The mixing layer height (MLH) is a key parameter for diagnosing turbulent mixing and pollutant dispersion within the atmospheric boundary layer; however, its estimated values can vary substantially depending on the applied methodology and observationa...
The mixing layer height (MLH) is a key parameter for diagnosing turbulent mixing and pollutant dispersion within the atmospheric boundary layer; however, its estimated values can vary substantially depending on the applied methodology and observational data. In this study, 53 high-frequency radiosonde soundings conducted over Seosan, Korea, from 23 to 25 January 2025, together with co-located continuous ceilometer observations, were used to systematically compare and evaluate 12 MLH retrieval methods under identical observational conditions.
The analyzed methods include thermodynamic radiosonde-based approaches (BRM1, BRM2, PM1, PM2, Heffter, and Liu–Liang), gradient-based methods (PTg, RHg, qg, and Ng), a ceilometer-based MLH (QC_MLH), and an ERA5 reanalysis-derived MLH. The analysis was stratified by time of day (daytime and nighttime), boundary-layer regime (stable, convective, and neutral boundary layers), and PM₂.₅ and PM₁₀ concentration levels. To focus on physically representative mixing layers, cases dominated by aerosol multiple layers or residual layers were excluded, and outliers exceeding ±1000 m relative to BRM1 were removed, resulting in 32 valid cases for the final analysis.
The thermodynamic MLH methods exhibited small biases and errors relative to BRM1, with bias, MAE, and RMSE generally within ±50–150 m, and showed strong - 64 correlations (R > 0.9) under daytime convective boundary-layer conditions. In contrast, the gradient-based methods showed large and inconsistent over- or underestimations, particularly in the presence of multi-layer aerosol structures. The ceilometer-based QC_MLH showed weaker direct agreement with ations with PM ₂.₅ and PM₁₀ concentrations, highlighting its sensitivity to aerosol dispersion depth rather than thermodynamic mixing.