The JPX Dashboard tracks aircraft operations, estimated noise levels, and curfew-period operations at East Hampton Airport (KJPX) in East Hampton, New York. It is an independent civic data project built and maintained by Zarata Media LLC under the direction of its founder, Marc Frons. Its purpose is to give residents, elected officials, and the public a clear, factual picture of airport activity and its effects on surrounding communities.

The dashboard does not advocate for a particular outcome. It presents data drawn from public and commercial sources, documents differences between published standards, voluntary commitments, and observed operations, and makes that record available for anyone to examine. Everything visible in the dashboard is explained on this page.

The dashboard includes dedicated analysis pages for the three aircraft categories of greatest community concern: Helicopters, Seaplanes, and Jets. Each page provides curfew-period operation tracking, route analysis, and activity trends specific to that category.

What we count. JPXWatch operation counts reflect flights observable through FlightAware and ADS-B. For transient traffic — arrivals from and departures to other airports — this is effectively a floor, since flights we cannot see would only raise the count. Local pattern work (training circuits) is undercounted, because FlightAware often records a full session as a single flight. Independent airport counts use different instruments and definitions, so the totals will not match exactly.

Three external sources feed the dashboard. Each covers a distinct part of the picture.

Operator Identification

The dashboard identifies aircraft operators through two sources. FlightAware provides operator names for commercial and charter flights — companies like NetJets, Blade, Flexjet, and VistaJet. For the remaining flights — primarily Part 91 private operations where FlightAware does not identify the operator — we cross-reference the aircraft's tail number (N-number) against the FAA Aircraft Registration Database, a public record maintained by the Federal Aviation Administration.

This dual approach resolves the registered owner for the vast majority of flights at KJPX. A small percentage of flights remain unresolved, typically due to temporary registrations, foreign-registered aircraft, or data entry discrepancies.

Data Freshness

Flight data updates multiple times per day via the FlightAware AeroAPI. The Data Through date in the dashboard header reflects the most recent data in the system. GPS ground track data is available for most flights, with the best coverage from mid-2025 forward. During the summer season, the dashboard's charts and flight lists default to the Last 7 Days view (the past 7 days plus today so far); in the off-season the default is a trailing 30-day window. Operator scorecards are the exception: they are always computed over a fixed trailing 12-month window, regardless of the selected time period — see the Operator Scorecards section.

The FAA Aircraft Registration Database is a public-domain dataset that links each N-number to its registered owner — whether that's an individual, an LLC, a corporation, or a trust. JPXWatch uses this data to identify the registered owner of aircraft where FlightAware does not provide an operator name.

What Is Shown

• Registered owner name — the entity (company, LLC, trust, or individual) listed as the owner in the FAA database
• Registrant type — Corporation, LLC, Individual, Partnership, Government, etc.
• Aircraft make and model — from the FAA aircraft reference data, when available

Privacy

The FAA registry includes the registered owner's mailing address. JPXWatch does not display individual owner addresses. For business entities (corporations, LLCs, trusts), the entity name alone provides sufficient identification. For individual owners, we display the name but suppress the address.

Data Source

Source: FAA Releasable Aircraft Database (registry.faa.gov). Public domain, no restrictions on use. Refreshed monthly.

JPXWatch uses the Town of East Hampton's published curfew as the primary standard for classifying nighttime operations. The Town publishes its curfew on its Flight Planning page and states it in a single line:

Curfew (All Aircraft)

10 p.m. to 7 a.m.

Source: Town of East Hampton — Flight Planning

The Town's published language is quoted verbatim. It contains no carve-outs for aircraft category, no weekend exception, and no separate tighter window for noisy aircraft. JPXWatch flags any operation with a wheels-up or wheels-down time at KJPX within the 10 PM – 7 AM Eastern window as a curfew period operation, regardless of aircraft type.

A published standard, not an enforceable rule

The Town publishes the curfew as a standard, but whether it can be enforced depends on federal law and the court orders affecting East Hampton Airport. Under the Airport Noise and Capacity Act of 1990 (ANCA), a public-use airport proprietor cannot impose noise or access restrictions without following a specific procedural process that culminates in FAA approval. East Hampton did not complete that process for the restrictions at issue. The Second Circuit's November 2016 decision in Friends of East Hampton Airport, Inc. v. Town of East Hampton, 841 F.3d 133 (2d Cir. 2016), addressed the Town's prior curfew and related ordinances on ANCA procedural grounds, and the Town's 2022 attempt to close and reopen the airport as a private-use facility was likewise blocked by the state courts on ANCA and SEQRA grounds. Those rulings remain in effect, and the Town is currently pursuing settlement discussions rather than continuing to defend the litigation.

In practical terms, operators may choose whether to observe the published window. There is no fine, no fee, no loss of airport access. JPXWatch records operations that occur within the published 10 PM – 7 AM window as curfew-period operations because the Town has published that window as the standard, not because JPXWatch has determined that a legally binding rule was violated. The dashboard's purpose is to show the public record of what happens at the airport relative to the standard the Town has chosen to publish.

Methodology update — April 2026

Prior to April 2026, JPXWatch classified operations against a 9 PM – 7 AM window drawn from the East Hampton Community Alliance's voluntary Pilot Pledge, a separately-published commitment by some operators. That window is tighter than the Town's published 10 PM curfew by one hour. In April 2026 we updated the primary standard to match what the Town itself publishes on its Flight Planning page, and re-ran historical data under the new threshold. Curfew counts on the dashboard decreased as a result, because operations landing or taking off between 9:00 PM and 9:59 PM are no longer classified as curfew operations under the published Town standard. In the same update, we corrected a separate issue where some historical curfew flags had been computed from estimated rather than observed flight times; flags derived from forecasts have been cleared and will no longer be written going forward.

Two other standards exist in public discussion of nighttime operations at KJPX and are worth naming:

• Pilot Pledge (9 PM – 7 AM). A voluntary commitment published by the East Hampton Community Alliance, an operator-affiliated organization. Tighter by one hour than the Town's published curfew. Full text below. JPXWatch used this as its primary standard prior to April 2026.
• 11 PM – 7 AM, fixed-wing voluntary (Sound Aircraft Services). The fixed-base operator at KJPX publishes an article describing a separate voluntary effort with fixed-wing operators: “The East Hampton Airport has been working with fixed-winged aircraft owners, pilots, and their passengers to observe the voluntary curfew from 11:00 PM to 7:00 AM.” This standard is one hour later than the Town's published curfew and is limited in scope to fixed-wing aircraft — it does not mention helicopters. This standard is documented here for completeness.
  Source: Sound Aircraft Services — Voluntary Curfew at the East Hampton Airport

We chose the Town's 10 PM – 7 AM window as the primary standard because it is what the municipal government of East Hampton — the airport's legal proprietor — has chosen to publish, and because it applies to all aircraft uniformly without category carve-outs. The Pilot Pledge and the SAS-documented 11 PM fixed-wing standard represent alternative community norms and operator practice, and are documented here as context. The dashboard's primary curfew count and grades are computed against the Town's published window.

The Pilot Pledge (alternative standard, documented for context)

The Pilot Pledge is the voluntary commitment JPXWatch previously used as its primary standard. It is published by the East Hampton Community Alliance, an operator-affiliated organization (not a Town of East Hampton government body), and includes a tighter 9 PM – 7 AM curfew window along with four other commitments. The full text is reproduced below for reference.

On the Daily (Summer) homepage, the “Typical for this week” tile places a day's operations count against a seasonal baseline rather than a single year-ago number. Single-day year-over-year comparisons are too noisy at a small airport — weather, day-of-week alignment, and one-off events swing them widely — so JPXWatch pools several days across several years instead.

How the baseline is pooled

For a given calendar date, we gather every day within ±3 days of that date — a seven-day window centered on it — from up to three prior years. For late May, for example, that pools roughly 21 days drawn from the same week in 2023, 2024, and 2025. Together they describe what this week usually looks like.

