Correction — September 16, 2026. An earlier version of this article described the 92.1 dB figure as "the loudest moment" of a delivery, attributed truck- and vacuum-cleaner loudness comparisons to the FAA's filing, cited the noise data to Appendix D, and described the draft's 1,000-flights-per-day figure as an authorized ceiling. Those statements were wrong. The 92.1 figure is a sound exposure level (SEL), which is not an instantaneous loudness measurement; the loudness comparisons do not appear in the filing; the noise data is in Appendix E, the Technical Noise Report; and the 1,000-flight volume is a proposal under FAA evaluation, not issued authority. The article below has been corrected throughout.

Tactical Summary

The FAA's draft environmental filing for Amazon's Baton Rouge delivery hub contains a number that travels fast and gets misread faster: a maximum A-weighted sound exposure level (SEL) of 92.1 for the delivery phase, in the 25-foot row of a table whose distance column is labeled, literally, "Distance to PADDC" — the filing's abbreviation for the Prime Air Drone Delivery Center. The table states its distances against that reference point; it does not report an aircraft-to-person range.[2]

The misreadings are the point of this article — an earlier version of it committed one. An SEL is not "how loud it got." The FAA defines SEL as all the acoustic energy of an individual noise event represented as if the event had occurred within one second — it captures both the magnitude and the duration of the event in a single number.[3] The maximum instantaneous level of an event is a separate metric, Lmax, which the Baton Rouge technical report evaluates separately alongside Leq and ambient levels.[2] A loudness talking point and a filed SEL table are answering different questions, and neither one settles the other.

That leaves a real story. The gap between public loudness talking points — anyone's — and the acoustic data in regulatory filings is where community acceptance for drone delivery will be argued, and the filings are public. Operators who can read them precisely — and quote them precisely — will have a much easier time in front of a city council than operators who paraphrase.

What the Filing Actually Reports

The noise data for the Baton Rouge Prime Air Drone Delivery Center lives in Appendix E of the draft environmental assessment, the Technical Noise Report.[2] (Appendix D, which an earlier version of this article cited, is the Section 106 resources and agency consultation appendix; it contains no noise tables.[4])

En route, the technical report gives a maximum-weight SEL of 63.7 dB with the aircraft at 200 feet.[2] At the delivery point, the profile changes: the draft EA describes the MK30 descending from en route altitude and dropping the package through payload doors from approximately 13 feet above the ground, with the whole delivery sequence — descent, drop, climb-out — taking roughly 61 seconds.[1] For that delivery event, Table 19 of the technical report lists the maximum A-weighted SEL as 92.1 in its 25-foot "Distance to PADDC" row, falling to 85.8 at 76 feet and 69.5 at 772 feet; the table's own footnote says the maximum SEL is derived from the report's underlying measurement table.[2] Takeoff and landing tables report maximum SELs of 90.5 and 91.8 at their 20-foot rows.[2]

Here is what you cannot honestly do with those numbers, and what the earlier version of this article did: treat 92.1 as the peak loudness of a hover, subtract the cruise figure from it to describe a "28-decibel spread" in perceived noise, or map either figure onto household appliances. SEL folds duration into the number — a longer event at the same level yields a higher SEL — so SEL values are not interchangeable with what a person standing in a yard hears at any instant. If you want the loudest-instant story, you need the Lmax data, and if you want the community-impact story, the analysis runs on DNL. The filing keeps these metrics distinct for a reason, and so should anyone quoting it.

What the filing does support saying: at the table's shortest listed distances, the delivery event deposits substantially more acoustic energy than an overhead pass at cruise altitude. That is unsurprising — the draft describes a hover-and-drop sequence lasting roughly a minute, flown far lower than a 200-foot transit. A cruise flyover and a delivery drop are different acoustic events measured on different rows of different tables, and the filed tables — not anyone's talking points — are what sit in a federal docket.

The Proposed Ceiling

The second thing the filing pins down is volume — as a proposal. The document's stated purpose is to evaluate the potential environmental impacts of the FAA's approval of the Proposed Action and the amendment of Prime Air's OpsSpec, and it says Prime Air "proposes to operate up to 1,000 MK30 drone delivery flights per operating day," up to 365 days a year, between 6 a.m. and 10:30 p.m., across a 7.5-mile radius and roughly 174 square miles.[1] The draft's noise analysis conservatively assumes up to 500 delivery overflights per day of a single location at that volume.[1]

Two cautions cut in opposite directions. The 1,000-flight figure is what is being requested and analyzed, not a volume the FAA has been shown to have granted — this draft is not a permit, and describing it as one overstates the record. At the same time, the draft's unfinished status does not establish that the site is flying without authority; what operating authority Baton Rouge holds sits in other instruments this filing does not display, and this article makes no claim about them either way. What operators should take from the document is the shape of the ask: complaint cycles in any market will eventually be argued against whatever volume ends up approved, not against today's tempo — which UAVHQ has not measured or verified for Baton Rouge. The same pattern is visible elsewhere on the FAA's environmental portal, where the Chicago Prime Air review proposes up to 1,000 flights per day from each hub. The environmental assessment is where the proposed operating envelope is written down, long before a neighborhood feels it.

