| |
 |
UNIVERSITY OF BUCHAREST
FACULTY OF PHYSICS Guest
2024-11-22 1:44 |
 |
|
|
|
Conference: Bucharest University Faculty of Physics 2006 Meeting
Section: Atmosphere and Earth Science; Environment Protection
Title:
Probabilistic forecast of thunderstorms using atmospheric sounding data - application to Bucharest area and surroundings
Authors:
Bogdan Antonescu, Adrian Tanase
Affiliation:
Romanian National Meteorological Administration
E-mail
bogdan.antonescu@meteo.inmh.ro
Keywords:
lightning, atmospheric sounding, PCA, logistic regression, probabilistic forecast
Abstract:
Forecasting the precise time and location of convective activity is an important goal for meteorologists, because thunderstorms affect day-to-day activities. An important goal is also to forecast the lightning activity, so that appropriate decisions can be made when thunderstorms threaten becomes important. Since 2003 the Romanian Lightning Detection Network (RLDN) provides real time lightning data for the Romanian territory and surroundings. Statistical models were developed to predict the probability of lightning around Bucharest for the months of May-September. The predictors were derived from atmospheric sounding data recorded daily at Afumati station. The selection of the predictors has been done using the Principal Component Analysis (PCA) method. Logistic regression was used to develop the forecast equation of probabilities for different types of lightning activities. The lightning and atmospheric sounding data from 2003 to 2004 were used to obtain the probabilistic equation, and the data from 2005 were used for testing. In this paper are presented the first results of this probabilistic forecasting technique applied at the Romanian National Meteorological Administration.
|
|
|
|