senior Judge.

Set forth below are the court’s findings of fact and conclusions of law following

the Phase II trial in Ideker v. United States. This takings action was brought by over four

hundred farmers, landowners, and business owners from six states claiming that changes

made by the United States Army Corps of Engineers (“Corps”) to its operations on the

Missouri River (“River”) after 2004 under the Missouri River Recovery Program

(“MRRP”) in order to comply with the Endangered Species Act (“ESA”) have resulted in the taking of flowage easements across their properties without just compensation in

violation of the Fifth Amendment to the United States Constitution.

In the Phase I trial, the court examined the claims of 44 representative plaintiffs to

determine whether they met some of the threshold requirements for establishing a taking

by intermittent flooding set forth by the Supreme Court in Arkansas Game & Fish

Commission v. United States, 568 U.S. 23, 34 (2012) (“Arkansas Game & Fish”). The

court in Phase I focused on whether the 44 representative plaintiffs could prove that some

or all of the flooding of their properties was caused by the Corps’ actions under the

MRRP to comply with the ESA and, if so, whether that flooding was a foreseeable

consequence of the Corps’ actions.1 The court also examined in Phase I whether any of

the flooding caused by the Corps was sufficiently severe to give rise to a potential taking.

Causation, foreseeability, and severity are three of the Arkansas Game & Fish factors the

court must consider in determining whether the government has taken a flowage

easement. 568 U.S. at 34, 39.

In the Phase I trial opinion, the court described the history of this case and made

extensive findings and conclusions regarding the Corps’ changes to the River under the

1 In Phase I, the court rejected the claims of all representative plaintiffs who sought compensation for a taking based only on the extraordinary flooding that occurred along the River in 2011. Ideker v. United States, 136 Fed. Cl. 654, 691 (2018). The court found that this flooding had nothing to do with ESA compliance; instead, it was caused by extraordinary flows into the Missouri River Mainstem Reservoir System’s reservoirs at levels beyond the System’s design. Id. As a consequence, the Corps was forced to release huge volumes of water from the System’s dams in order to protect the integrity of the System itself. Id. at 691-93. The court determined that the flooding in 2011 was necessary to save the System and did not give rise to a taking. Id. at 693. Many of the Phase I representative plaintiffs had a takings claim based only on flooding in 2011 and were accordingly not eligible to proceed to Phase II. See id. at 762-63.

2 MRRP. The court will not repeat in detail that history or those findings and conclusions

here. In brief, after considering all of the evidence presented in Phase I, the court

determined that 28 of the 44 representative plaintiffs had established that the changes

made by the Corps to the Missouri River Mainstem Reservoir System2 (“System

Changes”) and the River itself (“River Changes”) under the MRRP3 had foreseeably

caused increased and repeated flooding on their respective properties. Ideker Farms, Inc.

v. United States, 136 Fed. Cl. 654 (2018) (“Ideker I”); Ideker Farms, Inc. v. United

States, 142 Fed. Cl. 222 (2019) (“Ideker II”). More specifically, the court held that the

actions taken by the Corps when (1) releasing water from the Gavins Point Dam4 to

protect certain species during periods of high flows, and (2) undoing actions previously

taken by the Corps to make the River straighter and faster5 (and thus less flood prone)

resulted in more flooding6 than would have occurred on the properties of 28

2 The Missouri River Mainstem Reservoir System is the system of dams and reservoirs constructed by the Corps on the Missouri River under the Flood Control Act of 1944. Id. at 661- 62. 3 In its Phase I opinion, the court, as a shorthand, referred to the Corps’ River and System Changes as the MRRP. Id. at 665 (The MRRP “is the Corps’ umbrella program for returning the Missouri River to a more natural state to aid in the recovery of the Missouri River Basin ecosystem.”). In this opinion, the court continues to use the MRRP as the shorthand for these River and System Changes. 4 The Gavins Point Dam is one of six mainstem dams operated by the Corps on the River. Id. at 661. 5 The program that sought to make the river straighter, deeper, and faster is known as the Missouri River Bank Stabilization and Navigation Project (“BSNP”). Id. at 663 (describing the channelization and stabilization effect of the BSNP). 6 The court used the word “flooding” in the Phase I opinion as a shorthand for four different types of flooding: “overbank, levee overtopping, blocked drainage, and seepage.” Id. at 700. 3 representative plaintiffs without the System and River Changes for some or all of the

following years: 2007, 2008, 2010, 2013, and 2014. See Ideker I, 136 Fed. Cl. at 761-63.

The court further determined that 14 of the 28 had established that the increased flooding

attributable to the MRRP had occurred more than three times and was “severe.” Id. The

court relied primarily on the testimony of plaintiffs’ experts: Dr. Ronald Christensen

(regarding raised water surface elevations (“WSEs”)), Dr. Theodore Hromadka II

(regarding increased and more severe flooding based on raised WSEs), and Mr. Glenn

Tofani (regarding the effect of WSEs on levees and levee failure). See id. at 680-81, 761-

63.

In Phase II, the court asked the parties to identify a smaller number of

representative parties for the purpose of determining whether these plaintiffs could

establish the remaining Arkansas Game & Fish factors, and, if so, to provide evidence

regarding just compensation. The parties selected the tracts of three representative

plaintiffs for the Phase II trial. These are the Adkins property (Property 17), the Ideker

Farms, Inc. property (Property 33), and the Buffalo Hollow Farms, Inc. property

(Property 41). These properties are referred to as “representative,” “bellwether” or

“Phase II” properties by the witnesses in Phase II and throughout this opinion. In Phase

I, the court concluded that these three plaintiffs had already “established causation,

foreseeability, and severity” for the flooding of their properties in 2007, 2008, and 2010

for the Adkins property, in 2007, 2008, 2010, 2013, and 2014 for the Ideker property, and

The differences are explained in the Phase I opinion. See id. The court uses the term “flooding” the same way in this opinion. 4 in 2007, 2008, 2010, 2013, and 2014 for the Buffalo Hollow property. Ideker I, 136 Fed.

Cl. at 761-62.

The Phase II trial concerning these three representative tracts was held remotely

over three weeks beginning on July 20, 2020. The court heard testimony from 19

witnesses during the trial and received 1080 exhibits into evidence. Closing argument for

the Phase II trial was heard on September 22, 2020.

This opinion is divided into four sections: (1) causation for post-2014 flooding, (2)

takings liability for a permanent flowage easement under Arkansas Game & Fish, (3) the

date of the taking, and (4) just compensation. The court’s findings of fact and legal

conclusions are set forth in each section separately. As described in more detail below,

the court has concluded that the three representative plaintiffs have established that the

government has taken a permanent flowage easement across their representative tracts

and has determined the just compensation due to these three plaintiffs.

I. Causation for Post-2014 Flooding

The Supreme Court in Arkansas Game & Fish, 568 U.S. at 38-39, set forth the

standard for when government-induced intermittent flooding may give rise to a claim for

the taking of a flowage easement. Under Arkansas Game & Fish, the court must

undertake a multi-factor, “situation-specific factual inquiry,” id. at 31-32, considering

whether (1) the flooding was caused by the government, see id. at 38; see also Ridge

Line, Inc. v. United States, 346 F.3d 1346, 1356 (Fed. Cir. 2003) (holding the court must

determine “whether the [flooding] on the claimants[’] property was the predictable result

of the government action”); (2) “the degree to which the invasion is intended or is the

5 foreseeable result of the authorized government action,” Ark. Game & Fish, 568 U.S. at

39, (3) the “[s]everity of the interference,” id., (4) “time” and the duration of the flooding,

id. at 38, (5) “the character of the land at issue,” id. at 39, and (6) interference with “the

owner’s ‘reasonable investment-backed expectations’ regarding the land’s use,” id.

(quoting Palazzolo v. Rhode Island, 533 U.S. 606, 618 (2001)).

This first part of the Phase II trial opinion addresses whether the representative

plaintiffs have established causation for flooding that occurred on their respective

properties post-2014. In the Phase I trial, the court limited the evidence of flooding to

December 2014 so that the parties had a closing date for discovery. See Ideker I, 136

Fed. Cl. at 670. In the Phase II trial, the court allowed the three representative plaintiffs

to present evidence of any continued flooding post-2014, setting a cut-off date of

December 2018, again for trial preparation purposes. As discussed below, the three

representative plaintiffs contend that flooding on their properties is continuing post-2018

and, as permitted, presented evidence in support of that contention.

A. Scope of the Court’s Previous Findings on Causation and the Relevant Time Period

Before turning to the court’s causation findings regarding flooding on the three

representative plaintiffs’ properties post-2014, the court will address two threshold issues

raised in the parties’ Phase II post-trial briefs regarding the scope of the court’s holdings

during and after the Phase I trial. First, the parties disagree as to whether the court

previously determined that the Corps’ actions caused all of the flooding on the

representative properties for the years where the court found causation in Phase I, as the

6 representative plaintiffs contend, Pls.’ Br. at 59-90, 130-33, ECF No. 678, or whether the

Corps’ actions caused only additional flooding beyond what would have otherwise

occurred, as the government argues, Def.’s Br. at 12-20, 132-33, ECF No. 679.

The representative plaintiffs’ argument that the government caused all of the

flooding on their properties for the years where the court found causation is incorrect. It

is true that the court found that any flooding attributable to natural causes, as opposed to

the Corps’ actions, did not break the “chain of causation” between the MRRP and the

flooding on the representative plaintiffs’ properties. Pls.’ Br. at 133; Ideker I, 136 Fed.

Cl. at 670. However, the court did not go on to hold that all of the flooding on the

representative plaintiffs’ tracts was attributable solely to the MRRP. Rather, the court

concluded in Phase I that the Corps’ MRRP actions in certain instances made the

flooding more severe than it would have been had the MRRP not been in place.

Specifically, the court held that the “Corps’ River Changes have, together with the

Corps’ System Changes, caused WSEs to rise higher than they would have risen without

these Changes and that this rise in WSEs has led to more flooding or more severe or

longer flooding than would have occurred had these Changes not been made by the

Corps.” Id. at 696-97. Relying on the testimony of the Phase I plaintiffs and their

experts, for every year in which it found causation, the court concluded that the Corps’

actions under the MRRP caused “more severe flooding than would have occurred

without” the Corps’ actions. Ideker I, 136 Fed. Cl. at 729-30 (Adkins) (emphasis added);

id. at 747-49 (Ideker); id. at 757-58 (Buffalo Hollow). The only exception to this finding

was where the evidence established that certain levees would not have failed but for the

7 higher WSEs attributable to the MRRP. See, e.g., id. at 748 (for the Ideker property in

2010, “[t]he court finds, based on Mr. Tofani’s testimony, that flooding was due to levee

overtopping that would not have occurred without the System and River Changes”).

Apart from these exceptions, the court did not hold that the MRRP was the but-for cause

of all flooding on the representative plaintiffs’ properties, but only that it caused

additional or more severe flooding than would have occurred without the MRRP.

The representative plaintiffs’ reliance on their experts’ testimony from Phase I to

argue that the court found that the Corps was responsible for all flooding is not supported.

The representative plaintiffs first contend that Dr. Christensen and Dr. Hromadka opined

in Phase I that the MRRP was the cause of all of the flooding on the Phase I plaintiffs’

properties. Pls.’ Br. at 60-62. The court did not interpret the evidence in that way or

make such a finding. Rather, the court relied on Dr. Christensen, Dr. Hromadka, and Mr.

Tofani to hold that, generally, the MRRP caused more severe flooding in certain years.

Ideker I, 136 Fed. Cl. at 729-30 (Adkins); id. at 747-49 (Ideker); id. at 757-58 (Buffalo

Hollow).

The representative plaintiffs also appear to suggest that the WSEs presented by Dr.

Christensen in Phase I underestimate the extent of the flooding at the Adkins, Ideker, and

Buffalo Hollow properties, meaning that the MRRP could have caused all of the flooding

on the three representative plaintiffs’ properties. See Pls.’ Br. at 62-64. While Dr.

Christensen opined that the WSEs on plaintiffs’ individual properties could have been

greater than he estimated in his model, Dr. Christensen’s statement alone is not enough to

prove that the MRRP was responsible for all of the flooding on plaintiffs’ representative

8 tracts for the years in question. The court therefore agrees with the government that,

when examining the Arkansas Game & Fish factors as well as just compensation, the

court is tasked with determining whether the additional flooding caused by the MRRP is

sufficient to meet the standards for the taking of a flowage easement and, if so, the value

of the property interest it has taken.

However, the court disagrees with the government’s argument that “[b]ecause

Plaintiffs failed to present any evidence of the amount of incremental flooding, Plaintiffs

failed [to] prove either liability or just compensation for their claims.” Def.’s Br. at 133.

As explained below, the court finds that the representative plaintiffs have demonstrated

liability under the Arkansas Game & Fish factors here, building on the court’s Phase I

opinion that the MRRP caused more severe flooding on their properties. In fact, the

government’s own crop loss expert’s opinions demonstrate that the representative

plaintiffs’ properties experienced considerable crop losses based on the incremental

effects of MRRP flooding. See infra Part II. With respect to just compensation, the court

finds that the representative plaintiffs’ method of calculating the diminution in the fair

market value of their respective properties supplies a reasonable approximation of what

was taken as a result of the Corps’ actions under the MRRP. See infra Part IV.

Second, the parties disagree as to the court’s prior opinion regarding the relevant

time period over which to analyze the Arkansas Game & Fish factors. The government

presented evidence describing the flooding on the representative plaintiffs’ tracts that

occurred before the Mainstem System and BSNP (the program that made the river

straighter and faster, see infra n.5) were in place to argue that the character of the

9 representative plaintiffs’ land has not meaningfully changed with the introduction of the

MRRP. See Def.’s Br. at 134 (“In this case, Plaintiffs’ properties are located next to the

Missouri River and have been subject to flooding (and, at times, major flooding) at

unpredictable intervals throughout recorded history.”). The representative plaintiffs

argue that the proper time periods to compare are “the time period without the MRRP but

with the System and BSNP (not the beginning of time)” and “the time period with the

MRRP and the System and BSNP (the current period).” Pls.’ Resp. at 13, ECF No. 681.

The court agrees with the representative plaintiffs that in determining whether the

government can be liable for any increased flooding, the relevant time period begins after

the Mainstem System and BSNP were in place. This court has already determined in its

prior ruling denying the government’s motion to amend its answer that “the flood

protection provided by the Mainstem System and the BSNP is the baseline of flood

protection against which the additional flooding caused by the MRRP should be judged

for purposes of deciding both causation and government liability for any taking in this

case.” Ideker Farms, Inc. v. United States, 146 Fed. Cl. 413, 416 (2020) (“Ideker III”).

It is against this backdrop of its prior rulings that the court now turns to the claims

of the representative plaintiffs.

B. Phase I Findings of Fact on Causation and Foreseeability for the Three Representative Tracts

Before turning to the court’s Phase II findings on post-2014 causation for the

representative plaintiffs, the court will first review its Phase I findings for each

representative property.

10 1. The Adkins Property

The representative Adkins property is located along the Missouri River at River

mile markers 608-611 in Council Bluffs, Iowa. Tr. 46:23-47:2. The Adkins plaintiffs

own the property in fee simple. Ideker I, 136 Fed. Cl. at 729. The property is protected

on three sides by federal levees. Id.; see also Tr. 47:3-7 (“The Phase II tract is

surrounded on three sides, the west, south and east side, by Federal Levee L-627, and the

tie-back levee on Indian Creek.”).

The court determined in Phase I that the MRRP caused foreseeable and more

severe flooding on the Adkins representative tract in 2007, 2008, and 2010, but rejected

flooding claims for 2011, 2013, and 2014. Ideker I, 136 Fed. Cl. at 730-31.

The court found that, in 2007, flooding occurred in April and May during the

planting seasons. The Adkins claimed that 400 of their 1,044 acres flooded. About 50 to

60 percent of the property inside the levees was flooded for about three weeks. The court

found that flooding in 2008 was similar to 2007 in severity, duration, and damage. The

2008 flooding occurred in June and crops were lost. The court found that, in 2010,

flooding was similar to 2007 and 2008 but the water was higher and stayed longer.

Flooding again occurred in late June and crops were lost. Id. at 729-31.

Based on Dr. Hromadka’s testimony, the WSEs prepared and testified to by Dr.

Christensen, and the testimony of Mr. Tofani regarding water levels on the levees, the

court concluded in Phase I that the 2007, 2008, and 2010 flooding “was the result of

higher WSEs,” and that “the higher WSEs were caused by and were a foreseeable result

11 of the Corps’ System and River Changes which resulted in more severe flooding than

would have occurred without these Changes.” Id. at 729-30.

2. The Ideker Property

The Ideker representative property is located in Holt County, Missouri, at River

mile markers 510-12. Tr. 218:1-8; Ideker I, 136 Fed. Cl. at 747. The property is owned

in fee simple. Ideker I, 136 Fed. Cl. at 747. The Corps’ MRRP activities in the area

surrounding the Ideker property include (1) dike construction and notching near the

property; (2) dike alteration at River mile 512; and (3) the Corps’ operations at the

Thurnau Wildlife Refuge directly south of the property. Id. The Corning Levee District

levee separates the Thurnau Wildlife Refuge and the Ideker property. Id. The property is

also protected by a private levee built by the plaintiff. Id.

The court determined in Phase I that the Corps had caused foreseeable and more

severe flooding on the Ideker property for 2007, 2008, 2010, 2013, and 2014. Id. at 748-

49. The court rejected the Ideker flooding claims for 2011. Id. at 749.

The court found that in 2007, the entire Ideker farm flooded for 30 to 60 days

beginning in May. Floodwaters entered the farm from the south and east, coming from

the Thurnau Wildlife Refuge. The private levees held, but at that time there was no levee

on the east side of the farm. The floodwaters entering the farm from the Thurnau

Wildlife Refuge were estimated to be two to ten feet deep. Equipment was destroyed and

some land reclamation was necessary. The 55 acres riverward of the mainline levee were

flooded. The river cabin outside the levee flooded for the first time and required

extensive renovation. Id. at 747.

12 The court found that in 2008, flooding began in mid-June and lasted 30 to 45 days.

The Ideker levees held. Like in 2007, water entered from the Thurnau Wildlife Refuge.

A portion of the farm sustained flood damage due to seepage and blockage of the

drainage ditches. The drainage ditches on the property filled and flowed over onto some

crops, sustaining water from one inch to one foot deep. Some crops died but were re-

planted. Grain bins and irrigation equipment sustained damage. The 55 acres riverward

of the Ideker mainline levee flooded. The cabin on the property adjacent to the River was

flooded and required renovation. Id. at 747-48.

The court found that the flooding in 2010 was worse than in 2007 and 2008. Once

again, the high waters entered from the Thurnau Wildlife Refuge. Sandbagging occurred

to block the Thurnau waters from entering the farm. The efforts were successful until the

mainline levee breached on the northwest corner of the farm, leaving a 90-foot scour hole

destroying farmland. The entire farm was flooded for approximately 90 days with

floodwaters on the north side of the property reaching three feet deep and 11 to 13 feet

deep on the south end of the farm. The farm, home, structures, and equipment sustained

extensive damage. Up to five feet of sand was left on the 60- to 70-acre area in the

northwest of the farm. Grading equipment was used to remove sand and relocate and

rebuild the mainline levee on the west and south areas, widening it two to three feet and

raising it two feet. Additionally, levees were built on the east side of the farm to negate

floodwaters emptying into the farm from the east in the future. Id. at 748.

The court found that in 2013, flooding on the property began in May and lasted 30

to 45 days. Fifty-five acres on the riverward side of the mainline levee again flooded.

13 There were no levee breaches. The water was estimated at seven to eight feet on the

mainline levee on the west side and two to three feet over the riverbank. Id. at 749.

The court found that, in 2014, the 55 acres on the riverward side again flooded

much like 2013. Flooding began in June with the peak water in late June but drainage

was impacted periodically from June to October. Water reached the base of the mainline

levee on the west side and was estimated at one to two feet over the riverbank.

Floodwaters were not as deep as in 2013. Farming the acreage riverward of the levee

was determined to be no longer sustainable in light of the flood-prone River. Id.

Based on Dr. Hromadka’s testimony, Dr. Christensen’s WSE chart, and Mr.

Tofani’s testimony, the court found for 2007, 2008, 2013, and 2014 that the flooding

“was the result of higher WSEs,” and that “the higher WSEs were caused by and were a

foreseeable result of the Corps’ System and River Changes which resulted in more severe

flooding than would have occurred without these Changes.” Id. at 747-49. For 2010, the

court found based on Mr. Tofani’s testimony that “flooding was due to levee overtopping

that would not have occurred without the System and River Changes.” Id. at 748. Based

on Dr. Hromadka’s testimony and Dr. Christensen’s WSE chart, the court held that for

2010 “the higher WSEs were caused by and were a foreseeable result of the Corps’

System and River Changes which resulted in more severe flooding than would have

occurred without these Changes.” Id.

3. The Buffalo Hollow Farms Property

The Buffalo Hollow property is located in Doniphan County, Kansas, adjacent to

River mile markers 475-78. Tr. 142:12-15; Ideker I, 136 Fed. Cl. at 757. The property is

14 owned in fee simple and is protected by a private levee. Ideker I, 136 Fed. Cl. at 757.

