Understanding the warm core and circulation structure in the inner-core region of a tropical cyclone (TC) is crucial for predicting intensity changes. Recent aircraft observations of typhoons have advanced our knowledge of vertical structure through multi-point dropsonde deployments. However, because dropsonde observations are horizontally discrete, they are limited in their ability to resolve fine horizontal gradients of the warm core and the continuous azimuthal asymmetry of the wind field in a single case. In contrast, flight-level in situ measurements recorded onboard research aircraft can complement this limitation by providing continuous, high-temporal-resolution observations along the flight path. Under typhoon conditions, however, such measurements can be contaminated by errors associated with aircraft attitude changes, maneuvering, and airframe icing, making rigorous quality control (QC) indispensable.

In this study, focusing on Typhoon Nanmadol (2022; Typhoon No. 14) observed during the T-PARCII campaign, we developed a systematic QC framework for flight-level data and, together with dropsondes released during the same flight, used these observations to describe the horizontal structure of the upper-level warm core and wind field at an altitude of approximately 14.5 km (145–150 hPa).

The analysis targeted 16 September (developing stage) and 17 September (mature stage). To reduce uncertainty in the storm-center position, we first estimated the center using the brightness-temperature centroid derived from Himawari-8 infrared imagery and established a storm-relative coordinate system. We then applied QC to the flight-level data. For temperature, physical consistency between total air temperature and true airspeed was examined, and a time interval on 17 September exhibiting degraded consistency was identified as anomalous and excluded. For wind, segments affected by aircraft maneuvering were identified using roll and heading rate diagnostics, and a risk assessment based on high-frequency variability was applied to suppress spurious fluctuations.

The results indicate that the temperature structure on 16 September was relatively axisymmetric, whereas on 17 September pronounced warming near the storm center was observed together with a distinct asymmetric structure characterized by a positive temperature anomaly in the north–northwest (N–NW) quadrant. In the wind field, the maximum tangential wind speed increased from 28.0 m s⁻¹ to 39.6 m s⁻¹, while the radius of maximum tangential wind (RMW) contracted from 44 km to 24 km, indicating intensification and contraction of the inner core during the mature stage.

Environmental analysis based on GSM objective analysis data showed that the environmental vertical wind shear strengthened from 16 to 17 September and was directed toward the south-southeast. Accordingly, the warm anomaly observed in the N–NW quadrant on 17 September can be interpreted as an upshear warm anomaly that arises when a vortex tilts under vertical wind shear, consistent with the expected dynamical response of a sheared vortex. Consistent with this interpretation, Himawari-8 cloud-top height estimates reveal deeper convection on the shear-left side and cloud tops approximately 1 km lower on the shear-right side, indicating pronounced azimuthal asymmetry in convective structure under vertical shear.

In addition, previous studies have reported that inflow layers can appear above and below the primary outflow layer in the upper troposphere. Therefore, for aircraft observations conducted at a fixed flight altitude, quadrant-dependent differences in convective depth can lead to relative vertical displacements of the outflow layer. As a result, inflow was observed on the shear-left side, while outflow dominated on the shear-right side at the flight level. This interpretation explains the observed quadrant-dependent radial flow pattern on 17 September, with inflow to the east of the storm center and outflow to the west.

This study demonstrates that flight-level data subjected to appropriate and systematic quality control are highly effective for resolving inter-quadrant contrasts and steep horizontal gradients near the storm center that cannot be fully captured by dropsonde sampling alone. The results expand the observational utility of flight-level data in typhoon aircraft campaigns and contribute to a deeper understanding of structural changes in the upper troposphere associated with tropical cyclone intensification and vertical wind shear.