Winter v. Natural Resources Defense Council, Inc.

delivered the opinion of the Court.

“To be prepared for war is one of the most effectual means of preserving peace.” 1 Messages and Papers of the Presidents 57 (J. Richardson comp. 1897). So said George Washington in his first Annual Address to Congress, 218 years ago. One of the most important ways the Navy prepares for war is through integrated training exercises at sea. These exercises include training in the use of modern sonar to detect and track enemy submarines, something the Navy has done for the past 40 years. The plaintiffs, respondents here, complained that the Navy’s sonar-training program harmed marine mammals, and that the Navy should have prepared an environmental impact statement before commencing its latest round of training exercises. The Court of Appeals upheld a preliminary injunction imposing restrictions on the Navy’s sonar training, even though that court acknowledged that “the record contains no evidence that marine mammals have been harmed” by the Navy’s exercises. 518 F. 3d 658, 696 (CA9 2008).

The Court of Appeals was wrong, and its decision is reversed.

I

The Navy deploys its forces in “strike groups,” which are groups of surface ships, submarines, and aircraft centered around either an aircraft carrier or an amphibious assault ship. App. to Pet. for Cert. 316a-317a (Pet. App.). Seamless coordination among strike-group assets is critical. Before deploying a strike group, the Navy requires extensive integrated training in analysis and prioritization of threats, execution of military missions, and maintenance of force protection. App. 110-111.

Antisubmarine warfare is currently the Pacific Fleet’s top war-fighting priority. Pet. App. 270a-271a. Modern diesel-electric submarines pose a significant threat to Navy vessels because they can operate almost silently, making them ex*13tremely difficult to detect and track. Potential adversaries of the United States possess at least 300 of these submarines. App. 571.

The most effective technology for identifying submerged diesel-electric submarines within their torpedo range is active sonar, which involves emitting pulses of sound underwater and then receiving the acoustic waves that echo off the target. Pet. App. 266a-267a, 274a. Active sonar is a particularly useful tool because it provides both the bearing and the distance of target submarines; it is also sensitive enough to allow the Navy to track enemy submarines that are quieter than the surrounding marine environment.1 This case concerns the Navy’s use of “mid-frequency active” (MFA) sonar, which transmits sound waves at frequencies between 1 kHz and 10 kHz.

Not surprisingly, MFA sonar is a complex technology, and sonar operators must undergo extensive training to become proficient in its use. Sonar reception can be affected by countless different factors, including the time of day, water density, salinity, currents, weather conditions, and the contours of the sea floor. Id., at 278a-279a. When working as part of a strike group, sonar operators must be able to coordinate with other Navy ships and planes while avoiding interference. The Navy conducts regular training exercises under realistic conditions to ensure that sonar operators are thoroughly skilled in its use in a variety of situations.

The waters off the coast of southern California (SOCAL) are an ideal location for conducting integrated training exercises, as this is the only area on the west coast that is relatively close to land, air, and sea bases, as well as amphibious *14landing areas. App. 141-142. At issue in this case are the Composite Training Unit Exercises and the Joint Tactical Force Exercises, in which individual naval units (ships, submarines, and aircraft) train together as members of a strike group. A strike group cannot be certified for deployment until it has successfully completed the integrated training exercises, including a demonstration of its ability to operate under simulated hostile conditions. Id., at 564-565. In light of the threat posed by enemy submarines, all strike groups must demonstrate proficiency in antisubmarine warfare. Accordingly, the SOCAL exercises include extensive training in detecting, tracking, and neutralizing enemy submarines. The use of MFA sonar during these exercises is “mission-critical,” given that MFA sonar is the only proven method of identifying submerged diesel-electric submarines operating on battery power. Id., at 568-571.

