OPINION
RENDELL, Circuit Judge,
The Zoning Board of Adjustment of Par-amus, New Jersey (“ZBA”) appeals from the District Court’s grant of summary judgment to Appellees Sprint, T-Mobile and Omnipoint Communications in their challenge to the ZBA’s denial of a requested variance to a city zoning ordinance that prohibited the construction of monopoles to fill gaps in wireless service. Because the District Court correctly found that the ZBA’s denial violated the Teleeommunica-
tions Act of 1996 (“TCA”) and the New Jersey Municipal Land Use Law (“MLUL”), we will affirm.
Factual Background
Appellees filed a wireless tower siting application to fill a gap in their service in the Borough of Paramus, New Jersey. They proposed construction of a faux-tree “monopole” in one of two sites after investigating possible locations for its placement. Paramus had an ordinance which, among other things, prohibited the construction of cellular monopoles, defined as “[a]n antenna structure consisting of a single pole in commercial and residential zones.”
Sprint Spectrum v. Zoning Bd. of Adjustment of the Borough of Paramus,
21 F.Supp.3d 381, 383 (D.N.J.2014) (internal quotation marks, omitted). The ordinance also stated, however, that: “[t]he purpose of these regulations ... is to: ... [e]ncourage users of monopoles and antennas to locate them,
to the extent possible,
in areas where the adverse impact on the residential community is minimal, particularly to avoid adverse visual impacts upon residential dwellings.” (App. 1305) (emphasis added).
The TCA and MLUL
The TCA expressly preserves the authority of state and local governments to regulate land use and zoning, but places several substantive and procedural limits upon that authority when exercised in relation to personal wireless service facilities.
APT Pittsburgh Ltd. v. Penn Twp. Butler Cnty. of Pennsylvania,
196 F.3d 469, 473 (3d Cir.1999). One such substantive requirement is section 332(c)(7)(B) of the TCA, which states:
(i) The regulation of the placement, construction, and modification of personal wireless service facilities by any State or local government or instrumentality thereof — ...
(II) shall not prohibit or have the effect of prohibiting the provision of personal wireless services.
47 U.S.C. § 332(c)(7)(B). A state or local government has effectively prohibited the provision of wireless services where a carrier has demonstrated that (1) its facility will fill a significant gap in service, and (2) the manner in which it proposes to fill the significant gap in service is the least intrusive on the values that the denial sought to serve.
APT Pittsburgh Ltd.,
196 F.3d at 480. That is, “[a] local government may reject an application for construction of a wireless service facility in an under-served area without thereby prohibiting personal wireless services if the service gap can be closed by less intrusive means.”
Sprint Spectrum L.P. v. Willoth,
176 F.3d 630, 643 (2d Cir.1999). This requires
a
showing that a good faith effort has been made to evaluate less intrusive alternatives, which includes considerations of alternative sites, alternative tower designs, placement of antenna on existing structures, and “alternative system designs.”
APT Pittsburgh Ltd.,
196 F.3d at 480. The statutory bar against regulatory prohibition of wireless services is absolute, and does not anticipate deference to local findings.
Cellular Tel. Co. v. Zoning Bd. of Adjustment of the Borough of Ho-Ho-Kus,
197 F.3d 64, 71 (3d Cir.1999). Whether a state or local government’s regulation violates the effective prohibition provision of the TCA is reviewed de novo, and is not limited to the record compiled by the state or local authority.
APT Pittsburgh Ltd.,
196 F.3d at 475.
The TCA also places a second, procedural requirement on states and local governments in that same section:
(in) Any decision by a State or local government or instrumentality thereof
to deny a request to place, construct, or modify personal wireless service facilities shall be in writing and supported by substantial evidence contained in a written record.
47 U.S.C. § 332(c)(7)(B)(iii). This procedural protection applies to determinations of factual' issues made by state or local authorities when they apply state and local zoning law.
APT Pittsburgh Ltd.,
196 F.3d at 474. This subsection applies to decisions made solely on the basis of the factual record before the agency and are the subject of deferential substantial evidence review.
Id.
Substantial evidence means “such evidence as a reasonable mind might accept as adequate to support a conclusion.”
Ho-Ho-Kus,
197 F.3d at 71. If the record contains conflicting evidence, the fact-finder must adequately explain its reasons for rejecting or discrediting competent evidence.
See Benton v. Bowen,
820 F.2d 85, 88 (3d Cir.1987). The reviewing court’s task is to determine whether the decision, as guided by local law, is supported by substantial evidence.