The typical band is the interquartile range

The range shown on the tile is the interquartile range (IQR) — the 25th to 75th percentile of that pooled distribution, the middle half of comparable days. We use the IQR rather than the full minimum-to-maximum range because a single unusual day (a holiday surge, a weather diversion) can stretch min–max far beyond what is typical. The IQR describes the ordinary band; the full observed minimum and maximum are still computed and kept for context, but they are not what the tile displays.

Example — in late May the middle-half (IQR) band is roughly 43–87 operations, while the full observed range across the same pooled days runs about 22–170. The tile shows 43–87, because that is the typical band; the 170 was a single outlier day.

Positioned, not graded

A day's count is positioned against this band — shown as a marker relative to the typical range — not graded against it. A number above or below the band is described as a phenomenon, not a pass or a failure. This is consistent with JPXWatch's broader approach: we report what happened relative to a published or historical standard and leave the judgment to the reader.

Terms used throughout the dashboard, in plain English.

Published Helicopter Routes

East Hampton Airport (KJPX) has three published helicopter arrival and departure routes, established for noise abatement:

• Echo Route (North Shore) — Available for both arrivals and departures. Connects to the North Shore Helicopter Route. This is the airport's designated preferred noise abatement route. Waypoints must be flown precisely per the published procedure.
• November Route (South Shore) — Arrivals only. Connects from the South Shore Helicopter Route via Shinnecock Inlet. Voluntary; intended to spread traffic and reduce congestion on Echo.
• Sierra Route (South Shore) — Departures only. Available only when traffic permits (no fixed-wing traffic south of the airport). Connects to the South Shore Helicopter Route.

Source: East Hampton Town Airport, "Helicopter Arrival and Departure Routes," Rev. 3, 03/04/2025.

How Flights Are Classified

We use two methods to determine which route each helicopter used, depending on available data:

Track-Based Classification (higher confidence)

Approximately 37% of helicopter operations have detailed ADS-B flight track data — a series of GPS position reports showing the aircraft's actual path, altitude, and speed throughout its flight. For these flights, we compare the actual track against the published waypoint coordinates for each route:

• Echo waypoints: E1, E2, E3, E4 (with altitude checkpoints at 2,500 ft for departures, 3,500 ft for arrivals)
• November waypoints: Shinnecock Inlet, PECONIC, N1, N2, N3 (with 3,500 ft altitude checkpoint at N1)
• Sierra waypoint: S1 (departures proceeding south toward the South Shore Route)

A flight is classified to a route if its track passes within 1 nautical mile of the route's waypoints in the correct sequence and direction. Flights with track data that do not match any published route are classified as "Non-Standard."

Origin/Destination Estimation (lower confidence)

The remaining ~63% of helicopter operations do not have detailed track data, but we do know their origin or destination airport and whether they were arriving at or departing from JPX. We estimate the likely route based on this information:

• Arrivals from north or west (Manhattan heliports, Teterboro, White Plains, Farmingdale, etc.) → Estimated Echo (the preferred North Shore Route)
• Arrivals from South Shore origins (Gabreski/Westhampton) → Estimated November
• All departures → Estimated Echo (we cannot verify Sierra's "traffic permitting" condition without tower communications, so we conservatively default to the preferred route)

Estimated flights are never classified as "Non-Standard" because we cannot determine actual route adherence without track data.

Altitude Adherence

For flights with GPS track data, we check altitude at each published checkpoint waypoint. A flight is recorded as "within the published altitude" at a checkpoint if its altitude is within ±300 feet of the published (voluntary) altitude. Altitude adherence data is only available for track-based classifications, not for estimated flights.

What This Data Shows — and What It Doesn't

The purpose of this analysis is to provide the community with objective, data-driven insight into helicopter route usage patterns at East Hampton Airport to support informed discussion about noise impacts.

Data Sources

• Flight data: FlightAware AeroAPI — aircraft identification, flight tracks, origin/destination, aircraft type
• Route definitions: East Hampton Town Airport, "Helicopter Arrival and Departure Routes," Rev. 3, 03/04/2025
• Aircraft noise levels: FAA/EASA Part 36 type-certification data
• Helicopter type information: FAA aircraft registry, manufacturer published specifications

Seaplane Classification

Seaplanes are identified at the airframe (tail) level, not flight by flight. An amphibious aircraft does not switch between floats and wheels from one trip to the next, so once an aircraft is identified as a seaplane, all of its operations in eligible aircraft types are counted as seaplane — not only the legs that land on water. This avoids a per-flight flicker where the same aircraft would count as a seaplane on some trips and fixed-wing on others.

An aircraft is identified as a seaplane when any of the following holds:

1. Water-aerodrome operation (primary) — The aircraft completes at least one operation at a coded seaplane base, such as the NYC Skyports Seaplane Base, which is a water-only facility a wheeled aircraft cannot use. This signal needs no operator information.
2. Known seaplane operator — The aircraft is operated by, or flies under the callsign of, a company known to operate float-equipped aircraft at KJPX (for example, Tropic Ocean Airways).
3. Always-float type — The ICAO type designator is exclusively a float plane (DHC-2 Beaver, DHC-3 Otter).

Only candidate aircraft types are eligible for the first two signals — the Cessna 208 Caravan and Daher Kodiak, whose type codes are identical whether the aircraft is on wheels or amphibious floats. A type code alone is never sufficient: a Caravan with no water-aerodrome operation and no seaplane operator stays classified as fixed-wing. Aircraft we cannot confidently identify as seaplanes remain counted as fixed-wing — we deliberately under-count rather than over-count, moving an aircraft to the seaplane category only on a positive, defensible signal.

Key Seaplane Operators at KJPX

• Tropic Ocean Airways (Blade contractor) — Cessna 208 Caravan
• Tailwind Air — Cessna 208 Caravan (amphibious)
• FlyTheWhale — Cessna 208B Grand Caravan (amphibious)
• Acadian Seaplanes (Blade contractor) — Cessna 208 Caravan

Helicopter Corridor Proximity Analysis

The Seaplane Activity page includes an analysis of whether seaplane flight tracks intersect the published helicopter corridors (Echo, November, Sierra). This addresses a concern that seaplanes may be using routes designated for helicopters.

The Pilot Pledge commits that “all seaplane operators agree to standardize seaplane arrival and departure routes corresponding to the least noise-sensitive areas.” That commitment concerns seaplane routing generally and does not address the helicopter corridors; the corridor proximity analysis is a separate, descriptive measure of overlap.

Methodology:

• Helicopter corridor boundaries are defined from coordinates published by the Town of East Hampton.
• Corridors are buffered at 1.0 nautical mile from centerline.
• Each seaplane flight track is checked for intersection with corridor polygons.
• Flights with >25% of track positions inside a helicopter corridor are flagged as having significant overlap.
• Some corridor intersection is expected on standard approaches to Runway 28, as the November corridor crosses the approach path. Brief intersection during approach does not necessarily indicate a seaplane was using the helicopter corridor as its route.

Data Sources

• Flight tracks: FlightAware AeroAPI
• Corridor definitions: East Hampton Town Airport, "Helicopter Arrival and Departure Routes," Rev. 3, 03/04/2025

Jet Classification

Jets are identified by ICAO aircraft type designator and engine classification. All turbojet and turbofan-powered fixed-wing aircraft are classified as jets. Turboprop aircraft are tracked separately.

Unlike seaplanes, jet classification is unambiguous from the aircraft type code — no layered detection is required.

Aircraft Size Classes

Jets are certified under Part 36 Appendix B in Effective Perceived Noise Level (EPNdB) rather than the A-weighted decibels (dB(A)) used for smaller aircraft. EPNdB is a more sophisticated metric than dB(A) for jet noise specifically: it accounts for tone corrections (such as the distinct whine of a turbofan) and integrates noise over the duration of a flyover event. dB(A) does not capture the tonal character of jet noise as accurately.