Sharing the waters in the SOCAL operating area are at least 37 species of marine mammals, including dolphins, whales, and sea lions. The parties strongly dispute the extent to which the Navy’s training activities will harm those animals or disrupt their behavioral patterns. The Navy emphasizes that it has used MFA sonar during training exercises in SOCAL for 40 years, without a single documented sonar-related injury to any marine mammal. The Navy asserts that, at most, MFA sonar may cause temporary hearing loss or brief disruptions of marine mammals’ behavioral patterns.

The plaintiffs are the Natural Resources Defense Council, Inc., Jean-Michael Cousteau (an environmental enthusiast and filmmaker), and several other groups devoted to the protection of marine mammals and ocean habitats. They contend that MFA sonar can cause much more serious injuries to marine mammals than the Navy acknowledges, including permanent hearing loss, decompression sickness, and major behavioral disruptions. According to the plaintiffs, several mass strandings of marine mammals (outside of SOCAL) *15have been “associated” with the use of active sonar. They argue that certain species of marine mammals — such as beaked whales — are uniquely susceptible to injury from active sonar; these injuries would not necessarily be detected by the Navy, given that beaked whales are “very deep divers” that spend little time at the surface.

II

The procedural history of this case is rather complicated. The Marine Mammal Protection Act of 1972 (MMPA), 86 Stat. 1027, generally prohibits any individual from “taking” a marine mammal, defined as harassing, hunting, capturing, or killing it. 16 U. S. C. §§ 1362(13), 1372(a). The Secretary of Defense may “exempt any action or category of actions” from the MMPA if such actions are “necessary for national defense.” § 1371(f)(1). In January 2007, the Deputy Secretary of Defense — acting for the Secretary — granted the Navy a 2-year exemption from the MMPA for the training exercises at issue in this case. Pet. App. 219a-220a. The exemption was conditioned on the Navy adopting several mitigation procedures, including: (1) training lookouts and officers to watch for marine mammals; (2) requiring at least five lookouts with binoculars on each vessel to watch for anomalies on the water surface (including marine mammals); (3) requiring aircraft and sonar operators to report detected marine mammals in the vicinity of the training exercises; (4) requiring reduction of active sonar transmission levels by 6 dB if a marine mammal is detected within 1,000 yards of the bow of the vessel, or by 10 dB if detected within 500 yards; (5) requiring complete shutdown of active sonar transmission if a marine mammal is detected within 200 yards of the vessel; (6) requiring active sonar to be operated at the “lowest practicable level”; and (7) adopting coordination and reporting procedures. Id., at 222a-230a.

The National Environmental Policy Act of 1969 (NEPA), 83 Stat. 852, requires federal agencies “to the fullest extent *16possible” to prepare an environmental impact statement (EIS) for “every . . . major Federal actio[n] significantly affecting the quality of the human environment.” 42 U. S. C. §4332(2)(C) (2000 ed.). An agency is not required to prepare a full EIS if it determines — based on a shorter environmental assessment (EA) — that the proposed action will not have a significant impact on the environment. 40 CFR §§ 1508.9(a), 1508.13 (2007).

In February 2007, the Navy issued an EA concluding that the 14 SOCAL training exercises scheduled through January 2009 would not have a significant impact on the environment. App. 226-227. The EA divided potential injury to marine mammals into two categories: Level A harassment, defined as the potential destruction or loss of biological tissue (1 e., physical injury), and Level B harassment, defined as temporary injury or disruption of behavioral patterns such as migration, feeding, surfacing, and breeding. Id., at 160-161.

The Navy’s computer models predicted that the SOCAL training exercises would cause only eight Level A harassments of common dolphins each year, and that even these injuries could be avoided through the Navy’s voluntary mitigation measures, given that dolphins travel in large pods easily located by Navy lookouts. Id., at 176-177, 183. The EA also predicted 274 Level B harassments of beaked whales per year, none of which would result in permanent injury. Id., at 185-186. Beaked whales spend little time at the surface, so the precise effect of active sonar on these mammals is unclear. Erring on the side of caution, the Navy classified all projected harassments of beaked whales as Level A. Id., at 186, 223. In light of its conclusion that the SOCAL training exercises would not have a significant impact on the environment, the Navy determined that it was unnecessary to prepare a full EIS. See 40 CFR § 1508.13.