Corps construction activities near the property include the Wolf Creek Bend channel

widening at River mile 481 in 2006 and dike notching at River miles 492, 490, and 487.

Id. The court determined in Phase I that the Corps had caused foreseeable and more

severe flooding on the subject property for 2007, 2008, 2010, 2013, and 2014. Id. at 758.

The court rejected the Buffalo Hollow flooding claims for 2011. Id.

The court found that, in 2007, the flooding occurred in May, covering

approximately 25 percent of the property inside the levee and all but two or three acres

outside the levee. The River was three to four feet high against the levee and the water

depths inside the levee ranged from “spongy” soil to two feet. Crops were lost, there

were reduced crop yields, and pumping expenses were incurred. Id. at 757.

The court found that, in 2008, flooding occurred from May 30 to June 21. The

flooding covered approximately 50 percent of the property inside the levee and all but

two to three acres outside the levee. The River was four to five feet high against the

levee and the water depths inside the levee ranged from saturated soil to one to two feet

deep. Crops were lost, yields were reduced, and plaintiffs incurred pumping expenses

and clean-up costs. Id.

The court found that, in 2010, flooding occurred from June 11 to August 20.

Flooding covered about 75 percent of the property inside the levee and the entire property

that was outside the levee. The River was four to five feet against the levee and the water

depths inside the levee ranged from saturated soils to three feet deep. Crops were lost,

yields were reduced, pumping expenses and debris clean-up costs were incurred. Id.

15 The court found that in 2013, flooding occurred from late May to early June.

Almost all of the acres outside the levee were flooded with overbank flooding, and

seepage and/or blocked drainage occurred inside the levee. Crops were lost. The River

was about one to two feet high against the levee. The court found that flooding in 2014

was similar to 2013. Id. at 758.

For all of these years, the court found, based on Dr. Hromadka’s testimony, Dr.

Christensen’s WSE chart, and Mr. Tofani’s testimony, that the flooding “was the result of

higher WSEs,” and that “the higher WSEs were caused by and were a foreseeable result

of the Corps’ System and River Changes which resulted in more severe flooding than

would have occurred without these Changes.” Id. at 757-58.

C. Post-2014 Causation for the Phase II Representative Tracts

The court now turns to the post-2014 flooding on the three representative

properties. Based on the evidence discussed below, the court finds (1) that the Corps’

continued operation of the MRRP without change has caused more severe flooding in the

years 2016, 2017, and 2018 on the Adkins representative tract; 2015, 2016, 2017, and

2018 on the Ideker representative tract; and 2015, 2016, 2017, and 2018 on the Buffalo

Hollow representative tract than would have occurred without the MRRP and (2) that

because of the MRRP there is a new and increased flooding pattern on these

representative tracts and this new pattern of flooding is continuing into the future.

16 1. The Representative Plaintiffs’ Accounts of Increased Flooding and a Changed Flooding Pattern

The court begins with its review of the representative plaintiffs’ eyewitness

accounts of the increased magnitude and duration of flooding post-2014 and the changed

flooding patterns at their properties following implementation of the MRRP.

a. The Adkins property

Before the MRRP but after the Mainstem System and BSNP were in place, Mr.

Adkins testified that his property experienced some degree of flooding from seepage

and/or blocked drainage in 1967, 1984, 1993, and 1995. Tr. 57:8-16. However, Mr.

Adkins testified that “the flooding in the 1980s and 1990s” was “much different” from

the “MRRP flooding post-2004.” Tr. 58:11-15. According to Mr. Adkins, the “MRRP

flooding is different in how often it happens, its frequency, its severity, and just how long

it lasts . . . .” Id.

Regarding flooding on the property after 2014, Mr. Adkins testified that in 2016 a

high Missouri River in June and July caused blocked drainage and seepage on the

representative tract, and water backed up into the ditches causing wet fields and blocked

drainage. Crops were lost and yields were adversely impacted. Tr. 59:2-6. Mr. Adkins

further testified that flooding on the tract in 2017 was very similar to 2016. Tr. 59:7.

Mr. Adkins then testified that in 2018, crops were planted and looked good up

until the middle of June when, because of a high River, there was blocked drainage and

seepage, and all the crops outside the levee and within a half-mile inside the levee were

17 lost. He observed that the floodwaters remained on the property from June 2018 through

“Thanksgiving or so.” Tr. 59:8-60:5.

Mr. Adkins testified that the new pattern of flooding caused by the MRRP

beginning in 2007 has continued. Tr. 60:12-16; Tr. 86:25-87:4 (“The MRRP has

cause[d] the river to change. It has changed from the standpoint that its flooding patterns

have changed resulting in an increase in the frequency, severity and duration of the

flooding.”). He explained that the changes to the River from the MRRP are evidenced by

a change in the observed WSEs that now block the drainage ditches so that water cannot

runoff the property. Tr. 80:2-21. He also testified that before the MRRP the levee flood

gates and pressure relief wells on the property required little to no maintenance because

they were seldom used or needed. Tr. 80:22-81:1; Tr. 81:18-22. He stated that this is no

longer the case and that the flooding post-MRRP is very different from what it had been

before. Tr. 81:18-24; Tr. 82:5-6.

Based on his “personal observations of the river on a daily basis,” Mr. Adkins

testified “with confidence that we have a changed river since 2004.” Tr. 85:21-24. Mr.

Adkins stated that it is his “belief that the MRRP has altered the river’s flooding patterns

which will not go back to the way they were before unless the MRRP is terminated and

time passes.” Tr. 88:1-5. Because of this, Mr. Adkins “believe[s] the Phase II tract will

continue to be subjected under the MRRP flooding pattern of the river to inevitably

recurring and intermittent flooding on par with the flooding of the river from 2004

through 2014.” Tr. 88:12-18.

18 b. The Ideker property

Mr. Ideker testified that before the MRRP, his property flooded in 1952, 1962, and

1967 – prior to the Mainstem System becoming fully operational – and then in 1984 and

1993. Tr. 224:12-16; Tr. 261:8-13. However, Mr. Ideker stated that this flooding was

“nothing whatsoever like the flooding we have experienced since 2007,” after the

implementation of the MRRP. Tr. 224:17-20.

Mr. Ideker also testified about flooding after 2014. Mr. Ideker stated that in 2015,

overbank flooding was again experienced over the 55 acres on the west side of the farm

outside the levee, with blocked drainage inside the levee which adversely impacted farm

efforts. There were no crops planted on the 55 acres at the time. The new river home on

the acreage, now setting at a rebuilt higher elevation, was not damaged. Water again

reached the base of the levee and was estimated at one foot over bank. Six hundred to

700 acres of the interior farm was impacted by blocked drainage and seepage, and 50 to

60 percent of the farmable acreage flooded. Tr. 245:10-25.

Mr. Ideker stated that the flooding in 2016 was worse than in 2015. Overbank

flooding was experienced over the 55 acres on the west side of the farm outside the levee

in May through June. Due to blocked interior drainage and seepage, 200 acres of

cropland inside the levee were not planted. No crops were planted on the 55 acres

outside the levee next to the River. The interior farm was again impacted by blocked

drainage and seepage. Over 800 acres of farmable acreage flooded, or about 60 percent

of the farm. Tr. 246:1-16.

19 Mr. Ideker observed that in 2017 the flooding was similar to 2016. Overbank

flooding occurred on the 55 acres without reaching the interior of the new river home.

No crops were planted on the 55 acres. Blocked drainage and seepage again resulted,

adversely impacting farm efforts. Again, over 800 acres of the farmable acreage flooded,

or about 60 percent of the farm. Tr. 246:17-247:1.7

Regarding flooding in 2018, Mr. Ideker testified that there was a high River for

most of the year which periodically caused overbank flooding on 55 acres and created

severe problems inside the levee with blocked drainage and seepage. Planting was

hampered and the harvest was delayed until late 2018. The Ideker drainage tubes were

blocked over 300 days during 2018. Flooding interfered with efforts to farm the entire

year. Tr. 247:13-25.

Mr. Ideker stated that the changed pattern of flooding caused by the MRRP from

2007 to 2014 has continued and that this pattern is different than before 2007. Tr.

259:13-17; Tr. 263:17-264:1. Mr. Ideker testified that the “river since 2007 is

unpredictable,” and that the “severity, duration and frequency of the flooding is worse.”

Tr. 263:8-16; see also Tr. 267:16-18 (“This was not the situation prior to 2007, nothing

remotely close.”).

7 After experiencing repeated ongoing flooding in 2017, the ninth time in 11 years, the Idekers made the decision to terminate their efforts to farm the land themselves and leased the interior farm effective March 1, 2018, for $260 per acre. The 55 acres outside the levee, which includes the new river home, were not part of the lease. The property was under lease at the time of trial. Tr. 227:4-10; Tr. 247:2-12.

20 c. The Buffalo Hollow Farms property

Mr. Schneider8 testified that the Buffalo Hollow Phase II tract significantly

flooded in 1952 and 1967, prior to the Mainstem System becoming fully operational, and

then in 1973, 1984, and 1993. The 1973 and 1984 flood events took place before the

private levee on the tract was fully completed. Mr. Schneider testified that any other

flooding on the property from 1967 to 2007 was not significant and that even the 1973,

1984, and 1993 floods were nothing like what they have experienced since 2007. Tr.

146:22-147:4; Tr. 167:3-15. According to Mr. Schneider, “before the MRRP, the river

would raise slowly and drop quickly. Flooding was infrequent. Now, under the MRRP,

the flooding rises fast and it drops slowly. As a result, the inundation is more severe.”

Tr. 167:23-168:3.

Regarding flooding post-2014, Mr. Schneider described how, in 2015, the flooding

occurred on two separate occasions in June lasting up to 10 days. There was overbank

flooding outside the levee, with one to two feet of water up against the levee. Seepage

and blocked drainage occurred inside the levee. Crops were impaired or lost. On

December 16, water was over bank up against their levee and the Schneiders used a skid

loader to clean up the debris. Mr. Schneider stated that this event was very much like

those that occurred in 2013 and 2014. Tr. 159:20-160:7.

In 2016, Mr. Schneider stated that flooding took place May 10 through 31. As

before, seepage and blocked drainage resulted. There was standing water inside the

8 Mr. Schneider is the authorized representative of Buffalo Hollow Farms, Inc. Tr. 141:19-24. 21 levee. Crops had to be replanted and planting was delayed. Mr. Schneider described the

May flooding as “particularly significant,” but noted that there were other instances of

blocked drainage that took place in 2016. Tr. 160:8-16.

In 2017, Mr. Schneider stated that the River ran high, blocking drainage and

causing seepage problems during most of the summer. This was longer-lasting than the

previous two years, and extensive pumping efforts in May were undertaken and

continued. There was overbank flooding outside the levee. Mr. Schneider testified that

50 percent of the crops outside the levee were lost. Mr. Schneider likened the flooding in

2017 to that of “past years’ flooding,” which he ascribed to “a routinely high river.” Tr.

160:17-161:2.

Regarding flooding in 2018, Mr. Schneider testified that the River ran above

normal most of the year, blocking drainage for over 100 days in the summer, fall, and

into December. Harvest efforts were significantly impaired. Overbank flooding outside

of the levee occurred. Blocked drainage and seepage occurred again inside the levee

constantly for six months. Mr. Schneider testified that 70 percent of the crops outside the

levee were lost. Tr. 161:3-16.

Mr. Schneider testified that, based on his observation of the River on a daily basis,

the implementation of the MRRP has changed the flooding on his property, and that the

changed pattern of flooding is ongoing. Tr. 165:16-20; Tr. 166:15-20 (“[T]he river runs

higher than it did prior to 2004.”).

22 2. The Court Accepts Plaintiffs’ Accounts of Post-2014 Flooding

The court finds the representative plaintiffs’ descriptions of flooding on their

properties as described above to be reliable, and that the flooding as described by the

plaintiffs occurred as stated. The court also finds that the representative plaintiffs’

testimony regarding the change in flooding patterns on their properties post-2014 is

credible and reliable.

The government introduced crop insurance records to undermine the

representative plaintiffs’ recollections of flooding on their properties and plaintiffs’

testimony regarding the changed nature of the flooding. However, the court finds the

crop insurance evidence does not undermine the representative plaintiffs’ testimony.

To begin, crop insurance claims may be made for a number of reasons if

productivity is not achieved at the level of coverage provided. For example, claims can

be made for losses due to “excessive moisture,” in addition to claims for “flooding.” Tr.

2660:22-25 (Mr. Zanoni, a government witness who is a senior underwriter for the

federal crop insurance program, describing codes for losses). The government, not the

farmer, makes the decision on how to code a loss, and such codes do not describe the

source of the water that harms the crops. Tr. 2726:1-13; Tr. 2728:6-2730:21. The

government attempts to undermine the representative plaintiffs’ testimony by pointing

out that they made insurance claims for “excessive moisture” in years in which they did

not testify about any flooding. Def.’s Br. at 38-40. By counting claims for excessive

moisture as evidence of flooding events, the government also contends that the

representative plaintiffs either made a similar amount of crop insurance claims before and

23 after the MRRP, id. at 74 (arguing the Adkins farm made seven claims for excess

moisture prior to 2004 and eight claims for excess moisture between 2004 and 2018), or

failed to make claims for years in which they described flooding, id. at 76 (noting that

Mr. Ideker did not make a claim in 2013 through 2014).

The court finds, however, that crop insurance claims cannot be used to account for

flooding on the representative plaintiffs’ properties and thus do not undermine plaintiffs’

recollection. First, many of the insurance claims the government relies on are coded as

“excessive moisture,” see id. at 38-40 (describing claims), which usually references

severe rainfall onto the land and not flooding caused by a higher River. Tr. 168:9-14

(Schneider) (“A claim for excessive moisture, for example, can be due to rainfall and not

a rising river leading to flooding.”); Tr. 262:1-6 (Ideker) (“A crop insurance claim for

excessive moisture does not necessarily mean there was flooding from the river; it can be

due to rainfall.”); Tr. 2728:6-2730:21 (Zanoni) (testifying that the crop insurance records

“would not tell you” the source of the water that harmed the crops and that the “crop

insurance records alone cannot be used to confirm or not confirm a flood event in any

given year”). Because the crop insurance claims do not necessarily correlate to whether

River flooding occurred on the representative plaintiffs’ properties, the government’s

reliance on crop insurance records to undermine the credibility of plaintiffs’ testimony is

unpersuasive.

The government has not put forth any other evidence to undermine the

representative plaintiffs’ recollection. To the contrary, the government witnesses at the

Phase II trial each stated that they had no reason to doubt plaintiffs’ descriptions of

24 flooding on their properties. Tr. 1999:14-2000:3 (Jones); Tr. 1307:9-19, 1329:13-18

(Kelman); Tr. 1397:22-24 (Remus); Tr. 2268:23-2269:8 (Earles). In addition, the Corps’

historic flooding records generally support the representative plaintiffs’ description of

flooding on their properties prior to the MRRP. According to those records, historic

flooding occurred in 1952, 1973, 1978, 1984, and 1993. See Tr. 1531:7-1533:14

(testimony of Eric Shumate, Chief of the Hydrologic Engineering Branch in the Kansas

City District, Army Corps of Engineers). The Corps’ Omaha District Office post-flood

reports also included 1984 and 1993. Tr. 1453:7-22 (testimony of Daniel Pridal, Chief of

the River and Reservoir Engineering section of the Omaha District Corps of Engineers).

For all of these reasons, the court finds that the representative plaintiffs’

descriptions of the flooding on their properties and how the flooding has changed from

the period preceding the MRRP to after the MRRP was implemented are credible and

reliable. Taking this testimony together with the expert testimony described below, the

court finds that the MRRP has caused a pattern of increased flooding on the three

representative properties that will continue into the future.

3. The Plaintiffs’ Expert Dr. Mays Confirms a Finding of a Changed and Increased Flooding Pattern on the Representative Tracts post-MRRP

In support of the representative plaintiffs’ claim that the MRRP has caused a

changed flooding pattern resulting in more flooding than plaintiffs had experienced

before the MRRP was implemented, the plaintiffs presented the expert testimony of Dr.

25 Larry Mays.9 Dr. Mays is both a professional hydrologist and professional engineer. Dr.

Mays’ testimony corroborated the representative plaintiffs’ accounts of the additional

flooding they are experiencing beyond what would have been expected to happen on their

properties without the MRRP.

Dr. Mays was tasked with investigating whether flooding on the three

representative plaintiffs’ properties from 2015 through 2018 is consistent with the

changed flooding pattern demonstrated from 2007 through 2014 and is attributable to the

MRRP, and whether there is a heightened risk of flooding at the representative properties

due to the MRRP. Tr. 820:11-18. Before conducting an analysis of the post-2014

flooding, Dr. Mays examined the flooding that occurred from 2007 through 2014, and the

court’s findings regarding MRRP causation from the Phase I opinion. Tr. 820:20-821:1.

Dr. Mays opined based on the court’s findings that “[o]ne would logically expect that,

unless and until the river changes [caused by the MRRP] are deconstructed and the”

environmental priorities under the MRRP shifted, “any altered flooding pattern caused by

the MRRP changes would continue.” Tr. 836:12-17. In other words, because the MRRP

9 Dr. Mays is a registered professional hydrologist, a registered professional engineer in California, Arizona, and Texas, and is recognized as a leading authority on hydrology. Tr. 813:1-5; PX3443 (Mays CV). Dr. Mays’ extensive academic and professional experience spans 43 years, and he is currently a Professor Emeritus at Arizona State University, School of Sustainable Engineering and the Built Environment. Tr. 813:6-15. He is the author, co-author, or editor-in-chief of 24 books. Tr. 814:15-25. Dr. Mays has also consulted with many different government agencies and industries regarding flood control systems, which include the U.S. Army Corps of Engineers Waterways Experiment Station, the State of Texas Attorney General’s Office, and the American Water Works Association. Tr. 815:7-21. In recognition of his contributions to the field of hydrology, he has received numerous high honors and distinctions in the profession. Tr. 818:16-819:22.

26 changes are still in place, the flooding pattern recognized by the court in Phase I should

continue to persist. See Tr. 897:17-898:11 (“There has been no effort to deconstruct

physical channel changes or reinstitute a flood control first priority for require operations,

so a hydrologist would expect that new flooding pattern to persist.”). In this connection,

the court notes that the government admits that the MRRP is ongoing and has remained a

viable and active program at all times since 2014. Tr. 33:12-15 (admissions that the

MRRP “has continued and remained a viable and active program at all times since

2014”); see also Tr. 1400:11-13, 1402:16-17 (Remus); Tr. 1494:14-1495:14 (Pridal); Tr.

1538:9-11 (Shumate).

To determine whether the more severe flooding recognized by the court in the

Phase I opinion for the representative properties persists, Dr. Mays conducted three

quantitative studies and one qualitative study. Tr. 849:17-23; Tr. 891:16-892:1. First,

Dr. Mays analyzed the recorded annual peak discharges and corresponding gage heights

available at United States Geological Survey (“USGS”) gaging stations near each of the

representative properties. Tr. 850:12-22. This analysis allowed Dr. Mays to “illustrate

the patterns of flooding for each of the bellwether properties before and after the MRRP,

and demonstrate[] that the increased flooding of the properties is a continuance of the

newer post-MRRP flooding pattern.” Id.; see PX3413, PX3415, PX3419 (tables ranking

peak discharges and noting approximate return periods). Dr. Mays explained that the

available data would not allow him to measure the duration or depth of increased

flooding attributable to the MRRP. Tr. 851:3-11; Tr. 979:17-23. However, he explained

27 that the data provides “highly useful information regarding flooding patterns.” Tr.

851:10-11.

Specifically, Dr. Mays’ examined the peak discharges at the USGS gages near the

representative plaintiffs’ properties from 1980 through 2018. See Tr. 851:16-25. Peak

discharges “are the largest discharge that occurred in each of those respective years.” Tr.

852:6-8. Dr. Mays then sorted these peak discharges by magnitude. Tr. 851:24-25. Dr.

Mays noted whether each representative property experienced flooding in that year and

determined that, starting in 2007, every representative property started flooding during

years with peak flows at lower levels, where the properties had not historically flooded.

See Tr. 853:7-21. A review of this data led Dr. Mays to conclude that the pattern of

flooding in 2015 to 2018 is consistent with the post-MRRP flooding pattern after 2007.

Tr. 864:18-865:10 (summarizing conclusion for all properties).

Second, Dr. Mays conducted a frequency analysis to determine “ranges of return

periods for . . . USGS peak discharges” at the USGS gages closest to the three

representative properties. Tr. 867:5-19. Return periods represent “how frequently a

given flow is expected to occur or return at a particular location.” Tr. 866:19-22. For

example, Dr. Mays took the smallest peak discharges that resulted in flooding at the

Ideker property. Using USGS peak discharge data from 1970 to 2018,10 Tr. 868:19-21,

10 Dr. Mays examined the flows post-1970 to reflect the completion of the Mainstem System and BSNP. Dr. Mays also considered other year ranges to determine any potential impact in wet or dry cycles and concluded that his results for return periods were not unduly influenced by wet or dry cycles. Tr. 3019:18-3020:10.

28 Dr. Mays then calculated a return period of 1.5 years for discharges with this smallest

peak discharge range for the Ideker property. Tr. 870:16-22. In other words, discharges

within this range can be expected to occur every 1.5 years on the Ideker property. Tr.

