학술논문

An Automated Common Algorithm for Planetary Boundary Layer Retrievals Using Aerosol Lidars in Support of the U.S. EPA Photochemical Assessment Monitoring Stations Program.
Document Type
Article
Source
Journal of Atmospheric & Oceanic Technology. Oct2020, Vol. 37 Issue 10, p1847-1864. 18p.
Subject
*ATMOSPHERIC boundary layer
*ALGORITHMS
*MIXING height (Atmospheric chemistry)
*AEROSOLS
*AIR quality
*ENVIRONMENTAL protection
Language
ISSN
0739-0572
Abstract
A unique automated planetary boundary layer (PBL) retrieval algorithm is proposed as a common crossplatform method for use with commercially available ceilometers for implementation under the redesigned U.S. Environmental Protection Agency Photochemical Assessment Monitoring Stations program. This algorithm addresses instrument signal quality and screens for precipitation and cloud layers before the implementation of the retrieval method using the Haar wavelet covariance transform. Layer attribution for the PBL height is supported with the use of continuation and time-tracking parameters, and uncertainties are calculated for individual PBL height retrievals. Commercial ceilometer retrievals are tested against radiosonde PBL height and cloud-base height during morning and late-afternoon transition times, critical to air quality model prediction and when retrieval algorithms struggle to identify PBL heights. A total of 58 radiosonde profiles were used, and retrievals for nocturnal stable layers, residual layers, and mixing layers were assessed. Overall good agreement was found for all comparisons, with one system showing limitations for the cases of nighttime surface stable layers and daytime mixing layer. It is recommended that nighttime shallow stable-layer retrievals be performed with a recommended minimum height or with additional verification. Retrievals of residual-layer heights and mixing-layer comparisons revealed overall good correlations with radiosonde heights (square of correlation coefficients r2 ranging from 0.89 to 0.96, and bias ranging from approximately 2131 to 163m for the residual layer and r2 from 0.88 to 0.97 and bias from 2119 to 1101m for the mixing layer). [ABSTRACT FROM AUTHOR]