Seminar Organizing Committee
Corr. RAS V.N.Lykosov (Marchuk Institute of Numerical Mathematics of RAS, RCC MSU)
Doctor of Physics and Mathematics V.M. Stepanenko (RCC MSU, Geographical Faculty of Moscow State University)
A.V.Debolsky (RCC MSU, A.M. Obukhov Institute of Atmospheric Physics of RAS).
|Dmitriev Egor Vladimirovich||Marchuk Institute of Numerical Mathematics of RAS|
CLASSIFICATION OF LOCAL METEOROLOGICAL EVENTS IN THE URBANIZED COASTAL AREA FROM GROUND-BASED INSTRUMENTAL MEASUREMENTS
The problem is considered of atmospheric meteorological events’ classification such as sea breezes, fogs and high winds, in coastal areas. In-situ wind, temperature, humidity, pressure, turbulence etc. meteorological measurements are used as predictors. Local atmospheric events of 2013–2014 were analyzed and classified manually using data of the measurement campaign in the coastal area of the English Channel in Dunkirk (France). The results of that categorization allowed the training of a few supervised classification algorithms using the data of an ultrasonic anemometer as predictors. The comparison was carried out of the k-nearest neighbors classifier, support vector machine, and two Bayesian classifiers – quadratic discriminant analysis and Parzen-Rozenblatt window. The analysis showed that k-nearest neighbors and quadratic discriminant analysis classifiers reveal the best classification accuracy (up to 80% correctly classified meteorological events). At that, the latter classifier has higher calculation speed and is less sensitive to unbalanced data and the overtraining problem. The most informative atmospheric parameters for events recognition were revealed for each algorithm. The results obtained showed that supervised classification algorithms contribute to automation of processing and analyzing of local meteorological measurements.
Due to the self-isolation regime, the seminar will be held in the form of a webinar on the Zoom platform.
To participate in the webinar, we ask you to enter your data in the Google spreadsheet:
Instructions for installing and using the Zoom platform are available here:
For communication on all issues related to the work of the seminar, please contact the Scientific Secretary Andrey Vladimirovich Debolsky at email@example.com