The size-class ranges below are derived from EPNdB certification data, presented as plain dB on the dashboard for visual consistency across aircraft categories that use different certification units (Appendix B and H aircraft are certified in EPNdB; Appendix F, G, and J aircraft are certified in dB(A)). This labeling convention prioritizes consistency over per-category specificity. The full unit-convention treatment is documented in the Severity Levels section below.

Jets at KJPX range from very light jets to ultra-long-range aircraft, with significant variation in noise impact:

• Very Light Jet (VLJ): Eclipse 500, Phenom 100 (~70–75 EPNdB)
• Light: Citation CJ3, Phenom 300 (~75–80 EPNdB)
• Midsize: Citation XLS, Learjet 60 (~78–83 EPNdB)
• Super-Midsize: Citation Longitude, Challenger 350 (~80–85 EPNdB)
• Large: Gulfstream G450/G550, Falcon 900 (~83–88 EPNdB)
• Ultra-Long-Range: Gulfstream G650/G700, Global 7500 (~85–90 EPNdB)

Each 10 dB increase represents approximately a doubling of perceived loudness. A Gulfstream G650 on approach is roughly four times as loud to the ear as a Cessna 172.

Runway Corridor Analysis

Unlike helicopters and seaplanes, which have some flexibility in routing, jets are constrained to the runway centerline on approach and departure. IFR procedures, PAPI guidance, and ATC instructions keep them on a narrow, predictable path. This means jet noise is highly concentrated on a specific geographic band — and residents under that band bear a disproportionate burden.

The Jet Activity page analyzes flight tracks relative to the extended runway centerlines:

• Runway 28 approaches (the dominant arrival direction, since Runway 28 is the standard fixed-wing runway)
• Runway 10 departures
• Runway 16/34 operations (secondary crosswind runway)

Altitude adherence is also assessed: the Pilot Pledge asks turbojet pilots to maintain a minimum altitude of 1,500 ft in the pattern. The dashboard checks actual altitude profiles against this voluntary commitment.

Operator Categories

Jet operators at KJPX fall into distinct categories:

• Fractional operators (NetJets, Flexjet, WheelsUp, PlaneSense) — sell shares of aircraft, generating high-frequency repeat visits
• Charter operators — on-demand service
• Corporate/Private — individual or company-owned aircraft

This categorization helps distinguish high-volume commercial operations from occasional private flights.

Historical Context: Stage 3 Restrictions

The Town of East Hampton attempted to restrict Stage 3 jets (the oldest and loudest noise certification standard) in 2015–2017. Federal courts struck down these restrictions, ruling they violated grant assurances under the Airport Noise and Capacity Act. The Jet Activity page shows the current mix of Stage 3, Stage 4, and Stage 5 aircraft operating at KJPX.

Data Sources

• Flight data: FlightAware AeroAPI
• Aircraft type classification: ICAO Doc 8643
• Noise certification levels: EASA/FAA Part 36 type-certification data

Why Weather Matters for Airport Noise

Weather — especially wind direction — is the single biggest factor determining which runway is active and where aircraft fly. When the wind shifts, flight paths shift with it, and neighborhoods that were quiet one day can be under the flight path the next. Temperature, visibility, and atmospheric conditions also affect how far and how loudly sound travels on the ground.

The Weather Correlation Analysis page connects real weather observations to flight operations and noise data at East Hampton Airport, helping the community understand why certain days or seasons are louder than others.

What's on the Page

Current Conditions

Live weather data for the JPX area including temperature, wind speed and direction, sky condition, visibility, and flight category (VFR/MVFR/IFR). Updated automatically from NOAA aviation weather observations.

Wind Speed vs. Noise Level

Scatter plot showing the relationship between wind speed and average estimated noise level for each observation period. Higher winds can both change flight paths and affect how sound propagates to the ground.

Temperature vs. Operations

Scatter plot showing how temperature correlates with flight activity. At JPX, warm-weather weekends see significantly higher operations than cool weekdays — a pattern driven by seasonal Hamptons travel demand.

Operations by Weather Condition

Breakdown of observed flight operations by sky condition (Clear, Partly Cloudy, Overcast, Rain, Fog, Wind Advisory). Includes the average noise level associated with each condition. Operations tend to decrease in poor weather, but the noise per operation may vary.

Wind Rose — Operations by Direction

Polar chart showing how many operations occurred under each wind direction, color-coded by estimated noise level. This reveals which wind patterns correlate with the highest noise exposure. At JPX, summer's prevailing southwest winds sit at a wide angle to Runway 28 — more crosswind than headwind — so Runway 28, the standard fixed-wing runway, carries that traffic by preference, tolerating the crosswind rather than being favored by it, with departures heading west over Wainscott and nearby south-shore communities.

Key Insights

Auto-generated plain-English findings based on the data for the selected period. These highlight significant patterns — seasonal trends, weather-driven operational changes, and noise correlations.

“Today at JPX” on the Main Dashboard

The main dashboard includes a “Today at JPX” section showing current conditions and a brief explanation of what they mean for flight activity:

• Current temperature, sky condition, and wind speed/direction
• Which runway fixed-wing traffic is likely using, inferred from wind direction and the airport's published Runway 28 preference (the main runway is 10/28, roughly east–west, with a shorter 16/34 crosswind runway)
• Today's operations count with context (above/below average for the period)

This section updates live and provides a quick answer to the question “what's happening at the airport right now?”

How Wind Direction Affects Runway Usage

East Hampton Airport's main runway is Runway 10/28, oriented roughly east–west, with a shorter secondary crosswind runway, 16/34, running roughly northwest–southeast. For fixed-wing aircraft (jets and props), Runway 28 is the airport's published standard runway, used whenever winds allow. (Helicopters follow their own published routes and aren't covered here.)

• Winds from the west, or the prevailing summer southwest → Runway 28 (aircraft land from the east, depart to the west). A westerly favors Runway 28; the more common southwesterly is a crosswind that Runway 28 — the standard fixed-wing runway — tolerates and carries by preference, rather than being favored by it. Departures overfly Wainscott and communities to the west.
• Winds from the east or northeast → Runway 10 (aircraft land from the west, depart to the east). Departures overfly communities east of the airport, toward Montauk.
• Strong winds from the north or south → the 16/34 crosswind runway may come into use, when a north–south wind would be too strong a crosswind for Runway 10/28. Runway 16 sends traffic to the south, Runway 34 to the north.
• Light and variable winds (below ~5 knots) → Mixed runway usage. ATC may use either runway based on traffic flow, and flights may approach from either direction.

Understanding wind direction explains why noise patterns change day to day even when the total number of operations is similar.

Temperature Inversions

On some days, noise at ground level seems louder than usual even though operations haven't increased. This is often caused by a temperature inversion — a weather condition where a layer of warm air sits above cooler air near the surface. Normally, sound dissipates upward as it travels. During an inversion, sound waves are bent back toward the ground, effectively trapping and amplifying noise at the surface.

Temperature inversions are most common during calm, clear evenings and early mornings. Inversion analysis uses temperature profile data from the NWS radiosonde station at Upton, NY (OKX) as reference background. The Weather Correlation Analysis page does not currently surface inversion-period flags.

Flight Categories: VFR, MVFR, IFR, LIFR

The current conditions display includes a flight category badge. These categories are defined by the FAA based on visibility and cloud ceiling height. The definitions below are the single shared source used by every flight-category badge on the dashboard, so the live chips and this page can never drift apart:

• VFRVisual Flight Rules — Ceiling greater than 3,000 ft and visibility greater than 5 miles. Pilots navigate primarily by visual reference to the ground. Most general-aviation flights at KJPX operate under VFR.
• MVFRMarginal Visual Flight Rules — Ceiling 1,000–3,000 ft and/or visibility 3–5 miles. VFR flight is still possible but margins are reduced; pilots should exercise caution.
• IFRInstrument Flight Rules — Ceiling 500 to less than 1,000 ft and/or visibility 1 to less than 3 miles. Pilots must navigate by instruments and be IFR-rated. Traffic at KJPX is typically reduced.
• LIFRLow Instrument Flight Rules — Ceiling below 500 ft and/or visibility below 1 mile. The most restrictive conditions, with very limited operations.