Ho-Ho-Kus,
197 F.3d at 72 (citing
Omnipoint Corp. v. Zoning Hearing Bd. of Pine Grove Twp.,
181 F.3d 403, 408 (3d Cir.1999)). Therefore, “[t]he TCA itself does not provide the legal basis to deny an application to construct a personal wireless facility. That authority must be found in state or local law.”
Willoth,
176 F.3d at 644.
Under New Jersey law, zoning boards of adjustment may grant variances to local ordinances.
See
N.J.S.A. § 40:55D-70. A decision of a zoning board may be set aside only when it is “arbitrary, capricious or unreasonable,” but “[s]o long as the power exists to do the act complained of and there is substantial evidence to support it, the judicial branch of the government cannot interfere.”
Medici v. BPR Co.,
107 N.J. 1, 526 A.2d 109, 116 (1987). The New Jersey MLUL requires that “[l]ocal zoning officials ... weigh the positive and negative factors associated with a requested zoning variance and determine whether, on balance, those factors weigh in favor of granting or rejecting the request.”
Ho-Ho Kus,
197 F.3d at 72. “[T]he reviewing court’s task is to determine whether the findings of local officials concerning the positive, and negative factors, and their relative weight, is supported by substantial evidence.”
Id.
Thus, both federal and New Jersey law employ the substantial evidence standard.
Procedural History
Appellees applied to the ZBA for a variance to the ordinance in order to build the proposed monopole. The ZBA held seventeen public hearings on the potential impact of the monopole, hearing the views of experts on both sides, as well as possible alternatives. Rhoan Gordon, a radio frequency engineer, testified that the only alternative to a monopole would involve the utilization of short structures to satisfy the coverage gap, which would require at least a dozen such structures, and would not be a feasible design as it would require affixing “a huge box” on each such structure.
Sprint,
21 F.Supp.3d at 384. A member of the public then inquired as to whether a Distributed Antenna System (“DAS”) had been considered. DAS is a means of providing wireless coverage without the use of tall monopoles. The ZBA’s retained expert, Ross Sorci, a radio frequency engineer with twenty-nine years of relevant experience analyzing coverage and propagation issues, testified that a DAS would not be feasible, for a few reasons. While a monopole concentrates all wireless antennas onto a single tower, a DAS “distributes” those antennas across utility poles and other existing structures throughout the coverage area.
Id.
Mr. Sorci testified that a DAS system would
not be feasible because it would require the use of multiple structures and is more of a “spot solution” for small coverage gaps, and would not be suitable 'for covering a large area like the one in Paramus.
Id.
Appellees’ expert Glenn Pierson, a radio frequency engineer with twenty years of network design experience and who had designed a DAS network, testified that a DAS was not a viable alternative for the coverage gap in Paramus because of concerns regarding its reliability. He stated that because DAS nodes were interconnected, the whole system could suffer an outage if there was one fallen tree, ice storm, or car that hit a pole on which a DAS node was installed. He also testified that he was not aware of any DAS nodes with battery backup. He raised numerous other problems with a DAS — namely, that because a DAS system allows only for the triangulation of a user’s location to a thousand foot radius, first responders to an emergency would be looking for the emergency caller’s location for a longer time, that DAS systems posed worker safety concerns, including the potential for electromagnetic radiation exposure, and that DAS systems are more susceptible to dropped calls. Finally, Mr. Pierson testified that there would be various technical problems with locating multiple carriers on a single DAS antenna, and would require three or four boxes on each to satisfy the coverage gap.
The city, unhappy with the testimony of these two experts, hired David Maxson, a self-described “municipal wireless consultant” with no formal training regarding the placement or construction of wireless facilities, to testify at the hearings.
Id.
Mr. Maxson admitted that he is not a licensed professional engineer in New Jersey, has no educational experience in land use planning, is not a certified urban planner, is not a certified municipal engineer in New Jersey, and has taken just one non-graded course via CD-ROM on cellular CDMA technology. The ZBA permitted him to testify over the Appellees’ objections. Mr. Maxson stated that he had reviewed the land in question by looking on Google, and websites containing photographs of the buildings in the area, and that he had toured the area, but did not do a drive test or review the drive test data performed by the Appellees in the coverage gap. He concluded that, contrary to what the other experts stated, a DAS would in fact be feasible because the streets of Paramus were “well populated with utility poles,” a DAS system could be installed to cover the wireless gap.
Id.
at 385. However, Mr. Maxson agreed with Mr. Gordon that a DAS system would require numerous antennas to be installed up and down the road to provide the same service as a monopole. Additionally, in response to a question about how wireless service carriers would deal with an extended outage, Mr. Maxson testified that the use of a small generator, set on the street and chained to a utility pole for the duration of the power outage, would solve the problem.
