| |
 |
UNIVERSITY OF BUCHAREST
FACULTY OF PHYSICS Guest
2026-06-11 23:58 |
 |
|
|
|
Conference: Bucharest University Faculty of Physics 2026 Meeting
Section: Atmosphere and Earth Science; Environment Protection
Title:
Detection and climatological analysis of stratocumulus clouds in Romania using Cloudnet data
Authors:
Genica–Liliana SĂFTOIU (GOLEA) (1,2)
*
Affiliation:
(1) University of Bucharest, Faculty of Physics, PO Box MG-11, 077125, Bucharest, Romania
(2) “Horia Hulubei” National Institute for Nuclear Physics and Engineering, IFIN-HH, Reactorului 30, 077125, Bucharest-Magurele, Romania
E-mail
liliana.golea@nipne.ro
Keywords:
stratocumulus, Cloudnet, drizzle, cloud microphysics, machine learning
Abstract:
Stratocumulus clouds represent an important type of low clouds in atmospheric physics and climate studies, due to their significant influence on the Earth’s radiation balance, boundary-layer processes and cloud–aerosol interactions. The aim of this work was to develop and validate a cross-platform software application capable of detecting stratocumulus clouds from Cloudnet data and analysing their climatological, geometrical and microphysical characteristics for two Romanian observational sites: Bucharest and Galați. The study uses Cloudnet products covering the period March 2022 – February 2026 and focuses on the identification of stratocumulus profiles and episodes, together with the analysis of parameters such as cloud base, cloud top, cloud thickness, LWC, LWP, droplet effective radius and drizzle occurrence.
The application is based on a deterministic two-stage detection workflow. First, vertical profiles are classified using combined physical criteria derived from the Cloudnet target classification and from geometrical constraints specific to low-level stratocumulus clouds. Second, consecutive stratocumulus profiles are grouped into temporal episodes, while drizzle-related cases are identified using dedicated Cloudnet information and continuity criteria. The system automatically synchronizes auxiliary Cloudnet products and exports the results as CSV files, plots and statistical reports. In addition to the main detection workflow, selected AI/ML modules were applied to the exported results, including clustering of stratocumulus episodes, drizzle prediction, multivariate regression of LWP and DER, trend and change-point analysis, stratocumulus-to-cumulus transition prediction and survival analysis and transfer learning between the two stations.
The application identified 3.066 stratocumulus episodes, with frequencies of 5.65% in Bucharest and 6.89% in Galați. The obtained LWP and DER values were consistent with typical ranges reported for stratocumulus clouds, while the vertical LWC profile showed a pseudo-adiabatic structure. The comparison between the two stations highlighted regional differences in the frequency and microphysical properties of the detected episodes. These findings highlight the relevance of automated Cloudnet-based analysis for improving the characterization of low-level cloud regimes and for supporting future research on cloud variability and atmospheric processes over Romania.
Acknowledgement:
GLSG work was supported by the Romanian Nucleu Programme.
|
|
|
|