concurring and dissenting.
I concur in the Court’s opinion with respect to the claims against Harris County. But I dissent from the Court’s opinion with respect to the Harris County Flood Control District. Because the District has conclusively disproven the intent necessary to a takings claim (and a nuisance claim)— the only basis for waiver of governmental immunity asserted here — I would dismiss the claims against the District for lack of subject-matter jurisdiction.
Factual Background
Edward and Norma Kerr, together with approximately 400 other plaintiffs with homes in the upper White Oak Bayou area of northwestern Harris County, sued Harris County and the District (collectively, the Harris County entities) alleging that their homes flooded during storms in 1998, 2001, and 2002 because of the Harris County entities’ activities within the White Oak Bayou watershed. White Oak Bayou is a watershed of over 110 square miles that has long been identified as a high-risk area for flooding by the Federal Emergency Management Agency. According to FEMA’s 1976 flood maps, most of the homeowners’ properties were not located in the 100-year floodplain at the time of their construction in the mid-to-late 1970s and early 1980s. The homeowners asserted inverse condemnation and nuisance claims against the Harris County entities for “taking” their property without ade- ' quate compensation.
The District is a legislatively created special district, charged with “the control ... of the storm and flood waters, and the waters of the rivers and streams in Harris County and their tributaries for ‘flood control ... and other useful purposes.’ ” Act of May 10, 1937, 45th Leg., R.S., ch. 360, § 1, 1937 TEX. GEN. LAWS 714 (as amended). It is responsible for designing, building, and maintaining flood-control projects and infrastructure which are intend to reduce the risk of — but cannot always prevent — flooding. The importance of the District’s work is underscored by the fact that much of Harris County has been, and for the foreseeable future will continue to be, at risk for flooding during severe storms. This risk exists because flooding is inevitable when rainfall levels reach a certain number. The storm sewers cannot handle the rainfall, causing the storm water to pond and secondary drainage systems to surcharge.
In the early 1960s, almost 11 miles of White Oak Bayou were enlarged and lined with concrete by the Army Corps of Engineers. Flood control in the area was initially handled according to a 1965 study by Turner Collie & Braden, which called for the construction of drainage facilities throughout the watershed, storm sewers, drainage ditches, and other improvements *260to the bayou. At that time, the Harris County entities believed the bayou had sufficient capacity to accept runoff from planned development, making onsite detention facilities unnecessary.
The first detailed flood-insurance map was produced by the Army Corps in 1976. On that map, nearly all of the homeowners’ homes were built above the 100-year floodplain. The Corps completed its “Interim Report on Upper White Oak Bayou” the same year, concluding that the flooding problems there were “caused primarily by inadequate channel capacities of the streams.” The report observed that the flooding was “compounded by continuing urbanization which increases and accelerates the runoff from rainfall,” and recommended a number of improvements that focused primarily on channel enlargement and rectification of the lower bayou at a federal cost of about $50 million and a local cost of about $7 million. According to the final report, the District and the Harris County Commissioners Court “expressed favorable comments concerning the proposed project and have agreed to provide the necessary items of local cooperation.” The Corps, however, did not authorize any federal funding for construction.
In 1980, the District, as the local sponsor of federal flood-control projects, stated its objective to construct and maintain facilities “intended to minimize the threat of flooding.” The District stated that its objective could be best met in the long-term by “a continued program of improvement and extension of the District’s open channels.” However, the curtailment of federal funding had slowed open channel improvements, and as a result, the District recommended that its approval policy for new construction projects be modified to adopt “appropriate criteria for covering the design of supplemental storm water systems.” In additions to its concerns about the lack of federal funding, the District also questioned whether the Corps’ information and methodology were outdated. Out of these concerns, the District hired two engineering firms — Turner Collie & Braden and Pate Engineers — to update the computer models and develop a new method for determining rainfall/runoff characteristics and flooding. The Pate Plan was the result of this effort.