After these hearings, the ZBA denied the variance because “the substantial height of the proposed monopole and its placement -within the residential neighborhood would have a detrimental visual effect on the surrounding properties,” because Appellees “failed to investigate other less intrusive ways of providing coverage,” and because Appellees “did not put forth a good faith effort to explore and investigate alternative technology to provide coverage.”
Id.
at 387-88. The ZBA also found that Appellees “failed to meet their burden of showing that the benefits of the proposed improvements would substantially outweigh any possible detriment.”
Id.
at 388.
Appellees appealed the ZBA’s denial to the District Court, alleging that it violated the TCA and the MLUL. There were three different proceedings in the District Court, as the case was transferred to different judges (for reasons not relevant on appeal). As to the TCA, Judge Linares found that a significant gap in wireless coverage existed within the area presented, that the monopole proposed would adequately fill the gap, and that the Appellees had adequately considered alternative sites before arriving at the ones proposed. He denied summary judgment as to the TCA claim, however, because he concluded that genuine issues of material fact existed regarding the feasibility of a DAS system as a less intrusive alternative to the monopole. With respect to the Appellees’ claim that the ZBA’s denial was not based on “substantial evidence,” Judge Linares found that- the ZBA’s decision provided only generalized concerns regarding aesthetics, and that its concerns around the monopole’s threat of collapse and of ice accumulation were not realistic. He denied summary judgment to Appellees on their substantial evidence claim as well, because he concluded that factual issues remained regarding the feasibility of the DAS as an alternative. He noted that if a DAS is found not to be a feasible alternative to the monopole, the ZBA’s denial based on the Appellees’ alleged failure to investigate alternative, less intrusive means of providing coverage would not be supported by “substantial evidence.”
Judge Salas, following limited discovery into the technical aspects of an .outdoor DAS, adhered to Judge Linares’s earlier ruling that (a) a significant gap in wireless service exists in the area in question, (b) Appellees’ proposed facilities would fill the coverage gap, and (c) Appellees adequately considered alternative sites for the monopole. Like Judge Linares, Judge Salas denied summary judgment because genuine issues of fact remained regarding the feasibility of a DAS system. As for substantial evidence, she denied summary judgment as to Appellees, in order not to disturb Judge Linares’s ruling.
Finally, Judge McNulty held a bench trial limited to the issue of whether, for filling the gap in service, a DAS would be a feasible, less intrusive alternative to the proposed monopole. He heard testimony from three experts. The Appellees’ expert, radio frequency . engineer Richard Conroy, testified that to fill the coverage gap in Paramus, a dense multi-node network DAS would be required. Because a significant portion of the coverage gap area in Paramus consists of residential streets with dense trees, the utility poles on which the many DAS nodes would be installed would be at risk of damage from falling trees, and power outages due to storms. A monopole, on the other hand, would be relatively resistant to weather and other damage. While a monopole would permit signal rerouting in response to a fallen utility pole, he stated, DAS systems are typically not designed to permit such rerouting, While both systems would be affected by a power failure, every DAS node would need its own battery backup, whereas the monopole would only require one.
The ZBA’s proffered expert, Bruce Eisenstein, who conceded that he did not have practical experience designing a DAS or other cellular network, nevertheless concluded that a DAS would be feasible. While he conceded that the Appellees had conducted a good faith analysis of whether or not a DAS would be a feasible alternative in Paramus, he stated that qualified engineers could design a DAS system in such a way that a falling tree on one pole would not cause an outage of the entire system. When asked if he knew of any such DAS system currently in operation,
hbwever, he admitted that he did not. He also conceded that a DAS system would involve rapid “handoffs” that might degrade the reliability of service. He noted that a DAS system had successfully been set up in the Township of Harding, New Jersey, and the Township of Lower Mer-ion, Pennsylvania, but admitted that his opinion as to its feasibility in Paramus was conceptual, and that his analysis was not intended to be a design for a wireless network.
Finally, professional city planner David Karlebach testified that a monopole would be less intrusive and have less visual impact than placing DAS equipment on multiple utility poles. He described the design for a faux-tree monopole, and testified that while no one would mistake the pole for an actual tree, it would have the virtue of being confined in a single location. A DAS, on the other hand, would not rise high above the ground, but would create its visual impact at multiple locations throughout the service gap area.