870:23-871:1. Conducting this analysis for all properties, based on the dataset of peak

flows from 1970 to 2018, Dr. Mays determined that the return period for flooding tied to

the Corps’ MRRP actions at the Ideker, Buffalo Hollow, and Adkins representative tracts

were 1.5, 2.0, and 1.5 years, respectively. Tr. 871:11-12 (Ideker); Tr. 873:6-11 (Buffalo

Hollow); Tr. 874:20-25 (Adkins).

After a review of the peak discharge data, Dr. Mays concluded that “while the

bellwether properties used to flood relatively infrequently,” the “frequency of flooding

has dramatically changed.” Tr. 876:6-15. Now, “the MRRP-related flooding pattern that

started in 2007 causes flooding to occur significantly more frequently on the bellwether

properties compared to the pre-MRRP time period.” Tr. 877:2-7. In support of his

opinion, Dr. Mays noted that a 2003 Corps Flow Frequency Study,11 see DX1097,

DX1202 (appendices to the Flow Frequency Study), “obtain[ed] largely the same

results,” and that the 2003 Study “does not materially change” his conclusions. Tr.

877:16-22.

Third, Dr. Mays conducted a hydrologic risk analysis building off of his frequency

analysis. Dr. Mays’ hydrologic risk analysis was conducted to determine the likelihood

11 This study updated the flow and stage frequency for the Missouri River from Gavins Point Dam to the mouth at St. Louis. The study was a continuation of previous Corps efforts to update flood risk within the basin. Tr. 1448:2-7 (Pridal). 29 that, given a particular return period, a flood will occur in any given year using a

mathematical equation. Tr. 884:20-885:7; Tr. 886:2-20. Based on the frequency

analysis, Dr. Mays reiterated that probabilities of flooding reflect that, in the post-MRRP

flooding pattern, the “[b]ellwether properties cannot expect to go for even two years

without flooding, a marked departure from pre-MRRP flooding pattern.” Tr. 891:6-11.

Applying those return periods, Dr. Mays opined that for the Adkins and Ideker properties,

there was a 90 percent chance those properties would flood every two years, and a 75

percent chance the Buffalo Hollow property would flood every two years. Tr. 889:11-14

(Ideker); Tr. 890:6-9 (Buffalo Hollow); Tr. 890:20-23 (Adkins).

Finally, Dr. Mays conducted a qualitative risk analysis using a “generic risk

matrix.” Tr. 892:12-13. Dr. Mays opined that the representative properties were

experiencing “tolerable” risk pre-MRRP but “intolerable” risk post-MRRP because of the

frequent and more extensive flooding on the properties. Tr. 895:5-8.

The court finds that Dr. Mays’ analyses described above all confirm the

representative plaintiffs’ account of increased flooding post-2014 attributable to the

MRRP and establish that there has been a changed pattern of increased flooding

following the Corps’ implementation of the MRRP, as compared to the period of time

before the implementation of the MRRP but after the completion of the Mainstem System

and BSNP. Based on the plaintiffs’ and Dr. Mays’ opinion testimony, the court finds that

the government, in implementing the MRRP, has increased the frequency of flooding on

plaintiffs’ representative tracts as compared to the pre-MRRP period.

30 4. The Government’s Evidence Does Not Undermine, and at Times Supports, Plaintiffs’ and Dr. Mays’ Testimony

For the reasons that follow, the court finds that the government’s evidence

challenging Dr. Mays’ conclusions does not undermine the court’s finding that the

MRRP has permanently increased the frequency of flooding from what plaintiffs would

have experienced without the MRRP and that this increased flooding reflects a new and

ongoing pattern of flooding attributable to the MRRP at the representative properties.

First, the court rejects the government’s argument that to establish causation for

post-2014 and continued flooding, Dr. Mays was bound to rely on the same analysis the

plaintiffs used in Phase I, specifically, Dr. Christensen’s data and opinions regarding

changed WSEs. See Def.’s Br. at 65 (“Thus, to properly evaluate any ‘consistency’ with

the incremental flooding caused by the Corps’ actions from 2007 through 2014, Dr. Mays

would have to have evaluated for 2015-2018 the same factors that experts evaluated in

Phase I . . . .”); id. at 139 (arguing that plaintiffs “have not adduced competent evidence

that could show that the United States caused flooding experienced in those years under

the standard the Court laid out in Phase I”). Based on Dr. Christensen’s modeling

analysis, the court in Phase I had already determined that the MRRP resulted in increased

and more severe flooding from 2007 to 2014 for these representative plaintiffs. The

representative plaintiffs were not required to repeat this analysis in Phase II to prove that

the MRRP caused increased and more severe flooding post-2014. Dr. Mays’ testimony

persuasively builds off Dr. Christensen’s analysis and the court’s related findings in

Phase I. Dr. Mays explained that his opinion builds off of the court’s findings in Phase I

31 by examining whether there is a continued, changed flooding pattern based on those

findings. Tr. 819:25-823:18. As described above, he did this by demonstrating that

flooding was occurring on plaintiffs’ representative properties at lower peak discharges

than previously associated with flooding on those properties pre-MRRP. Tr. 820:20-

821:8. Having incorporated the court’s findings from Phase I into his analysis, the court

disagrees with the government that Dr. Mays’ analysis did not properly evaluate the

consistency of flooding from 2014 through 2018.

Second, the testimony of the government’s experts Dr. Robert Holmes, Mr.

Jonathan Jones, and Dr. Andrew Earles do not persuasively show that the MRRP has not

caused a new and ongoing pattern of increased flooding. See Def.’s Br. at 49-63 (Holmes

and Jones), 68-72 (Earles). The government first presented testimony from Dr. Holmes,12

who modeled daily WSEs at the representative tracts between 1950 and 2018. Tr.

1592:13-1593:2. Dr. Holmes used an interpolation method similar to that used by Dr.

Christensen in Phase I. Tr. 1592:13-1593:2 (“In assigning this task, [the] Department of

Justice asked me to use the interpolation model, similar to the approach used by Dr.

Christensen in Phase I . . . so I built an interpolation model similar to Dr. Christensen’s

but not exactly like Dr. Christensen’s to provide a consistent methodology for estimates

12 Dr. Holmes has spent his entire 33-year career with USGS, concentrating his efforts on the study of river hydrodynamics. Tr. 1583:3-7. Dr. Holmes currently serves as the Chief of the Branch of Hydrodynamics for the USGS. Tr. 1584:4-6. He previously served as the agency’s National Flood Hazard Coordinator. Tr. 1584:16-17. Dr. Holmes earned a B.S. and M.S. from the University of Missouri-Rolla (now the Missouri University of Science and Technology, where he is on faculty) and a Ph.D. from the University of Illinois. Tr. 1586:14-22, 1587:25- 1588:6; Tr. 1588:22-1589:4.

32 for that entire time period.”); Tr. 1594:2-6 (“Although I was asked by DOJ to use an

interpolation method similar to Christensen in Phase I, I did not use the exact manner or

relationship that Dr. Christensen did. But, rather, I built my model from the ground

up.”); Tr. 1688:6-1690:11 (comparing Dr. Holmes’ interpolation model with Dr.

Christensen’s interpolation model).13

Dr. Holmes modeled WSEs at the representative properties using historic daily

WSEs at USGS gages, historic daily discharges at those gages, and water surface profiles

from the Corps’ 2003 Flow Frequency Study. Tr. 1594:7-24. A water surface profile is a

measure of how WSEs change along a reach of river. Tr. 1595:2-4.

To measure WSEs at a point at the representative properties, Dr. Holmes

interpolated between gages upstream and downstream of the properties using a “proration

scheme.” Tr. 1631:7-13; see also Tr. 1647:22-1654:9 (describing methodology behind

“proration scheme”). This proration interpolation model was built using water surface

profile data from the Corps’ 2003 Flow Frequency Study. Tr. 1648:1-9. That 2003

Study in turn used bathymetric and terrain data from 1994 to 1999. See Tr. 1721:24-

1722:10; PX3702 (excerpt of Phase I government witness expert rebuttal report stating

13 The results of Dr. Holmes’ analysis produced similar results to Dr. Christensen’s analysis in Phase I. Tr. 1689:2-6. Dr. Holmes confirmed that the approach taken by Dr. Christensen to model the WSEs at the representative tracts in Phase I was sound. Tr. 1700:4-8 (“The interpolation model has its warts, it has periods of underprediction and overprediction, but, you know, I can’t guarantee that I would do a lot better with the HEC-RAS model.”); Tr. 1700:12-17 (“I find that the interpolation model used in this analysis provides overall reasonable estimates of daily water surface elevations at the representative Plaintiff properties from 1950 to 2018.”). Dr. Christensen, however, did not have later Corps studies and thus explained that his modeling of WSEs at the representative properties was an underestimate. See Tr. 2252:23-2253:13 (discussing Phase I Tr. 4769:4-10).

33 that the Corps’ profiles are “based on bathymetric and terrain data from 1994 to 1999 and

calibrated to historical data from 2003 and earlier”). Using the 2003 Study, Dr. Holmes

“back calculated what would be the proration value that would allow [him] to match” the

Corps’ data. Tr. 1648:18-24. Dr. Holmes also used historic water surface profile data

from the Corps’ records regarding historic floods to “back compute[] what the proration

value of [the] model would have to be” to simulate the flood elevation at the plaintiffs’

properties for these historic floods. Tr. 1650:19-1652:7. Dr. Holmes then used this

proration model and the WSEs at the upstream and downstream bounding gages to

compute the WSEs on the plaintiffs’ properties on a given day. Tr. 1664:6-1666:9.

Dr. Holmes also conducted assessments of how his model performed at other

stream gages on the River near the plaintiffs’ representative properties to compare the

model estimates against the real world observed data at those gages. Tr. 1674:21-1681:2;

see DX6065, DX6066, DX6067 (plotting differences between modeled and actual data).

At the Brownville gage closest to the Ideker property, for example, Dr. Holmes opined

that the average difference between his model and the actual data at the Brownville gage

was 0.9 feet, meaning that, on average, the model is underestimating the elevation by 0.9

feet. Tr. 1680:17-1681:9. Although acknowledging that the model was not perfect,

based on his assessments of the model, Dr. Holmes opined that the model “provides

overall reasonable estimates of daily water surface elevations at the representative

Plaintiff properties from 1950 to 2018.” Tr. 1700:12-17.

34 Using Dr. Holmes’ WSEs and other data, Mr. Jones14 was then tasked with

determining whether the “high-water events that occurred at the representative properties

from 2004 to 2018 were of comparable frequency, magnitude and duration to events

experienced historically.” Tr. 1838:17-21. To analyze frequency, Mr. Jones compared

Dr. Holmes’ modeled WSEs against key threshold elevations on the representative tracts,

such as the top of the bank and the drain outlets on the properties. Tr. 1843:22-1845:4.

With this comparison, Mr. Jones determined the number of days that the modeled WSEs

exceeded the key threshold elevations from 1950 to 2018 on the representative tracts. Tr.

1849:23-1850:5, 1853:11-17 (Adkins); Tr. 1863:20-1864:12 (Ideker); Tr. 1866:24-

1867:16 (Buffalo Hollow). Mr. Jones concluded that “before 2004, clusters of high-

water events occurred during 15-year periods at every representative property . . . that

were similar to the number of events experienced during the 2004 to 2018 period.” Tr.

1872:11-17.

To analyze duration and magnitude, Mr. Jones used the same data to “rank all

years between 1967 and 2018, in terms of the duration of time that each threshold

elevation was exceeded in each year.” Tr. 1909:17-1910:19. Mr. Jones also calculated

“the percentage of days that each threshold elevation was exceeded for 1983 to [19]97,

14 Mr. Jones is the CEO of Wright Water Engineers. DX4840-A (Jones CV). Mr. Jones holds B.S. and M.E. degrees from the University of Virginia. Id. He is a registered professional engineer in nineteen states. Id. He is also a registered professional hydrologist with the American Institute of Hydrology and a diplomate of water resources engineering with the American Academy of Water Resources Engineers. Id.

35 2004 to [20]18, and then 1967 to 2018.”15 Id. Mr. Jones concluded that the “1983 to

[19]97 period saw high-water events of comparable magnitude and durations to the high-

water events that occurred from 2014 to [20]18 at each representative property.” Tr.

1927:12-17. In other words, according to Mr. Jones, “[n]o overarching trend has been

identified at any of the representative properties for critical elevations that show one

period to have consistently higher duration events than the other.” Tr. 1927:18-22.

Based on Dr. Holmes’ and Mr. Jones’ analyses, the government argues that “high

water events of similar frequency, duration, and magnitude” to the post-MRRP flooding

“occurred at the representative properties between 1967 and 2004.” Def.’s Br. at 49. For

the following reasons, the court disagrees.

To begin, the court notes that neither Dr. Holmes nor Mr. Jones evaluated or tied

their analyses to any actual flooding at the plaintiffs’ representative properties. Dr.

Holmes stated that his modeling was not for the purpose of predicting or determining

when flooding occurred on the representative properties. Tr. 1766:22-1768:5. Mr. Jones

also did not do any analysis to determine instances of actual flooding at the representative

properties and testified that he was not offering any opinions on whether actual flooding

occurred at any of the representative properties. Tr. 2007:20-2008:9; Tr. 2010:22-25; Tr.

2061:15-16; Tr. 3013:23-25 (Dr. Mays’ rebuttal testimony noting that Mr. Jones’ analysis

15 Mr. Jones also compared “annual stage hydrographs for the relevant post-2004 years” against “historical stage hydrographs qualitatively,” Tr. 1910:14-19, and found general “comparability,” Tr. 1927:5-10, of WSE trends on plaintiffs’ properties during select high water years pre- and post-MRRP.

36 “completely ignores the bellwether Plaintiffs’ testimony about when they are actually

flooding”).

In addition, Dr. Holmes’ modeled WSEs – relied on by Mr. Jones – were based on

pre-MRRP data, even though more recent data was available. As noted above, to create

his modeled WSEs, Dr. Holmes relied on water surface profiles sourced from the Corps’

2003 Flow Frequency Study, which Dr. Holmes stated are “the primary way that a model

reflects channel geometry and morphology at localized spots within a river reach.” Tr.

1705:25-1706:9. That 2003 Study used data from 1994 to 1999 to create water surface

profiles. Tr. 1720:19-1722:23; see also 1714:2-1715:11 (use of data for Adkins);

1715:12-23 (use of data for Ideker); 1715:24-1716:13 (use of data for Buffalo Hollow).

Because the Corps’ actions on the MRRP began after 2004, the data from the 2003 Flow

Frequency Study does not reflect any changes resulting from the implementation of the

MRRP and does not paint an accurate picture of how the River has changed since then or

how the MRRP has affected flooding at the representative properties.16

Moreover, Dr. Holmes’ model relied on the 2003 Study data despite the fact that

the Corps had prepared updated post-2004 water surface profiles that could have yielded

16 The court notes that in Phase I, Dr. Christensen relied on the Corps’ 2003 Flow Frequency Study to support his analysis of WSEs. Dr. Christensen, however, testified that he was underestimating flooding at individual properties precisely because he did not have more recent data on the River changes post-MRRP. See Tr. 1719:24-1721:11; see also Tr. 2552:23-2253:13 (discussing Phase I Tr. 4769:4-10). In Phase I, the court found that Dr. Christensen provided reliable estimates for purposes of the causation analysis, but explained that this model was conservative. Ideker I, 136 Fed. Cl. at 696 (noting that Dr. Christensen’s “interpolation of water surface levels was probably conservative because it would not reflect higher WSEs at a particular property”). In other words, Dr. Christensen’s interpolation model provided reliable minimum estimates of the effect of the MRRP.

37 a more accurate analysis. See PX3718 (comparing 2003 Study with later studies);

PX3716 (Corps’ updated May 2015 HEC-RAS Report, Appendix D). In Phase I, the

government’s expert Mr. Mark Woodbury opined in his rebuttal report that use of this

more recent data would have produced more accurate results. See PX3702 (excerpts

from Phase I Woodbury rebuttal report, noting “updated models” that “were built using

bathymetry data from after the 2011 flood to best represent current state of the channel

bed”); Tr. 1722:2-23. But Dr. Holmes testified that he was not given access to any of the

updated hydraulic modeling that incorporates the modifications to the River channel

made pursuant to the MRRP. Tr. 1722:2-23.

Dr. Mays persuasively explained that, by relying on the 2003 Study data in his

proration scheme, Dr. Holmes’ model predictably underestimated the WSEs in the post-

MRRP world because it did not incorporate post-MRRP changes to the geometry and

roughness of the river. Tr. 2992:12-2993:12. To show the model’s tendency to

underpredict, Dr. Mays separated out and analyzed how Dr. Holmes’ model performed

only during days where the representative properties experienced MRRP flooding. Tr.

2994:14-3000:15. Dr. Mays found that, of 50 data points at the Brownville gage, nearby

Ideker, 49 showed under-predictions with an average under-prediction of almost two feet.

Tr. 2997:9-11; see also PX3496 (plotting differences between modeled and observed

WSEs for the range of MRRP flooding).

At the Oregon gage, which is across the River from Buffalo Hollow, Dr. Mays

compared Dr. Holmes’ modeled WSEs at Buffalo Hollow with actual recorded WSEs at

the Oregon gage for the three years where data was available (2009, 2010, and part of

38 2011). Dr. Mays opined that Dr. Holmes’ model “is consistently off by over three feet

during high flow conditions.” Tr. 3001:18-3002:3. Dr. Mays further testified that the

mile distance between the Oregon gage and the Buffalo Hollow property would not

explain this discrepancy. Tr. 3004:17-3005:2. These differences measured in feet

between Dr. Holmes’ model and the actual recorded data are troubling, where plaintiffs

in Phase I testified that inches of water can be the difference between flooding and no

flooding, for example, by blocking drainage pipes into the River. See PX3710 at 2873

(Mar. 2017 Phase I Tr.); PX3711 at 4164 (Apr. 2017 Phase I Tr.). Dr. Holmes did not

disagree with this testimony. Tr. 1765:14-1766:11.

In addition, Dr. Holmes acknowledged that the gages he examined are not capable

of picking up specific WSE changes associated with MRRP projects nearest to the

plaintiffs’ representative tracts. Tr. 1629:3-10 (“[W]hile the streamgages – or a

streamgage may not sense the exact effect from a MRRP project miles downstream near a

bellwether property, it does sense those MRRP projects nearest to the gage.”). For

reference, the Adkins tract is five miles downstream from the nearest gage, the Ideker

tract is 10 miles downstream from the nearest gage, and the Buffalo Hollow tract is 20

miles downstream from the nearest gage. Tr. 846:3-8 (Mays, on Adkins); Tr. 841:10-16

(Mays, on Ideker); Tr. 843:24-844:1 (Mays, on Buffalo Hollow). And each tract is near

multiple BSNP structures, MRRP sites, and dike notches that can impact flooding. See

Tr. 847:2-11 (Mays, on Adkins); Tr. 842:16-20 (Mays, on Ideker); Tr. 845:9-23 (Mays,

on Buffalo Hollow). While Dr. Holmes believed that the model’s failure to pick up

increases at the representative properties would be made up for by the gages picking up

39 different WSE increases from MRRP projects closer to the upstream and downstream

gages, he did no study of how much WSE increases result from different MRRP projects

or types of projects, nor did he study where MRRP projects are relative to the gages. Tr.

1751:10-1752:5. Dr. Mays persuasively testified in Phase II, and the court accepted Dr.

Christensen’s testimony in Phase I, that, for this reason, an interpolation model like Dr.

Holmes’ would likely underpredict WSEs at the representative tracts. Tr. 2983:1-10;

Ideker I, 136 Fed. Cl. at 696 (citing Dr. Christensen’s explanation that “his interpolation

of water surface levels was probably conservative” because none of the MRRP projects

are located at a specific gage location).

Because Mr. Jones relied on Dr. Holmes’ analysis, which likely underpredicted

WSEs at the representative properties, the court further finds that Mr. Jones’ opinions are

unpersuasive. For these reasons, the court finds that Mr. Jones’ and Dr. Holmes’

analyses do not call into question the court’s findings based on the plaintiffs’ testimony

and Dr. Mays’ opinion testimony.

The court also finds the opinions of Dr. Andrew Earles,17 the government expert

who challenged Dr. Mays’ opinions, to be unpersuasive. Dr. Earles was tasked with

providing a rebuttal to Dr. Mays’ expert opinions. Tr. 2171:4-6. Dr. Earles did not

perform his own frequency study. Tr. 2270:2-5.

17 Dr. Earles is the vice-president of water resources for Wright Water Engineers. Tr. 2172:21- 24. Dr. Earles holds a Bachelor’s degree in civil engineering from Stanford University and Master’s and Ph.D. degrees in civil and environmental engineering from the University of Virginia. Tr. 2172:11-15. 40 Dr. Earles focused much of his criticism of Dr. Mays’ opinion regarding flow

frequency on the ground that Dr. Mays had failed to follow all aspects of a USGS

publication known as Bulletin 17C to develop his flood flow frequency curves and

calculate his recurrence intervals. See Tr. 2199:1-4 (testifying that Dr. Mays “did not

conduct any of the necessary procedures” for his analysis). Bulletin 17C “provides the

standard guidelines for conducting flow frequency analysis in the United States.” Tr.