The purpose of the Pate Plan was to “to alleviate flooding from the 100-year storm event along White Oak Bayou and allow full development of the watershed” and to ensure that “new land developments [did] not increase the flooding potential for other properties in the watershed.” The Pate Plan was to be funded by “District resources,” contributions by municipal utility districts, and a $8,000 per acre impact fee imposed on private developers for properties of less than ten acres. The Pate Plan had five components, including three regional storm-water-detention facilities and earthen channel improvements. Two of the stated purposes of the regional facilities were to: (1) “reduce existing flood levels while not allowing flood level increases from new developments” and the proliferation of small detention basins on each development; and (2) collectively fund and install regional basins, which are more effective and reliable. The total cost for the final improvements would be more than $66 million. Because the funds were not immediately available, however, the Pate Plan recommended a series of interim measures to address flooding at a cost of $25 million. That recommendation was based on the assumption that 2,500 acres of new development would occur and help fund the project. The County adopted the Pate Plan and authorized the District to commence work implementing it in 1984.
After a downturn in the economy stalled the anticipated funding, Pate En*261gineers conducted an interim study “to define an initial channel improvement project to eliminate existing flooding upstream of the existing concrete lined channel.” This 1985 study included the design of the initial project and the establishment of rights-of-way where the planned improvements were to be constructed. Land acquisition began the next year, and construction on the first detention facility began in 1988. The District acknowledged in correspondence with the Corps that the District would lose federal participation in the project by moving forward with its own plan; nevertheless, the District remained committed to constructing the first phase of the Pate Plan because “timely implementation of the regional project [was] critical.”
In 1989, the Pate Plan had not been fully implemented and severe storms resulted in the flooding of over 200 homes near White Oak Bayou. In response to a homeowner complaint, the District responded that it was “very aware of the house flooding potential” in some subdivisions along White Oak Bayou. To “greatly reduce” this risk, the District was “in the process of implementing the initial phase” of the Pate Plan and had already purchased regional detention sites, adjusted bridges, and approved construction plans but was awaiting the condemnation of various rights-of-way. The District anticipated this work would be completed in the near future, allowing it to seek construction bids for “the first channel rectification segment.”
In 1990, the District ordered a new study of the White Oak Bayou by Klotz Associates. The reason for the new study is disputed. The District asserts that the 1989 flooding caused the District to reevaluate the Pate Plan. According to its director, Michael Talbott, the 1989 flood revealed that the engineering model used to develop the Pate Plan “was not accurate. The computer model failed to predict unexpected flooding in lower White Oak Bayou and the District was forced to reevaluate the underlying premises of the Pate [P]lan.” The homeowners assert — but did not prove — that the new study was conducted to avoid a tax rate increase.1
Klotz Associates determined that the Plate Plan “need[ed] to be modified to provide even more protection” because the flood flows and levels were higher than those shown in the Pate Plan. Klotz recommended two earthen channel improvements and a detention storage facility, which would have “eliminated most of the flooding” in one of the nine subdivisions where the homeowners’ properties are located. Klotz completed a first report in 1992 in which it updated the conditions existing at the time. After reworking the computer model and applying more advanced computer software, Klotz determined that there were “significantly higher flows and flood levels than in” the previous flood-hazard survey.
The parties dispute the effectiveness of the Klotz Plan. Talbott stated in his affi*262davit that Klotz’s “multistage study” recommended a regional flood-control plan, including some features that “could be implemented over a shorter-term horizon than the original [Pate] [P]lan. The features of that near-term plan ... [were] similar to but more extensive than the ... Pate Plan.” The Klotz Plan also called for the District to require on-site detention for future development. The homeowners in their brief describe the Klotz Plan as “less effective,” “downsized,” and a “reduced mitigation plan” because it implemented only a portion of the channel improvements in the Pate Plan and did not protect against 100-year floods.2
Construction began on another detention facility that was not part of the Pate Plan but was recommended by Klotz. Over the course of the next several years, the District continued construction and acquired additional land for the project. Klotz submitted a “Final Recommended Plan for White Oak Bayou” in April 1998, which the District submitted to FEMA. Tropical' Storm Frances — the storm that first lead to the homeowners’ claims — occurred five months later. FEMA produced new floodplain maps in 1999. On those maps, the homeowners’ houses were located in the 100-year floodplain.
