A United Airlines Boeing 737-900 flying from San Francisco to Newark encountered turbulence during descent on February 10, 2024, injuring a flight attendant who was still on their feet in the cabin. The National Transportation Safety Board classified the event as an accident and published a final investigation report detailing the forces involved, the evidence collected, and the circumstances that left crew members vulnerable. The case puts a spotlight on a persistent gap in aviation safety: cabin crew face a sharply higher risk of turbulence injuries than seated, belted passengers, and the tools available to predict invisible rough air remain limited.
What Happened on Flight 1890
United Airlines Flight 1890, designated UAL1890, was on approach to Newark Liberty International Airport when the aircraft struck severe turbulence. The NTSB’s final report confirmed that flight attendants aboard the Boeing 737-900 sustained serious injuries during the encounter. Because the injuries met the federal threshold for “serious,” the NTSB treated the turbulence event as an accident rather than a routine incident, triggering a full investigation with multiple evidence streams.
Investigators drew on the aircraft’s digital flight data recorder, cockpit voice recorder, air traffic control communications, and pilot reports filed by other aircraft in the area. That combination of data sources allowed the board to reconstruct the sequence of events with high fidelity, from the crew’s last routine cabin activity through the turbulence encounter and the immediate aftermath. The breadth of evidence is typical of NTSB turbulence probes, where the goal is to determine not just what happened but what information was available to the crew and how the encounter unfolded.
Why Cabin Crew Bear the Brunt
Turbulence injuries are statistically rare relative to the number of flights operated each year, yet they fall disproportionately on flight attendants. The reason is straightforward: passengers spend most of a flight seated and belted, while cabin crew routinely stand, walk the aisle, and handle service carts during phases of flight when turbulence can strike without warning. The Associated Press, drawing on NTSB data, has noted that cabin crew are overrepresented among those hurt in turbulence over multi-year tracking periods.
That pattern challenges a common assumption in post-incident coverage, which tends to focus on passenger experience. In practice, the people most at risk are the ones whose job requires them to be unrestrained during exactly the flight phases when sudden jolts are hardest to anticipate. Airlines have bracing protocols and seatbelt-sign policies, but those depend on pilots receiving timely information about rough air ahead, and that information is not always available.
Clear-Air Turbulence and the Forecasting Gap
The NTSB distinguishes between convective turbulence, which is associated with visible storm clouds and can often be detected by onboard weather radar, and clear-air turbulence, which occurs in cloudless skies and is invisible to standard radar equipment. Clear-air turbulence is generated by wind shear along jet stream boundaries and other atmospheric dynamics that leave no visual or radar signature for pilots to avoid. This distinction matters because it determines whether a crew had any realistic chance of advance warning.
Pilot reports, known as PIREPs, are one of the primary tools for flagging turbulence along a route. When one aircraft hits rough air, the crew radios a report that can be relayed to following traffic. But PIREPs are reactive by nature: they describe turbulence that has already been encountered, and they depend on voluntary reporting. If no aircraft has recently flown through a turbulent patch, subsequent flights enter it blind. The NTSB investigation of Flight 1890 included PIREP data as part of its evidence base, suggesting investigators examined whether adequate reports were available to the crew before descent.
How the NTSB Builds a Turbulence Case
The board’s formal investigation process for turbulence accidents follows the same general framework used for crashes and mechanical failures. Investigators secure the flight data recorder and cockpit voice recorder, obtain air traffic control audio and radar tracks, and collect weather data and NOTAMs (Notices to Air Missions) that were active at the time. For turbulence cases specifically, the flight data recorder is critical because it captures vertical acceleration values that define whether turbulence was light, moderate, severe, or extreme.
The NTSB’s public docket system makes many of these primary materials available once an investigation is complete, including factual reports, specialist analyses, and air traffic control packages. The board also maintains an online library of accident reports that allows researchers and journalists to compare turbulence events across years, operators, and aircraft types. These records form the backbone of trend analysis on turbulence injuries and help identify recurring vulnerabilities, such as cabin-service timing or communication gaps between cockpit and cabin.
Safety Data Systems and Their Limits
Beyond individual investigations, the Federal Aviation Administration operates the Aviation Safety Information Analysis and Sharing program, known as ASIAS, which collects and analyzes safety data from across the U.S. aviation system. ASIAS incorporates voluntary safety reporting and flight data programs that can capture turbulence encounters even when they do not result in injuries or formal accident classifications. The program is designed to detect patterns early, so regulators and industry can act before a cluster of minor events escalates into a serious accident.
A significant limitation, however, is that many of the data streams feeding ASIAS are protected and de-identified to encourage candid reporting by airlines and crew members. That protection serves a valid purpose: without confidentiality guarantees, carriers and pilots might underreport incidents to avoid regulatory scrutiny or reputational harm. But it also means the public and independent researchers cannot easily access granular turbulence injury data tied to specific flights, routes, or operators, making it harder to evaluate how well specific safety interventions are working.
The FAA also runs the Service Difficulty Reporting System, and its public updates offer a window into certain mechanical and operational issues that airlines flag to regulators. While this database is not focused on turbulence per se, it illustrates how targeted reporting can surface recurring safety concerns. By contrast, turbulence-related injuries to cabin crew often appear only in de-identified datasets or in NTSB accident files when they cross the “serious” threshold, leaving a gray zone of lesser injuries that are harder to track systematically.
Protecting the People on Their Feet
The Flight 1890 accident underscores a basic asymmetry: the people tasked with keeping passengers safe are themselves exposed to the greatest risk when the cabin suddenly lurches. In response, airlines may refine procedures for suspending service earlier in anticipated rough air, reinforcing the authority of flight attendants to sit down and secure the cabin when they judge it necessary. Some carriers have also adjusted training to emphasize the dangers of clear-air turbulence, which can strike even on seemingly smooth days when passengers may question why the seatbelt sign remains on.
Yet procedural changes can only go so far without better forecasting tools. Efforts to integrate real-time turbulence measurements from aircraft sensors, combined with more sophisticated atmospheric models, aim to give dispatchers and pilots more precise guidance on where and when to expect rough air. Until those systems are more mature and widely deployed, accidents like the one that injured the United Airlines crew will continue to test the balance between maintaining service and prioritizing crew safety.
For regulators and airlines, the lesson from Flight 1890 is less about a single descent into Newark than about a systemic exposure. Turbulence will remain an unavoidable feature of high-altitude flight, but the burden does not have to fall so heavily on the people walking the aisle. Strengthening data-sharing frameworks, refining cabin procedures, and investing in better prediction tools all point toward the same goal: ensuring that when the air turns rough without warning, the crew members responsible for everyone else’s safety are not the ones paying the highest price.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
