United Airlines Flight 1890

United Airlines flight 1890 (UAL1890) encountered turbulence during descent into Newark Liberty International Airport (EWR) which resulted in three flight attendants receiving serious injuries. The encounter occurred within a localized region of moderatetosevere turbulence near flight level (FL) 215 (21,500 feet). The flight was conducted under the provisions of Title 14 Code of Federal Regulations Part 121 as a scheduled domestic passenger flight from Los Angeles International Airport (LAX), Los Angeles, California to EWR.

The flight was in instrument meteorological conditions (IMC) and descending along an air traffic control (ATC) assigned route, toward the Hancock (HNK) VOR. The turbulence encounter resulted in significant cabin vertical accelerations, with quick access recorder (QAR) data showing fluctuations between -0.55 g and +1.82 g over a period of several seconds.

Although the flight crew conducted standard preflight turbulence assessments using Weather Services International (WSI) and SkyPath tools, these products depicted only light to moderate turbulence along the planned route. The dispatcher similarly reported only one moderate turbulence report, from a regional jet on descent into EWR, with no updated concerns during the subsequent hours of the flight.

Throughout cruise and initial descent, the flight crew continued monitoring WSI and SkyPath, and neither indicated significant turbulence activity along their descent path. The captain ensured that the onboard radar was active for descent; however, no returns indicating convective or turbulent weather were noted by the flight crew.

ATC did not provide any real-time weather advisories during the period the airplane was under Boston Center’s control, despite receiving a moderate turbulence report from ACA548 (an uninvolved airplane) about three minutes prior to the encounter. ACA548 was operating ahead of UAL1890, on the same route and reported moderate turbulence to the Boston Air Route Traffic Control Center (ZBW ARTCC), R24C controller. The controller did not acknowledge or disseminate this report to other aircraft, including UAL1890, as required by Federal Aviation Administration (FAA) procedures for Pilot Weather Reports (PIREPs) solicitation and dissemination. The controller stated during their interview they did not hear the PIREP. UAL1890 subsequently encountered turbulence in approximately the same area. A turbulence PIREP was transmitted to the flight at 1541 eastern standard time (EST), although it could not be conclusively determined if the crew saw or read the message.

Meteorological analysis revealed that there was a cold front in the vicinity of the accident site along with a relatively strong jet stream of 150 knots near 30,000 ft. The cold front and associated upper level trough provided the support for the formation of a band of rain showers and embedded thunderstorms in western and central New York that was moving eastward with time. The tops of the convection ranged from 21,000 to 23,000 ft in the vicinity of the accident site based on the Geostationary Operational Environmental Satellite number 16 (GOES-16) imagery and pilot statements. Lightning was noted in the convective line encountered by the accident flight in the GOES-16 lightning flash imagery about 20 to 30 minutes prior to the encounter. Figure 1 presents the GOES-16 visible image at 1546 EST at 3X magnification with the accident site highlighted with a red square. There was cloud cover indicated above the accident site with the cloud cover moving from west to east. Figure 2 presents the GOES-16 infrared image at 1546 EST (closest image to the accident time) at 8X magnification with the accident site highlighted with a red square. Cloud cover was indicated above the accident site with lower brightness temperatures (green colors; higher cloud tops) located above and to the northeast through south of the accident site.

Figure 1. GOES-16 visible image at 1546 EST at 3X magnification.

Figure 2. GOES-16 infrared image at 1546 EST at 8X magnification.

In addition, the High-Resolution Rapid Refresh (HRRR) model soundings showed a large change of wind speed between 17,000 ft (around a 45-knot wind) and 21,000 ft (around 110-knot wind) and this caused wind shear and turbulence at the tops of the rain shower/thunderstorm activity. The sounding analysis also indicated moderate or greater clear-air turbulence (CAT) between 18,000 and 21,000 ft near the time and location of the encounter. The turbulence at the tops of the cloud cover is visible on the GOES-16 imagery as evidence by the transverse banding.

UAL1890 checked in with the R24A controller at 1535:10, while level at FL270, but did not receive an immediate response because the controller was occupied with other coordination tasks. As the airplane continued descending under successive clearances, ATC did not pass along the moderate turbulence PIREP issued by ACA548 at 1540:38. This omission reduced the crew’s situational awareness regarding developing turbulence in the same airspace.

When UAL1890 encountered turbulence descending through FL210, the airplane was already committed to the assigned crossing restriction near the HNK VOR. The crew reported the encounter as “severely moderate turbulence,” and ATC also did not respond to that advisory. Shortly thereafter, when handed to the R05 controller, the facility advised of only “light to moderate turbulence,” further indicating that ATC personnel had not fully integrated or disseminated the earlier PIREP information.

QAR data show that autopilot and autothrottle remained engaged throughout the encounter, and the airplane remained within design structural limits. As shown in figure 3, the turbulence generated rapid vertical accelerations and short-duration negative-g forces sufficient to lift unsecured crewmembers from the cabin floor. At the time of the event, the captain had recently re-illuminated the seatbelt sign and instructed the flight attendants to prepare the cabin for landing, a phase during which some flight attendants were completing final safety tasks and not seated. As a result, three crew members were injured. Figure 3. Time history plot presenting accelerations and attitude control parameters from 1535:00 EST to 1607:00 EST, including activity observed near 1543 EST. A peak EDR value of 0.431 was noted as the accident aircraft was in IMC at 1543 EST while descending through both the convective updrafts in the rain shower and thunderstorms and the wind shear layer in the jet stream. This value corresponds to moderate to severe turbulence for a “heavy” class of aircraft.

The evidence indicates the flight crew performed standard preflight and in-flight turbulence risk mitigation, used available tools, and operated the airplane within procedures. The turbulence encountered was not forecast to be significant at the precise altitude, time, and location of the airplane by the operator’s weather provider. Further, the most relevant real-time warning—the moderate turbulence PIREP issued minutes prior—was not disseminated to the flight crew by ATC in accordance with FAA requirements.

The location of the three seriously injured flight attendants (FAs) were two positioned in the aft galley area and one located in the mid-cabin aisle near door 3. FA 2 was lifted into the ceiling and forcefully thrown to the cabin floor, resulting in a spinal fracture. FA 3 was thrown to the ceiling, lost consciousness, struck a beverage cart, and landed heavily on the left leg, sustaining a shattered femur, crushed spinal vertebrae, and a head injury with bleeding. FA 9 was thrown upward and then into galley structures, lost consciousness, and sustained facial lacerations and a diagnosed intracranial bleed.

Verbatim NTSB analysis and probable cause from the NTSB dataset