The homeowners’ properties experienced flooding during Tropical Storm Frances and two later storm events, Tropical Storm Allison in June 2001 and an unnamed storm in October 2002. According to Andrew Yung, an expert for the District, Tropical Storm Frances saw average rainfall in the Upper White Oak Bayou watershed between a 50- and 100-year frequency and a rainfall of “greater than a 100-year event in the immediate vicinity of the Plaintiffs’ homes.” Steve Fitzgerald, the District’s chief engineer, also stated that Tropical Storm Frances “was an unprecedented storm that produced extremely heavy rainfall.” The homeowners’ expert, Larry Mays, stated that Tropical Storm Frances resulted in flooding approximating a 10-year event. With respect to Tropical Storm Allison, Young stated that it brought rainfall greater than a 100-year frequency; Mays, in contrast, opined that it was a 50-year event. The third storm did not exceed a 100-year storm, but was still a serious and unusual storm. According to Yung, it brought rainfall between a 25-year and a 50-year storm event; according to Mays, it was a 10-year event.
Pertinent Procedural Background
The homeowners brought takings and nuisance claims against the Harris County entities. Their claims focus on three acts: (1) the approval of upstream development, (2) the failure to complete the Pate Plan, and (3) the construction of a transitional flood-control structure (i.e., a dam).
The Harris County entities challenged each of these claims in a combined plea to the jurisdiction and motion for summary judgment. The plea was supported by affidavits from Talbott and Melvin Spinks, P.E., an expert in water resources engineering and civil engineering. As stated by Talbott, the District did not proceed with the Pate Plan because the 1989 flooding demonstrated that the Pate Plan “was based upon faulty information and would not have worked.” Talbott averred that the Klotz studies in the 1990s — which were before the second and third storms at issue here — developed a new model that was “more sensitive to observed storm events and additional factors not contained in the [earlier] computer model.” Talbott fur*263ther stated that the earlier model was based upon certain federal policy decisions that required the exclusion of data regarding “a substantial amount of’ development at the time. The District approved the Klotz Plan based on the engineers’ certification that the project complied with applicable regulations and would not increase downstream runoff. The revisions to the Pate Plan resulted in improvements to alleviate flooding, which “greatly exceeded the work called for in the original Pate Plan,” and additional design and construction work. At the time of its plea, the District had spent over $70 million improving the bayou, including the completion of ten regional detention facilities and the purchase of land for ten more.
The homeowners’ response included an affidavit and reports from their expert Mays, a professor of civil and environmental engineering at Arizona State University. Mays stated in his report that the failure to implement the Pate Plan “for the segment of White Oak Bayou adjacent” to the homeowners’ properties and approval of “thousands of additional acres of development ... without proper storm-water management measures” caused the 1989 flooding of the homeowners’ properties. In his affidavit, Mays averred that the Harris County entities approved development upstream of the homeowners’ properties knowing that (1) development causes increased runoff downstream, (2) insufficient capacity existed in the bayou adjacent to the approved upstream development to handle the increased runoff, and (3) the developers of the upstream area had not provided sufficient mitigation for the increased runoff. He further stated that the flooding from the three storms resulted from the approval of the upstream development and the failure to implement the entire Pate Plan, not unprecedented rainfall or inadequate local drainage systems within the subdivisions.
More specifically, Mays stated that only three of the Pate Plan’s five components were implemented, leaving portions of the work that “would have alleviated” flooding of the homeowners’ properties unfinished. Mays averred that the full implementation of the Pate Plan would have mitigated the adverse impact from upstream development along White Oak Bayou, but the District chose not to implement the phase of the Pate Plan that “would have prevented flooding of [the homeowners’ properties] up to and including a 100-year flood event.” Mays further stated that Klotz had to revise the models developed for the Pate Plan because the District permitted additional land development without adequate storm-water-runoff mitigation. Finally, based on the Corps’ 1976 study, Mays stated that the Harris County entities knew that approving “unmitigated upstream development ... would be substantially certain to result in increased flooding along the bayou in the vicinity of the [homeowners’] properties.”
The trial court denied the Harris County entities’ motion and they appealed, challenging all three bases for the homeowners’ claims. The homeowners do not contend on appeal that the construction of the dam increased flooding on their property or amounted to a taking of their property.
Immunity
The Harris County entities have governmental immunity. Governmental immunity includes two distinct principles: immunity from suit and immunity from liability. See Tex. Dep’t of Parks & Wildlife v. Miranda, 133 S.W.3d 217, 224 (Tex.2004). While immunity from liability is an affirmative defense, immunity from suit deprives a court of subject-matter jurisdiction. See id. The Harris County entities’ *264governmental immunity deprives the trial court of subject-matter jurisdiction over this action absent a proper waiver of immunity. See id.; Tex. Dep’t of Transp. v. Jones, 8 S.W.3d 636, 638 (Tex.1999). Governmental immunity is waived for constitutional takings claims. See Tex. Const, art. I, § 17; Gen. Servs. Comm’n v. Little-Tex Insulation Co., Inc., 39 S.W.3d 591, 598 (Tex.2001).