Judge McNulty concluded that, considering all of the evidence, the DAS was not a feasible alternative to the monopole proposed because it would not provide “the same level of robust, reliable coverage as a monopole under the prevailing conditions in Paramus,” and because it had significant reliability concerns.
Id.
at 395. He was not persuaded by Dr. Eisenstein’s rebuttal of Mr. Conroy’s testimony regarding the feasibility of a DAS. Taking Judge Linares and Judge Salas’s findings that (a) there existed a significant gap in wireless service in Paramus, (b) that the proposed monopole would fill that gap, and (c) that Appel-lees had considered alternative sites for the monopole, he added that Appellees had made a good faith effort to identify and evaluate less intrusive alternative system designs to the monopole, and that the DAS was not such a less intrusive, feasible alternative. Therefore, he found that the ZBA’s denial constituted an “effective prohibition” of wireless service, in violation of § 332(e)(7)(B)(i)(II) of the TCA.
As to the issue of whether the ZBA’s denial was based on substantial evidence, Judge McNulty noted that previous rulings in the case had narrowed the inquiry to the question of whether the DAS was a feasible alternative to the monopole; - that is, whether or not the ZBA’s decision was supported by substantial evidence depended on the answer to this question. Based on the record before the ZBA, Judge McNulty concluded that the DAS was not a feasible alternative to the monopole, and therefore, denying coverage due to the possibility of implementing a DAS in place of a monopole was not a decision supported by “substantial evidence.” Additionally, adhering to what Judge Linares had concluded, Judge McNulty found that since the monopole would be very tall, but confined to one small area, whereas the DAS would be shorter but dispersed in multiple locations, from a visual standpoint, there was “no clear aesthetic winner.”
Id.
at 396. Therefore, denial of the variance application on the basis of aesthetics was also not supported by “substantial evidence.”
Standard of Review
On appeal from a bench trial, we review a district court’s findings of fact for clear error,
Scully v. U.S. WATS Inc.,
238 F.3d 497, 505-06 (3d Cir.2001), and its conclusions of law de novo,
U.S. v. Coggins,
986 F.2d 651, 654 (3d Cir.1993). Clear error is only present if the District Court’s findings are “completely devoid of minimum evidentiary support displaying some hue of credibility or bears no rational relationship to the supportive evidentiary data.”
Berg Chilling Sys. v. Hull Corp.,
369 F.3d 745, 754 (3d Cir.2004).
Discussion
1. ZBA’s decision was an “effective prohibition” on wireless service.
The District Court correctly concluded that the ZBA’s denial of the variance requested was an “effective prohibition” under the TCA because a significant gap in wireless coverage existed within the area presented, the monopole proposed would adequately fill the gap, and Appel-lees had adequately considered alternative sites and systems before arriving at the ones proposed. On appeal, the ZBA does not contest the District Court’s finding that there is a significant gap in the area where Appellees seek to build a monopole, or that the monopole would fill that gap. The ZBA also does not contest that Appel-lees considered alternative locations for the monopole. The ZBA disputes the District Court’s finding on the feasibility of the DAS as an alternative system to the monopole, -and therefore argues that the monopole is not the least intrusive means to fill the significant gap in wireless service that exists in Paramus.
As a preliminary matter, the ZBA’s argument that the Appellees did not conduct a good faith investigation into the feasibility of DAS was contradicted by the ZBA’s own expert, Mr. Eisenstein. More importantly, however, the monopole is in fact the least intrusive means, because the DAS was found not to be a feasible system for the coverage area in question. During the ZBA hearings, radio frequency engineers Rhoan Gordon, Ross Sorci and Glenn Pier-son all agreed that DAS was not a viable alternative to a monopole because it raised serious reliability concerns. Only David Maxson, whose credentials were questionable, at best, believed that a DAS would be feasible in Paramus. During the' bench trial, radio frequency engineer Richard Conroy also testified that a DAS system raised serious reliability concerns. Professional city planner David Karlebach testified that a monopole would be less intrusive and have less visual impact than a DAS system. The ZBA’s proffered expert, Bruce Eisenstein, who conceded that he did not have practical experience designing a DAS or other cellular network, testified that it would be feasible and offered suggestions for how outages could be avoided. Taking these views into consideration, the District Court concluded that a DAS system is infeasible because it is more susceptible than a monopole to outages due to falling trees, less flexible and therefore less able to cover multiple carriers, and designed to cover a smaller gap than required. The opinions of the experts who testified at the ZBA hearings and the bench trial below weighed heavily against the feasibility of a DAS system in the coverage area, particularly when considering that the experts who had actual experience with DAS systems harbored that view. We review these factual findings for clear error, and the ZBA has not demonstrated before us that they were “completely devoid of minimum evidentia-ry support.”