At JPX, the published helicopter routes (Echo, November, and Sierra) are designed for VFR conditions. Under IFR or lower, those published VFR routes are not flyable — pilots must use instrument procedures or divert, and most helicopter operations are reduced or suspended.

Lower flight categories generally mean fewer operations but do not eliminate them — instrument-rated pilots and aircraft can still operate in reduced conditions.

Weather Observations

Weather data displayed on the Weather Correlation page is sourced from hourly METAR observations at KJPX (East Hampton Airport, the primary station) and KFOK (Francis S. Gabreski Airport, Westhampton, ~12 miles west) as a fallback for gaps in KJPX's reporting.

KJPX is a non-federal AWOS-3 (Automated Weather Observing System), which provides full sensor coverage including precipitation discrimination. Its observations are typically reported every ~20 minutes. KFOK is a federal ASOS (Automated Surface Observing System) and reports every ~5 minutes, though full parsed values (temperature, dewpoint, humidity) are only included in the standard hourly METAR.

Observations are sourced from the Iowa Environmental Mesonet, an academic archive maintained by Iowa State University that ingests from NOAA's authoritative feeds (Unidata IDD, NCEI ISD, MADIS One Minute ASOS).

When KJPX has missing observations (typically due to brief sensor outages or maintenance windows), the page falls back to KFOK data for the affected hours or days. The percentage of fallback usage is shown in each chart's source attribution. Across the backfilled range (2020-01-01 onward), KJPX covers 98.2% of days and KFOK fills the remaining 1.8% (42 days).

Data Sources

• Selected-period weather correlation: historical METAR observations from KJPX (primary) with KFOK (Westhampton) as fallback; sourced from the Iowa Environmental Mesonet (NOAA-derived).
• Live current conditions chip: real-time METAR from NOAA Aviation Weather Center (aviationweather.gov), no API key required. Used for the live conditions chip and the Raw METAR tooltip; sourced separately from the historical path.
• Temperature inversion analysis: NWS radiosonde station OKX (Upton, NY) upper-air soundings; reference background only. Inversion-period flags are not currently surfaced on the page.
• Flight data: FlightAware AeroAPI.
• Noise estimates: FAA/EASA Part 36 type-certification profiles, applied to each operation based on aircraft type, altitude, and distance.

What It Shows

The Carbon Emissions module estimates the CO₂ produced by flight operations at East Hampton Airport based on aircraft type, fuel burn rates, and flight duration. It provides:

• Total CO₂ for the selected period across all observed operations
• Per-operation average in kilograms
• Car-year equivalent — total emissions translated into the equivalent number of average passenger cars driven for one year, using EPA's figure of 4.6 metric tonnes CO₂ per car per year
• Top emitters by aircraft type — which aircraft types contribute the most total CO₂, ranked by aggregate emissions rather than individual flight impact
• Category breakdown — share of total emissions from jets, fixed wing, helicopters, and unclassified aircraft

How Emissions Are Calculated

Each operation's CO₂ estimate follows a three-step calculation:

1. Aircraft type → fuel burn rate

Every aircraft type has a known approximate fuel consumption rate based on its engine model and configuration. We use published fuel burn data from the ICAO Aircraft Engine Emissions Databank and manufacturer specifications to assign a fuel burn rate (in kg per hour) to each aircraft type observed at JPX.

2. Fuel burn × flight duration → total fuel consumed

Flight duration is sourced from FlightAware data for each operation. For operations without precise duration data, we use average flight times for typical JPX routes (e.g., approximately 35–45 minutes for NYC-Hamptons helicopter shuttles).

3. Fuel consumed × emission factor → CO₂

We apply standard emission conversion factors:

• Jet A-1 fuel (used by turbine aircraft — jets and turboprops): 3.16 kg CO₂ per kg of fuel burned
• Avgas (used by piston aircraft): 3.10 kg CO₂ per kg of fuel burned

These conversion factors are established by the International Civil Aviation Organization (ICAO) and are consistent with those used by the FAA, EASA, and EnviroSuite's carbon emissions module.

What These Estimates Include — and What They Don't

These estimates represent the direct flight-phase carbon footprint of airport operations. Actual total climate impact is higher when upstream and non-CO₂ effects are included.

Accuracy and Limitations

Carbon emissions estimates are approximate. Actual fuel consumption varies based on specific engine variant within a type family, payload, weather conditions (headwinds increase fuel burn), and pilot technique and ATC routing.

For community-level environmental intelligence, these estimates provide a reliable indicator of the relative scale and trend of aviation carbon emissions at JPX. They are suitable for understanding patterns and comparing periods, but should not be used as precise accounting for regulatory or offset purposes.

Data Sources

• Aircraft type and flight duration: FlightAware AeroAPI
• Fuel burn rates: ICAO Aircraft Engine Emissions Databank; manufacturer published specifications
• Emission conversion factors: ICAO standard factors (3.16 kg CO₂/kg Jet A-1; 3.10 kg CO₂/kg avgas)
• Car-year equivalence: U.S. EPA figure of 4.6 metric tonnes CO₂ per average passenger car per year

Data Source

Flight data is sourced from FlightAware's AeroAPI, which aggregates FAA radar, ADS-B, and airline data. Where track data is available, JPXWatch uses observed FlightAware ground-track data rather than modeled or synthetic paths. Track coverage varies by aircraft, date, and FlightAware data availability; see Known Limitations. Flight data updates multiple times per day.

Aircraft Noise Levels

Aircraft certification data is collected under standardized procedures defined in 14 CFR Part 36. Depending on aircraft category and certification appendix, the reported values may include takeoff/flyover, approach, sideline/lateral, or other prescribed measurement conditions. JPXWatch maps these published or derived aircraft-type reference values to ICAO type codes and uses them as the source-noise input for its comparative model. JPXWatch's propagation model uses a reference distance of 1,000 feet (304.8 meters) for its attenuation calculations, applying inverse-square geometric spreading, with atmospheric absorption applied as a simple linear function of total slant distance. JPXWatch maintains an EASA-derived noise profile mapping for 57 ICAO aircraft type codes, covering the helicopter, jet, and fixed-wing types most commonly observed at KJPX. Aircraft types not present in the mapping receive a category-average estimate based on whether they are classified as helicopter, jet, or fixed-wing.

Every recorded operation receives a modeled noise value. Across operations in the trailing twelve months, 84.2% use a type-specific certified reference value where one exists, 8.2% fall back to an aircraft-class average where no type-specific reference is available, and 7.6% use a default value for the small remainder. No operation is excluded from noise computation. (Resolver-cascade coverage as of June 2026.)

JPXWatch uses publicly available aircraft noise reference data from ICAO NoisedB, EASA TCDSN, FAA AC 36-1H, and, for certain residual helicopter types without published certified values in those sources, FAA/Volpe/Technical Center ROSAP measurement reports. Some aircraft types require fallback or residual treatment.

Certification units vary by aircraft category — A-weighted decibels (dB(A)) for most propeller aircraft and helicopters, and Effective Perceived Noise Level (EPNdB) for jets, as detailed in the Jets section.