Although the homeowners assert both a takings claim and a nuisance claim, they rely exclusively on their constitutional takings claim to establish a waiver of the Harris County entities’ immunity from suit. The trial court’s subject-matter jurisdiction over this suit is therefore contingent on the jurisdictional requisites of a takings claim, whether asserted as a traditional takings claim or as a nuisance claim that “rises to the level of a constitutional taking under Article I, Section 17.” See City of Dallas v. Jennings, 142 S.W.3d 310, 312 (Tex.2004) (holding that city retained immunity from nuisance claim because plaintiffs did not establish constitutional taking and did not assert separate waiver of immunity for nuisance claim). Thus, if either of the Harris County entities has disproved a jurisdictional requisite of a takings claim, this Court must dismiss this action — including both the homeowners’ takings claim and their nuisance claim — with respect to that defendant. See, e.g., Hearts Bluff Game Ranch, Inc. v. State, 381 S.W.3d 468, 491 (Tex.2012) (dismissing action for want of subject-matter jurisdiction because claimant could not establish a viable takings claim); Gulf Coast Waste Disposal Auth. v. Four Seasons Equip., Inc., 321 S.W.3d 168, 175-76 (Tex.App.-Houston [1st Dist.] 2010, no pet.) (dismissing takings claim when evidence negated intent element and, regarding takings claim, observing that “the government retains immunity absent intentional conduct”). This is true regardless of whether the issue was raised in the Harris County entities’ plea to the jurisdiction or their motion for summary judgment. See, e.g., City of Denton v. Paper, 376 S.W.3d 762, 764 (Tex.2012) (dismissing suit on immunity grounds in interlocutory appeal from denial of summary judgment); Miranda, 133 S.W.3d at 234 (dismissing suit on immunity grounds in appeal from denial of plea to jurisdiction).
Intent
To recover on a takings claim under Article I, Section 17 of the Texas Constitution, a property owner must establish that the governmental entity: (1) engaged in a specific act causing the taking, damaging, or destroying of private property; (2) engaged in the act intentionally, i.e., either knowing that the specific act was causing identifiable harm or knowing that specific property damage is substantially certain to result; and (3) took the property for a public use. See Sw. Bell Tel., L.P. v. Harris Cnty., 267 S.W.3d 490, 495 (Tex.App.-Houston [14th Dist.] 2008, no pet.). Thus, in a takings claim, the plaintiff must prove intent, causation, and public use. See City of San Antonio v. Pollock, 284 S.W.3d 809, 812 (Tex.2009) (dismissing takings and nuisance claims due to lack of evidence that city knew its actions were substantially certain to cause harm); City of Keller v. Wilson, 168 S.W.3d 802, 808 (Tex.2005) (holding that intent element requires proof that governmental entity (1) knows that specific act is causing identifiable harm and intends that harm or (2) knows that specific property damage is substantially certain to result from authorized government action); Tarrant Reg’l Water Dist. v. Gragg, 151 S.W.3d 546, 554 (Tex.2004) (affirming jury verdict for landowner whose property was repeatedly flooded after upstream construction of water-supply reservoir); Jennings, 142 *265S.W.3d at 312 (dismissing takings and nuisance claims because there was no evidence that city knew plaintiffs’ home would be damaged or that damage was substantially certain).
To prevail on the intent element, it is not enough for the homeowners to show that the acts themselves were intentional. Jennings, 142 S.W.3d at 813. Rather, the homeowners must prove that the Harris County entities “knew (not should have known) that flooding was substantially certain.” City of Keller, 168 S.W.3d at 829. The governmental entity’s “intent” must be established under a “heightened intent standard.” City of Arlington v. State Farm Lloyds, 145 S.W.3d 165, 168 (Tex.2004). “Substantially certain” does not mean flooding is possible, at increased risk, or even more likely than not. See Pollock, 284 S.W.3d at 821 (“The governmental entity’s awareness of the mere possibility of damage is no evidence of intent”); Gulf Coast Waste Disposal Auth., 321 S.W.3d at 175 (holding that “allegations demonstrat[ing] awareness ... that an increased risk existed” did not satisfy “necessarily incident to or a consequential result of’ intent standard in takings case). A government entity has knowledge that flooding is substantially certain to occur “only when the damage is ‘necessarily an incident to, or necessarily a consequential result of the [entity’s] action.’” Pollock, 284 S.W.3d at 821 (quoting Jennings, 142 S.W.3d at 314).