The Paperwork Loose End

There is also a docket-status observation worth recording precisely — with the where and the when attached, because absence claims are only as good as their scope. When UAVHQ checked the FAA's drone-environmental portal (faa.gov/uas/advanced_operations/nepa_and_drones) on September 15, 2026, the Baton Rouge draft EA was listed with a comment period that closed July 22, 2026; the Omaha draft carried the same closing date; and no final EA or Finding of No Significant Impact for either site appeared in the portal's completed-reviews section. That is a statement about one public page on one date, nothing more. The portal's silence is not evidence about what authority the operation holds, and the absence of a final document there does not establish that none exists anywhere.

Meanwhile, the FAA has changed the process for everyone who comes next. The agency's national Programmatic Environmental Assessment for Part 135 drone delivery lets future markets be checked against a standing document; a proposal that fits inside it may not need its own environmental assessment or comment window, though the FAA's decision document commits the agency to continue site-specific historic-property and biological analyses until programmatic agreements with the SHPOs and the U.S. Fish and Wildlife Service are in place. Fitting is a qualified noise calculation, not a flat rule: the evaluated reference case is 1,150 average annual daily deliveries with hubs at least 600 feet from noise-sensitive land uses, but the document itself provides that quieter aircraft or hubs far from airports may support higher volumes, that lower-volume hubs take shorter setbacks from a table in its appendix, and that operators using the PEA must submit vehicle noise data with their applications. We broke down that framework in our analysis of the national programmatic EA's setback and delivery-cap reference case. Baton Rouge came in under the older site-specific system; noise data like its Appendix E tables is exactly the kind of fleet measurement evidence the national framework's analysis is built on.

Why Operators Outside Amazon Should Care

If you run or are planning any delivery, DFR, or infrastructure program that flies over homes, Baton Rouge is a preview of your community meeting. Four takeaways:

  1. Your EA is your deposition. Every acoustic claim you make in public will be checked against your own filing. Brief your spokespeople on the SEL tables — including what SEL is and is not — rather than the marketing sheet. The safest public statement is the one that quotes the filing directly, metric names included.
  2. Name the metric, name the phase. Cruise, approach, and delivery hover are different acoustic events with different durations, which is exactly why the technical report carries ambient, Leq, Lmax, and SEL as separate columns. A single loudness talking point cannot summarize that profile, and audiences eventually notice. People tend to forgive noise they were told about accurately; they organize against noise they were promised away.
  3. Route design is noise policy. Where your corridors sit relative to homes determines how many people experience your acoustic footprint, and "altitude is the regulator's call" is an answer that is technically defensible and rhetorically costly at a public meeting. Operators who proactively bring routing and altitude options to the FAA, rather than pointing at the agency, will own the narrative in their markets.
  4. Plan against the proposed ceiling. Community sentiment measured at today's tempo tells you little about life at the full proposed volume. If your business case needs everything you asked for, your acceptance strategy has to survive everything you asked for.

Noise is not a PR problem to be phrased around. It is a siting, routing, and flight-profile engineering problem, and it is rapidly becoming the binding constraint on drone delivery in populated areas. The operators who treat acoustic data as core documentation — disclosed, defined, and defended like any other performance number — are the ones who will still be flying at scale in five years.

For continuing coverage of how delivery noise, siting rules, and the national programmatic framework evolve, the UAVHQ dispatch on Substack is where we plan to track this beat. If you want a structured way to organize your own program's documentation for regulators and community meetings, the Part 108 BVLOS Readiness Playbook is a planning aid built around exactly that discipline; it does not and cannot promise approval outcomes.

Sources

  1. FAA Draft Environmental Assessment: Amazon Prime Air Package Delivery Operations in Baton Rouge (faa.gov) — proposed action at PDF p. 7 / printed p. 1-1; delivery profile at PDF p. 25 / printed p. 3-11; en-route overflight assumption at printed p. 3-26
  2. FAA Baton Rouge EA, Appendix E: Technical Noise Report (faa.gov) — Tables 19–20 at PDF p. 66 / report p. 25; takeoff/landing at Table 18; en-route SEL at main report Table 3
  3. FAA, Fundamentals of Noise — SEL and Lmax definitions (faa.gov)
  4. FAA Baton Rouge EA, Appendix D: Section 106 Resources and Agency Consultation (faa.gov) — cover page