2196:17-19. Dr. Earles opined that Dr. Mays’ analysis “incorrectly applied Bulletin 17C

methods to regulated data without making any adjustments,” and that, as a result, Dr.

Mays’ calculations were unreliable. Tr. 2207:6-9.

On cross-examination, Dr. Earles admitted, however, that Bulletin 17C expressly

states there is no national guidance on how to develop flood frequency curves for

regulated rivers. Tr. 2324:24-2325:4; see also PX3749 at 35-36 (USGS Bulletin 17C).

There is no dispute that the Missouri River is a regulated river. Tr. 2322:12-2324:21. Dr.

Mays opined that Bulletin 17C need not be used to calculate reliable recurrence intervals

and merely provides guidelines. Tr. 3015:20-3016:13.

Dr. Earles also challenged Dr. Mays’ analysis because he did not follow the steps

for developing flow frequency relationships used in the Corps’ 2003 Study. Tr. 2218:12-

2221:12. In its study, the Corps developed regulated and unregulated flow relationships

to quantify the value of flood protection that the Mainstem System provides as well as to

incorporate both regulated and unregulated historical data in its analysis. Tr. 2220:16-25.

However, Dr. Mays opined that it was unnecessary to take these steps in this case, and

that to do so would only add uncertainty to his analysis, which included exclusively

41 regulated river data. Tr. 3018:11-3019:12. Dr. Mays also noted that the Corps itself does

not always rely on such regulated and unregulated flow relationships when it is analyzing

exclusively regulated data. Tr. 3017:7-3018:7.

Dr. Earles also criticized Dr. Mays’ recurrence intervals because they differed

from the recurrence intervals outlined in the Corps’ 2003 Study. Tr. 2238:5-2242:2.

However, the Corps’ 2003 Study used data from 1897 to 1997, while Dr. Mays used data

from 1970 to 2018. Tr. 2291:25-2292:3. Dr. Earles stated that he had no knowledge of

the effect of the use of the different datasets on the recurrence interval calculation. Tr.

2295:18-24 (“I have not had an opportunity to do a comparison of a dataset properly

analyzed from 1970 to 2019 with the flow frequency study.”); Tr. 2296:19-23 (Dr. Earles

agreeing that he had “not done any analysis to see what the effect is on recurrence

interval calculation if one includes water years subsequent to . . . 2011”).

Dr. Earles further criticized Dr. Mays for using outdated and inappropriate

“regional skew data” in his analysis, which produced a “less accurate fit of the data.” Tr.

2213:12-20; Tr. 2218:5-10. But, again, Dr. Earles did not conduct his own flood

frequency study and did not opine what the effect of using this outdated data would be.

Dr. Mays persuasively concluded that, even if any of Dr. Earles’ proposed extra steps

were added to his methodology, it is impossible that the effect would have been

sufficiently material to alter his conclusions about a changed flooding pattern along the

River due to the MRRP. Tr. 3020:16-3021:9 (Dr. Mays stating that “[n]one of Dr.

Earles’ criticisms affect my confidence in my conclusions” and observing that Dr. Earles

42 did not “offer any analysis to quantify how my thresholds would have been affected had I

incorporated into my analysis any of the changes he opined I should have made”).

Dr. Earles also criticized Dr. Mays for not aligning his analysis with the

representative plaintiffs’ testimony regarding flooding. However, the only example Dr.

Earles offered was Mr. Adkins’ testimony regarding whether flooding occurred in 1995

or 1996. Tr. 2253:25-2254:17; see also Tr. 2304:14-2312:19 (discussing Dr. Mays’

testimony regarding flooding on the Adkins property and affirming that the Adkins

testimony is the only example of “improper reliance” on plaintiffs’ testimony regarding

historical flooding included). Dr. Mays addressed and corrected this issue in his opening

testimony. See Tr. 2304:14-2312:19 (cross examination of Dr. Earles discussing Dr.

Mays’ opening testimony).

Finally, Dr. Earles criticized Dr. Mays’ “calculations about the changes in

flooding patterns and future flooding” for using small sample sizes and for failing to

account for “wet and dry cycles in the Missouri Basin.” Tr. 2258:5-10. However, Dr.

Mays credibly testified that he performed separate analyses on multiple different time

frames and was thus “able to test whether I was getting unreliable results that were

unduly influenced by a wet or dry cycle or some other time-limited phenomenon.” Tr.

3019:15-3020:14.

The court therefore finds that Dr. Earles’ criticisms of Dr. Mays do not undermine

Dr. Mays’ opinions. The court agrees that Dr. Mays was not bound to follow all of the

steps in Bulletin 17C or those taken by the Corps in the 2003 Study and, while offering

criticisms of Dr. Mays’ calculations, the court notes that Dr. Earles did not provide his

43 own flow frequency study and could not say whether his criticisms would have had a

material effect on Dr. Mays’ conclusions. Dr. Mays also provided reasoned responses to

Dr. Earles’ criticisms. The court finds Dr. Mays’ opinions more persuasive.

Finally, the government’s own hydrology expert Dr. Jeffery Bradley, an expert

identified for trial but never called,18 stated in his deposition testimony that flooding

events have become more common and of longer duration in the post-MRRP timeframe,

as the representative plaintiffs claim. See PX3740 at 78-83 (excerpt from June 19, 2020

Bradley Depo.); Tr. 2078:8-2084:7 (reading excerpts of Bradley deposition testimony

into the record). Although Dr. Bradley characterized the increased recurrence intervals as

“minor,” Dr. Bradley’s testimony is nonetheless consistent with Dr. Mays’ opinions.

PX3740 at 80. Dr. Bradley also opined that Dr. Holmes’ use of outdated water surface

profiles was not best practice, consistent with Dr. Mays’ criticisms of Dr. Holmes and

Mr. Jones. PX3738 at 30-34 (excerpt from June 19, 2020 Bradley Depo.); PX3739 at 36

(excerpt from June 19, 2020 Bradley Depo.).

Having found that the government’s evidence challenging the representative

plaintiffs’ lay testimony and expert evidence of increased flooding post-2014 is not

persuasive (and even, at times, supports plaintiffs’ claims), the court concludes that the

plaintiffs have met their burden of demonstrating that the MRRP has caused increased

18 Although the government did not present Dr. Bradley as a trial witness, the court acknowledges that Dr. Bradley’s deposition testimony constitutes an admission against the government and is admissible as such. See Tr. 2491:7-2492:13.

44 flooding for the years at issue in both Phase I and Phase II and has caused a changed

pattern of increased flooding on the plaintiffs’ representative properties.

II. Liability for a Permanent Flowage Easement Under the Remaining Arkansas Game & Fish Factors

Against the backdrop of the above findings on causation, the court now turns to

whether the flooding described above satisfies the criteria for establishing a taking of a

flowage easement under Arkansas Game & Fish. In that case, the Supreme Court

identified the following factors that the court must consider in determining whether

government action that causes intermittent flooding amounts to the taking of a flowage

easement: severity, duration, intent or foreseeability, character of the land, and

reasonable investment-backed expectations. See Ark. Game & Fish, 568 U.S. at 38-39;

see also Ridge Line, 346 F.3d at 1355-56 (considering intent or foreseeability and “the

nature and magnitude of the government action”). Although these factors are related, the

court will examine them separately to determine whether the representative plaintiffs

have established each.19

A. Severity

One factor that the court must consider in the liability determination under

Arkansas Game & Fish is the “[s]everity of the interference.” 568 U.S. at 39. This factor

requires the court to assess whether the intermittent flooding at issue was “substantial”

enough to rise to the level of a taking. Ridge Line, 346 F.3d at 1355. In doing so, the

19 Although the government in closing argument suggested that plaintiffs need not establish each factor in order to show a taking occurred, Tr. 3452:18-3453:20, the court finds that the requisite showing for all six factors has been met and therefore does not reach that issue. 45 court must determine whether “after the [government action] began the flooding lasted

for significantly longer periods of time and had much more serious consequences than the

flooding” during the period before the government action. Ark. Game & Fish v. United

States, 736 F.3d 1364, 1374 (Fed. Cir. 2013). In other words, the court must determine

whether or not “the asserted intrusion was within a range that the property owner could

have reasonably expected to experience in the natural course of things.” Id. at 1375.

Although “it may often be difficult to say, in the abstract, whether a particular intrusion is

severe or only incremental in nature” the “consideration of the effects of the intrusion on

the property owner will often make that distinction easier to draw.” Id.

While the government acknowledges that the court found that flooding on the

three representative properties caused by the MRRP was “severe,” the government

contends that “the severity of any additional incremental flooding” caused by the MRRP

is “relatively small,” and thus does not meet the “severity” prong of the Arkansas Game

& Fish test. Def.’s Br. at 148. The court disagrees. Based on the testimony of the

representative plaintiffs and Dr. Mays, the court concluded above that the Corps’ ongoing

implementation of the MRRP has caused a new pattern of increased flooding on the

representative properties beginning in 2007. The court further finds that the Corps’

actions have resulted in more severe flooding than plaintiffs experienced pre-MRRP.

Following implementation of the MRRP, the return periods associated with flooding are

now likely to occur every 2 years. Tr. 889:11-14 (Mays, on Ideker); Tr. 890:6-9 (Mays,

on Buffalo Hollow); Tr. 890:20-23 (Mays, on Adkins); Tr. 891:6-9 (“The probabilities of

flooding reflect that, in the new flooding pattern, the bellwether properties cannot expect

46 to go for even two years without flooding.”). Moreover, the plaintiffs’ testimony

established that flooding on the plaintiffs’ representative tracts is far more frequent and

damaging than they had experienced before implementation of the MRRP, see supra Part

I.B & I.C, and is outside the “range that [the representative plaintiffs] could have

reasonably expected to experience,” Ark. Game & Fish, 736 F.3d at 1375.

In addition, as discussed in more detail below, the post-MRRP flooding has had

considerable effects on the plaintiffs’ crop yields. See infra Part II.E. This is a “serious

consequence[]” of the MRRP flooding, further evidencing that the flooding was “severe.”

Ark. Game & Fish, 736 F.3d at 1374. The court therefore concludes that the plaintiffs

have met their burden of establishing that the increased and repeated flooding attributable

to the MRRP is sufficiently severe for each representative plaintiff to demonstrate the

Arkansas Game & Fish severity factor.

B. Duration

The time and duration of the government invasion is also an important

consideration in takings cases based on intermittent flooding. See Ark. Game & Fish,

568 U.S. at 38-39. This factor is often relevant in the context of a temporary taking, as

was the case in Arkansas Game & Fish. In this case, however, both parties agree that if

the court finds that a taking occurred, that taking is permanent. Def.’s Br. at 158; Pls.’

Br. at 119-20. This court has previously held that where the government action that

causes intermittent flooding will continue into “the foreseeable future,” the time and

duration factor is not measured by the length of time that water inundates the plaintiffs’

properties, but by the government’s “permanent right to inundate the property.” In re

47 Upstream Addicks and Barker (Texas) Flood-Control Reservoirs, 146 Fed. Cl. 219, 250

(2019). In other words, the permanent nature of the intermittent flooding means that the

duration factor “weighs in favor of plaintiffs.” Id. Nevertheless, in this case, the court

will consider both the duration of the more severe flooding attributable to the MRRP each

year, as well as the permanent nature of the MRRP itself.

As discussed above, the court heard undisputed testimony from the plaintiffs that

the flooding attributable to the MRRP lasted of sufficient duration, each year, to impact

their farming operations. See supra Part I.B & I.C. In addition, as discussed in more

detail below, the increased flooding attributable to the MRRP caused the representative

plaintiffs to lose crops that they would not have otherwise lost absent the incremental

effects of the MRRP. See infra Part II.E. The court thus concludes that the flooding each

year was of a duration substantial enough for the duration factor to weigh in plaintiffs’

favor.

Moreover, these significant invasions of increased flooding are not temporary or

isolated events. The representative plaintiffs have shown that more severe flooding will

often recur for the foreseeable future. The government concedes that the MRRP is

ongoing. See, e.g., Tr. 34:7-16 (admitting that the government has not removed or closed

any of the chutes constructed and/or opened in conjunction with the MRRP since 2014);

Tr. 1400:11-13, 1402:16-17 (Remus, confirming that the MRRP “still exists”); see also

Pls.’ Br. at 55-56. The plaintiffs can now anticipate that they will experience more

frequent flooding periods from increased WSEs than they would have experienced

without the MRRP. See infra Part I.C. The permanent nature of the increased flooding

48 also indicates that the duration factor weighs in favor of finding a taking. See In re

Upstream Addicks and Barker, 146 Fed. Cl. at 250.

C. Intent and Foreseeability

Arkansas Game & Fish also requires the court to evaluate “the degree to which the

invasion is intended or is the foreseeable result of authorized government action.” 568

U.S. at 39. An action is foreseeable if “the government should have predicted or foreseen

the resulting injury.” Moden v. United States, 404 F.3d 1335, 1343 (Fed. Cir. 2005). An

injury “may not be foreseeable if an intervening cause breaks the chain of causation.” Id.

at 1344.

In Phase I, the court determined that the increased flooding on plaintiffs’

properties was the foreseeable result of the System and River Changes made under the

MRRP. Ideker I, 136 Fed. Cl. at 761-62 (listing plaintiffs that have “established

causation, foreseeability and severity”); id. at 705 (“The rising WSEs described by Dr.

Hromadka and Dr. Christensen are the result of known processes that change the

hydrology and hydraulics of a river. When the Corps took combined actions to make the

River shallower and slower, rising WSEs were a natural, direct, and probable

consequence of the Corps’ actions.”); see also id. at 729-30 (Adkins), 747-49 (Ideker),

757-58 (Buffalo Hollow), 690 (finding that Dr. Christensen’s testimony was sufficient to

show foreseeability), 696-97 (concluding that, in light of Dr. Christensen’s and Dr.

Hromadka’s testimonies, the System and River Changes have “caused WSEs to rise

higher than they would have risen without these Changes and that this rise in WSEs has

led to more flooding or more severe or longer flooding than would have occurred had

49 these Changes not been made by the Corps”). Dr. Mays’ testimony regarding the

predictable increase in flooding because of the MRRP confirms the court’s Phase I

finding. Tr. 835:12-21 (testifying that it “makes perfect sense from a hydrological

perspective” that higher WSEs are the “predictable result” of the MRRP); Tr. 833:1-7

(testifying that “[t]he evidence has established that the river changes by the Corps . . .

have had the effect of raising Missouri River’s water surface elevations in periods of high

flows”); Tr. 833:8-834:4 (testifying that “[w]hether the Corps took combined actions to

make the river shallower and slower, arising water surface elevations were a natural,

direct, and probable sequence of the Corps’ actions”).

While the government continues to dispute that the MRRP has caused any change

in flooding, see, e.g., Def.’s Br. at 139 (“The flooding that occurred between 2007 and

2014 did not significantly alter the character of the land” and “was consistent with the

historical record”), the government did not present any specific evidence on

foreseeability in Phase II to counter the court’s Phase I finding that the flooding changes

on plaintiffs’ properties following implementation of the MRRP were a foreseeable

consequence of the Corps’ actions under the MRRP. As noted above, the government

also concedes that it is continuing to implement the MRRP and will do so into the future.

The court therefore concludes that the plaintiffs have shown that continued increased

flooding on the plaintiffs’ representative tracts is a foreseeable consequence of the Corps’

implementation of the MRRP.

50 D. Character of the Land

The court must also determine in the takings analysis whether the increase in

flooding “was great enough to change the character” of the land. Ark. Game & Fish, 736

F.3d at 1371. “The character of the land in government flooding cases is usually defined

by whether, inherently, the property is ‘especially susceptible to flooding.’” In re

Upstream Addicks and Barker, 146 Fed. Cl. at 248 n.18.

The government argues that the Corps’ actions under the MRRP did not

significantly alter the “character” of the representative plaintiffs’ land. Def.’s Br. at 139

(“The flooding that occurred between 2007 and 2014 did not significantly alter the

character of the land” and “was consistent with the historical record”). A great deal of

the evidence presented by the government in Phase II was centered on this factor. The

government presented as its first witness Dr. Ari Kelman, an environmental historian.

See Tr. 1246:18-1250:10. Dr. Kelman testified that properties next to the Missouri River

have always been subject to extensive flooding. Tr. 1253:12-15. In support, Dr. Kelman

presented evidence of flooding that occurred on the Missouri River beginning in July

1867 through the end of the twentieth century. See generally Tr. 1254:7-1266:20. Dr.

Kelman’s opinions were based solely on the review of documents concerning historical

flood events identified in public records; he did not consult the plaintiffs regarding their

experience of flooding on the representative tracts before and after the MRRP. Tr.

1312:24-1313:7.

The court finds that Dr. Kelman’s testimony is largely irrelevant. The court has

previously determined that, in evaluating whether the government is liable for a taking in

51 this case, the relevant time frame is from the period after the Mainstem System and

BSNP were completed, in the 1970s, and then the period following implementation of the

MRRP in 2004. Ideker III, 146 Fed. Cl. at 416. Given these bounds, the court finds

evidence regarding the character of the land that predates the Mainstem System and

BSNP to be of little value.

Moreover, the evidence presented by the representative plaintiffs in Phase II and

described above establishes that the character of plaintiffs’ land has changed. There is no

dispute that the representative plaintiffs’ land flooded prior to the MRRP; the plaintiffs’

themselves so testified. However, the court has concluded that the changes implemented

by the Corps under the MRRP caused more severe and frequent flooding than the

representative plaintiffs have historically experienced. The court therefore concludes that

the “character of the land” factor weighs in plaintiffs’ favor. See In re Upstream Addicks

and Barker, 146 Fed. Cl. at 286 n.18 (“That plaintiffs’ properties may be susceptible to

flooding during extreme weather events is of some relevance, but it is independent from

the fact that plaintiffs’ properties . . . only flooded in this case because of the

government’s construction of the” dams.).

The government’s additional argument that the character of the land has not

changed because the plaintiffs’ representative tracts were in a FEMA designated flood

hazard zone before the MRRP is without merit. See Def.’s Br. at 136. There is no

dispute that the plaintiffs’ properties are within a flood zone. The issue in this case is

whether the changes made by the Corps to implement the MRRP for the purpose of

protecting and promoting habitat for endangered species has increased the flooding on

52 plaintiffs’ property such that it is now experiencing more frequent and more severe

flooding post-MRRP. In Arkansas Game & Fish, the land at issue was also in a flood

zone, but the Supreme Court recognized that a taking can nonetheless occur. See 568

U.S. at 26-29 (noting that, prior to the government action, the area at issue flooded

approximately 65 days per year during the relevant period).

In sum, the court finds that the increased frequency, severity, and duration of

flooding post-MRRP demonstrates that the MRRP changed the character of the

representative tracts of land. It cannot be the case that land that experiences a new and

ongoing pattern of increased flooding does not undergo a change in character. The court

therefore finds that the plaintiffs have met their burden of demonstrating that the Corps’

actions under the MRRP are “great enough to change the character” of the representative

properties. Ark. Game & Fish, 736 F.3d at 1371.

E. Reasonable Investment-Backed Expectations

Finally, the court must consider a property owner’s “reasonable investment-

backed expectations regarding the land’s use” in the takings inquiry. Ark. Game & Fish,

568 U.S. at 39. An objective standard applies. Chancellor Manor v. United States, 331

F.3d 891, 904 (Fed. Cir. 2003). The “burden is on the owner to establish a reasonable

investment-backed expectation in the property at the time [the owner] made the

investment.” Cienega Gardens v. United States, 503 F.3d 1266, 1288 (Fed. Cir. 2007)

(citing Forest Props., Inc. v. United States, 177 F.3d 1360, 1367 (Fed. Cir. 1999)). The

court considers whether a plaintiff had expectations that were backed by investments,

whether those expectations were reasonable, and whether the government action

53 interfered with those expectations. Cienega Gardens, 503 F.3d at 1288-89; Palazzolo,

533 U.S. at 618.

The court notes that the concept of “reasonable investment-backed expectations”

is traditionally considered in a regulatory, not physical, takings analysis. Love Terminal

Partners, L.P. v. United States, 889 F.3d 1331, 1345 (Fed. Cir. 2018) (“The reasonable,

investment-backed expectation analysis is designed to account for property owners’

expectation that the regulatory regime in existence at the time of their acquisition will

remain in place, and that new, more restrictive legislation or regulations will not be

adopted.”). However, the Supreme Court in Arkansas Game & Fish recognized that

“reasonable investment-backed expectations” is also an appropriate factor to consider in a

flooding takings case. 568 U.S. at 39. The Supreme Court applied this factor to the

flooding before it, explaining that the flooded area at issue in Arkansas Game & Fish had

flooded previously, but that the flooding involved in the takings claim was of a different

kind than the plaintiffs had previously encountered. Id. Likewise, the question in this

case is whether plaintiffs had a reasonable expectation that the flooding pattern on their

properties after the Mainstem System and BSNP were in place would continue, whether

that expectation was backed by investment in plaintiffs’ properties, and whether that

expectation was interfered with by the Corps’ actions under the MRRP.

