Estimating Boundary Layer Height from LiDAR Data under Complex Atmospheric Conditions Using Machine Learning

Abstract

Reliable estimation of the atmospheric boundary layer height (ABLH) is critical for a range of meteorological applications, including air quality assessment and weather forecasting. Several algorithms have been proposed to detect ABLH from aerosol LiDAR backscatter data. However, most of these focus on cloud-free conditions or use other ancillary instruments due to strong interference from clouds or residual layer aerosols. In this paper, a machine learning method named the Mahalanobis transform K-near-means (MKnm) algorithm is first proposed to derive ABLH under complex atmospheric conditions using only LiDAR-based instruments. It was applied to the micro pulse LiDAR data obtained at the Southern Great Plains site of the Atmospheric Radiation Measurement (ARM) program. The diurnal cycles of ABLH from cloudy weather were detected by using the gradient method (GM), wavelet covariance transform method (WM), K-means, and MKnm. Meanwhile, the ABLH obtained by these four methods under cloud or residual layer conditions based on micropulse LiDAR data were compared with the reference height retrieved from radiosonde data. The results show that MKnm was good at tracking the diurnal variation of ABLH, and the ABLHs obtained by it have remarkable correlation coefficients and smaller mean absolute error and mean deviation with the radiosonde-derived ABLHs than those measured by other three methods. We conclude that MKnm is a promising algorithm to estimate ABLH under cloud or residual layer conditions.

Diurnal and Seasonal Variability of the Atmospheric Boundary-Layer Height in Marseille (France) for Mistral and Sea/Land Breeze Conditions

“…The ability of this algorithm to retrieve the ABLH for both high and low signal-to-noise ratio instruments has already been demonstrated [74]. Another possibility would be to use machine learning [52,75]. In [75], the authors developed an algorithm based on this method to retrieve the ABLH in complex stratification and cloudy situations and compared it with other methods such as the Haar wavelet method.…”

“…Another possibility would be to use machine learning [52,75]. In [75], the authors developed an algorithm based on this method to retrieve the ABLH in complex stratification and cloudy situations and compared it with other methods such as the Haar wavelet method. They showed that machine learning presented the best results of all the methods tested, which makes it promising.…”

Midday Boundary-Layer Collapse in the Altiplano Desert: The Combined Effect of Advection and Subsidence

“…Vertical profiles of T , P, R H, U and W D were obtained to characterize h, the jumps at the top of the boundary layer ( θ , q) and lapse rates above the ABL (γ θ , UAV flights were performed every 30 min from 0900 to 1200 LT and up to 500 m above ground level (a.g.l) during the same day. The observed boundary-layer heights were estimated from these observations using the gradient method (Liu et al 2022). The method consists of analyzing the gradient of the potential temperature profile, which is virtually zero in the mixed layer but shows a strong θ gradient at the top of the ABL due to the temperature jump in the transition zone to the free atmosphere (Marques et al 2018;Lothon et al 2009).…”

“…Fig.15ABL evolution estimated from the CLASS model experiments defined in Table1, compared to drone and radiosonde observations estimated using the gradient method(Liu et al 2022), and the WRF model results estimated using the bulk Richardson number method(Min et al 2020). The vertical dashed line indicates the time when the regional air mass arrives at the Salar del Huasco.…”

Atmospheric Boundary Layer Height: Inter-Comparison of Different Estimation Approaches Using the Raman Lidar as Benchmark

“…The determination of the ABLH in cloudy conditions is characterized by possible interferences from cloud layers [22]. Machine learning methods have been proposed to estimate ABLH from elastic lidar data under complex weather conditions [23], with results revealing a sensitive reduction in the potential biases affecting ABLH estimates [24]. In general, the combined use of more sensors and approaches is preferred, especially in unstable weather conditions [25].…”