Given the distribution of aerosol particles and moisture is affected by mixing processes amongst others, attenuated backscatter profiles can be exploited to help track the recent history of ABL dynamics. Layer boundaries can be detected if aerosol properties differ between the atmospheric layers examined. The most pronounced layer edge is usually the ABLH because aerosol concentrations and humidity tend to be significantly higher in the ABL than in the air above. But also within the ABL, mixing dynamics and advection can lead to contrasting aerosol properties between different layers so that often the height of the mixed layer (MLH) can be determined based on aerosol backscatter profile data.The performance of aerosol-based layer height retrieval algorithms significantly depends on the quality of the attenuated backscatter analysed. To exploit the signals from different ALC models most effectively, the ABL testbed applies instrument-specific corrections and applies algorithms tailored to the respective measurement capabilities. STRATfinder for ALC data with relatively high signal-to-noise ratio and CABAM for data with relatively low signal-to-noise ratio, respectively.
Corrections
To account for instrument-specific artefacts, the following preprocessing procedures are implemented:
- Lufft CHM15k: To account for the temperature-dependance of the optical overlap, a dynamic overlap model is derived for every sensor and laser module following the procedure of Hervo et al. (2016).
- Vaisala CL31 and CL51: Near range artefacts and instrument-related background are corrected for following the method outlined by Kotthaus et al (2016). In addition, vertical and temporal smoothing and a subsequent SNR filtering is applied Kotthaus et al (2016).
STRATfinder
STRATfinder (Kotthaus et al. 2020) tracks MLH and ABLH, providing a full picture of the boundary layer structure over a 24 h period at 1 min resolution. Layers are traced using the graph-based ‘pathfinder’ (deBruine et al. 2017) approach which determines the optimised path between two points in a two-dimensional field of weights by minimising a cost function. Regions with low weights are more likely to guide that path. In STRATfinder, weights fields are defined based on the vertical gradient in attenuated backscatter, with low values where strong negative gradients occur. In addition, the high SNR allows for the temporal variance introduced by entrainment at the top of the mixed layer to be exploited. As variance is particularly strong at the top of the mixed layer during the morning growth and during daytime when air from the residual layer or the free troposphere is being mixed into the lowest atmospheric layer, a second field of weights is defined by the inverse of the attenuated backscatter variance. The search regions for ABLH and MLH are restricted by pre-defined limits.
STRATfinder is built on the heritage of STRAT-2D/STRAT+ (Morille et al. 2007, Haeffelin et al. 2012, Pal et al. 2013) and pathfinderTURB software (Poltera et al. 2017). Compared to STRAT-2D/STRAT+, STRATfinder’s main advantages are the implementation of the pathfinding algorithm and the definition of reasonable boundaries for the search region. The main advantage of STRATfinder compared to pathfinderTURB is its simpler, more user-friendly setup and its applicability to a 24 h period (pathfinderTURB only has daytime MLH detection capability).
STRATfinder is developed by Simone Kotthaus (Institut Pierre Simon Laplace, Centre National de la Recherche Scientifique) with contributions by Melania Van Hove (IPSL), Marc-Antione Drouin (LMD-IPSL), Martial Haeffelin (IPSL), Maxime Hervo (MeteoSwiss), and Alexander Haefele (MeteoSwiss).
STRATfinder development received funding support from IPSL, MeteoSwiss, ACTRIS, ICOS, and E-PROFILE.
STRATfinder is written in MATLAB (tested with version 2017b) and licensed under the GNU General Public License v3.0. The code is accessible on gitlab.
CABAM
The CABAM algorithm (Kotthaus and Grimmond 2018, Kotthaus et al. 2020), tracks the MLH and additional layers forming the residual layer, while taking into account the presence of clouds. Following the detection of significant vertical gradients in the attenuated backscatter profiles, layers are connected using a dynamic decision-tree with rules varying through the day. Periods adversely affected by precipitation are flagged automatically. Automatic layer detection and attribution are performed at the raw resolution of the data, and final results are block-averaged to 15 min (time ending). MLH results were successfully evaluated against independent reference measurements (Kotthaus and Grimmond 2018) and compared to mixing heights derived from Doppler lidar profiles (Kotthaus et al. 2018).
CABAM has been developed by Simone Kotthaus and Sue Grimmond at the Department of Meteorology, University of Reading, UK. Further improvements are made by Simone Kotthaus, Melania Van Hove, Martial Haeffelin, and Marc-Antoine Drouin at Institut Pierre Simon Laplace (IPSL), Ecole Polytechnique, Centre National de la Recherche Scientifique (CNRS), France.
CABAM is written in R (tested with version 3.4.4) and licensed under the GNU General Public Licence v3.0. The code is accessible on gitlab.
Quality control
While both CABAM and STRATfinder include some internal quality control filters, e.g. to account for complex conditions during precipitation, an additional automatic quality control procedure is developed to remove physically unreasonable layer results. The quality control is implemented to remove MLH that reveal a clear under- or overestimation or show unreasonable temporal inconsistencies (e.g. during rainfall periods). MLH is removed if values experience a very strong increase (> 300 m within 15 minutes) at night. Overestimation may also occur if shallow layers were not detected and the residual layer is mistaken for the mixed layer at night. To filter out those faulty results, absolute thresholds are applied which vary by season.
Underestimation of MLH occurs for CABAM results derived from observations obtained with the Vaisala CL31 or CL51, mostly during daytime, buoyant conditions. Different instrument-related artefacts result in erroneous gradients in the fields of attenuated backscatter that can confuse the detection algorithm at times. Such faulty results are removed based on criteria that assess the temporal variability of the detected MLH through the course of the day and also its relation to the cloud base height, in case boundary layer clouds are present. STRATfinder results obtained from CHM15k observations are also filtered out if temporal variations appear physically unreasonable.