The representative plaintiffs argue that based on the pre-MRRP flooding patterns

of the River, the plaintiffs had reasonable expectations of being free from the increased

flooding caused by the MRRP. Pls.’ Br. at 146-52. Plaintiffs contend that “no

reasonable River Basin stakeholder, including the Plaintiffs, would have ever predicted

54 that the Corps, after decades of honoring a pre-emptive priority of flood control and after

spending billions of dollars to tame the River in order to induce people and businesses to

settle and invest in the Basin, would suddenly change cause and intentionally alter the

existing flooding patterns of the River to benefit the ecosystem and fish and wildlife.” Id.

at 150. Plaintiffs further argue that “there is also evidence in the record establishing that

each of the Phase II Plaintiffs invested in their property interests relying on their

reasonable expectations of flooding based upon the pre-MRRP flooding patterns.” Id. at

152.

The government contends that the representative plaintiffs have not carried their

burden of demonstrating reasonable investment-backed expectations because “it is not

clear” from the plaintiffs’ testimony “what specific expectations” the plaintiffs had about

the government’s actions or whether any investments were actually made based on those

expectations. Def.’s Br. at 141. The government also argues that because the

representative plaintiffs’ “land has always been subject to flooding, any investment-

backed expectations were not interfered with” by the Corps’ actions. Id. The

government contends that the representative plaintiffs could not have reasonably

expected that the Corps to continue its pre-2004 operations of the Mainstem System and

BSNP without change when the Corps had changed its operations in the past. Id. at 144.

The government points out that the Corps under federal law has been authorized since the

1980s to complete projects to mitigate BSNP habitat losses, and that these federal laws

were publicly available and widely known within the Missouri River Basin. Id. at 147.

55 For the reasons that follow, the court concludes that the plaintiffs have established

that they had reasonable investment-backed expectations that the pre-MRRP flooding

pattern would continue, and that the Corps’ actions under the MRRP interfered with those

expectations.

1. The Government’s State Law Arguments are Waived and Fail for Lack of Proof

As an initial matter, the court rejects the government’s argument, raised for the

first time in the government’s Phase II post-trial brief, that the plaintiffs’ reasonable

investment-backed expectations should be framed by the Corps’ rights as an upstream

riparian landowner under state law. See Def.’s Br. at 138-39 (in the context of the

character of the land factor), 143 (reasonable investment-backed expectations factor).

The government briefly argues in its post-trial brief that under the state law of Kansas,

Missouri, and Iowa, “those owning land along a river must anticipate that other

landowners will make reasonable use of their properties and the benefits and burdens of

riparian ownership will be shared with their neighbors.” Def.’s Br. at 139. The plaintiffs

contend, and the court agrees, that to the extent the government’s argument is a defense

to any takings liability for flooding plaintiffs’ properties, that defense has been waived.

Pls.’ Resp. at 41-43; see also RCFC 12(h)(2); Kontrick v. Ryan, 540 U.S. 443, 459 (2004)

(“A defense or objection that is not raised by motion or in the responsive pleading is

waived unless it is protected by Rules 12(h)(2) or 12(h)(3) or by the successful invocation

of the liberal amendment policy of Rule 15.” (quotation omitted)).

56 But even if not waived, the court finds that this argument fails for lack of proof.

Whether the Corps’ actions in implementing the MRRP even qualify as rights of a

riparian landowner under state law is unclear. Even if the government could make a

claim based on its rights as a riparian landowner, the government presented no evidence

and scant argument on this issue. For example, some of the footnoted cases cited by the

government in support of this argument apply a “reasonable use” rule in determining

whether an upstream landowner may take actions on that landowner’s own property even

though those actions may result in flooding damages to a neighboring property. See

Bettinger v. City of Springfield, 158 S.W.3d 814, 818 (Mo. Ct. App. 2005). The

government did not present any evidence describing what riparian rights it was exercising

and whether its actions met this “reasonableness” standard. The court has no basis for

determining whether the government’s actions would qualify under state riparian laws or

how these laws apply to the facts in this case.

The court will now turn to the evidence presented regarding reasonable

investment-backed expectations and the extent to which the government has interfered

with those expectations.

2. The Plaintiffs Made Substantial Investments in Farming Their Representative Properties in Reliance on the Flood Protection Provided by the Mainstem System and BSNP

Mr. Adkins testified that he grew up and has spent his life at or near the Phase II

tract. Tr. 48:14-17. Over time, the Adkins family invested in acquiring additional tracts

of land for farming. Tr. 76:6-7. Robert and Betty Adkins, Sr. settled on the Phase II tract

57 in 1948. Tr. 76:4-5. Mr. Adkins, his brother, and his parents all built homes next to each

other on the northern portion of the tract. Tr. 48:16-27. Mr. Adkins explained that after

the Mainstem dams were built, the Adkins family proceeded to invest millions of dollars

in the development and farming of the Phase II tract and the other land they own near the

River. Tr. 78:11-79:14. He testified that but for the government’s assurances that the

Mainstem System and the BSNP would provide enhanced flood protection, the Adkins

family would not have made these investments. Id. He testified that the Adkins family

relied on the government’s representations about flood protection. Id.

b. The Ideker property

Although Mr. Ideker conceded that the Ideker representative tract when acquired

was prone to flooding, he testified that the Ideker family acquired the property with the

expectation that the flood protection the government promised to provide with the

Mainstem System, the levees, and the BSNP would allow the property to be productive.

Tr. 221:19-222:19. He testified that the Idekers relied upon many government

publications and representations that flooding would be substantially reduced. Id. Mr.

Ideker further explained that the government encouraged citizens to invest in the

farmland that the Corps’ actions made possible. Tr. 220:20-24; Tr. 223:17-24; see also

PX1 (cross-section photo of Ideker property comparing September 1934 with March

1977). In reliance, the Ideker family began investing heavily to develop the land for

farming. Tr. 223:17-24.

Mr. Ideker testified that his family built private levees to protect their investment

from flooding. Tr. 225:10-226:14. He stated that the levees represent an investment of

58 millions of dollars. Id. Additionally, he testified that Ideker Farms has invested

hundreds of thousands of dollars in equipment, including irrigation equipment, in

planting beans and corn, and harvesting crops. Id. He explained that drainage ditches

and two drainage conduits were constructed at significant cost. Id. He also testified as to

the wells dug to irrigate the farm. Id. He testified that these investments in the farm were

prompted by the belief that the land would be suitable and productive for farming, as well

as for their recreational use and enjoyment without repetitive, atypical flooding they are

now experiencing due to changes in River management by the Corps. Id.

Mr. Ideker testified that it costs the Idekers about $750 per acre to plant a crop of

corn, and $550 per acre to plant a crop of beans. Tr. 226:15-24. The Idekers planted an

average of 645 acres of corn and 733 acres of beans each year for an average total planted

acreage of 1,378 acres. Id. Their total investment for planting each year averaged

$483,750 for corn and $403,150 for beans, for a total of $886,900. Id. Mr. Ideker stated

that this represents a substantial business investment. Id.

After the MRRP flooding began in 2007, Mr. Ideker testified as to the steps the

farm has taken to prevent the flooding and protect their property. Tr. 227:11-228:10.

These efforts included the rocking of the riverbank to abate erosion and preserve the

levee and river cabin situated close to the river. Tr. 228:1-10. It also included the repair

of the existing levees breached or damaged as a result of the flooding, as well as the

building of new levees on the east side of the property, all at substantial expense. Tr.

227:11-25.

59 c. The Buffalo Hollow Farms property

Mr. Schneider testified that his family originally acquired the Phase II tract in

1962, believing that the new publicized flood protection to be provided by the

government through new dams and reservoirs and the BSNP would allow the ground

adjacent to the river to be productive and otherwise free from the frequent flooding that

had been experienced up until that time. Tr. 145:8-20.

Title to the representative property was transferred to the Schneider’s family-

owned Buffalo Hollow Farms in July 2009. Tr. 171:20-172:4. Buffalo Hollow Farms

absorbed Buffalo Hollow Ranch. Id. No money was exchanged for the transfer of title.

JX8 (Jan. 27, 2020 Schneider Depo. at 19:13-21, 20:18-21:14).20

Mr. Schneider testified that his family has invested millions of dollars in the land

since acquisition. Tr. 146:6-12. He explained that but for the many public assurances by

the government regarding flood control and enhanced flood protection beginning during

the last century, the Schneiders and Buffalo Hollow Farms would not have invested in the

property as they have done. Id. Based upon the Schneiders’ experience with the Phase II

tract prior to 2007, they expected the farm to be productive each year and experience wet

years approximately once per decade. Tr. 147:12-18.

Mr. Schneider testified that it costs Buffalo Hollow Farms $670 per acre to plant a

crop of corn on 579 acres, and $450 per acre to plant a crop of beans on 248 acres,

20 The government argued that there cannot be any reasonable investment-backed expectations regarding property that is “inherited.” Def.’s Br. at 40. The court disagrees. As discussed above, Mr. Schneider continues to farm the subject property and his undisputed testimony states that he and his family have invested millions in the farm. 60 exclusive of labor costs. Tr. 149:11-150:4. The Schneiders’ total investment for crops

per year, exclusive of labor costs, is approximately $500,000. Id. Mr. Schneider stated

that this investment was made in reliance on the flood control Buffalo Hollow

experienced prior to 2007. Id. Mr. Schneider further explained that the substantial

investment of millions of dollars the Schneiders have put in the land for farming since

acquisition makes it economically difficult to simply quit trying to farm the ground. Id.

The Schneiders’ expectations for their Phase II tract remained the same until at least

2011. Tr. 151:14-19. He testified that the flooding impacts of the MRRP and the

deprioritization of flood control came as a surprise to them. Id.

Based on the testimony of the representative plaintiffs summarized above, the

court finds that each of the plaintiffs has shown that they had an investment-backed

expectation that they would continue to farm their properties under the pre-MRRP flood

pattern. The representative plaintiffs own their farmland, invest yearly in crop production

to make use of the land, have an interest in the land gaining value, spend money to

prevent flooding, and remediate their land after flooding. The representative plaintiffs

made these investments to make their farms productive with the expectation that the pre-

MRRP flood patterns would continue.

3. Plaintiffs’ Expectation that the Representative Properties Would Not Be Subject to Increased Flooding is Reasonable

The court further finds that the representative plaintiffs’ expectations are

supported by evidence and this court’s prior opinions and are therefore reasonable. The

plaintiffs in both Phase I and Phase II presented government publications that

61 demonstrate that the United States expected people to be protected from flooding along

the River following the construction of the Mainstem System and BSNP. See Pls.’ Br. at

35-38; see, e.g., Ideker I, 136 Fed. Cl. at 661 (noting that the federal government

determined it was in the national interest for the Missouri River “to be controlled for

purposes of human settlement and as a resource to support economic development”

(citing PX16 at PLTF-00003157, PLTF-00003098)). For example, a U.S. Department of

Interior Overview Report dated March 2005 states that after “closing of the dams, the

vast lands were cleared for agricultural production . . . As time passed, the idea that these

lands were flood-free caused developers to move in, thus supplementing the demands for

bank stabilization projects.” PX18 at PLTF00005755; see also PX16 at PLTF-00003229

(2002 National Resource Council publication stating “[h]owever, the land between

federal levees and the river has been farmed, and expectations consequently arose to

protect this land, as well as those lands behind the levees . . . .”). The Phase II

representative plaintiffs testified that they relied on the government’s assurances about

flood protection when investing in their properties. See Tr. 78:17-79:14 (Adkins); Tr.

221:19-222:19 (Ideker); Tr. 146:6-12 (Buffalo Hollow).

It was reasonable for the plaintiffs to expect the prioritization of flood protection

to continue, and for the prioritization of habitat protection under the ESA to come as a

surprise to the plaintiffs. As this court has previously held, “the changes made to the

Corps’ River and Mainstem [S]ystem after the court order requiring the Corps’

compliance with the ESA increased flooding to a degree that would not have been

contemplated when the River and Mainstem System structures were planned.” Ideker II,

62 142 Fed. Cl. at 232 (emphasis added). Separately, the court has explained that the actions

taken by the Corps to comply with the ESA represent a “significant change in the focus

of the work the Corps was doing in managing the River—from flood control to River

restoration work.” Ideker I, 136 Fed. Cl. at 668. It is true, as the government argues, that

the Corps must comply with federal laws like the ESA and that the Corps was free to

change its operations of the Mainstem System and BSNP, and has in fact done so in the

past. Def.’s Br. at 46; see also, e.g., Tr. 1362:20-1363:16 (Remus) (citing PX6 ¶ 5-23)

(noting that as early as 1960, the Corps explained in its master manual that its operation

of the mainstem system could not “be expected to remain entirely fixed in the future”).

But it was nonetheless reasonable for the representative plaintiffs to view the post-MRRP

flooding as “unexpected” given the Corps’ significant priority change from flood

protection to species protection after 2004. Cf. In re Upstream Addicks and Barker, 146

Fed. Cl. at 261 (“[I]t is not the case that a takings claim must fail simply because a

property owner acquired land while on notice that a taking . . . had the potential to

occur.” (quotation and internal alterations omitted)).

The government also argues that Dr. Kelman’s and Mr. Jones’ testimony, along

with government documents introduced by other government witnesses, establish that the

representative plaintiffs’ expectations regarding flooding on their properties was not

reasonable because plaintiffs’ land has always been subject flooding risk. Def.’s Br. at

43-49 (discussing Dr. Kelman’s testimony and documents), 59-62 (discussing Mr. Jones’

review of other documents); see also id. at 144 (“Plaintiffs’ land was subject to flooding

before and after [the] Mainstem System, and before and after 2004-2006 changes.”).

63 However, as stated above, the court gives Dr. Kelman’s testimony little weight. Dr.

Kelman conducted no interviews with the representative plaintiffs regarding their

expectations or what information they had available when forming their expectations,

despite acknowledging that such information is relevant to his opinions. Tr. 1312:24-

1313:7. On cross-examination, Dr. Kelman admitted that the representative plaintiffs’

personal observations for the period of time they lived on the river and how that was

managed are relevant and important considerations. Tr. 1310:21-1311:8.

Mr. Jones also performed a literature review of historical documents on

government communications regarding flood risk, but has no training as a historian or in

any social sciences. Tr. 2092:14-2093:6. The government contends that this literature

review supports the proposition that plaintiffs should have known that their land “has

always been subject to flooding.” Def.’s Br. at 141; see also id. at 59-60 (discussing Mr.

Jones’ literature review). As an initial matter, the court has already rejected the

government’s argument that the nature of the flooding on representative plaintiffs’

properties has not changed. See infra Part I.C. Notwithstanding that, the court concludes

that Mr. Jones’ testimony does not show that the representative plaintiffs’ expectations

were unreasonable. Mr. Jones acknowledged during cross examination that after the

BSNP was constructed, farm development investment followed, and that reasonable

investment-backed expectations can arise from enhanced flood control. Tr. 2100:1-

2101:13. Mr. Jones also acknowledged that an “important part” of his conclusions about

plaintiffs’ expectations was his belief that those expectations would have been formed

based in part on flooding patterns that predated completion of the dams, which is not the

64 relevant time period. Tr. 2067:7-2071:1. For these reasons, plaintiffs’ expectations that

there would not be increased flooding after the BSNP and Mainstem System was in place

are reasonable.

4. The Corps’ MRRP Actions Interfered with the Representative Plaintiffs’ Reasonable Investment-Backed Expectations

Finally, the court finds that the Corps’ MRRP actions have interfered with the

representative plaintiffs’ reasonable investment-backed expectations to be able to farm

and invest in their property. The government argues that the “productivity” and “value”

of the representative plaintiffs’ land has not “changed considerably as a result of the

MRRP changes.” Def.’s Br. at 146. However, this assertion is undermined by the

government’s expert on crop losses. The court finds that the testimony of the

government’s crop loss expert, Dr. Robert Evans, confirms that the MRRP has resulted in

increased flooding which in turn has led to lower crop yields and therefore a drop in

productivity and value.21

Specifically, Dr. Evans calculated for years 2007, 2008, 2010, 2013, and 2014 the

but-for or incremental impacts of the MRRP on the three representative plaintiffs’ corn

and soybean crop yields. See Tr. 2524:13-2529:11; see also DX7007-223, DX7007-224,

DX7007-225 (demonstrative summaries of opinions on yield impact for claim years).

These lost yields were then valued by the government’s expert Dr. David Sunding22 at

21 Dr. Evans’ qualifications and opinions are discussed in more detail in Part IV, infra. 22 Dr. Sunding’s qualifications and opinions are discussed in more detail in Part IV, infra.

65 $1.3 million dollars. Tr. 2854:9-18; see also DX6033-0030 (table of crop loss

calculations using Dr. Evans’ report). Dr. Sunding also testified that Dr. Evans’ analysis

demonstrated a 12 percent average drop in crop yield across the three representative

properties for the crops Dr. Evans considered. Tr. 2879:12-19. While the plaintiffs argue

that Dr. Evans’ study undervalues the impact of the MRRP on crop yields, see Pls.’ Br. at

81, Dr. Evans’ work nonetheless establishes that but for the MRRP the representative

plaintiffs would have seen greater crop yields. The losses described by Dr. Evans and Dr.

Sunding are significant, and with more frequent flooding, the losses are continuing.

Accordingly, the court finds that the representative plaintiffs’ have met their burden of

establishing that the MRRP interfered with their reasonable investment-backed

In sum, for the reasons discussed above, the three representative plaintiffs have

established all of the factors set forth by the Supreme Court in Arkansas Game & Fish for

proving the taking of a flowage easement. Both parties agree that if the court finds a

taking here, that taking is of a permanent flowage easement,23 given the ongoing nature

of the MRRP. See Def.’s Br. at 158; Pls.’ Br. at 119-20. The court now turns to the issue

of just compensation.

23 Arkansas Game & Fish involved a temporary taking by intermittent flooding. As discussed in Part III, infra, the court concludes (and the parties agree) that it is appropriate to apply the Arkansas Game & Fish factors for the taking of a permanent flowage easement by intermittent flooding as well. See Tr. 3390:8-14; Tr. 3424:13-3425:5 (closing arguments). 66 III. Date of Taking

Having determined that the United States is liable for a taking of a permanent

flowage easement for the three representative plaintiffs under the Arkansas Game & Fish

factors, the court now turns to the question of just compensation. As an initial matter, the

parties disagree as to the appropriate date of taking, which implicates not only how just

compensation is valued but whether plaintiffs’ claims are time-barred. The court

therefore addresses this issue first.

The representative plaintiffs argue that to determine the date of taking, the court

must apply the stabilization doctrine in United States v. Dickinson, 331 U.S. 745 (1947)

and the doctrine approved by the Federal Circuit in Arkansas Game & Fish, which allows

the court to group multiple years of flooding together so long as they are sufficiently

related. According to the representative plaintiffs, the period of taking is one that begins

in 2007 but stabilizes for accrual and valuation purposes on December 31, 2014. Pls.’ Br.

at 157-64. The representative plaintiffs point out that the permanent nature of the

flowage easement was not confirmed until 2020, when the government admitted that the

MRRP was not going to be terminated. Pls.’ Br. at 162-63. However, the representative

plaintiffs contend that the accrual and valuation date under the unique facts and

circumstances of this case is December 31, 2014, when the effects of the MRRP were

sufficiently stabilized as shown by the evidence. Id.

December 31, 2014 is also the date this court set as a “cut-off” date as to what

flooding would be considered in addressing the representative plaintiffs’ claims for just

compensation. See Ideker I, 136 Fed. Cl. at 670 (“The year 2014 was selected as the cut-

67 off year for purposes of proving flooding by the Corps’ System and River Changes;

however, some plaintiffs have continued to experience flooding.”). The representative

plaintiffs therefore chose December 31, 2014 as the date of taking “to be conservative

with their demand and consistent with the court’s approach to trial.” Pls.’ Br. at 163.

The government contends that for the representative plaintiffs, the appropriate date

of taking is 2007, the first year that atypical flooding at the three representative properties

occurred. Def.’s Br. at 151. Based on this date, the government argues that the

representative plaintiffs’ claims are time-barred under the six year statute of limitations

because they filed their complaint on March 5, 2014. Def.’s Br. at 152-53 (citing 28

U.S.C. § 2501). Alternatively, the government argues that the representative plaintiffs

have failed to meet their burden to provide any evidence of a credible date of taking.

Def.’s Br. at 153-54.

A. The Stabilization Doctrine

In general, where the United States permanently takes an interest in real property,

the date of a taking is the date the United States enters into possession of the property.

United States v. Dow, 357 U.S. 20, 22 (1958). This date is generally also “the date as of

which the land is to be valued.” Id. However, where flooding results from a “continuing

process of physical events” set in motion by the United States, the date of taking is set

when the property has experienced flooding and the situation has “stabilized.” Barnes v.

United States, 538 F.2d 865, 873 (Ct. Cl. 1976) (discussing Dickinson, 331 U.S. at 748-

49).

68 The Supreme Court held in Dickinson v. United States that where the “source of

the entire [takings] claim . . . is not a single event[, but] is continuous,” such as a series of

floods, a claim does not arise “until the situation becomes stabilized.” 331 U.S. at 748-

49. This is because “when the Government chooses not to condemn land but to bring

about a taking by a continuing process of physical events, the owner is not required to

resort either to piecemeal or to premature litigation to ascertain the just compensation for

what is really ‘taken.’” Id. at 749. Under Dickinson, a claim becomes “stabilized” where

the “consequences of inundation have so manifested themselves that a final account may

be struck.” Id.