Part 36 certification data is the foundational noise source used throughout aviation noise regulation. The FAA's Aviation Environmental Design Tool (AEDT) — the agency's required model for domestic regulatory noise analyses under current FAA environmental policy, including FAA Order 1050.1G (effective June 30, 2025, superseding 1050.1F) — computes noise contours using Noise-Power-Distance (NPD) curves derived from certification flight test data (AEDT Technical Manual §4.2; FAA ASCENT Center of Excellence). ICAO Doc 9911, the international standard for computing noise contours around airports, is explicitly “designed to make full use of” the Aircraft Noise and Performance (ANP) database assembled by manufacturers in collaboration with noise certification authorities. AEDT is Doc 9911-compliant. JPXWatch uses aircraft type-reference noise data drawn from the same general class used in aviation noise modeling, applying a simplified public-data propagation model rather than full NPD segmentation.

Altitude: MSL to AGL Correction

FlightAware reports aircraft altitude as MSL (Mean Sea Level) — the standard used by aircraft altimeters and air traffic control. However, noise impact at the ground depends on the aircraft's height above the listener, known as AGL (Above Ground Level).

JPXWatch converts MSL altitude to estimated AGL by subtracting a ground elevation appropriate to the listener's location. The default is the airport field elevation of 55 feet MSL (the FAA-standard reference for KJPX, per FAA Form 5010). Where higher-elevation zones along the South Fork moraine have been characterized, the model applies a location-specific override — currently for the moraine ridge near Town Line Road in Wainscott (132.5 ft MSL). Planned improvements include extending location-specific elevation data to additional flight-path-sensitive areas and eventually integrating per-location terrain data from the USGS National Elevation Dataset; status is tracked in the methodology revision log.

Terrain across Long Island's South Fork is relatively flat, with elevations typically ranging from 20 to 80 feet MSL, with some higher points along the glacial moraine running through Wainscott and East Hampton. The constant-elevation approach is a reasonable approximation for this terrain, though it does not capture per-property differences. A helicopter reported at 500 feet MSL over terrain at 55 feet elevation is calculated as 445 feet above the ground.

Distance Attenuation

Noise decreases with distance from the source. JPXWatch estimates received noise at ground level using slant distance — the true straight-line distance from the listener to the aircraft, calculated as:

Slant distance = √(ground distance² + altitude²)

Where ground distance is the horizontal distance from the listener's location to the nearest point on the aircraft's GPS track, and altitude is the AGL-corrected height at that point.

Sound attenuation is modeled using geometric spreading (inverse-square law: noise decreases by 6 dB per doubling of distance) plus atmospheric absorption (approximately 0.5 dB per 1,000 feet — a linear approximation; AEDT itself uses frequency-dependent atmospheric absorption methodology that JPXWatch currently does not implement). The estimated noise at the receiver is:

Received noise (dB) = Reference noise − 20 × log₁₀(slant distance / reference distance) − (0.5 × slant distance / 1,000)

This is a first-order source-propagation model informed by established aviation-noise modeling principles, but it is not an AEDT implementation and should not be read as regulatory contour methodology. Specific planned improvements are documented in the methodology revision log.

Two limitations in JPXWatch's modeled event estimates are worth surfacing explicitly, both inherited from the broader methodology family rather than specific to JPXWatch.

First, JPXWatch currently does not apply lateral attenuation or ground-impedance corrections. The model does not adjust for how sound propagation differs over grass, pavement, buildings, water, or other surfaces — a category of limitation that affects regulatory tools as well: AEDT's more advanced lateral-attenuation methodology was derived over grass-covered terrain and the AEDT 4a Technical Manual acknowledges that it can under-predict noise levels over acoustically hard surfaces (Section 4.3.5, footnote xix). This is particularly relevant at JPX, where many helicopter and seaplane routes cross Mecox Bay, Three Mile Harbor, and Long Island Sound.

Second, JPXWatch's modeling does not separately account for airframe noise as distinct from propulsion noise. The AEDT 4a Technical Manual identifies this as a current gap (Section 11.2.2): “one of the gaps in the current version of AEDT is the inability to separate the effects of airframe noise from propulsion system noise.” JPXWatch shares this limitation with the regulatory state of practice.

These limitations argue against treating JPXWatch's noise estimates as monitor-equivalent measurements at any specific listener location. They are appropriately precise for relative comparison — comparing one operator to another, one aircraft type to another, one neighborhood's exposure to another's — and that's how the dashboard is designed to be read.

JPXWatch's methodology is under active development. Specific planned improvements are documented in the methodology revision log. The methodology described on this page reflects the current state as of the most recent revision-log entry.

Severity Levels

85 dB is a project-defined accountability threshold used by JPXWatch. It corresponds to the upper edge of the typical operations band at KJPX (most operations fall in the 75–84 dB range) and flags operations meaningfully above the prevailing acoustic baseline. It is not an FAA regulatory standard, and JPXWatch does not claim parity with formal noise monitoring systems such as Casper Noise Lab or Envirosuite, which use calibrated Type 1 microphones at known ground locations. JPXWatch exists to provide community-level transparency in the absence of installed monitoring infrastructure at KJPX. JPXWatch does not advocate for a specific airport-access or regulatory outcome; it does identify data-quality improvements — including calibrated ground-based noise monitoring — that would improve public understanding of aircraft noise.

JPXWatch's approach — correlating aircraft type with noise characteristics using certification data — is consistent with the methodology used by established airport noise transparency platforms worldwide. WebTrak (Envirosuite/Brüel & Kjær), deployed at JFK, LaGuardia, Newark, Reagan National, LAX, and across Australia's Airservices network, integrates flight tracking with aircraft-type-based noise data. Casper Noise serves Amsterdam Schiphol, London Gatwick, and Tampa International. Noise-Map.com estimates aircraft noise globally using ADS-B flight data and aircraft-type-based propagation modeling. None of these systems claim to replace calibrated ground monitoring; all use aircraft type identification — the same certification-linked variable — as a core input for noise characterization.

Why some displays use dB(A), plain dB, or EPNdB

JPXWatch's dashboard surfaces noise values from three sources, each with its own native unit. Jet estimates display in EPNdB because the underlying data for jets is computed from Part 36 Appendix B EPNdB certification readings. Helicopter and small fixed-wing estimates display in dB(A) because the underlying data is LAmax at 1,000 ft — an A-weighted measurement, either recorded as FAA LAmax field measurements (source verification in progress; see the June 2026 provenance entry in the methodology revision log) or converted from cert EPNdB. Cross-category aggregations — mixed-fleet bands and fleet-wide averages — use plain dB as a unit-neutral fallback. Physical sensor measurements such as the reference levels shown in the sensor levels legend use dB(A) per the IEC 61672 standard for sound level meters. The unit shown on each display reflects what JPXWatch's source data uses, not an editorial choice about what's “right.”

Relationship to Peer-Reviewed Methodology

Feng, Zhou, Zeng & Ding (2023), “Review on Metrics and Prediction Methods of Civil Aviation Noise” (Int. J. Aeronaut. Space Sci. 24:1199–1213, full text, doi:10.1007/s42405-023-00609-0), classify aviation noise prediction models into three families: best-practice, scientific, and machine-learning. JPXWatch draws from the same broad source-propagation family as FAA AEDT, the legacy INM, and ECAC Doc 29, treating each aircraft as a noise point source. Unlike those regulatory models, JPXWatch does not perform full Noise-Power-Distance segmentation, lateral attenuation adjustment, or full flight-path integration; it is a first-order approximation suitable for comparative analysis, not regulatory contour production.

We pair DNL with a Number-Above-threshold count, surfaced as the Address-Level Noise Events metric. Supplementing energy-averaged metrics with event counts is the approach recommended under EU Directive 2002/49/EC for noise environments dominated by short-duration events — the profile of a seasonal general-aviation airport, where annualized averages understate community impact.