Shortly after the Navy released its EA, the plaintiffs sued the Navy, seeking declaratory and injunctive relief on the grounds that the Navy’s SOCAL training exercises violated *17NEPA, the Endangered Species Act of 1973 (ESA), and the Coastal Zone Management Act of 1972 (CZMA).2 The District Court granted plaintiffs’ motion for a preliminary injunction and prohibited the Navy from using MFA sonar during its remaining training exercises. The court held that plaintiffs had “demonstrated a probability of success” on their claims under NEPA and the CZMA. Pet. App. 207a, 215a. The court also determined that equitable relief was appropriate because, under Ninth Circuit precedent, plaintiffs had established at least a “ ‘possibility’ ” of irreparable harm to the environment. Id., at 217a. Based on scientific studies, declarations from experts, and other evidence in the record, the District Court concluded that there was in fact a “near certainty” of irreparable injury to the environment, and that this injury outweighed any possible harm to the Navy. Id., at 217a-218a.

The Navy filed an emergency appeal, and the Ninth Circuit stayed the injunction pending appeal. 502 F. 3d 859, 865 (2007). After hearing oral argument, the Court of Appeals agreed with the District Court that preliminary injunctive relief was appropriate. The appellate court concluded, however, that a blanket injunction prohibiting the Navy from using MFA sonar in SOCAL was overbroad, and remanded the case to the District Court “to narrow its injunction so as to provide mitigation conditions under which the Navy may conduct its training exercises.” 508 F. 3d 885, 887 (2007).

On remand, the District Court entered a new preliminary injunction allowing the Navy to use MFA sonar only as long as it implemented the following mitigation measures (in addition to the measures the Navy had adopted pursuant to its MMPA exemption): (1) imposing a 12 nautical mile “exclusion *18zone” from the coastline; (2) using lookouts to conduct additional monitoring for marine mammals; (3) restricting the use of “helicopter-dipping” sonar; (4) limiting the use of MFA sonar in geographic “choke points”; (5) shutting down MFA sonar when a marine mammal is spotted within 2,200 yards of a vessel; and (6) powering down MFA sonar by 6 dB during significant surface ducting conditions, in which sound travels further than it otherwise would due to temperature differences in adjacent layers of water. 530 F. Supp. 2d 1110, 1118-1121 (CD Cal. 2008). The Navy filed a notice of appeal, challenging only the last two restrictions.

The Navy then sought relief from the Executive Branch. The President, pursuant to 16 U. S. C. § 1456(c)(1)(B), granted the Navy an exemption from the CZMA. Section 1456(c)(1)(B) permits such exemptions if the activity in question is “in the paramount interest of the United States.” The President determined that continuation of the exercises as limited by the Navy was “essential to national security.” Pet. App. 232a. He concluded that compliance with the District Court’s injunction would “undermine the Navy’s ability to conduct realistic training exercises that are necessary to ensure the combat effectiveness of... strike groups.” Ibid.

Simultaneously, the Council on Environmental Quality (CEQ) authorized the Navy to implement “alternative arrangements” to NEPA compliance in light of “emergency circumstances.” See 40 CFR § 1506.11.3 The CEQ determined that alternative arrangements were appropriate because the District Court’s injunction “create[s] a significant and unreasonable risk that Strike Groups will not be *19able to train and be certified as fully mission capable.” Pet. App. 238a. Under the alternative arrangements, the Navy would be permitted to conduct its training exercises under the mitigation procedures adopted in conjunction with the exemption from the MMPA. The CEQ also imposed additional notice, research, and reporting requirements.