The Court of Claims applied Dickinson in Barnes, 538 F.2d at 873, where the

United States’ actions at the Gavins Point and Fort Randall Dams caused farmland

flooding from 1969 to 1975. The Barnes court set the date of taking not at the

approximate date of the first flood, but at November 30, 1973, reasoning that by then,

“the permanent character of intermittent flooding could fairly be perceived.” 538 F.2d at

873.

Likewise, in Cooper v. United States, the Federal Circuit applied Dickinson in a

temporary takings case to determine when the government’s taking of timber by

intermittent flooding occurred. 827 F.2d 762, 764 (Fed. Cir. 1987). The Federal Circuit

explained that “the critical question is: when did the destruction of trees become

sufficiently stabilized so that the owner could determine the amount of timber taken?”

Id. The Federal Circuit stated that, although trees began to die on the plaintiff’s property

69 in 1979, “the extent of the destruction was not ascertainable until 1984,” when the

plaintiff filed suit. Id.

More recently, the Federal Circuit held in an erosion case that a situation becomes

stabilized when the physical process caused by government action has “substantially

encroached the parcels at issue and the damages [are] reasonably foreseeable.” Banks v.

United States, 741 F.3d 1268, 1272-73 (Fed. Cir. 2014) (quoting Boling v. United States,

220 F.3d 1365, 1373 (Fed. Cir. 2000)). The Federal Circuit has also applied Dickinson to

hold, in a case alleging that the government’s actions in draining a lake caused

uncontrolled vegetation growth, that a taking accrues when all the events which fix the

United States’ alleged liability have occurred and the harmed party knows or should have

known their existence. Nw. La. Fish & Game Pres. Comm’n v. United States, 446 F.3d

1285, 1290 (Fed. Cir. 2006). In that case, the Federal Circuit held that the proper accrual

date occurred when the damages from the vegetation overgrowth became “quantifiable

and present,” even where the plaintiff had estimated damages years earlier. Id. at 1291.

Where stabilization is an issue, “it is the uncertainty surrounding the permanent

nature of the taking, and not the uncertainty surrounding the ultimate extent of the[]

damage, that is critical in determining whether the situation has stabilized.” Boling, 220

F.3d at 1372. Stabilization occurs where “environmental forces have substantially and

permanently invaded the private property such that the permanent nature of the taking is

evident and the extent of the damage is reasonably foreseeable.” Id. at 1371. Selecting

the date of taking under Dickinson and its progeny is a question of fact that is dependent

on the unique circumstances of a particular case. Dickinson, 331 U.S. at 748-49; Barnes,

70 538 F.2d at 873. Determining the date is “a practical matter and not a technical rule of

law.” Dickinson, 331 U.S. at 749.

Applying these standards, the question in this case is therefore: when did the

flooding consequences of the MRRP claimed by the representative plaintiffs become

sufficiently stabilized? In answering this question, the court considers when the nature of

the flooding caused by the MRRP could be reasonably perceived, the foreseeability of the

damages from this flooding, when the plaintiffs were or should have been aware of the

MRRP-caused flooding on their properties, and the practicalities presented by the

circumstances of this case.

B. The Taking Did Not Accrue in 2007

Applying the Dickinson stabilization doctrine to this case, the court first rejects the

government’s contention that the taking accrued in late 2007, when the atypical flooding

on the representative plaintiffs’ properties first became apparent. See Def.’s Br. at 151.

The MRRP’s effect on the River had not “stabilized” as of 2007, and it is unreasonable to

expect the representative plaintiffs to have known of the MRRP’s ongoing effect on their

property after the first atypical flooding event.

To begin, Dickinson and the cases that follow reject the idea that the date of taking

is the first flooding event because, at that time, the nature and the extent of the flooding at

issue is still unknown. Dickinson, 331 U.S. at 769 (holding that a claim does not accrue

“as soon as [the property owner’s] land is invaded” because of the “uncertainty of the

damage”); see also Cooper, 827 F.2d at 764 (holding that the taking did not accrue when

the timber in question was first damaged by flooding in 1979, but in 1984 when the

71 extent of the injuries was “ascertainable”); Barnes, 538 F.2d at 873 (setting the date of

taking after five years of flooding). Likewise, here, at the time of the first atypical flood

in 2007, there was no reason to know whether that flooding would continue in a way that

would be so substantial and frequent as to constitute a taking. Boling, 220 F.3d at 1372

(“[R]equiring the plaintiffs to sue immediately upon the initial encroachment of their land

is too rigid an application of the stabilization principle.”).

Rather, not until the plaintiffs had experienced multiple years of intermittent

flooding would a taking have accrued. For example, in Barnes, the Court of Claims held

that government-caused intermittent flooding did not amount to a taking of a permanent

flowage easement until after years of flooding that was expected to continue into the

future. See 538 F.2d at 872. The Court of Claims therefore did not set the date of taking

at the date of the first flood in 1969, but five years later in 1973, when the “permanent

character of intermittent flooding could fairly be perceived.” Id. at 873. A similar

finding is warranted here, where the representative plaintiffs have presented evidence of

years of intermittent flooding that began in 2007 and is expected to continue.

Moreover, as discussed in the Phase I decision, the Corps’ projects under the

MRRP continued past 2007, into 2014. See Ideker I, 136 Fed. Cl. at 669 (“Corps studies

explain that as of 2014, the Corps had undertaken 1,697 dike notching actions, 354 major

modification actions, 63 dike lowering actions, 36 dike extension actions, 39 side-channel

chute actions, 20 revetment chute actions, 14 backwater actions, and 3 channel widening

actions.”). The continued construction projects under the MRRP years after 2007

undermine the government’s argument that the situation had “stabilized” by then.

72 Evidence in the Phase II record also supports the representative plaintiffs’ contention that

the River continued to change between 2003 and 2015 near the representative plaintiffs’

tracts. PX3718 (comparing “roughness” values in the Corps’ 2003 Flow Frequency

Study and May 2015 Missouri River Unsteady HEC-RAS Model Calibration Report);

PX3716 (Corps’ May 2015 HEC-RAS Report Appendix D). These studies show

increased “roughness” in the River adjacent to the Adkins and Ideker properties. See

PX3716 at USACE2571084 (roughness values); PX3718 at USACE0231783 (roughness

values); Tr. 1729:10-1734:20 (Dr. Holmes, discussing these documents). Although these

studies use two different modeling methods, see Def.’s Resp. at 31 n.28, the studies

nonetheless recognize changes in the River post-2007, see PX3716 at USACE2571059.

In addition, it is unreasonable to expect the representative plaintiffs to have known

of the cause of the increased flooding by 2007 given the complex nature of the hydrology

of the River. See Ideker I, 136 Fed. Cl. at 702-05 (describing Dr. Hromadka’s

testimony). As the representative plaintiffs have testified, and many, if not all, of the

Phase I plaintiffs testified, the cause of the flooding beginning in 2007 was not known

until later. See Pls.’ Resp. at 30. The representative plaintiffs specifically explained that

they were “not aware until sometime later [of] the MRRP, the deprioritization of flood

control and the changes that had been made to the river and river management in 2004.

The ongoing flooding was much different than before.” Tr. 82:2-6 (Adkins). And,

although they “beg[a]n questioning the cause of flooding in our region” during “the 2010

flood event,” the representative plaintiffs did not until March 2014 file suit “after getting

confirmation from our expert in the latter half of 2013, the Government’s actions in

73 conjunction with the MRRP were causing the flooding we were experiencing.” Tr.

150:22-151:13 (Schneider).

The court rejects the government’s contention that the evidence demonstrates that

the representative plaintiffs should have known of the cause of the increased flooding in

2007. The government points to testimony by plaintiffs regarding observed changes in

the River at various dates before the plaintiffs’ chosen date of 2014. See Def.’s Br. at 152

(citing, for example, testimony by Mr. Adkins that the 2007 flooding was atypical), 154

(citing testimony from the Phase II and Phase I plaintiffs regarding changes in the River

since 2004). While changes in the River may have occurred earlier, the representative

plaintiffs reasonably did not know the cause or character of the River changes and the

foreseeable extent of the damages during those earlier times. See Nw. La. Fish & Game,

446 F.3d at 1291-92 (holding that the fact that the plaintiffs estimated damages before the

extent of potential harms became known did not set accrual at an earlier date than when

the potential harms actually occurred and subsequently “stabilized”).

The government further argues that this court has already held that the 2007

flooding at issue on the representative plaintiffs’ properties was the foreseeable result of

the MRRP, a required element of plaintiffs’ takings claims. Def.’s Br. at 152. Yet, this

does not mean that the impact of the MRRP had stabilized by 2007, or that the nature of

the takings claims had become known by the first atypical flood. See Barnes, 538 F.2d at

873 (setting the date of taking as November 30, 1973, but noting that the relevant

flooding “stretch[ed] back to 1969,” when the first flood occurred).

74 For all of these reasons, the court disagrees with the government’s proposed 2007

date of taking.

C. The Accrual and Valuation Date is December 31, 2014

The court instead agrees with the plaintiffs that, under the unique facts and

circumstances of this case, the accrual and valuation date for a permanent flowage

easement is December 31, 2014. First, as noted above, the MRRP construction activities

relied on by the representative plaintiffs continued through 2014. See Ideker I, 136 Fed.

Cl. at 669. This suggests that the full effect of the MRRP would not have “stabilized”

until these construction activities were completed – in other words, when the events

fixing the United States’ liability occurred – supporting plaintiffs’ December 31, 2014

accrual date.24 See Nw. La. Fish & Game, 446 F.3d at 1291-92.

Second, although they were aware of atypical flooding earlier, it is the combined

effect of the years of intermittent flooding between 2007 and 2014 that supports the

taking of a permanent flowage easement for the three representative plaintiffs.

“Adopting a date of taking must often be done in a somewhat imprecise manner,” and the

court is satisfied that based on the facts of this case for these three plaintiffs, “the

permanent character of intermittent flooding could fairly be perceived” by the end of

24 Although the Corps plans to continue to construct new projects under the MRRP, see, e.g., Tr. 1469:6-1471:22 (regarding minor modifications and yet-to-be-completed spawning habitat); Def.’s Br. at 73 (noting that the planned future MRRP construction will take place downstream of the three representative properties), plaintiffs do not base their takings claims on this new construction, except, as discussed above, to demonstrate that the MRRP is ongoing. 75 2014. Barnes, 538 F.2d at 873 (finding a date of taking for a permanent flowage

easement after five years of intermittent flooding).

Third, the representative plaintiffs testified that while they began to notice changes

in the river before 2014 and began to question those changes in 2010, they could not

reasonably have known that the MRRP caused those changes before consulting with their

experts and ultimately filing suit. This, combined with the unique circumstances and

“practical[ities]” of the case, further supports a date of taking of December 31, 2014.

Dickinson, 331 U.S. at 749. As discussed above, the changes to the river caused by the

MRRP are complex. See Ideker I, 136 Fed. Cl. at 702-05 (describing Dr. Hromadka’s

testimony). As the Federal Circuit has held, a taking does not accrue until all the events

which fix the United States’ alleged liability have occurred and the harmed party was

aware of or should have been aware of their existence. Nw. La. Fish & Game, 446 F.3d

at 1290. It is reasonable for the representative plaintiffs not to have known the cause of

the flooding on their properties until consulting an expert. Although the plaintiffs

consulted their expert in late 2013 and filed their lawsuit on March 5, 2014, the court

concludes that it is appropriate to use December 31, 2014, the court’s original “cut-off”

date for plaintiffs’ flooding claims, as the date of taking.

In sum, the court concludes that the end of 2014 best represents when all of the

events fixing the United States liability occurred, when the intermittent flooding on the

representative plaintiffs’ properties became sufficiently permanent in nature, and when

the plaintiffs knew or should have been aware of the nature and extent of the MRRP-

76 caused flooding. The court therefore agrees with the representative plaintiffs that

December 31, 2014 is the appropriate date of accrual and valuation in this case.

D. The Taking Began in 2007

Finally, the court agrees with the representative plaintiffs that it is appropriate to

continue to treat the increased flooding attributable to the Corps’ MRRP actions prior to

2014 as a continuous flood as it did in Phase I in determining when the taking began,

based on the Federal Circuit’s remand decision in Arkansas Game & Fish. See Pls.’ Br.

at 158-64. In that case the Federal Circuit, on remand from the Supreme Court, found

that where the Corps’ yearly actions between 1993 and 2000 causing intermittent

flooding on the land in question “were directed to a single purpose,” it was appropriate to

analyze the intermittent flooding “as having lasted for seven years.” 736 F.3d at 1370.

In other words, it was appropriate in that circumstance to treat the intermittent flooding as

a single flood for purposes of the takings analysis, rather than treating each flood as an

individual, separate action. Id.

Pursuant to the Federal Circuit’s remand decision, in Phase I the court found here

that “(1) the Corps’ System and River Changes were made for a single purpose; (2) the

cumulative and combined effects of the System and River Changes made for that single

purpose led to higher WSEs than would have existed without the System and River

Changes; and (3) the higher WSEs led to flooding, or more severe flooding on the

property owned or farmed by that individual plaintiff than the flooding the plaintiff

would have experienced without the Corps’ System and River Changes.” Ideker I, 136

Fed. Cl. at 674. Applying these facts to the reasoning in the Arkansas Game & Fish

77 remand decision, the court held that “it is proper to consider the series of changes made

by the Corps for a single purpose” in the causation analysis, id. at 674, rather than isolate

each individual Corps action under the MRRP and connect that action to each flooding

event on each plaintiffs’ individual property, as the government sought, id. at 673.

The court now finds that this same reasoning extends to determining the period of

the taking. The atypical flooding caused by the MRRP began in 2007, as part of a series

of floods caused by the Corps’ actions under the MRRP that is continuing today. Under

the Federal Circuit’s remand decision in Arkansas Game & Fish, the court treats this

series of intermittent floods collectively in determining the period and nature of the

taking and in evaluating just compensation. See 736 F.3d at 1370.

The government suggests that, where the court has determined an accrual date in

2014, treating the taking as beginning in 2007 is inconsistent with Barnes, where the

Court of Claims held that any flooding damages “sustained prior to the date of taking are

the product of tortious invasions” and are thus unrecoverable. 538 F.2d at 874; Def.’s Br.

at 154, 159. The Barnes court reasoned that “not until the date of taking,” when it

became clear that prior flooding would continue and that the government therefore had

taken a permanent flowage easement, “did these several tortious invasions ripen to the

extent necessary to confer on the defendant a flowage easement.” Id. at 874. According

to the government, any damages claimed by the plaintiffs prior to the 2014 accrual date

are tort damages beyond this court’s jurisdiction. Def.’s Br. at 154, 159.

It is well-settled that tortious flood damages claims are not within the court’s

jurisdiction. Ridge Line, 346 F.3d at 1355. However, since Barnes, the Supreme Court

78 has held that government-induced intermittent flooding could constitute a taking, rather

than a tort, even where the flooding is temporary in nature. Ark. Game & Fish, 568 U.S.

at 27. In doing so, the Supreme Court reversed the Federal Circuit’s decision, which

relied in part on Barnes, that the government’s actions “at most created tort liability.”

Ark. Game & Fish Comm’n v. United States, 637 F.3d 1366, 1378, 1379 (Fed. Cir. 2011).

On remand from the Supreme Court, the Federal Circuit held that the seven years of

intermittent flooding caused by the government in that case could be treated as one flood

in the takings analysis. 736 F.3d at 1370. The Federal Circuit upheld the trial court’s

determination that this intermittent flooding (which had a single purpose of providing

water to farmers), taken together, constituted a taking. Id. at 1381.

In this case, there has been a permanent, rather than a temporary, taking, based on

intermittent flooding related to the single purpose of the MRRP. The parties agree that

the Supreme Court’s decision in Arkansas Game & Fish is applicable not only to a

temporary taking due to intermittent flooding, but to the taking of a permanent flowage

easement due to intermittent and continuing flooding. See supra n.23. The court finds no

reason not to apply the rationale of Arkansas Game & Fish in this case, even though the

intermittent flooding here – unlike the intermittent flooding in Arkansas Game & Fish –

will continue into the future. Applying Arkansas Game & Fish, the court finds it

appropriate to hold that the accrual date of intermittent flooding resulting in a permanent

flowage easement may be later than the date that the taking began. The court therefore

treats the taking here as beginning in 2007, when flooding attributable to the MRRP

79 began, and rejects the government’s argument that any floods prior to the 2014 accrual

date were tortious invasions outside this court’s jurisdiction.

Based on the foregoing, the court determines that, while the taking began in 2007,

the representative plaintiffs’ claims did not accrue until December 31, 2014. The

representative plaintiffs have therefore brought their claims within the six-year statute of

limitations applicable to this case. In addition, as discussed below, December 31, 2014 is

the appropriate date for valuing just compensation for the permanent flowage easement

taken.

IV. Just Compensation

Having determined that the taking began in 2007 and accrued in 2014, the court

next turns to just compensation. Where there is a physical taking under the Fifth

Amendment, the court must award just compensation. See Yee v. City of Escondido, 503

U.S. 519, 527 (1992) (holding that the government must pay just compensation when it

“requires [a] landowner to submit to the physical occupation of his land”); see also Banks

v. United States, 721 F. App’x 928, 940 (Fed. Cir. 2017) (“It cannot be the case that acres

of property are lost due to erosion and the value of that total property is not affected.”).

Just compensation must be paid in an amount that puts the property owner in “as good a

position pecuniarily as he would have occupied if payment had coincided with the

appropriation.” Kirby Forest Indus., Inc. v. United States, 467 U.S. 1, 10 (1984). The

court does not require proof of the “precise amount” of just compensation due, but only a

“reasonable approximation,” Ark. Game & Fish, 736 F.3d at 1379, which “requires more

than a guess, but less than absolute exactness,” Precision Pine & Timber Inc. v. United

80 States, 596 F.3d 817, 833 (Fed. Cir. 2003). “[J]ust compensation must be measured by

an objective standard that disregards subjective values which are only of significance to

an individual owner.” United States v. 50 Acres of Land, 469 U.S. 24, 35 (1984).

The representative plaintiffs’ claims for just compensation in this case involve

valuing the permanent flowage easement based on the diminution in land value due to the

increased flooding pattern on the representative tracts as measured on the date of accrual

in 2014, as well as the crop and other losses, including land and levee restoration,

sustained from 2007 to 2014. The court first addresses the valuation of the permanent

flowage easement, and then turns to whether plaintiffs are entitled to lost crop and other

damages prior to the 2014 date of accrual. Finally, the court addresses the proper interest

rate.

A. Valuation of Just Compensation for the Taking of the Permanent Flowage Easement

In this case, the United States has taken a flowage easement across the

representative plaintiffs’ properties. Where the United States takes less than an entire

parcel, the “before-and-after” valuation method is generally the simplest and perhaps the

most widely-used approach that “serves to lessen the pitfalls and problems that arise

when a series of factors affecting value are added together to arrive at a total severance

damage determination.” Ga. Pac. Corp. v. United States, 640 F.2d 328, 336 (Ct. Cl.

1980). Under the before-and-after method, “just compensation” is the difference between

the fair market value of the whole parcel immediately before the taking and the remainder

after the taking. United States v. Miller, 317 U.S. 369, 376 (1943); see also Otay Mesa

81 Prop., L.P. v. United States, 670 F.3d 1358, 1363-64 (Fed. Cir. 2012) (“Where the

property interest permanently taken is an easement, the ‘conventional’ method of

valuation is the ‘before-and-after’ method, i.e., ‘the difference between the value of the

property before and after the Government’s easement was imposed.’” (quoting United

States v. Va. Elec. & Power Co., 365 U.S. 624, 632 (1961)).

However, there is no one approved approach for determining what injuries and

losses are compensable; rather, the determination must be based on the particular facts of

each case. Hendricks v. United States, 14 Cl. Ct. 143, 149 (1987) (holding “the concept

of just compensation cannot be reduced to formula nor can it be confined to inexorable

rules”) (internal citation and quotation marks omitted); Ridge Line, 346 F.3d at 1354-55

(holding that the trial court erred in finding that damages could not be demonstrated

simply because the plaintiff did not provide appraisals of its land value before and after

the taking). The trial court has considerable discretion in determining the methodology to

be used in calculating just compensation. Innovair Aviation, Ltd. v. United States, 82

Fed. Cl. 567, 568 (2007). Regardless of the valuation method used, “a judge may award

damages, even if [the judge] does not fully credit that party’s methodology.” Banks, 721

F. App’x at 940 (quotation omitted).

Ultimately, the court must determine whether the damages were shown “to a

reasonable approximation” of the value of what was taken from the plaintiffs. Ark. Game

& Fish, 736 F.3d at 1379. For the reasons set forth below, the court finds that the

representative plaintiffs’ damages calculations meet that standard.

82 1. The Representative Plaintiffs’ Method of Before-and-After Valuation Offers a Reasonable Approximation of the Diminution in Fair Market Value of Their Properties

The representative plaintiffs relied on three experts, Mr. Leo Smith, Mr. Tim

Keller, and Dr. Bruce Babcock, to determine the diminution in the fair market value of

their tracts due to the imposition of the permanent flowage easement. As discussed

below, the plaintiffs did not use the traditional “before-and-after” appraisal method to

determine just compensation. However, the court finds the plaintiffs experts’ opinions

credible and persuasive, and that the representative plaintiffs’ method, under the unique

circumstances of this case, provides a reasonable approximation of the diminution in fair

market value of the representative plaintiffs’ tracts.

a. The court finds the testimony of plaintiffs’ appraisers credible and reliable

First, the representative plaintiffs relied on the testimony of local appraisers, Mr.