Peer-reviewed research supports the broader premise that aircraft certification categories and aircraft-type noise characteristics produce useful relative comparisons against measured noise trends. Rindfleisch et al. (2024), in the largest AEDT validation study to date, compared model predictions against ground-level measurements for 200,000 arrival trajectories at San Francisco International Airport over 12 months (Journal of the Acoustical Society of America 155(3):1928). A multi-model comparison by Meister et al. (2021) tested three noise programs (sonAIR, FLULA2, and AEDT) against several thousand single-flight measurements at Zurich and Geneva airports (Aerospace 8(12):388). At Amsterdam Schiphol, Simons et al. (2022) confirmed that aircraft categorized by noise certification class tracked measured noise trends over six years — validating that certification categories correlate with real-world noise behavior.

Certification noise levels provide useful relative indicators of aircraft-type noise output. JPXWatch's primary analytical value is comparative: identifying which aircraft types, operators, and time periods produce more or less noise impact at the community level, not certifying the precise decibel level at any listener's property. These studies support the broader use of aircraft-type and certification-linked data for comparative noise analysis. They do not validate JPXWatch's KJPX-specific estimates, which are not calibrated against local ground monitors.

The remaining gap to a full best-practice implementation is ground-truth calibration against Type 1 (IEC 61672-1 Class 1) reference microphones, which JPXWatch does not currently operate. Modeled levels should be read as directionally accurate estimates, not certified measurements.

Within-type variation

The event-level noise values on the scorecards are derived from Part 36 type-certification data — a standardized reference measurement taken under specified test conditions for a representative aircraft of each type.

Actual noise produced by a specific aircraft at KJPX may differ from its type-certification reference. Aircraft within the same ICAO type code can span decades of production with different engines, propellers, and airframes. Some older aircraft predate Part 36 certification entirely. Many types are delivered with multiple engine or propeller options that produce different noise profiles, and aftermarket modifications (STCs for engines, propellers, or mufflers) can further change an aircraft's noise signature. Operating technique, aircraft weight, and meteorological conditions also affect real-world noise.

Event level values should therefore be read as a comparative reference — useful for distinguishing a typical helicopter from a typical jet from a typical piston single — not as a prediction of what any specific aircraft produced on any specific flight. There are currently no official ground-based noise monitors at KJPX; all noise values are modeled.

The Noise Impact Score (0–100) is a composite metric that combines three factors to quantify the overall noise burden on the community. Each subscore is computed independently from per-day operation counts and then combined as a weighted average.

Fleet Mix Score (40% weight)

Measures the proportion of high-noise aircraft in the operations mix. Helicopters always count as high-noise. Jets count as high-noise only when their modeled event-level noise meets or exceeds 85 dB at 1,000 ft reference distance.

min(100, (helicopters + loud_jets) / total_operations × 500)

The 500 multiplier is calibrated so that 20% high-noise operations yields a Fleet Mix subscore of 100. At 10% high-noise, the subscore is 50; at 5%, it is 25.

Concentration Score (30% weight)

Measures operations density. Higher daily operation counts produce higher subscores.

min(100, average_operations_per_day × 2)

The 2 multiplier is calibrated so that 50 operations per day yields a Concentration subscore of 100. At 25 operations per day, the subscore is 50.

Time-of-Day Score (30% weight)

Measures the share of operations occurring during the Town of East Hampton's published 10 PM – 7 AM curfew period. Non-curfew flights (including in-flight operations whose actual landing or takeoff time has not yet been observed) are excluded from this count under strict equality on the curfew flag.

min(100, curfew_operations / total_operations × 1000)

The 1000 multiplier is calibrated so that 10% curfew operations yields a Time-of-Day subscore of 100. At 5% curfew, the subscore is 50; at 1%, it is 10.

Composite Noise Impact Score

The three subscores are combined as a weighted average:

round(fleet_mix × 0.40 + concentration × 0.30 + time_of_day × 0.30)

The result is a single 0–100 number reflecting the overall noise burden from operations during the period.

Severity Tiers

• 0–39 — LOW. Minimal aircraft-noise burden. Quiet fixed-wing traffic predominates.
• 40–64 — MODERATE. Noticeable aircraft-noise burden. Elevated helicopter, jet, or curfew-period activity.
• 65–84 — HIGH. Significant aircraft-noise burden. Heavy concentration of louder aircraft and/or curfew operations.
• 85–100 — SEVERE. Upper portion of observed scores. Reflects sustained heavy traffic of louder aircraft, often combined with curfew or shoulder-hour activity.

The tier boundaries are project-defined and may be adjusted as the underlying score distribution changes. Tier assignment is computed from the score; it is not an editorial judgment about the operating period itself.

What Is DNL?

Day-Night Average Sound Level (DNL) is the FAA's primary metric for assessing aircraft noise impact around airports. It represents a 24-hour cumulative energy average of all aircraft noise events, with a 10 dB penalty applied to flights between 10 PM and 7 AM ET to account for the greater disruption of nighttime noise.

The FAA uses DNL in Part 150 noise compatibility studies to determine which areas around an airport experience "significant" noise exposure. In Part 150 noise compatibility planning, 65 dB DNL is the conventional threshold at which residential land use is generally treated as incompatible absent sufficient noise reduction. JPXWatch references this threshold for context only. JPXWatch does not produce FAA-approved Part 150 contours and does not determine whether any property qualifies for sound insulation or other regulatory treatment.

How JPXWatch Computes DNL

JPXWatch estimates DNL for each address using flight track data from FlightAware and noise reference levels from EASA/FAA Part 36 aircraft type-certification data. For each flight that passes within the relevant analysis radius:

1. The aircraft's type-certification noise level is looked up from Part 36 reference data.
2. The closest slant distance from the listener/grid point to the aircraft track is computed using AGL-corrected altitude, as described in the Noise Estimation Methodology above.
3. The reference value is adjusted using geometric spreading and a static atmospheric absorption coefficient.
4. Sound Exposure Level is estimated using track-derived exposure duration — the time represented by FlightAware track points within the analysis radius.
5. Flights between 10 PM and 7 AM receive the standard 10 dB nighttime penalty.
6. All events are energy-averaged over the full season (Memorial Day through Labor Day) to produce the seasonal DNL value.

The model does not currently apply lateral attenuation, terrain/DEM correction, or live weather adjustment; all three are tracked as methodology improvements.

What This Model Does Not Capture

Why DNL May Understate Impact at Seasonal Airports

DNL is computed by averaging noise energy over a 24-hour period, with a 10 dB penalty applied to nighttime operations. This averaging mathematically dilutes the impact of individual loud events when they're spread thin over a quiet day.

At a high-traffic metropolitan airport with hundreds of operations daily, the cumulative energy of many flights produces a DNL that meaningfully reflects the experienced noise environment. At a low-traffic seasonal airport like JPX — where daily operations typically range from a handful in winter to a few dozen on summer weekends — the same averaging produces a DNL that may understate the experienced loudness of individual overflights. Each individual event may be loud (80–95+ dB at close range for helicopters and jets at low altitude), but the long quiet intervals between events pull the 24-hour average down substantially.

The result: most impacted addresses near KJPX produce estimated DNL values below the FAA's 65 dB threshold, with small zones modeled at or above it under the current propagation model — one immediately around the KJPX airfield, and a second, small zone at Gabreski Airport in Westhampton (KFOK), where KJPX–Gabreski shuttle flights operate at low altitude. This does not mean noise outside those zones is insignificant. It means DNL is poorly calibrated for seasonal, low-frequency airports where the problem is intense individual events, not continuous background exposure.

This limitation has been recognized in noise research literature. Supplementary single-event metrics — including Sound Exposure Level (SEL), Maximum Sound Level (Lmax), and counts of events above specific thresholds — are commonly used to characterize noise environments where DNL alone is insufficient. JPXWatch surfaces both DNL and event-based metrics for this reason; reading them together gives a more complete picture than either alone.

