Tracking ABL height with VHF radar: Diurnal and seasonal variability with its link to convection

The atmospheric boundary layer (ABL) is primarily characterised by its diurnal variability, which is crucial in exchanging heat, momentum, moisture, and chemical components between the surface and the free atmosphere. Despite its significance, studies on the evolution of the ABL height with sufficiently high temporal precision are quite few in the tropical region. Our study paves the way to utilise radar data to explore high temporal variability, which is not feasible with the limited availability of conventional radiosonde profiles. This study employs very high frequency (VHF) radar to examine the diurnal and seasonal variability in ABLH using a novel method known as the Normalised Standard Deviation Method (NSDM) spanning over seven years from 2018 to 2024 at a tropical coastal station, Cochin (10.04° N, 76.33° E) in India. This method identifies the intensity of turbulence within the ABL using the profiles of signal-to-noise ratio, wind variability and spectral width. To guarantee accurate profiling, clear-sky and partly clear-sky days were chosen, thereby removing contamination due to rain. Depending on data availability, the dataset includes both continuous profiles of varying durations (24- h or less) and time specific snapshots at selected synoptic hours (06:00, 09:00 and 12:00 UTC), covering over 800 days of observations. Radar-based measurements clearly demonstrate ABLH changes during the day, showing significant temporal variability, as it is influenced by factors such as topography, solar radiation, surface roughness and other surface forcings. Although no discernible seasonal shift is observed in the timing of maximum ABLH, pronounced differences occur in its magnitude. Spectral width values are low at night, indicating weak turbulence within the stable boundary layer and show increased values during the day as convection develops. The seasonal dependence is distinct: boundary layer development is deepest during the pre-monsoon season, with mean peak values of 1.6 ± 0.6 km, followed by winter (1.45 ± 0.2 km), post-monsoon (1.3 ± 0.4 km) and monsoon with the shallowest ABLH (1.2 ± 0.1 km) during clear sky conditions. The study further makes a comparison of the ABLHs derived from reanalysis data such as MERRA2, IMDAA and ERA5. This comparison is aimed at validating the reliability and consistency of the observed values in different atmospheric conditions. Additionally, surface density shows a strong and consistent inverse relationship with ABLH, making it a useful proxy when profile observations are unavailable. Further, the application of machine learning (ML) methods demonstrates that non-linear models outperform linear methods in predicting ABLH. Our results emphasise the robustness of radar-based ABLH estimates, establishing their suitability for regional-scale investigations of boundary layer processes. Given their relatively high temporal resolution, radar observations serve as a reliable benchmark for validating model outputs and reanalysis products. A key strength of this study is its ability to capture vertical reflectivity and turbulence structures, significantly improving its usefulness for convection studies. However, challenges persist under overcast or rainy conditions, which require further investigation.

Keywords: VHF Radar; Radiosonde; ABLH; Diurnal variability; Seasonal evolution