In light of these actions, the Navy then moved to vacate the District Court’s injunction with respect to the 2,200-yard shutdown zone and the restrictions on training in surface ducting conditions. The District Court refused to do so, 527 F. Supp. 2d 1216 (2008), and the Court of Appeals affirmed. The Ninth Circuit held that there was a serious question regarding whether the CEQ’s interpretation of the “emergency circumstances” regulation was lawful. Specifically, the court questioned whether there was a true “emergency” in this case, given that the Navy has been on notice of its obligation to comply with NEPA from the moment it first planned the SOCAL training exercises. 518 F. 3d, at 681. The Court of Appeals concluded that the preliminary injunction was entirely predictable in light of the parties’ litigation history. Ibid. The court also held that plaintiffs had established a likelihood of success on their claim that the Navy was required to prepare a full EIS for the SOCAL training exercises. Id., at 693. The Ninth Circuit agreed with the District Court’s holding that the Navy’s EA — which resulted in a finding of no significant environmental impact — was “cursory, unsupported by cited evidence, or unconvincing.” Ibid.4

The Court of Appeals further determined that plaintiffs had carried their burden of establishing a “possibility” of irreparable injury. Even under the Navy’s own figures, the court concluded, the training exercises would cause 564 physical injuries to marine mammals, as well as 170,000 disturb*20anees of marine mammals’ behavior. Id., at 696. Lastly, the Court of Appeals held that the balance of hardships and consideration of the public interest weighed in favor of the plaintiffs. The court emphasized that the negative impact on the Navy’s training exercises was “speculative,” since the Navy has never before operated under the procedures required by the District Court. Id., at 698-699. In particular, the court determined that: (1) The 2,200-yard shutdown zone imposed by the District Court was unlikely to affect the Navy’s operations, because the Navy often shuts down its MFA sonar systems during the course of training exercises; and (2) the power-down requirement during significant surface ducting conditions was not unreasonable because such conditions are rare, and the Navy has previously certified strike groups that had not trained under such conditions. Id., at 699-702. The Ninth Circuit concluded that the District Court’s preliminary injunction struck a proper balance between the competing interests at stake.

We granted certiorari, 554 U. S. 916 (2008), and now reverse and vacate the injunction.

Ill

A

A plaintiff seeking a preliminary injunction must establish that he is likely to succeed on the merits, that he is likely to suffer irreparable harm in the absence of preliminary relief, that the balance of equities tips in his favor, and that an injunction is in the public interest. See Munaf v. Geren, 553 U. S. 674, 689-690 (2008); Amoco Production Co. v. Gambell, 480 U. S. 531, 542 (1987); Weinberger v. Romero-Barcelo, 456 U. S. 305, 311-312 (1982).

The District Court and the Ninth Circuit concluded that plaintiffs have shown a likelihood of success on the merits of their NEPA claim. The Navy strongly disputes this determination, arguing that plaintiffs’ likelihood of success is low because the CEQ reasonably concluded that “emergency *21circumstances” justified alternative arrangements to NEPA compliance. 40 CFR § 1506.11. Plaintiffs’briefs before this Court barely discuss the ground relied upon by the lower courts — that the plain meaning of “emergency circumstances” does not encompass a court order that was “entirely predictable” in light of the parties’ litigation history. 518 F. 3d, at 681. Instead, plaintiffs contend that the CEQ’s actions violated the separation of powers by readjudicating a factual issue already decided by an Article III court. Moreover, they assert that the CEQ’s interpretations of NEPA are not entitled to deference because the CEQ has not been given statutory authority to conduct adjudications.

The District Court and the Ninth Circuit also held that when a plaintiff demonstrates a strong likelihood of prevailing on the merits, a preliminary injunction may be entered based only on a. “possibility” of irreparable harm. Id., at 696-697; 530 F. Supp. 2d, at 1118 (quoting Faith Center Church Evangelistic Ministries v. Glover, 480 F. 3d 891, 906 (CA9 2007); Earth Island Inst. v. United States Forest Serv., 442 F. 3d 1147,1159 (CA9 2006)). The lower courts held that plaintiffs had met this standard because the scientific studies, declarations, and other evidence in the record established to “a near certainty” that the Navy’s training exercises would cause irreparable harm to the environment. 530 F. Supp. 2d, at 1118.