Berg Chilling,
B69 F.3d at 754. Therefore, the District Court correctly concluded that the ZBA’s denial of the variance was an “effective prohibition” under the TCA.
The ZBA argues that the District Court improperly placed the burdens of proof and persuasion on the ZBA regarding the issue of the feasibility of the DAS, and therefore, Appellees succeeded without demonstrating that the monopole is the least intrusive means to fill the coverage gap. This is not what the District Court did. The feasibility of a DAS system was extensively reviewed, and rejected, as explained by the Appellees’ experts who testified at the ZBA hearings and during the bench trial regarding why a DAS would not work in the Paramus coverage gap.
In fact, as the District Court, correctly noted, “[Appellees] do not bear the burden of proving that every potential alternative, no matter how speculative, is unavailable. The proper inquiry for an effective prohibition claim is whether a
good faith effort
has been made to identify and evaluate less intrusive alternatives.... ”
Sprint,
21 F.Supp.3d at 397 (internal quotation marks omitted). A good faith effort was made to consider DAS, and it was rejected for valid reasons.
2. The ZBA’s decision was not supported by “substantial evidence.
”
The District Court correctly found that the ZBA’s decision to deny the variance was not based on “substantial, evidence.” The ZBA denied the variance application due to the negative criteria presented by the construction of the monopole: it believed that the Appellees had failed to investigate less intrusive ways of providing coverage — namely, a DAS system — and it believed that the monopole would have a detrimental visual effect on the surrounding properties in the neighborhood.
As to the feasibility of the DAS, Judge McNulty evaluated the testimony in the ZBA hearing record from Mr. Gordon and Mr. Sorci, both of whom agreed that the DAS was not a feasible solution for the coverage gap in Paramus. Specifically, he rejected the ZBA’s conclusion that this alternative had not been explored. Mr. Maxson’s opinion that a DAS would be feasible was greatly suspect, given his questionable credentials and lack of expertise to evaluate the feasibility of a DAS solution. It is clear to us that, the views expressed in the hearings alone demonstrated that a DAS was not a feasible alternative to the monopole.
The ZBA argues that the bench trial improperly influenced Judge McNulty’s conclusion that a DAS is not feasible. The ZBA argues that because a “substantial evidence” claim is reviewed for whether the evidence in the record before the ZBA was sufficient, and because Judges Linares and Judge Salas denied summary judgment to Appellees based on questions of fact that remained on the feasibility of the DAS, Judge McNulty’s conclusion that the ZBA’s decision was not based on “substantial evidence” was error. We do not agree that the bench trial improperly influenced the District Court’s conclusion on this matter. As Judge McNulty stated, “[t]he expert testimony in this Court has educated me as to the technical issues, and in that way has influenced my conclusions
as to the meaning of the evidence the Board heard,
as well as my conclusions as to what the presentation may have lacked.”
Sprint,
21 F.Supp.3d at 398 (emphasis added).
Additionally, the ZBA’s reliance on the visual impact of the monopole does not represent “substantial evidence” to support the denial of the variance because, as Judge McNulty noted, there was “no clear aesthetic winner” between a DAS and the proposed monopole.
Id.
at 396. The ZBA’s denial of the variance was likewise unsupported by the New Jersey MLUL, which-requires that such a denial be supported by substantial evidence.
The ZBA next argues that the substantial evidence and summary judgment standards are the same; that is, because Judges Linares and Judge Salas denied summary judgment to Appellees below on the issue of whether the DAS was a feasible alternative to the monopole, this necessitates finding that the ZBA’s decision to deny the variance based on the possibility of a DAS was based on “substantial evidence.” ’ However, this is simply not the case. While the substantial evidence standard requires weighing evidence on the merits of a claim, the summary judgment
standard does not. As to substantial evidence, we have stated that “if the record as a whole contains conflicting evidence, the fact-finder must adequately explain its reasons for rejecting or discrediting competent evidence.”
Ho-Ho-Kus,
197 F.3d at 71-72. Conflicting evidence on material facts makes summary judgment inappropriate.
Finally, as the ZBA never brought up their argument that a federal court injunction requiring it to grant a variance to Appellees would result in unconstitutional commandeering before the District Court, we consider it waived.
Conclusion
Because the ZBA’s denial of Appellees’ zoning variance violates the TCA’s “effective prohibition” language and was not based on “substantial evidence” as required by both the TCA and MLUL, we affirm the District Court.