There is also a temporal component of intent: “the government’s knowledge must be determined as of the time it acted, not with benefit of hindsight.” Id. at 821. The homeowners’ burden includes producing “evidence of ‘objective indicia of intent’ showing the [Harris County entities] knew identifiable harm was occurring or substantially certain to result” irom their actions. City of Keller, 168 S.W.3d at 830. In evaluating the objective evidence of a governmental entity’s intent, a court should evaluate any evidence regarding what the entity was told about the consequences of its actions. Id. at 829.
A. The District disproved the necessary intent for a takings claim
The homeowners contend that the Harris County entities did not present admissible evidence negating the intent element of the takings claim. More specifically, the homeowners contend that Talbott’s affidavit regarding intent is conclusory and only relates to the District, not the County. Alternatively, the homeowners contend that they presented sufficient contrary evidence to raise an issue of fact with respect to intent.
1. Talbott’s affidavit did not establish the County’s lack of intent
The homeowners are correct that Tal-bott’s twenty-three-page affidavit does not address the County’s intent; it addresses only the District’s intent. The County therefore did not satisfy its burden to establish that it lacked the requisite intent, the homeowners did not have to produce any evidence to raise a fact issue as to intent,3 and the trial court’s summary judgment in favor of the County cannot be affirmed on this ground. I therefore concur with the Court’s disposition' of the homeowners’ claims against the County.
2. Talbott’s affidavit did establish the District’s lack of intent
Talbott’s affidavit established the District’s lack of intent. Talbott, a licensed civil engineer and hydrologist, began working for the District as a floodplain hydrolo*266gist and was steadily promoted until he reached his current position as director. Talbott detailed the District’s attempts to reduce the risk of flooding in Harris County and categorically stated that the District knew only of the general possibility of flooding (which he described as “a flooding risk”), but did not know that its specific actions would cause flooding. He stated that to intentionally cause flooding would be “contrary to the [District’s] fundamental reason for existence and its policies and values.” Regarding the area in question, Talbott averred that the District’s goal for the past twenty-four years has been to implement a regional flood-damage-reduction plan for White Oak Bayou, and the District has “greatly. exceeded the work that was originally called for in the 1984 Pate Plan and has spent more than $70 million” on studies and construction “to decrease the risk of flooding along White Oak Bayou.” He described in detail aspects of the District’s work in White Oak Bayou. He further stated, “I can say without reservation that the District did not intentionally cause flooding of Plaintiffs’ properties” and that the District’s work was not “performed with substantial certainty that such work would cause the flooding of Plaintiffs’ properties or that flooding would be necessarily incident to or a consequential result of the District’s work.”
Talbott explained that the District does not' attempt to eliminate the risk of flooding; indeed, that would be impossible in many areas of Harris County.
[T]he District cannot prevent all flooding. The area that is called White Oak Bayou is a natural geographical gateway and watershed that has transported storm water runoff long before the Allen Brothers landed in Houston and has flooded many times throughout history. In simple terms, a “watershed” is the region of land whose water drains into a body of water such as White Oak Bayou. Plaintiffs’ homes were built in this naturally occurring watershed near the natural waterway with its naturally occurring floodplain. The District is well aware of the flooding potential of this watershed for the reasons stated above (flat topography, clay soils, abundant rainfall, and history of flooding); the District is not aware of any flooding caused by actions of the District.
Talbott also explained that the District’s knowledge about flooding risks evolved after the Pate Plan was adopted because more information became known and the various disciplines involved in designing and implementing flood-risk-reduction plans — e.g., hydrology, hydraulics, civil engineering, geography, geotechnical engineering, surveying, and computer science — advanced in technology and in their understanding of the dynamics of flooding. Thus, the District relied on computer software developed by the Corps and private engineering firms in trying to predict “how stormwater might act under certain conditions.” He stated that the Pate Plan was not implemented because “it was based on knowledge and assumptions available at one point in time that were later proven to be scientifically inaccurate by actual storm events and further analysis by outside engineering firms.”