The Navy challenges these holdings, arguing that plaintiffs must demonstrate a likelihood of irreparable injury— not just a possibility — in order to obtain preliminary relief. On the facts of this case, the Navy contends that plaintiffs’ alleged injuries are too speculative to give rise to irreparable injury, given that ever since the Navy’s training program began 40 years ago, there has been no documented case of sonar-related injury to marine mammals in SOCAL. And even if MFA sonar does cause a limited number of injuries to individual marine mammals, the Navy asserts that plaintiffs have failed to offer evidence of species-level harm that *22would adversely affect their scientific, recreational, and ecological interests. For their part, plaintiffs assert that they would prevail under any formulation of the irreparable injury standard, because the District Court found that they had established a “near certainty” of irreparable harm.

We agree with the Navy that the Ninth Circuit’s “possibility” standard is too lenient. Our frequently reiterated standard requires plaintiffs seeking preliminary relief to demonstrate that irreparable injury is likely in the absence of an injunction. Los Angeles v. Lyons, 461 U. S. 95, 103 (1983); Granny Goose Foods, Inc. v. Teamsters, 415 U. S. 423, 441 (1974); O’Shea v. Littleton, 414 U. S. 488, 502 (1974); see also 11A C. Wright, A. Miller, & M. Kane, Federal Practice and Procedure §2948.1, p. 139 (2d ed. 1995) (hereinafter Wright & Miller) (applicant must demonstrate that in the absence of a preliminary injunction, “the applicant is likely to suffer irreparable harm before a decision on the merits can be rendered”); id., at 154-155 (“[A] preliminary injunction will not be issued simply to prevent the possibility of some remote future injury”). Issuing a preliminary injunction based only on a possibility of irreparable harm is inconsistent with our characterization of injunctive relief as an extraordinary remedy that may only be awarded upon a clear showing that the plaintiff is entitled to such relief. Mazurek v. Armstrong, 520 U. S. 968, 972 (1997) (per curiam).

It is not clear that articulating the incorrect standard affected the Ninth Circuit’s analysis of irreparable harm. Although the court referred to the “possibility” standard, and cited Circuit precedent along the same lines, it affirmed the District Court’s conclusion that plaintiffs had established a “ ‘near certainty’ ” of irreparable harm. 518 F. 3d, at 696-697. At the same time, however, the nature of the District Court’s conclusion is itself unclear. The District Court originally found irreparable harm from sonar-training exercises generally. But by the time of the District Court’s final decision, the Navy challenged only two of six restrictions *23imposed by the court. See supra, at 17-19. The District Court did not reconsider the likelihood of irreparable harm in light of the four restrictions not challenged by the Navy. This failure is significant in light of the District Court’s own statement that the 12 nautical mile exclusion zone from the coastline — one of the unchallenged mitigation restrictions— “would bar the use of MFA sonar in a significant portion of important marine mammal habitat.” 530 F. Supp. 2d, at 1119.

We also find it pertinent that this is not a case in which the defendant is conducting a new type of activity with completely unknown effects on the environment. When the Government conducts an activity, “NEPA itself does not mandate particular results.” Robertson v. Methow Valley Citizens Council, 490 U. S. 332, 350 (1989). Instead, NEPA imposes only procedural requirements to “ensur[e] that the agency, in reaching its decision, will have available, and will carefully consider, detailed information concerning significant environmental impacts.” Id., at 349. Part of the harm NEPA attempts to prevent in requiring an EIS is that, without one, there may be little if any information about prospective environmental harms and potential mitigating measures. Here, in contrast, the plaintiffs are seeking to enjoin — or substantially restrict — training exercises that have been taking place in SOCAL for the last 40 years. And the latest series of exercises were not approved until after the defendant took a “hard look at environmental consequences,” id., at 350 (quoting Kleppe v. Sierra Club, 427 U. S. 390, 410, n. 21 (1976); internal quotation marks omitted), as evidenced by the issuance of a detailed, 293-page EA.