The Nighttime Penalty

DNL applies a 10 dB penalty to flights between 10 PM and 7 AM — the FAA's standard nighttime window for acoustic exposure modeling. This happens to coincide with the Town of East Hampton's published curfew of 10 PM – 7 AM, but the two standards serve different purposes and are computed independently: DNL is an acoustic weighting used for noise exposure modeling under Part 150, while the curfew is an operational standard for when aircraft should not be flying.

The Address-Level Noise Events metric — a count of individual flights whose modeled noise at the listener's address met or exceeded a project-defined threshold — is currently under methodology review. Detailed guidance will return to this section once the review of how the threshold should be defined relative to acoustic standards for community noise is complete.

What Is NoiseRank?

NoiseRank shows where your address falls on the spectrum of aircraft noise exposure relative to the broader East End community during East Hampton Airport's summer season (Memorial Day through Labor Day).

Your NoiseRank percentile compares your address's estimated DNL against a reference grid of evenly spaced points covering the KJPX impact zone — from Hampton Bays to Montauk, and from the South Fork across to the North Fork. A ranking of "Top 5%" means your address experiences more estimated aircraft noise than 95% of locations in the impact zone.

The Reference Grid

Rather than comparing only among self-selected users (which would introduce selection bias — people in noisy areas are more likely to check), NoiseRank compares against a pre-computed grid of points spaced 100 meters apart across the entire impact zone. The grid:

• Covers the geographic area defined by the convex hull of all observed flight tracks, plus a buffer
• Excludes points over open water
• Includes approximately 310,000 grid points (the exact count is displayed on the NoiseRank card)
• Is recomputed at the end of each summer season, and updated weekly during the live season

Each grid point has its own estimated DNL, computed using the same methodology described in the DNL section above. Your address's DNL is compared against this distribution to produce your percentile rank.

The Five-Band Scale

Why Low DNL Can Coexist with High NoiseRank

It is possible — and at KJPX, common — for an address to show a DNL well below 65 dB while also ranking in the top percentiles for NoiseRank. This is not a contradiction. It reflects two facts:

1. DNL is an absolute metric. As explained above, DNL at a seasonal airport will be low in absolute terms because the 24-hour averaging dilutes a small number of loud events across long quiet periods.
2. NoiseRank is a relative metric. It compares your address against all other locations in the impact zone. The vast majority of the East End — interior farmland, areas distant from flight paths, the North Fork — experience near-zero aircraft noise. Even modest flight activity puts an address near the airport in the upper percentiles, because most of the grid gets even less.

The Noise Exposure Heatmap on the Noise Map page makes this visually clear: the East End is overwhelmingly green, with concentrated yellow, orange, and red zones near the airport and under flight paths. An address in the red zone will have a high NoiseRank but may still fall below the FAA's 65 dB DNL threshold.

Both metrics are correct and useful. DNL anchors the data to the FAA's regulatory framework. NoiseRank shows where you stand relative to your community. Together with the Address-Level Noise Events count, they provide a fuller picture: cumulative seasonal exposure, frequency of address-level events, and relative exposure compared with the broader impact zone.

Data Source

Flight tracks: FlightAware AeroAPI. Noise reference levels: EASA/FAA Part 36 type-certification data. Grid computation runs against locally stored data in the JPXWatch database — it does not make live API calls to FlightAware.

What Is the Noise Exposure Heatmap?

The Noise Exposure layer on the Noise Map page shows estimated cumulative aircraft noise across the KJPX impact zone, visualized as a color overlay on the map. It uses the same reference grid and DNL computation that powers NoiseRank, rendered geographically.

This is different from the Flight Density heatmap, which shows where aircraft fly. The Noise Exposure heatmap shows where noise accumulates on the ground — how sound radiates outward from flight paths and compounds over a season.

Color Scale

The color scale was redesigned in May 2026 to make the spatial gradient across the East End visible. The previous scale used categorical DNL reference levels (45 / 55 / 65 dB) for the fill, which collapsed most of the populated East End into a single tier. More broadly, DNL is the FAA's standard cumulative noise metric, but it was designed for airports with relatively steady year-round operations. At a strongly seasonal airport like KJPX — where summer operations substantially exceed winter — averaging across an entire year understates the seasonal peak when most flights occur. The current visualization reports DNL over the active summer season only, and uses a color scale tuned to where East End exposure actually varies.

The fill is a continuous warm gradient calibrated to roughly 25 to 70 dB DNL, so the spatial pattern across communities reads visibly rather than flattening into a single tier — including the 65 dB cells, which now read distinct from the 55–65 field rather than saturating with it. The reference thresholds remain on the map as labeled contour lines.

Red is reserved for the 65 dB DNL regulatory contour, so it stays visually distinct from the modeled gradient. Street names and place labels render above the overlay to remain legible.

How to Read the Map

Deeper amber means higher modeled seasonal DNL. The warm core follows the northeast–southwest approach and departure corridor through East Hampton Village, Wainscott, and Northwest Woods; communities along that axis read warmer than off-corridor areas at similar distances from KJPX. This corridor-vs-off-corridor pattern, which the previous categorical scale could not resolve, is what the continuous ramp is designed to surface.

Paler yellow covers most of the populated East End, where modeled DNL sits in the 25 to 45 dB range — below the FAA's 65 dB residential-incompatibility threshold, but not zero.

The boundary line (dashed) shows the extent of the computed grid. Areas outside this boundary were not analyzed — not because they have zero noise, but because observed flight tracks do not extend there.

A small inset map shows the airport area in detail, where the 55 and 65 dB DNL contours around the runway sit. The 65 dB residential-incompatibility contour is concentrated at the KJPX airfield; a separate, small 65 dB area also appears at Gabreski Airport, where KJPX–Gabreski shuttle flights operate at low altitude. Because the grid models recorded flight tracks from KJPX operations wherever they fly — not a fixed radius around the airfield — exposure can appear at other airports KJPX aircraft frequently use. The inset preserves the airfield-area detail.

Time Period

The heatmap currently shows data from the most recent completed summer season (Memorial Day through Labor Day). During the live summer season, it updates weekly to show season-to-date cumulative noise exposure. A “Summer 2025 (Historical)” label — or, in-season, a “Season to Date” label — is always displayed on the map.

Flight Track Color Coding

The Routes view on the Community Noise Map color-codes flight tracks by aircraft category. This provides immediate visual differentiation of traffic patterns:

Aircraft-type coloring makes the map immediately informative: helicopter corridors, seaplane routes, and jet approach paths are visually distinct without requiring any interaction.

A compact legend on the map doubles as a toggle control — click any category to show or hide its tracks. This works in conjunction with the aircraft category cards above the map.

Limitations

At a glance

Each scorecard shows an Observed-Operations Score (0–100 — higher scores mean fewer deductions from observed operations) or an honest not-scored label, plus six supporting fields. All fields are derived from FlightAware flight records and JPXWatch's modeled event-level noise estimates, over one fixed trailing 12-month window. Field definitions:

• Score. A number from 0 to 100 summarizing the operator's observable behavior at KJPX. Higher is better. The full formula is published below.
• Flights. The total number of flights by this operator in the window.
• Curfew. The count of flights recorded during the voluntary 10 PM – 7 AM curfew period that do not have an accepted exception.
• Curfew %. The percentage of the operator's flights within the curfew period. Displayed only when the operator has at least 5 flights; otherwise, a dash is shown to avoid misleading percentages from small samples.
• Shoulder-hour operations. The count of flights during the hour before curfew (9–10 PM) and the hour after curfew (7–8 AM). Displayed as context; shoulder-hour operations do not affect the score.
• High-noise. The number of flights whose modeled event-level noise meets or exceeds 85 dB at 1,000 ft reference distance — a JPXWatch-defined modeled high-noise event. This is a project-defined single-event screening level, not an FAA standard, and distinct from the FAA's 65 dB DNL cumulative exposure threshold.
• Event level. The typical modeled noise level for an overflight from this operator, based on FAA/EASA Part 36 aircraft certification data and distance-based propagation assumptions. A modeled comparative value, not a ground-monitor reading.