Talbott’s statements were not concluso-ry. A conclusory statement is one that does not provide the facts on which it is based. Arkoma Basin Exploration Co. v. FMF Assocs. 1990-A, Ltd., 249 S.W.3d 380, 389 n. 32 (Tex.2008).4 Talbott’s affidavit included the factual basis for his *267statements regarding the District’s intent. A governmental entity, unlike a natural person, can only testify through its employees or representatives. See Waterman S.S. Corp. v. Ruiz, 355 S.W.3d 387, 402 (Tex.App.-Houston [1st Dist.] 2011, no pet.) (stating that legal entity’s representative may testify to facts regarding entity’s activities); Martinez v. Hays Const., Inc., 355 S.W.3d 170, 178 (Tex.App.-Houston [1st Dist.] 2011, no pet.) (same).
Because the District satisfied its burden of proof, the burden to establish intent shifted to the homeowners.
B. The homeowners did not create a fact issue on the District’s intent
In the trial court, the homeowners relied upon three categories of conduct by the Harris County entities that they claim were substantially certain to result in the flooding of their homes: (1) failure to fully implement the Pate Plan; (2) approval of upstream development; and (3) construction of a dam. I address each one separately.
1. Failure to implement the Pate Plan
The District’s failure to implement the Pate Plan cannot, without more, satisfy the intent element of the homeowners’ takings claim. The intent element cannot be satisfied by demonstrating that the governmental entity acted with negligence. City of El Paso v. Ramirez, 349 S.W.3d 181, 187 (Tex.App.-El Paso 2011, no pet.) (“A governmental entity’s failure to act, even in the face of evidence that curative measures are necessary to prevent future damage, rise only to the level of a negligence claim.”). A governmental entity’s knowledge “of the potential for” flooding, and “its subsequent failure to take measures to prevent such flooding,” is insufficient to satisfy this element. Id. Mere negligence does not constitute “an unconstitutional taking” because that “would be an anomalous result if ... an entity otherwise generally entitled to immunity for negligence[] were subject to liability for something less than intentional behavior.” Gragg, 151 S.W.3d at 554. Even “a finding of gross negligence does not supply the requisite intent to sustain liability of a governmental entity for a constitutional takings claim.” Karnes City v. Kendall, 172 S.W.3d 624, 629 (Tex.App.-San Antonio 2005, pet. denied). And evidence that the governmental entity’s actions caused flooding does not establish the requisite intent. Ahart v. Tex. Dep’t of Transp., No. 14-05-00027-CV, 2006 WL 2167223, at *4 (Tex.App.-Houston [14th Dist.] Aug. 1, 2006, pet. denied) (mem. op.).
In City of Keller, the Court held that the city’s failure to comply with its master drainage plan in approving a developer’s drainage plans did not. show intent to cause flooding damage. 168 S.W.3d at 829. In City of El Paso, the court of appeals held that the city’s failure to comply with its landfill permit and new landfill regulations did not show intent to cause flooding damage. 349 S.W.3d at 187. Additionally, prior warnings from the Texas Commission on Environmental Quality, the city’s own expert report, and previous experience with erosion and runoff did not show intent to cause flooding damage. Id. at 185-86; see also City of Van Alstyne v. Young, 146 S.W.3d 846, 850 (Tex.App.-Dallas 2004, no pet.) (concluding that city’s knowledge of prior pump problems was not same as knowledge that its decision not to replace pumps would cause flooding of plaintiffs home); City of Del Rio v. Felton, No. 04-06-00091-CV, 2007 WL 247655, at *7 (Tex.App.-San Antonio Jan. 31, 2007) (mem. op.) (holding that knowledge that watering of park caused damage to neighboring property did not establish intent to take plaintiffs’ property).