As explained in the next section, even if plaintiffs have shown irreparable injury from the Navy’s training exercises, any such injury is outweighed by the public interest and the Navy’s interest in effective, realistic training of its sailors. A proper consideration of these factors alone requires denial of the requested injunctive relief. For the same reason, we *24do not address the lower courts’ holding that plaintiffs have also established a likelihood of success on the merits.

B

A preliminary injunction is an extraordinary remedy never awarded as of right. Munaf, 553 U. S., at 689-690. In each case, courts “must balance the competing claims of injury and must consider the effect on each party of the granting or withholding of the requested relief.” Amoco Production Co., 480 U. S., at 542. “In exercising their sound discretion, courts of equity should pay particular regard for the public consequences in employing the extraordinary remedy of injunction.” Romero-Barcelo, 456 U. S., at 312; see also Railroad Comm’n of Tex. v. Pullman Co., 312 U. S. 496, 500 (1941). In this case, the District Court and the Ninth Circuit significantly understated the burden the preliminary injunction would impose on the Navy’s ability to conduct realistic training exercises, and the injunction’s consequent adverse impact on the public interest in national defense.

This case involves “complex, subtle, and professional decisions as to the composition, training, equipping, and control of a military force,” which are “essentially professional military judgments.” Gilligan v. Morgan, 413 U. S. 1, 10 (1973). We “give great deference to the professional judgment of military authorities concerning the relative importance of a particular military interest.” Goldman v. Weinberger, 475 U. S. 503, 507 (1986). As the Court emphasized just last Term, “neither the Members of this Court nor most federal judges begin the day with briefings that may describe new and serious threats to our Nation and its people.” Boumediene v. Bush, 553 U. S. 723, 797 (2008).

Here, the record contains declarations from some of the Navy’s most senior officers, all of whom underscored the threat posed by enemy submarines and the need for extensive sonar training to counter this threat. Admiral Gary *25Roughead — the Chief of Naval Operations — stated that during training exercises:

“It is important to stress the ship crews in all dimensions of warfare simultaneously. If one of these training elements were impacted — for example, if effective sonar training were not possible — the training value of the other elements would also be degraded . . . .” Pet. App. 342a.

Captain Martin May — the Third Fleet’s Assistant Chief of Staff for Training and Readiness — emphasized that the use of MFA sonar is “mission-critical.” App. 570-571. He described the ability to operate MFA sonar as a “highly perishable skill” that must be repeatedly practiced under realistic conditions. Id., at 577. During training exercises, MFA sonar operators learn how to avoid sound-reducing “clutter” from ocean floor topography and environmental conditions; they also learn how to avoid interference and how to coordinate their efforts with other sonar operators in the strike group. Id., at 574. Several Navy officers emphasized that realistic training cannot be accomplished under the two challenged restrictions imposed by the District Court — the 2,200-yard shutdown zone and the requirement that the Navy power down its sonar systems during significant surface ducting conditions. See, e. g., Pet. App. 333a (powering down in presence of surface ducting “unreasonably prevent[s] realistic training”); id., at 356a (shutdown zone would “result in a significant, adverse impact to realistic training”). We accept these officers’ assertions that the use of MFA sonar under realistic conditions during training exercises is of the utmost importance to the Navy and the Nation.

These interests must be weighed against the possible harm to the ecological, scientific, and recreational interests that are legitimately before this Court. Plaintiffs have submitted declarations asserting that they take whale watching trips, observe marine mammals underwater, conduct scien*26tifie research on marine mammals, and photograph these animals in their natural habitats. Plaintiffs contend that the Navy’s use of MFA sonar will injure marine mammals or alter their behavioral patterns, impairing plaintiffs’ ability to study and observe the animals.