Who receives a score

Scores attach to demonstrated agency over the flight decision. An operator receives a score only when its flights carry commercial operator attribution in FlightAware data — the entity that scheduled or dispatched the flight — and only when it has at least 12 operations in the trailing 12 months (about one per month; below that, a single event dominates any rate and no score is meaningful; the card reads “Insufficient operations to score”).

Everyone else gets the same statistics with no score:

• FAA-registered owners. Where no commercial operator is attributed, JPXWatch may display the FAA-registered owner. Registry ownership alone cannot distinguish an owner-operator from a passive lessor, so registered-owner cards carry full statistics but are never scored. The registered owner may differ from the person or entity operating a specific flight.
• Unconfirmed identities. Operations that follow an observed seaplane service pattern without a confirmed operator are grouped as “Unknown seaplane operator” and are not scored.
• Privacy-masked identities. Flights attributed through a privacy service that masks the operator at the source are not scored.

Some scored cards display a brand name rather than a legal name (for example, “NetJets”); brand display names correspond to registered operators, and any registry-evidenced consolidations behind a card are documented in JPXWatch's audit record. Scores do not consolidate distinct legal entities under shared ownership: separately registered companies receive separate cards, even where they may be commonly owned.

How the Observed-Operations Score is computed

The score is calculated using a deductible-balance model: every scored operator starts at 100, and points are deducted for observable events in the trailing 12-month window, subject to a limit on how much any single category can reduce the score. Higher scores mean fewer deductions from observed operations. Two categories feed the score: curfew-period operations and fleet noise. Nothing else does.

Curfew-period operations (up to 60 points). A curfew-period operation is a flight recorded during the Town of East Hampton's published voluntary 10 PM – 7 AM curfew window, with no accepted exception on file. Flights with pending or unresolved exception status do not count. The deduction escalates with repetition: a first curfew operation costs the same for every operator. A second costs more. From the third onward, the penalty follows the share of the operator's flights that occur during curfew hours. The exact schedule:

Range boundaries include the lower bound and exclude the upper (a rate of exactly 2% falls in the “2% to under 5%” row).

Fleet noise (up to 25 points). The share of the operator's flights whose modeled event level meets or exceeds JPXWatch's 85 dB threshold — a JPXWatch-defined modeled high-noise event, derived from aircraft type-certification data (see the noise methodology section); it is not an FAA single-event standard:

Fleet choice materially influences the score — up to 25 points of it. Conduct determines the rest. Because the fleet-noise deduction is capped at 25, an operator with no curfew-period operations never scores below 75, no matter what they fly: no operator's score falls to the bottom of the scale on equipment alone.

What no longer feeds the score. Two inputs from the previous formula were removed in the July 2026 revision (see the methodology revisions log):

• Shoulder-hour operations (9 – 10 PM and 7 – 8 AM, the hours adjacent to the curfew) are still counted and displayed, but no longer deduct points. Operating legally adjacent to the curfew is compliance with it, not a violation of it — and recorded times near an hour boundary carry enough imprecision that penalizing the adjacent hours would punish measurement noise.
• Average event-level noise no longer carries its own deduction. The old formula deducted twice for the same underlying fact — a loud fleet raised both the high-noise count and the fleet average. Fleet composition now affects the score in exactly one place: the fleet-noise deduction above. The average event level remains displayed as context.

Reading the scores. A score of 100 means zero deductible events: no curfew-period operations and no flights at or above the 85 dB modeled threshold in the trailing 12 months. Scores cluster at specific values — most single-curfew-operation operators score exactly 85, most loud-fleet operators with clean curfew records score exactly 75 — because operators with identical conduct profiles receive identical scores. That is deliberate: the score has no hidden adjustments, and any reader can recompute any operator's score from the counts shown on their card and the two tables above.

Guarantees the formula is built to keep (each is checkable against the published tables):

• Zero deductible events always scores 100.
• An operator with no curfew-period operations never scores below 75.
• A single curfew-period operation costs exactly 15 points for every operator and can never take a score below 60 on its own.
• A score below 60 requires at least two curfew-period operations.
• Repeated curfew-period operations at a high rate reach the full 60-point deduction regardless of fleet.
• Shoulder-hour operations never affect the score.

Time period

Every scorecard is computed over one fixed window: the trailing 12 months — 365 days — ending on the most recent complete day, Eastern Time. The end date is printed on the card. All fields — Flights, Curfew, Curfew %, Shoulder, High-noise, Event level, and the score itself — are computed over that same window.

The window is not adjustable. The dashboard's date-range selector and fleet tabs do not change how any score is computed; the fleet tabs only filter which operators are listed. One operator has one score, computed over all of their flights at KJPX in the trailing 12 months — which always contains exactly one full summer high season, so operators with different seasonal patterns are compared over like periods.

Scores change over time. Each day, a new day of operations enters the window and the oldest day ages out — so a score can change even on days when the operator didn't fly. Scores can also change when historical flight records are corrected (for example, through data reconciliation or an accepted correction). The most recent day in the window may still be filling in until the morning data update completes. A score is therefore a statement about the trailing 12 months as of the date printed on the card; scores observed on different dates are not directly comparable snapshots, and a screenshot without its end date loses that context.

The dashboard does not retain or publish scores from previous dates. Each score shown is a live computation from the current flight records.

What the score is not

Because these distinctions matter to how the score should be read, they are stated here explicitly:

• The score is not an FAA-issued determination or any form of regulatory classification. The FAA does not issue operator noise scores at KJPX or anywhere else.
• The score is not a safety rating. It measures noise-related operating behavior, not operational safety.
• The score is not a judgment about an operator's intent, professionalism, or character. It is derived from observable events in the FlightAware data.
• The score does not reward or punish flight volume. Deductions are keyed to a flat first-event cost and to the share of an operator's flights involved — not to raw event counts — so an operator with fifty flights and an operator with two thousand can earn the same score when their observable behavior is comparable as a share of their flying.
• The score is not a measure of cumulative community noise exposure, such as DNL. Metrics for community exposure — modeled DNL estimates and relative percentile rank — are presented separately on the DNL map and NoiseRank views.

Data sources

All inputs to the score come from two sources:

• FlightAware AeroAPI (Standard tier) provides flight records, timestamps, operator identification, aircraft types, and track data.
• FAA and EASA Part 36 aircraft type-certification data provide reference noise values per aircraft type, which are combined with flight geometry to produce modeled event-level noise estimates.

No physical noise monitors currently measure sound at KJPX; all noise values are modeled rather than measured. This limitation is inherent to the current dashboard and is acknowledged throughout the site.

Submitting a correction

If an operator believes the dashboard has misattributed a flight, misidentified an aircraft type, or made any factual error affecting their scorecard, they can submit a correction through the “Submit a correction about this operator” link on their scorecard. Corrections are reviewed and resolved publicly through the site's corrections log.

Intellectual honesty requires stating what this dashboard cannot do alongside what it can.

JPXWatch maintains a public record of material corrections to data and methodology changes that affect reported operations. This ensures users can track how reported figures and analytical approaches have evolved over time.

Material revisions to data and methodology — including the April 2026 Noise Impact Score correction, SEVERE tier adjustment, unified noise source-level data, approach-phase offset correction, ADS-B proximity verification of operations, inbound-diversion exclusion, and the April 2026 unit-convention pass (Track A: per-flight value labels, band legend unification, and About-page methodology consistency) — are maintained on a dedicated page: Methodology & Data Revisions. The April 11, 2026 curfew methodology realignment is documented in the Curfew Methodology section above.

JPXWatch is designed to provide accurate, sourced information about airport operations. The following guidance is for anyone who may cite this data in formal settings — town board meetings, press inquiries, regulatory submissions.

Questions about methodology or data should be submitted using the feedback form in the dashboard.