Leo Smith and Mr. Tim Keller. Mr. Smith and Mr. Keller were not tasked with

performing before-and-after property appraisals. Rather, Mr. Smith and Mr. Keller were

tasked with valuing the representative plaintiffs’ tracts as of the December 31, 2014 date

of taking; in other words, to appraise the plaintiffs’ property after the imposition of the

permanent flowage easement. Tr. 316:10-317:2 (Smith); Tr. 453:15-454:2 (Keller). Mr.

Smith and Mr. Keller were also tasked with providing Dr. Babcock additional appraisal

data needed to perform his analysis (described below). Tr. 319:5-8 (Smith); Tr. 454:10-

15 (Keller). Although the government had originally planned to call its own appraisal

83 witness, that witness was withdrawn before trial, and the government did not offer its

own appraisals of the plaintiffs’ properties.

The court finds the appraisals of Mr. Smith and Mr. Keller credible and reliable.

Both Mr. Smith and Mr. Keller are local, experienced appraisers. Mr. Smith has been a

licensed real estate appraiser for over 21 years and is a licensed certified general real

property appraiser in Iowa and Nebraska. Tr. 320:3-17; see also PX3031 (Smith CV).

Mr. Keller has been a licensed real estate appraiser for over 27 years, in Kansas,

Missouri, Nebraska, and Colorado. Tr. 455:15-456:19. Both Mr. Smith and Mr. Keller

previously completed appraisals for the federal government, including for the U.S. Army

Corps of Engineers. Tr. 321:22-322:13 (Smith); Tr. 545:7-546:9 (Keller). Mr. Smith and

Mr. Keller are familiar with Missouri River bottom farmland, the type of property which

they were asked to appraise. Tr. 323:3-7 (Smith); Tr. 460:4-10 (Keller). The appraisals

were developed in conformity with the Uniform Standards of Professional Appraisal

Practice. Tr. 320:19-24 (Smith); Tr. 457:1-14 (Keller). Both appraisers testified that

they complied with “all applicable ethical and professional standards in conducting” the

appraisals. Tr. 321:19-21 (Smith); Tr. 457:23-25 (Keller).

Mr. Smith appraised the Adkins representative tract using the comparable sales

approach. This approach uses sales considered to have been made in the most similar

market areas and with the most similar land characteristics as the Adkins tract, near the

December 31, 2014 valuation date. Tr. 323:13-324:1. Mr. Smith split the Adkins

property into four appraisals after considering the highest and best use of the property.

Tr. 317:19-318:11. Only the two Adkins farmland tracts, referred to as the farmland

84 inside and outside the levee, are relevant here. Tr. 1013:1-8 (Babcock). Mr. Smith also

verified these two appraisals using an income approach and a cost approach. Tr. 324:16-

24. In conducting his appraisals, Mr. Smith received information including the Adkins

tract profile, visited with Mr. Adkins, and visited the Adkins representative tract. Tr.

323:7-12.

The Adkins representative tract inside the levee is approximately 754.52 acres of

improved farmland of which 698.03 acres are tillable and 47.92 are timberland. Tr.

325:13-16. The appraised market value using the sales comparison approach of the

Adkins representative tract inside the levee was $6,700 per acre or $5,055,000 as of

December 31, 2014. Tr. 331:5-10; see also PX3037 (opinion of value document using

different approaches). In performing the comparable sales appraisal, Mr. Smith made

adjustments for characteristics including building improvements and soil quality

differences and noted that several of the comparable sales had slow drainage issues like

the Adkins farmland inside the levee. Tr. 330:19-331:3.

The Adkins representative tract outside the levee is 221.57 acres of unimproved

farmland with 98.08 tillable acres and 72.28 acres enrolled in the Forest Reserve. Tr.

335:17-22. The appraised fair market value of the Adkins representative tract outside the

levee, using the sales comparison approach, was $2,300 per acre or $510,000 as of

December 31, 2014. Tr. 340:16-19; see also PX3055 (opinion of value document using

different approaches). In conducting the appraisal using the comparable sales approach,

Mr. Smith made adjustments for the timing of the sales and land mix. Tr. 339:24-340:4.

85 Mr. Keller appraised the Ideker and Buffalo Hollow tracts using a sales

comparison approach, as verified by an income approach. Tr. 460:4-461:22. The

comparable sales analysis focused on comparable sales of properties with similar

characteristics that occurred near the time of the December 31, 2014 valuation date. Tr.

461:2-4. Mr. Keller “received information including the tract profiles, met and visited

with” the Schneiders and Mr. Ideker, and “visited each of the representative tracts on

several occasions . . . .” Tr. 460:11-16. In conducting the appraisals, Mr. Keller

researched, among other things, flood zone status, tax data, crop history, soil data,

comparable sales data, and the plaintiffs’ summary of the River’s flooding history. Tr.

459:22-460:3.

Mr. Keller appraised the Ideker representative tract at $8,250 per acre, or

$12,330,000 as of December 31, 2014. Tr. 478:19-23; Tr. 480:3-13; Tr. 483:8-12; see

also PX3273 (Ideker sales comparison conclusions). Mr. Keller appraised the Buffalo

Hollow representative tract at $8,000 per acre or $6,230,000 for the bottomland, $2,500

per acre or $1,240,000 for the upland, for a total value of $7,470,000. Tr. 470:2-14; Tr.

483:1-7; see also PX3266 (Buffalo Hollow sales comparison conclusions). Both of these

values fell within the range in prices for Missouri River bottomland in Northeast Kansas

and Northwest Missouri of $6,500 per acre to $8,250 per acre. Tr. 483:14-17.

Based on their qualifications and testimony, the court finds Mr. Smith and Mr.

Keller’s appraisals of the representative plaintiffs’ tracts persuasive and adopts those

valuations.

86 b. The court finds the representative plaintiffs’ diminution in value analysis a credible method of measuring just compensation

Next, the plaintiffs relied on the econometric analysis of Dr. Bruce Babcock, an

agricultural economist,25 who, using farmland sales data obtained from Mr. Smith and

Mr. Keller and interviews of the representative plaintiffs, estimated the diminution in the

fair market value of the representative plaintiffs’ properties due to the MRRP. Tr.

1001:3-1002:3. To do so, Dr. Babcock measured “the extent to which the fair market

values of the Phase II tracts in the actual world . . . that . . . includes the MRRP

permanent flowage easements . . . are less than the hypothetical fair market values of

those Phase II tracts . . . in the but-for world (without the MRRP permanent flowage

easement).” Tr. 1016:19-1017:6.

Dr. Babcock’s approach used actual farmland sales data in the Missouri River

Basin, both before and after the MRRP, to determine how the MRRP on average affected

farmland price. Tr. 1026:2-6. Farmland price data was obtained from Mr. Smith and Mr.

Keller, plaintiffs’ appraisers. Tr. 1039:14-19 (referencing Mr. Smith); Tr. 1048:3-8

(referencing Mr. Smith and Mr. Keller). Mr. Smith compiled 20 years of arms-length

land sales of Missouri River bottomland in Nebraska and Iowa. Tr. 358:2-7 (Smith). He

25 Dr. Babcock is an agricultural economist and is currently a Professor of Public Policy at the University of California at Riverside. Tr. 1002:9-15. Dr. Babcock received his Bachelor’s degree from the University of California at Davis in 1980, a Master’s degree from the University of California at Davis in 1981, and his Ph.D. in Agricultural and Resource Economics from the University of California at Berkeley in 1987. Tr. 1003:19-1004:4. For most of his career, Dr. Babcock held various roles at Iowa State University; most notably, he was the director of a public policy research center called The Center for Agricultural and Rural Development, for which he was the director from 1998 to 2011. Tr. 1002:16-22; see generally PX3350 (Babcock CV). 87 also gathered twenty years of land sales data for both the East and West Nishnabotna

River Basins, a tributary of the Missouri River. Tr. 359:21-360:3 (Smith). Mr. Keller

collected data from Kansas, Missouri, and Nebraska. Tr. 3204:19-3205:22 (Babcock).

Dr. Babcock limited the data to arms-length, commercial agricultural transactions, and

ruled out land sales where the price was inflated by developmental value. Tr. 1048:3-19.

Dr. Babcock also limited the farmland data along the Missouri River Basin to the state of

Iowa. Tr. 1143:8-1144:9; Tr. 1148:21-1149:1. Dr. Babcock excluded data from other

states because of the small number of observations and because the other states would

have impeded his ability to use the Iowa soil quality index as an explanatory variable in

his analysis. Tr. 3205:3-3207:16.

Using this data, Dr. Babcock conducted a regression analysis that allowed for a

change in farmland prices to occur in 2011. Tr. 1027:8-16. Dr. Babcock used this 2011

“switch point” or “intercept” variable to measure the effect of the MRRP, reasoning that

by 2010, the effects of the MRRP would have been reflected in farmland prices. Tr.

1061:5-1062:4. Dr. Babcock’s regression analysis also controlled for other variables that

determine farmland prices, such as technology costs, soil quality, and the amount of

tillable acreage. Tr. 1033:5-12; Tr. 1036:22-1037:23.

To control for whether some factor other than the MRRP may have impacted

farmland value, Dr. Babcock’s model also incorporated farmland data from the

Nishnabotna River Basin. Tr. 1039:14-1040:9. The Nishnabotna is a tributary of the

Missouri River. Tr. 359:23-360:1 (Smith). The farmland along the Nishnabotna is

unaffected by the MRRP but otherwise similar to the farmland along the Missouri River

88 Basin. Tr. 1027:17-3 (“I also got additional farmland sales data from a nearby river basin

that’s not impacted by the MRRP, and I used that data . . . as a check.”); Tr. 1046:14-

1047:24 (description of regression equation). Although Dr. Babcock did not initially

label it as such, this type of regression model is known as a “difference-in-difference”

model. Tr. 3168:21-3170:15 (Dr. Babcock equating his model with a “difference-in-

difference” model). The first “difference” in the model measures the change in farmland

prices in 2011 along the River, which Dr. Babcock uses to measure the impact of the

MRRP. The second “difference” measures the change in 2011 on the Nishnabotna River.

See Tr. 1050:17-1053:17. The difference between these farm price changes measures the

impact of the MRRP. See Tr. 1055:8-1060:13.

The results of Dr. Babcock’s initial regression analysis indicated that if the MRRP

had not affected farmland prices, the Missouri River Basin farmland prices would have

been, on average, 25.7 percent higher. Tr. 1055:13-16; see also PX3389 (demonstrative

sales data plots); PX3356 (model results chart); PX3390, PX3391, PX3392

(demonstrative graphs visualizing the regression results). After discovering that he

erroneously included three land sales, Dr. Babcock re-ran the analysis, which increased

the average diminution in value to 26.9 percent. Tr. 1059:20-1060:10. Dr. Babcock

opined that this change was statistically significant, meaning that he could be confident

that the change was not zero. Tr. 1057:19-1058:3. In contrast, Nishnabotna River Basin

farm prices would have been 9 percent lower had there been no change in 2011. Tr.

1055:16-18. However, this change in farmland prices in the Nishnabotna River Basin

89 was not statistically significant. Tr. 1058:9-1059:6. Dr. Babcock therefore treated this

shift – the second “difference” – as if it had not occurred. Tr. 1059:3-6.

Dr. Babcock also performed a check on the robustness of the model to determine

whether allowing the change in farmland prices to occur in 2011, rather than some other

year, was appropriate. Tr. 1060:14-1064:6. Dr. Babcock determined that the highest

diminution in value occurred in 2010, and opined that that was the correct year of the

diminution in value for his statistical model. Tr. 1063:24-1064:6. Dr. Babcock further

testified that his result aligned with his interviews with the representative plaintiffs and

Dr. Mays’ analysis. Tr. 1061:17-25. He testified that flooding through the late 2010 time

period would have increased the knowledge that something had changed among local

farmers and outside buyers. Tr. 1061:13-16. When he first talked to the representative

plaintiffs, they told him that they first became aware that something had changed on the

river in the first decade in the late 2000s. Tr. 1061:18-25. And they all said they were

aware of the changed river by 2011. Tr. 1062:1-4.

Dr. Babcock then took that average diminution in value of 26.9 percent and

translated it into a tract-specific diminution in value. Tr. 1065:13-1069:9. Based on a

review of the Phase I opinion, his conversations with the representative plaintiffs, and Dr.

Mays’ analysis, Dr. Babcock concluded that the representative plaintiffs had greater-than-

average increased flood losses. Id. Dr. Babcock then incrementally increased the

diminution in value of the Adkins tract to 27.5 percent, and of the Buffalo Hollow and

Ideker tracts to 30 percent, which he believed to be a “conservative” estimate of the

diminishment in their respective fair market values. Tr. 1069:10-23. Based on its review

90 of Dr. Babcock’s approach and Dr. Babcock’s qualifications, the court finds Dr.

Babcock’s method of just compensation persuasive and reliable.

c. The government’s arguments do not undermine plaintiffs’ method

The government contends that the representative plaintiffs’ method of valuing the

diminution in the fair market value of their properties should be rejected on three general

grounds: (1) Mr. Jones and Mr. Keller’s appraisals of plaintiffs’ properties are unreliable;

(2) Dr. Babcock’s method of calculating the before-and-after valuation of the plaintiffs’

properties is improper; and (3) assorted additional criticisms of Dr. Babcock’s model.

Def.’s Br. at 83-100, 158-59. The court determines that the government’s objections and

criticisms do not undermine Dr. Babcock’s approach.

First, the government raises several issues with Mr. Smith and Mr. Keller’s

appraisals of the representative plaintiffs’ tracts as of December 31, 2014. However, the

issues raised by the government do not undermine the reliability of the appraisals. For

example, the government raises issues with the appraisal of development and residential

tracts on the Adkins property. Def.’s Br. at 84-85 (criticizing Mr. Smith’s reliance on the

city of Council Bluff’s 2030 development plan). But this criticism is irrelevant, as Dr.

Babcock’s ultimate opinions regarding diminution in value apply to only two of the

appraisals Mr. Smith provided, both of which are for farmland (inside and outside the

levee). Tr. 1013:1-8; Tr. 1014:3-8.

As another example, the government argues that Mr. Smith and Mr. Keller did not

appropriately adjust the comparable sales for flood frequency or size. Def.’s Br. at 86-88.

91 However, in determining just compensation, mathematical certainty is not required. See,

e.g., Otay Mesa Prop., L.P. v. United States, 779 F.3d 1315, 1323 (Fed. Cir. 2015)

(requiring “more than a guess, but less than absolute exactness” (quoting Precision Pine,

596 F.3d at 833)). The government did not offer its own appraisals of the properties or

provide the proper method of doing so, nor did the government provide a sense of

whether these adjustments would have made a large or small difference in the appraisals.

While it could be true that Mr. Smith and Mr. Keller could have made certain

adjustments to their appraisals, the court finds that none of the issues raised by the

government fatally undermine the credibility and reliability of Mr. Smith and Mr.

Keller’s opinions. Even if the appraisals lack perfection, they are the best evidence

available. For these reasons, the court rejects the government’s arguments regarding the

reliability of the appraisals of plaintiffs’ properties.

Second, the government argues that because the plaintiffs did not conduct a

standard before-and-after valuation as called for by the Yellow Book,26 the representative

plaintiffs’ method of measuring diminution in value is improper. Def.’s Br. at 83, 87. As

an initial matter, the Yellow Book applies to only those appraisals being conducted at the

request of the government. See Hardy v. United States, 141 Fed. Cl. 1, 32 (2018) (“[A]s

a matter of law, [Plaintiffs’ appraiser] was not bound by the Yellow Book. The Yellow

Book applies only to appraisers hired by the federal government for condemnation

26 The Yellow Book is shorthand for the Uniform Appraisal Standards for Federal Land Acquisitions, developed and adopted by the Interagency Land Acquisition Conference and published by the Appraisal Institute to guide appraisals associated with federal land acquisitions. See Tech. College of the Low Country v. United States, 145 Fed. Cl. 408, 426 (2019). 92 purposes; it is not mandatory with respect to appraisers not hired by the government.”).

The plaintiffs’ appraisers, therefore, were not bound by the Yellow Book.

In addition, the court disagrees with the government’s argument to the extent that

the government suggests that a standard before-and-after valuation method by an

appraiser must be used in determining just compensation. Def.’s Resp. at 35, ECF No.

682 (“The correct way of determining just compensation in a takings case is through a

straightforward before-and-after analysis of fair market value, where an appraiser uses

comparable properties and other techniques to determine the subject property’s value

immediately before and immediately after the date-of-taking.”). The Federal Circuit has

stated that in assessing just compensation, the court has the flexibility to consider

alternative approaches from the standard before-and-after method using comparable

sales. See, e.g., Banks, 721 F. App’x at 940. As the government itself acknowledges,

“courts are willing to employ non-traditional valuation methods when necessary.” Def.’s

Resp. at 35. The court does not fault plaintiffs for seeking the expert opinion of a

qualified agricultural economist, rather than an appraiser, to evaluate the effects of the

MRRP on plaintiffs’ properties.

Moreover, the model by Dr. Babcock is not a wholesale departure from the before-

and-after valuation method. Evidence was presented showing that farmland prices were

generally increasing in the Missouri River Basin from the 1990s through 2013, meaning

that the value of the flowage easement taken by the government may not have been

reflected in a standard before-and-after appraisal. See Tr. 1028:25-1030:1; see also

PX3386 (demonstrative plotting farmland price data on the Missouri River). In light of

93 that, the court finds Dr. Babcock’s method proper. As Dr. Babcock opines in his rebuttal

testimony, his model measures the average before-and-after effect of the MRRP on

properties along the Missouri River Basin, which he then adjusts to plaintiffs’ respective

properties. Tr. 3212:4-3215:18. While criticizing certain aspects of Dr. Babcock’s

model, the government’s rebuttal expert, Dr. Sunding, did not testify that the “difference-

in-difference” model could not be used for this purpose.

Third, the court finds that none of the problems the government raises with Dr.

Babcock’s model through the testimony of its rebuttal expert, Dr. Sunding,27 are

persuasive. Def.’s Br. at 89-100. For example, the government argues that Dr. Babcock

did not perform a difference-in-difference analysis and offered “flaw[ed], nonstandard”

work product as a result. Id. at 89, 95; see also Tr. 2759:4-6 (Dr. Sunding testifying that

“Dr. Babcock uses a nonstandard statistical framework and misinterprets his results”). In

particular, Dr. Sunding criticizes Dr. Babcock for not using the “second difference” in

forming his opinion. See Tr. 2763:22-24 (“[T]he way economists would describe what

Dr. Babcock is doing is he’s estimating a first difference.”); Tr. 2764:17-24

(“Observations from the Nishnabotna . . . don’t directly influence [Dr. Babcock’s]

estimated treatment effect, because . . . Dr. Babcock is only estimating a first difference

27 Dr. Sunding is a professor of agricultural and resource economics at the University of California at Berkeley. He has been on the faculty there since 1992, specializing in agricultural economics, applied econometrics, natural resource economics, and law and economics. Tr. 2736:12-19. He also has been a visiting professor at Stanford in the Woods Institute of the Environment. He has won a number of research awards over the course of his career. In 2009, he was named the inaugural Thomas J. Graff Professor in the College of Natural Resources at Berkeley in the area of natural resource economics, with an emphasis in the economics of water resource. He has served two terms as chair of Berkeley’s Department of Agricultural and Resource Economics. Tr. 2737:7-2739:10; DX6033-0031 (Sunding CV). 94 for the Missouri Basin parcels, which is why I say that Dr. Babcock didn’t use a

difference-in-difference framework.”); see also Tr. 2769:19-2770:6 (summarizing the

difference-in-difference analysis). However, Dr. Babcock explained that he dropped this

“second difference” as statistically insignificant, which resulted in a more conservative

estimate of diminution in value. Tr. 3179:24-3181:16.

Dr. Sunding also criticizes Dr. Babcock for relying on plaintiffs’ appraisers and

for failing to include certain explanatory variables and data points from other states in the

model. See Tr. 2755:6-2756:3 (criticizing Dr. Babcock for not using a “random sample”

of data); Tr. 2752:20-2754:9 (criticizing Dr. Babcock for not collecting his own data and

not including the “entire population of sales in his dataset” as he claims to have done);

Tr. 2814:22-2815:13 (criticizing Dr. Babcock for not using all of the information he was

armed with in his analysis); Tr. 2824:9-2825:17 (criticizing Dr. Babcock for basing his

analysis on four Iowa-based counties that do not contain any of the representative

properties and are distanced from the representative properties). Yet, the court has

already determined plaintiffs’ appraisers to be reliable and credible. In addition, Dr.

Babcock persuasively opined that the additional variables added by Dr. Sunding are

themselves statistically insignificant and added no explanatory power to the model. See

Tr. 3195:10-3197:24. Dr. Babcock also persuasively testified as to why he excluded

certain data points, including the fact that other states had small and unbalanced datasets

for purposes of the before-and-after valuation, and that limiting the dataset to relevant

Iowa sales was appropriate in light of the superior explanatory variables that were limited

to Iowa, such as the Iowa soil quality measure. Tr. 3205:18-3210:14.

