Filters
RCIN and OZwRCIN projects

Search for: [Abstract = "ldentification of monthly precipitation patterns is an importance issue for climate studies. lt is also necessary for development of rainfall erosivity factor annual distribution curves. Establishing of monthly precipitation distributions for different Polish gauging stations by means of Self\-Organizing Feature Maps \(SOFM\) was the first study aim. The attempt to use SOFM for classification of analyzed stations into typical precipitation regions was the second study aim. The research was based on database consisted of monthly precipitation totals from years 1961 \- 1980 for a total number of 103 gauging stations located in the whole country. For needs of local monthly precipitation patterns identification SOFM networks of simple architecture 12\:12\-4\:1 were used. Additional SOFM network of 12\:12\-10\:1 architecture was developed for classification of local average monthly precipitation patterns into 10 separate classes. SFOM networks proved to be useful for calculation of average monthly precipitation distributions on the network of gauging stations. lt was observed that implementation of SFOMs instead simple average value calculation made the data processing less sensitive for extreme events. Also topological map developed for classification of analyzed gauging station with respect to their previously calculated average monthly precipitation distributions into precipitation regions performed well. lt was able to divide logically stations into groups of similar precipitation patterns having an explanation in local climate conditions. Moreover the arrangement of topological map nodes was generally clear and understandable due to neighbourhood of identified precipitation regions. All this suggests that SFOM networks could find practical implementations for climate precipitation studies, optimization of the gauging station networks and development of R\-factor annual distributions in Poland."]

Number of results: 1

Items per page:

This page uses 'cookies'. More information