*268As the City of El Paso court explained, even when there is evidence that a governmental entity knew that the specific property damage was substantially certain to result from government inaction, it is not sufficient if it “fails to account for the State’s general immunity from liability for negligence, and risks public payment for damage for which the public received no benefit.” 349 S.W.3d at 186-87. Like the intent evidence in City of El Paso, the homeowners’ intent evidence with regard to the District’s failure to implement the Pate Plan is evidence of negligence here, but it does establish the intent necessary for a takings claim. See id. “A governmental entity’s failure to act, even in the face of evidence that curative measures are necessary to prevent future damage, rise[s] only to the level of a negligence claim.” Id. at 187.5
2. Approval of development6
The homeowners argue that their expert affidavit from Mays creates a fact issue on intent with respect to the District’s approval of upstream development. Mays opined:
[The Harris County entities] approved development upstream of [the homeowners’ properties]
• Knowing that development causes increased runoff downstream[;]
• Knowing that there was insufficient capacity in the bayou adjacent to [the homeowners’ properties] to handle any increased runoff from upstream development^ and]
• Knowing that no mitigation or insufficient mitigation was implemented by the developers associated with their upstream developments[.]
The District responds that even if Mays’s assertions were true, they do not provide evidence of the requisite intent because Mays’s theory that the upstream developers failed to include adequate detention was based on an assumption that had no “evidentiary support.” But even if we assume that all three of these statements are supported by the summary judgment evidence, the District’s knowledge that runoff would increase and that the bayou could not handle the increase does not equate to knowledge that specific property damage was substantially certain to occur. For example, the runoff could fill the streets or yards but cause no property damage.
Regardless, Mays’s assertion regarding the adequacy of the developers’ mitigation efforts is conclusory. Inadequate detention cannot be assumed “merely because flooding subsequently occurred.” Mays conceded that he did not know which developments upstream of the homeowners’ *269homes were in fact developed with no or insufficient mitigation. The District also points out that it required new development to include either adequate onsite detention facilities or, for tracts of less than ten acres, a $8,000 per acre impact fee to be used for construction of larger regional detention facilities. The District presented evidence that it intended to raise money via the impact fees to provide regional detention and flood-control measures, which the District believed would be more effective. Regardless of whether the District was correct or was negligent in its implementation of such regional measures, this evidence negates the homeowners’ reliance on the mere absence of detention facilities in new construction to establish intent.
The District was in the midst of an ongoing flood control program. It was necessary, therefore, for the District to know that the flooding of the homeowners’ properties was substantially certain to occur as a result of additional development before the District completed its revised program. To demonstrate such knowledge, the homeowners had to show: (1) the amount of additional development that occurred in the time between the adoption of the Pate Plan and each of the three flood events; (2) the construction undertaken in that interim period both pursuant to and in lieu of the Pate Plan; and (3) the County was substantially certain the totality of the construction work would be insufficient. Instead, the homeowners offered proof that the entirety of the Pate Plan was not followed and that the Klotz Plan was less effective than the Pate Plan.
The homeowners did not present any “evidence that the [District] knew that the [homeowners’] property was being damaged or that damage was a necessary consequence” in the delay in completing the flood-control measures. Pollock, 284 S.W.3d at 821. The District certainly knew that it was possible that the additional development in the area would increase runoff in the bayou and that a rainfall could occur that would cause some flooding,7 but its “awareness of the mere possibility of damage is no evidence of intent.” Id. The homeowners had no evidence that the District’s decisions made downstream flooding before completion of flood control measures “inevitable.” Gragg, 151 S.W.3d at 555; see also Jennings, 142 S.W.3d at 315 (it is insufficient to show that the governmental entity’s action “sometimes results” in the claimed property damage). Just as in Pollock, the damage the homeowners claim — flood water flowing onto their land from upstream development— “is neither necessarily incident to or a consequential result of the [approval of the other developments].” “It can be prevented,” id., and the District established that it was in the process of taking measures to prevent it. The homeowners’ true complaint is that the District could have prevented or lessened the risk of damage to their property but failed to do so. Such allegations might support a negligence claim but, absent evidence that the District took affirmative action that it knew would cause the damage to the homeowners’ properties, will not support a takings claim against a governmental entity.
The homeowners also contend that the District’s intent was demonstrated by evidence of reoccurrence: three floods in five years. Recurrence “is a probative factor” in determining whether the claimed prop*270erty damage is “substantially certain to occur.” Gragg, 151 S.W.3d at 555. But only one of the three floods — the 1989 flood — occurred before the County approved the developments in question. One flood does not establish recurrence. See Pollock, 284 S.W.3d at 821 (Tex.2009) (“The government’s knowledge must be determined as of the time it acted, not with benefit of hindsight”).