95 Dr. Sunding further criticizes Dr. Babcock of “data mining” in choosing the 2011

date as the “switch point” in his model, purportedly to cherry pick the year that yielded

“the largest diminution in property value.” Tr. 2798:21-2799:6. Dr. Sunding argues that

in determining the “switch point,” Dr. Babcock should have conducted local media

searches to determine when market participants would have been made aware of the

MRRP or flooding. Tr. 2797:12-2798:10; Def.’s Br. at 96. But, as explained above, Dr.

Babcock persuasively opines that the 2011 date is based on both a robustness check and

his interviews with the plaintiffs, which highlighted their overall understanding of the

flooding. Tr. 3190:1-3191:11. While more limited than a local media search, the court

finds Dr. Babcock’s method behind selecting 2011 as his “switch point” reliable.

Finally, as Dr. Babcock points out, the government’s own expert on crop damages,

Dr. Evans, confirms the plausibility of Dr. Babcock’s diminution in value. As discussed

in the liability section of this opinion, the government relied on the testimony of Dr.

Robert Evans28 to provide an alternative valuation of crop loss, should the court award

damages for lost crops. In brief, Dr. Evans, relying on WSEs from plaintiffs’ Phase I

expert, Dr. Christensen, and the government’s Phase II expert, Dr. Holmes, used

28 Dr. Robert Evans is the United States’ expert witness in agricultural engineering. He has over thirty-five years of experience conducting agricultural water table management research and advising farmers about strategies to manage on-farm water to optimize agricultural production. Tr. 2337:9-14. Dr. Evans has over 50 years of first-hand experience managing excess water on his family farm, a farm comprised of inherently wet, poorly drained soils. Tr. 2337:21-23; Tr. 2343:25-2443:3. He has Ph.D., M.S., and B.S. in Biological and Agricultural Engineering from North Carolina State University. Tr. 2338:20-23. Dr. Evans has served as a Professor Emeritus at North Carolina State University since 2017 and head of the Department of Biological and Agricultural Engineering from 2006 until 2014. Tr. 2339:7-12. He has taught or co-taught courses on DRAINMOD to engineers and other professionals across the United States, including the mid-west. Tr. 2342:15-19. 96 DRAINMOD, a water management simulation model, to isolate and measure the

incremental impact of the MRRP on plaintiffs’ crop yields from 2004 to 2015. Tr.

2356:15-22; Tr. 2378:5-8; Tr. 2344:12-19; Tr. 2345:23-2346:3; Tr. 2428:7-13. As

recounted by Dr. Sunding, Dr. Evans estimated an average 12 percent drop in production

for corn and soybeans on the representative plaintiffs’ properties. Tr. 3181:21-3182:3

(Dr. Babcock summarizing Dr. Sunding’s testimony). As opined by Dr. Babcock, this

drop in production translates roughly into a 27 percent decrease in fair market value,

which is in line with Dr. Babcock’s estimated 26.9 percent average decrease attributable

to the MRRP. Tr. 3186:16-24.

For these reasons, the court finds that Dr. Babcock’s method in determining the

before-and-after effect of the MRRP reasonably approximates the diminution in value of

the representative properties. Notably, the government has not offered an alternative

method of calculating the diminution in value, only criticisms of the plaintiffs’ method.

This was not enough, particularly in light of Dr. Babcock’s reasoned responses to Dr.

Sunding’s criticisms.

2. The Court Adopts Dr. Babcock’s Diminution in Value Calculations for the Plaintiffs’ Representative Tracts

Having concluded that Dr. Babcock’s method in determining the average effect of

the MRRP on Missouri River Basin properties is appropriate, the court now turns to the

diminution in value of plaintiffs’ representative properties. To determine the diminution

in fair market value of plaintiffs’ representative properties, Dr. Babcock increased his

“conservative” estimate of diminution in value to 27.5 percent for Adkins, and 30 percent

97 for Buffalo Hollow and Ideker. Tr. 1069:10-20. Dr. Babcock based his increases on the

court’s opinion in Phase I that the representative plaintiffs had greater than average

increased flood losses, his discussion with plaintiffs, and Dr. Mays’ analysis. Id.

Although the government through the testimony of Dr. Sunding argues that Dr.

Babcock’s analysis is “subjective and unscientific,” Tr. 2835:12-16, the court agrees with

the plaintiffs that Dr. Babcock’s approach represents a reasonable approximation of the

representative plaintiffs’ losses.

Just compensation “should be carefully tailored to the circumstances of each

particular case” and “should be based on an assessment of precisely what the government

takes from a landowner.” Otay Mesa, 670 F.3d at 1368. It was therefore entirely

appropriate for Dr. Babcock to translate his average diminution in value to the plaintiffs’

representative properties. While not precise, Dr. Babcock’s approach amounts to “more

than a guess,” which is all that is called for in evaluating just compensation. Precision

Pine, 596 F.3d at 833.

Moreover, Dr. Babcock’s analysis of the government’s expert on crop loss, Dr.

Evans, supports a higher diminution in value for the plaintiffs’ representative tracts. As

noted above and persuasively explained by Dr. Babcock in rebuttal, Dr. Evans testified

that plaintiffs’ properties experienced decreased crop yields during the relevant flooding

years. See Tr. 3186:16-21 (Dr. Babcock noting that Dr. Evans estimated a 12 percent

drop in crop yields during the relevant period). After looking at this data “more

careful[ly],” Tr. 3186:25-3187:3, Dr. Babcock opined, using Dr. Evans’ data, that “the

fair market value would be 37 percent higher but-for the MRRP using my rebuttal report

98 estimates and using Dr. Evans’ estimate.” Tr. 3188:4-16. The court is therefore satisfied

that Dr. Babcock’s more conservative estimates are appropriate for plaintiffs’

representative properties.

Finally, the court rejects the government’s assertion that the representative

plaintiffs have failed to carry their burden of proof regarding diminution in value, and

that, therefore, the plaintiffs’ should be awarded no damages. Tr. 3513:1-3 (government

arguing that the court should award “either zero or nominal damages”). Here, the

government’s expert Dr. Evans acknowledges that the MRRP caused a reduction in crop

yields across plaintiffs’ properties. It cannot be the case that this reduction in crop yields

did not negatively affect the value of the representative properties, which are utilized as

farmlands. The government did not provide its own estimate of the diminution in fair

market value. And the plaintiffs have presented a credible methodology for calculating

damages to a reasonable approximation. In these circumstances, the court will award

damages. See Banks, 721 F. App’x at 940-41 (“It cannot be the case that acres of

property are lost to erosion and the value of that total property is not affected.”).

For these reasons, the court holds, based on its review of the evidence, that the

diminution in value of the plaintiffs’ representative tracts are as follows: Adkins, 27.5

percent; Ideker, 30 percent; Buffalo Hollow, 30 percent. Translated into dollar values,

the court values just compensation for the taking of a permanent flowage easement across

the representative plaintiffs’ properties as follows: Adkins, $1,530,268; Ideker,

$3,698,887; Buffalo Hollow, $1,868,928. Tr. 1012:5-10 (Ideker), 1012:20-25 (Buffalo

Hollow), 1013:13-15 (Adkins).

99 B. Compensation for Lost Crops and Other Costs

Having determined the value of the permanent flowage easements, the court next

turns to the question of what, if any, other damages plaintiffs are entitled to due to the

flooding of their properties caused by the MRRP. As discussed below, the court

concludes – with one exception – that all other damages the plaintiffs seek are

consequential damages and are therefore not compensable.

The Fifth Amendment provides the right to just compensation for property taken,

not “damages” that are the “proximate result” of the government’s action. Yuba Nat.

Res., Inc. v. United States, 904 F.2d 1577, 1581 (Fed. Cir. 1990) (“It is a well settled

principle of Fifth Amendment taking law . . . that the measure of just compensation is the

fair value of what was taken, and not the consequential damages the owner suffers as a

result of the taking.”); see also Kirby Forest, 467 U.S. at 15 (just compensation is “the

fair market value of the property on the date it is appropriated” and “[u]nder this

standard, the owner is entitled to ‘what a willing buyer would pay in cash to a willing

seller’ at the time of the taking.” (internal citations omitted)); United States v. General

Motors, 323 U.S. 373, 380 (1945) (“[T]hat which is taken or damaged is the group of

rights which the so-called owner exercises in his dominion of the physical thing, and that

damage to those rights of ownership does not include losses to his business or other

consequential damage.” (footnote omitted)). Thus, “not all losses suffered by the owner

are compensable under the Fifth Amendment.” U.S. ex rel. Tennessee Valley Auth. v.

Powelson, 319 U.S. 266, 281 (1943).

100 In the case of a partial taking, compensation may include the diminished value to

the remaining portion of land, called “severance damages,” as well as the value of the

portion of land taken. Miller, 317 U.S. at 375-76; Ga. Pac. Corp., 640 F.2d at 336. “The

cost to cure – or the cost of mitigating damages caused by the taking – provides an

alternative means of quantifying severance damages.” Childers v. United States, 116

Fed. Cl. 486, 497 (2013).

As discussed above, there exist alternative ways to assess and calculate the

compensation owed to a landowner for governmental interference with the use of

property. See Otay Mesa, 670 F.3d at 1369 (holding “there can in principle be an

appropriate alternative valuation measure to the ‘before-and-after’ method in a given

takings case”); Ridge Line, 346 F.3d at 1359 (“damages may be assessed based on Ridge

Line’s cost in constructing prudent flood control measures”); Vaizburd v. United States,

67 Fed. Cl. 499, 501-02 (2005) (“the ‘cost of cure’ approach to recovery in a takings case

is an alternative to computing damages through diminished market value”). But those

alternatives should not be used to award damages above and beyond what was taken.

Here, plaintiffs are alleging damages above and beyond the value of the flowage

easement that the government has taken. In particular, the plaintiffs allege through the

testimony of their damages expert, Dr. Bateman,29 crop losses and lost profits based on

reduced yields, damage to structures, damages to equipment, flood prevention expenses,

29 Dr. Bateman received his B.S. in economics from the University of Utah in 1960, and his Ph.D. in economics from MIT in 1965 with a special emphasis on price theory as applied to financial and agricultural commodity markets, econometrics. Tr. 563:19-564:2; PX3214 (Dr. Bateman’s CV). 101 and flood reclamation expenses. Tr. 577:3-578:8; Tr. 610:25-614:4; Tr. 573:23-574:2;

Tr. 728:4-21. However, these are consequential damages that are an indirect result of the

taking of the flowage easement. It is improper to both claim compensation for

diminution in value and claim compensation for these consequential damages. See Yuba

Nat. Res., Inc., 904 F.2d at 1581-82 (“[Plaintiff’s] claim for the difference in the value of

the gold during the taking period and after the taking is precisely the kind of claim for

consequential damages—here, lost profit—that is not an appropriate element of just

compensation for the temporary taking of the property.”); Childers, 116 Fed. Cl. at 600

(noting consequential damages include “loss of business, relocation expenses, and the

like”); R.J. Widen Co. v. United States, 357 F.2d 988, 993-94 (Ct. Cl. 1966) (holding that

there was no Fifth Amendment taking for “incidental spoliation of the plaintiff’s

inventory and equipment, the reduction or loss of its good will and profits, and the

expenses incurred in having to readjust its manufacturing operations”).

Relying on Dickinson, the plaintiffs argue that the government is liable for both

“the land which it permanently floods as well as that which inevitably washes away as a

result of that flooding.” Pls.’ Resp. at 29 (quoting Dickinson, 331 U.S. at 750).

However, plaintiffs quote Dickinson out of context. In discussing “that which inevitably

washes away as a result” of flooding, the Court in Dickinson was addressing additional

land that had eroded as a consequence of government-induced flooding, not crops,

structures, or equipment incidentally damaged by the flooding. Plaintiffs reliance on

Dickinson is thus misplaced.

102 Plaintiffs also cite Ridge Line for the proposition that the court must award “all

damages, past, present, and prospective.” Pls.’ Br. at 119 (quoting Ridge Line, 346 F.3d

at 1359); Pls.’ Resp. at 29. However, Ridge Line does not discuss consequential

damages, but instead affirms that flood control measures may be used as an alternative

method of calculating just compensation. 346 F.3d at 1358-59. This selective quote in

Ridge Line does not support plaintiffs’ claim for lost crops and other incidental

damages.30 For these reasons, the court will not award the damages claimed by plaintiffs

above and beyond the diminution in value of their property.31

The court will, however, award severance damages for the repair of the Ideker

levee in 2010. See Tr. 238:3-13 (Mr. Ideker describing levee repair efforts). In the Phase

I opinion, the court held that in 2010, the flooding on the Ideker property was “due to

levee overtopping that would not have occurred without the System and River Changes.”

Ideker I, 136 Fed. Cl. at 748 (citing PX2554). In other words, the destruction of this

30 Moreover, even if the court were to consider these additional damages, just compensation would only include an amount “proportionate to the government’s quantitative contributions” to the damage incurred. Ridge Line, 346 F.3d at 1359. Unlike the government’s expert Dr. Evans, plaintiffs have not apportioned crop losses (or other costs) to the increment of flooding caused by the Corps’ actions under the MRRP. See Tr. 3283:18-24. The Court of Claims in Barnes also recognized that the plaintiffs there could possibly be compensated for the value of “mature” crops destroyed on the date of taking, but not “immature” crops. 538 F.2d at 874. The representative plaintiffs have not presented any evidence with specificity regarding this distinction between immature and ready-to-harvest crops that would allow the court to make an award. 31 The government argues that even if the plaintiffs are entitled to just compensation for crop losses, just compensation requires offsetting those losses by payments received by the plaintiffs based on various federal-backed crop insurance programs. Def.’s Br. at 165-67. Because the court has declined to award the plaintiffs damages for crop loss, the court does not address this argument. 103 levee is entirely attributable to the government’s actions. In addition, the court finds that

the repair of the levee is properly considered as a severance damage assumed to protect

the remainder of the property. See Ideker I, 136 Fed. Cl. at 747 (“The property is

protected by a private levee built by plaintiff.”). Compensation for this damage may be

measured by the mitigation cost of rebuilding the levee. See Childers, 116 Fed. Cl. at

497; Miller, 317 U.S. at 375-76. This levee repair occurred during the period of taking,

which, as discussed above, the court finds began in 2007. The court therefore holds that

the Ideker plaintiffs are entitled to the cost of levee repair in 2010 in the amount of

$1,032,338 without interest. Tr. 238:3-13 (Ideker); Tr. 661:4-13 (Bateman) (describing

$823,123 to rebuild the levee and $209,225 in fuel costs, which together add to

$1,032,338).

C. Interest Rates

Finally, the court must determine the appropriate interest rate to apply to the just

compensation award described above. In doing so, “the court is to choose an interest rate

that puts the property owner in as good a financial position as if the compensation were

given concurrently with the taking.” Textainer Equip. Mgmt. Ltd. v. United States, 115

Fed. Cl. 708, 719 (2014) (citing Kirby Forest, 467 U.S. at 10); see also NRG Co. v.

United States, 31 Fed. Cl. 659, 670 n.8 (1994) (observing that “the point” of determining

interest rates “is not which approach yields a higher or lower payment, but rather which

approach is the more accurate measure of the economic harm to property owners”).

The “guiding principle” used to determine the appropriate interest rate to apply to

takings claims is the “Prudent Investor Rule,” which examines “how a reasonably

104 prudent person would have invested the funds owed by the government to produce a

reasonable return while maintaining safety of principal.” Sears v. United States, 124 Fed.

Cl. 730, 734-35 (2016) (internal quotation marks and alterations removed). Notably,

prior courts have recognized that the Prudent Investor Rule contemplates investments

with “minimal risk,” and have rejected proposed rates based on volatile investments.

Sears, 124 Fed. Cl. at 735 (rejecting rate premised on a diversified mutual fund with

volatility of 7.4 percent for failing to comport sufficiently with the “minimal risk”

criterion).

Here, the parties disagree on the proper marker for investments that comply with

the Prudent Investor Rule. The representative plaintiffs present two options for the

court’s consideration. Plaintiffs first argue that a mix of stocks and bonds evinces the

appropriate interest rate. In support, plaintiffs offered testimony from plaintiffs’ damages

expert, Dr. Bateman, who recommended a bond equity combination utilizing the iShares

Core U.S. Aggregate Bond Exchange Traded Funds (the “ETF”) by Blackrock and the

SPY Fund, which was created in 1993 by State Street Corporation in cooperation with the

American Stock Exchange and provides a rate of return of 3.5 to 4 percent. Tr. 629:19-

24. Dr. Bateman argued that the ETF best adheres to the Prudent Investor Rule as its

mixture of bonds and stocks speaks to “the principles of diversity and balance” which are

“key to managing risk.” Tr. 627:23-628:3. As an alternative, the representative plaintiffs

propose using the Moody’s Composite Index of Yields on Aaa Long-Term Corporate

Bonds (the “Moody’s rate”), which provides a rate of return of 1.7 percent. Tr. 618:13-

619:17. Dr. Bateman acknowledges that the Moody’s rate “has been recognized by the

105 Court of Federal Claims as an appropriate index for satisfying the [Prudent Investor

Rule].” Tr. 624:23-625:1.

The government disagrees with both of plaintiffs’ proposed rates, maintaining

instead that government bonds provide the appropriate measure of interest. In support,

the government offered the testimony of Dr. Sunding, who opined that, from an

economics perspective, the proper interest rate to apply is the yield on one-year Treasury

bonds (described as the “T-bond rate”), a “risk-free rate of return.” Tr. 2842:12-15. Dr.

Sunding maintains that using a risk-free rate of return to calculate interest is appropriate

here because “[t]he conversion of past losses into current period damages doesn’t require

the Plaintiffs to take on any additional risk.” Tr. 2843:4-9. Indeed, Dr. Sunding states

that doing otherwise would be “inappropriate as a matter of economics.” Id.

While the court acknowledges that others have tied the interest rate to the T-bond

rate in other takings cases, “the prudent-investor rule does not require that a reference be

made only to a rate of interest on Treasury securities where the United States is the

defendant.” Indep. Park Apartments v. United States, 61 Fed. Cl. 692, 717 (2004), rev’d

on other grounds, 449 F.3d 1235 (Fed. Cir. 2006). Indeed, other courts have recognized

a recent “consensus” reached among the Court of Federal Claims “that the Moody’s rate

is the appropriate benchmark by which to award delay damages” for takings claims.

Tech. College of the Low Country v. United States, 147 Fed. Cl. 364, 370 (2020) (quoting

Hardy v. United States, 138 Fed. Cl. 344, 356 (2018)); see also Adkins v. United States,

Nos. 09-503L, 02-241L, 09-158L, 2014 WL 448428, at *2 (Fed. Cl. Feb. 4, 2014)

(applying the Moody’s rate to delay damages arising from a 2003 taking); Biery v. United

106 States, Nos. 07-693L, 07-675L, 2012 WL 5914521, at *4 (Fed. Cl. Nov. 27, 2012)

(applying the Moody’s rate to delay damages arising from a 2007 taking); Sears, 124

Fed. Cl. at 736-37 (applying the Moody’s rate to delay damages arising from a taking that

occurred after February 2016). The court sees no reason to stray here. Accordingly, the

court concludes that the Moody’s rate is the appropriate measure of interest. See Pitcairn

v. United States, 547 F.2d 1106, 1124 (1976) (“[L]ong-term corporate bond yields are an

indicator of broad trends and relative levels of interest rates. They cover the broadest

segment of the interest rate segment.”).

The court further finds that it is appropriate to compound interest annually. As

repeatedly acknowledged by this court and others, “[p]rudent investment practices

compel the compounding of interest” as “no prudent, commercially reasonable investor

would invest at simple interest.” Tech. College of the Low Country, 147 Fed. Cl. at 370

(quoting Brunswick Corp. v. United States, 36 Fed. Cl. 204, 219 (1996)). But, the parties

dispute the timeline on which interest should be compounded. The representative

plaintiffs maintain that interest should be compounded quarterly; the government

maintains that interest should be compounded annually. Neither presents substantive

arguments to support its approach. Without meaningful guidance, the court elects to

follow its prior decisions in Textainer, 115 Fed. Cl. at 719, Adkins, 2014 WL 448428 at

*2, and Biery, 2012 WL 5914521 at *5, and concludes that interest in this case should be

compounded annually.

The interest rates adopted by the court should be applied for the Ideker levee as of

the date that those costs were incurred and for the diminution in fair market value as of

107 December 31, 2014, the accrual and valuation date. See Tech. College of the Low

Country, 147 Fed. Cl. at 367 (“Any delay in payment of . . . just compensation entitles

the property owner ‘to interest thereon sufficient to ensure that he is placed in as good a

position pecuniarily as he would have occupied if the payment had coincided with the

appropriation.’” (quoting Kirby Forest, 467 U.S. at 10)).

V. CONCLUSION

Based on the foregoing, the court finds that the government’s actions under the

MRRP and the flooding of the three representative plaintiffs’ properties constitutes the

taking of a permanent flowage easement under the Fifth Amendment, and that the

plaintiffs are entitled to just compensation for that taking. The parties are directed to file

a joint status report with a proposed judgment consistent with this opinion, to include the

amount of just compensation due to plaintiffs with interest, by January 15, 2021.

IT IS SO ORDERED.

s/Nancy B. Firestone NANCY B. FIRESTONE Senior Judge