While we do not question the seriousness of these interests, we conclude that the balance of equities and consideration of the overall public interest in this case tip strongly in favor of the Navy. For the plaintiffs, the most serious possible injury would be harm to an unknown number of the marine mammals that they study and observe. In contrast, forcing the Navy to deploy an inadequately trained antisubmarine force jeopardizes the safety of the fleet. Active sonar is the only reliable technology for detecting and tracking enemy diesel-electric submarines, and the President— the Commander in Chief — has determined that training with active sonar is “essential to national security.” Id., at 232a.

The public interest in conducting training exercises with active sonar under realistic conditions plainly outweighs the interests advanced by the plaintiffs. Of course, military interests do not always trump other considerations, and we have not held that they do. In this case, however, the proper determination of where the public interest lies does not strike us as a close question.

C

Despite the importance of assessing the balance of equities and the public interest in determining whether to grant a preliminary injunction, the District Court addressed these considerations in only a cursory fashion. The court’s entire discussion of these factors consisted of one (albeit lengthy) sentence: “The Court is also satisfied that the balance of hardships tips in favor of granting an injunction, as the harm to the environment, Plaintiffs, and public interest outweighs the harm that Defendants would incur if prevented from using MFA sonar, absent the use of effective mitigation *27measures, during a subset of their regular activities in one part of one state for a limited period.” Id., at 217a-218a. As the prior Ninth Circuit panel in this case put it, in staying the District Court’s original preliminary injunction, “[t]he district court did not give serious consideration to the public interest factor.” 502 F. 3d, at 863. The District Court’s order on remand did nothing to cure this defect, but simply repeated nearly verbatim the same sentence from its previous order. Compare 530 F. Supp. 2d, at 1118, with Pet. App. 217a-218a. The subsequent Ninth Circuit panel framed its opinion as reviewing the District Court’s exercise of discretion, 518 F. 3d, at 697-699, but that discretion was barely exercised here.

The Court of Appeals held that the balance of equities and the public interest favored the plaintiffs, largely based on its view that the preliminary injunction would not in fact impose a significant burden on. the Navy’s ability to conduct its training exercises and certify its strike groups. Id., at 698-699. The court deemed the Navy’s concerns about the preliminary injunction “speculative” because the Navy had not operated under similar procedures before. Ibid. But this is almost always the case when a plaintiff seeks injunctive relief to alter a defendant’s conduct. The lower courts failed properly to defer to .senior Navy officers’ specific, predictive judgments about how the preliminary injunction would reduce the effectiveness of the Navy’s SOCAL training exercises. See Wright & Miller §2948.2, at 167-168 (“The policy against the imposition of judicial restraints prior to an adjudication of the merits becomes more significant when there is reason to believe that the decree will be burdensome”).

The preliminary injunction requires the Navy to shut down its MFA sonar if a marine mammal is detected within 2,200 yards of a sonar-emitting vessel. The Ninth Circuit stated that the 2,200-yard shutdown zone would not be overly burdensome because sightings of marine mammals *28during training exercises are relatively rare. But regardless of the frequency of marine mammal sightings, the injunction will greatly increase the size of the shutdown zone. Pursuant to its exemption from the MMPA, the Navy agreed to reduce the power of its MFA sonar at 1,000 yards and 500 yards, and to completely turn off the system at 200 yards. Pet. App. 222a-230a. The District Court’s injunction does not include a graduated power-down, instead requiring a total shutdown of MFA sonar if a marine mammal is detected within 2,200 yards of a sonar-emitting vessel. There is an exponential relationship between radius length and surface area (Area = % r2). Increasing the radius of the shutdown zone from 200 to 2,200 yards would accordingly expand the surface area of the shutdown zone by a factor of over 100 (from 125,664 square yards to 15,205,308 square yards).

