Answers are grounded in this paper and cite their sources.
Bayesian method improves classification reliability of dangerous meteorological phenomena in cumulonimbus clouds, highlighting turbulence effects.
Cumulonimbus clouds and accompanying meteorological phenomena such as rainfalls, thunderstorms, hail are dangerous weather phenomena for aviation. These phenomena pose the greatest danger during takeoff and landing operations of aircraft. The existing classification of such dangerous meteorological phenomena in weather radars is based on data on the altitude distribution of the radar reflectivity of the atmosphere, air temperature and has its drawbacks. It is characteristic that all legislatively accepted criteria for classifying meteorological phenomena from clouds to tornadoes, naturally based on long-term observations and in this sense inspiring respect, are developed for each phenomenon separately and are of a somewhat heuristic nature. Purpose: to develop a method for classifying dangerous meteorological phenomena of cumulonimbus clouds in weather radars, which will increase the reliability of the classified meteorological phenomena. The paper shows that significant shortcomings of the criteria used to classify dangerous meteorological phenomena in weather radars include: the absence of information on atmospheric turbulence in the classification features and the use of its own criterion for each phenomenon. It is obvious that the criteria for classifying hazardous meteorological phenomena should be optimized in the following areas: the use of information on the altitude distribution of maximum values of not only reflectivity, but also atmospheric turbulence in the classification criteria; the classification should be based on a single selected criterion for distinguishing statistical hypotheses, which is implemented in the article. To solve the problem of classifying hazardous meteorological phenomena of cumulonimbus clouds, the Bayesian approach is substantiated under conditions of complete a priori uncertainty. A classification method is proposed based on the maximum likelihood criterion using sufficient statistics of reflectivity and atmospheric turbulence. Based on the analysis of the information content of classification features for one-dimensional and multidimensional densities, it was concluded that the required reliability of classification is achieved with the simultaneous use of all four features: maximum values of reflectivity and atmospheric turbulence, as well as their distribution by height. For a specific statement of the problem, an algorithm for classifying dangerous meteorological phenomena of cumulonimbus clouds was developed. The selected criterion for distinguishing statistical hypotheses and the introduction of information on atmospheric turbulence into the criteria for classifying dangerous meteorological phenomena associated with cumulonimbus clouds will improve the reliability of the classified meteorological phenomena in weather radars.
No takes yet. Share an insight, caveat, or question.
Vasiliev et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: