Observations of Nocturnal Temperature Inversions During Dust Events in Beijing Using Multi-Mode Lidar

Objective Dust storms frequently influence northern China in spring and can perturb the vertical thermodynamic structure through radiative and dynamical processes. Nocturnal temperature inversions are commonly observed during dust episodes, yet their formation and persistence mechanisms in downstream urban receptor regions remain insufficiently constrained by observations. This is because conventional temperature measurements often lack the temporal and vertical resolution needed to resolve the co-evolution of dust-layer structure and temperature stratification. Beijing represents a typical receptor environment where dust loading is generally lower than in source regions and synoptic disturbances occur frequently. Under these conditions, the relative roles of dust-layer radiative effects, vertical mixing, and synoptic forcing in shaping nocturnal inversions can differ markedly from those in desert source areas. This work aims to provide observational evidence and a process-oriented interpretation for nocturnal inversion formation during springtime dust events in Beijing by jointly tracking aerosol vertical structure and temperature profiles at high resolution and by comparing multiple events that exhibit distinct dust-layer stability and inversion behavior. Methods A coordinated multi-mode lidar observation strategy was implemented at the Beijing Institute of Technology to obtain simultaneous vertical information on aerosols and temperature during spring 2025 dust events. A Raman-Mie lidar operating at 354.7nm was used to retrieve nighttime temperature profiles from pure rotational Raman signals by employing two rotational Raman channels that sample high and low rotational quantum-number populations. To ensure temperature retrieval capability under strong aerosol extinction, the optical system adopted narrowband filtering and strong out-of-band suppression to mitigate elastic-scattering leakage into the rotational Raman channels, and enhanced transmitter power was used to improve signal levels. Raw signals were preprocessed through background subtraction, temporal averaging, vertical accumulation and smoothing, and quality control to remove low-signal-to-noise-ratio and cloud-contaminated data. Temperature uncertainty was quantified using signal-to-noise-ratio-based error propagation. Independent validation was performed by comparison with radiosonde observations and satellite temperature products during nighttime overpasses to assess retrieval reliability beyond calibration times. To characterize dust layers and aerosol type, a fluorescence-Raman-Mie polarization lidar provided particle depolarization ratio and fluorescence capacity, enabling aerosol classification in a two-parameter space and supporting identification of dust-dominated layers and their geometric boundaries. Dust loading was additionally represented by aerosol optical depth integrated over the identified dust-layer depth. Auxiliary datasets included near-surface particulate matter concentrations for event detection and classification, ERA5 reanalysis for synoptic circulation and dynamical diagnostics, and HYSPLIT 72-hour backward trajectories initialized at multiple heights to infer transport pathways. Frontal activity was objectively identified from reanalysis temperature fields with morphological constraints. The gradient Richardson number was used to evaluate dynamical stability and the potential for turbulence and vertical mixing, thereby linking dust-layer top stability to the nocturnal mixing environment. Four representative dust events were selected based on particulate matter thresholds, dust dominance confirmed by lidar classification, dust-layer thickness requirements, and sufficient continuous temperature observations to allow robust comparisons among cases. Results and Discussions The coordinated lidar observations show that nocturnal inversions during spring dust events in Beijing are mainly elevated features in the free troposphere rather than shallow near-surface inversions. Across the investigated period, the diagnosed boundary layer height generally remained below about 1.5km, while inversion structures, when present, predominantly occurred above 2km and most frequently within 2.0 - 4.0km. In events with sustained inversions, the inversion-base height exhibited a clear temporal correspondence with the dust-layer top height, indicating that the inversion altitude is not random but is closely linked to the upper boundary of the dust layer. This coupling was most evident when the dust-layer top remained relatively stable in height. Notably, the event without an observed inversion showed a continuously descending dust-layer top together with a less stable layer-top structure, which likely prevented the persistence of a stability maximum at a fixed altitude. Dynamical diagnostics further support that weak vertical mixing is a necessary background condition for inversion development. Periods with enhanced dynamical stability near the dust-layer top were associated with reduced turbulent exchange, allowing temperature gradients to build and persist near the layer boundary. Synoptic forcing modulated the inversion behavior among cases. One event developed a pronounced inversion without an overlapping mid-level frontal signal, suggesting that dust-layer structure combined with weak mixing can be sufficient. In contrast, other events showed frontal stratification overlapping the inversion altitude, consistent with a scenario in which frontal thermodynamic structure can precondition or strengthen the inversion and dust-related effects can act as an additional enhancer. Overall, the multi-case comparison indicates that dust-layer top stability, rather than dust loading alone, plays a primary role in controlling whether an elevated nocturnal inversion forms and persists in Beijing during spring dust episodes. Conclusions This study demonstrates that coordinated multi-mode lidar observations can resolve the coupled evolution of dust-layer vertical structure and nighttime temperature profiles in Beijing under dust conditions. Elevated nocturnal inversions were mainly observed in the 2.0 - 4.0km altitude range and were clearly separated from the boundary layer, emphasizing their free-tropospheric nature. Sustained inversions were systematically linked to stable dust-layer top heights, and the inversion-base height exhibited strong temporal correspondence with the dust-layer top height, indicating tight vertical coupling between aerosol-layer structure and thermodynamic stability. Dynamical analyses show that weak vertical mixing is conducive to dust-layer top stabilization and supports inversion growth and persistence. Frontal activity can further modulate inversion strength when it overlaps the relevant altitude range, though it is not required for inversion formation. The results collectively suggest a coupled mechanism for Beijing's springtime receptor environment in which weak mixing stabilizes the dust-layer top, persistent radiative cooling near the layer top strengthens the temperature gradient to form an inversion, and the established inversion further suppresses mixing to maintain stability. These findings provide observation-based constraints for understanding nocturnal thermodynamic structure evolution during dust episodes in Beijing and offer guidance for future studies that aim to quantify radiative contributions and improve high-resolution modeling of dust-meteorology interactions.