The lower courts did not give sufficient weight to the views of several top Navy officers, who emphasized that because training scenarios can take several days to develop, each additional shutdown can result in the loss of several days’ worth of training. Id., at 344a. Limiting the number of sonar shutdowns is particularly important during the Joint Tactical Force Exercises, which usually last for less than two weeks. Ibid. Rear Admiral John Bird explained that the 2,200-yard shutdown zone would cause operational commanders to “lose awareness of the tactical situation through the constant stopping and starting of MFA [sonar].” Id., at 332a; see also id., at 356a (“It may take days to get to the pivotal attack in antisubmarine warfare, but only minutes to confound the results upon which certification is based”). Even if there is a low likelihood of a marine mammal sighting, the preliminary injunction would clearly increase the number of disruptive sonar shutdowns the Navy is forced to perform during its SOCAL training exercises.

The Court of Appeals also concluded that the 2,200-yard shutdown zone would not be overly burdensome because the Navy had shut down MFA sonar 27 times during its eight *29prior training exercises in SOCAL; in several of these cases, the Navy turned off its sonar when marine mammals were spotted well beyond the Navy’s self-imposed 200-yard shutdown zone. 518 F. 3d, at 700, n. 65. Vice Admiral Samuel Locklear III — the Commander of the Navy’s Third Fleet— stated that any shutdowns beyond the 200-yard zone were voluntary avoidance measures that likely took place at tactically insignificant times; the Ninth Circuit discounted this explanation as not supported by the record. Ibid. In reaching this conclusion, the Court of Appeals ignored key portions of Vice Admiral Locklear’s declaration, in which he stated unequivocally that commanding officers “would not shut down sonar until legally required to do so if in contact with a submarine.” Pet. App. 354a-355a. Similarly, if a commanding officer is in contact with a target submarine, “the CO will be expected to continue to use active sonar unless another ship or helicopter can gain contact or if regulatory reasons dictate otherwise.” Id., at 355a. The record supports the Navy’s contention that its shutdowns of MFA sonar during prior training exercises only occurred during tactically insignificant times; those voluntary shutdowns do not justify the District Court’s imposition of a mandatory 2,200-yard shutdown zone.

Lastly, the Ninth Circuit stated that a 2,200-yard shutdown zone was feasible because the Navy had previously adopted a 2,000-meter zone for low-frequency active (LFA) sonar. The Court of Appeals failed to give sufficient weight to the fact that LFA sonar is used for long-range detection of enemy submarines, and thus its use and shutdown involve tactical considerations quite different from those associated with MFA sonar. See App. 508 (noting that equating MFA sonar with LFA sonar “is completely misleading and is like comparing 20 degrees Fahrenheit to 20 degrees Celsius”).

The Court of Appeals also concluded that the Navy’s training exercises would not be significantly affected by the requirement that it power down MFA sonar by 6 dB during *30significant surface ducting conditions. Again, we think the Ninth Circuit understated the burden this requirement would impose on the Navy’s ability to conduct realistic training exercises.

Surface ducting is a phenomenon in which relatively little sound energy penetrates beyond a narrow layer near the surface of the water. When surface ducting occurs, active sonar becomes more useful near the surface but less useful at greater depths. Pet. App. 299a-300a. Diesel-electric submariners are trained to take advantage of these distortions to avoid being detected by sonar. Id., at 333a.

The Ninth Circuit determined that the power-down requirement during surface ducting conditions was unlikely to affect certification of the Navy’s strike groups because surface ducting occurs relatively rarely, and the Navy has previously certified strike groups that did not train under such conditions. 518 F. 3d, at 701-702. This reasoning is backwards. Given that surface ducting is both rare and unpredictable, it is especially important for the Navy to be able to train under these conditions when they occur. Rear Admiral Bird explained that the 6 dB power-down requirement makes the training less valuable because it “exposes [sonar operators] to unrealistically lower levels of mutual interference caused by multiple sonar systems operati