3. Construction of a dam8
Finally, Mays states in his affidavit that the District “constructed a dam downstream of [the homeowners’ properties] that was intended to ensure that [the homeowners’ properties] received no benefit from that portion of the [regional flood-control] plan that had been constructed downstream.” But Mays’s affidavit and reports do not provide any factual basis for this assertion of the District’s intent. See, e.g., City of Keller, 168 S.W.3d at 830 (requiring claimant to produce “objective indicia of intent” that governmental entity “knew identifiable harm was occurring or substantially certain to result”) (quoting Gragg, 151 S.W.3d at 555).
The District, on the other hand, presented evidence that it “never intended that the [dam would] cause flooding” and “never had any information that showed with substantial certainty that the temporary [structure would cause' flooding or that flooding would be necessarily incident to or necessarily a consequential result of [ ] building and maintaining the [structure.” Talbott testified that transition control structures such as the dam used by the District, “are not known to increase any flooding risk” and “are commonly used in many waterways and have been built across Harris County, other counties, communities, and cities across the United States, and the United States Army Corps of Engineers has used transition control structures on many of their projects.” He further testified that the dam used by the District was “typical” and was a temporary measure until channel work continued upstream. Thus, the homeowners failed to rebut the District’s evidence that it lacked the requisite intent for a takings claim and the attendant waiver of governmental immunity.
Conclusion
Because the District has presented evidence establishing its lack of the requisite intent for a takings claim — the sole basis for the homeowners’ asserted waiver of the District’s governmental immunity — and the homeowners have not presented contrary evidence raising an issue of fact on intent, I would dismiss the homeowners’ claims against the District. I therefore dissent from the portion of the Court’s opinion that affirms the trial court’s order denying the plea to the jurisdiction with respect to the District. I concur in the remainder of the Court’s disposition.
. According to the deposition testimony of Art Storey, the District’s executive director, the tax rate at that time was five cents per hundred dollars of assessed valuation. Storey testified that he was informally asked by some commissioners for his opinion on the tax rate and that he recommended raising it above 5 cents. Storey testified that he "initially identified requirements that could have cost as much as 10 cents” and that he was successful in getting the rate raised but not to that level. He also testified that, as "[t]imes started to improve,” there was agreement regarding a lower rate. The homeowners did not present any evidence that Klotz was instructed to develop a less expensive alternative to the Pate Plan or what programs, if any, were part of Storey's "initial requirements” that were not implemented because of the lower tax rate.
. The homeowners’ record citations indicate that the Klotz Plan stated it used the 10-year frequency plan for its analysis. That plan is not in the record.
. The County relied on Talbott’s affidavit but his affidavit does not address the County’s intent nor does it show how he would know the County’s intent.
. For expert witnesses, a statement is also considered conclusory when the stated basis for the opinion does not actually support the opinion. Pollock, 284 S.W.3d at 818.
. The homeowners identify the District’s intentional act as the implementation of the Klotz Plan in lieu of the Pate Plan. But the homeowners cannot convert the District’s failure to implement the Pate Plan into an affirmative action by pairing it with the District’s use of the Klotz Plan. They have not argued or presented evidence tending to show that the implementation of the Klotz Plan caused them property damage that would not have occurred if the District had not implemented any flood control measures.
. The Harris County entities have not argued that a governmental entity cannot be liable for approving a developer's plans. We therefore do not consider that issue. See City of Keller, 168 S.W.3d at 810 (reserving this question "for another day”); but see id. at 833-34 (O’Neill, J., concurring) (stating that "[b]e-cause the primary responsibility for a development’s design rests with the developer, and because the plat-approval process does not transfer such responsibility to the municipality, mere plat approval cannot be a basis” for liability and that "the developer's defective drainage design,” rather than the city's approval of the plat was the proximate cause of the flooding as a matter of law).
. The very existence of a 100-year floodplain is an acknowledgment that a flood is likely to occur once every 100 years. As explained by Talbott, a 100-year flood is "a flood that statistically should occur once in 100 years, or more accurately ... [is] a 1% chance of being equaled or exceeded in any given year.”
. The homeowners argued this theory in their summary-judgment briefing but not in their appellate briefing.