Calculation of Climate Loads Design Values According to the Probability Model of Annual Maximum Series

Abstract

The aim of this work is to improve a method for determining the characteristic values of climatic loads according to a probabilistic model of the annual maxima sequence, by choosing a rational type of generalized extreme value distribution law. An analysis is provided regarding the suitability of using four types of distributions for describing a data collection of maximum values of climatic loads. Using example data from the meteorological stations of Ukraine, it is found that for coefficients of variation smaller than 0.85–1.0, it is advisable to use the double exponential Gumbel distribution (generalized extreme value distribution type-I), and at higher values of the coefficient of variation, it is advisable to use the Weibull distribution (generalized extreme value distribution type-III). Recommendations are provided for considering the accuracy in the estimations of the characteristic values of loads according to the probabilistic model for the annual maximum value series.

Long-term forecasting of thermal and humidity actions on buildings

“…The modern practice of climate action standardisation has been reduced to the establishment of characteristic and design values based on the statistical analysis of the results of previously performed meteorological observations. Sequences of annual minima, maxima, or mean annual values of the corresponding climatic characteristics are often used for this purpose [1][2][3][4]. The final stage of climate action study involves the generalisation of data from individual weather stations to reflect the territorial variability of actions in different ways [5,6] and development of regulations that determine loads and impacts during the design of buildings and structures.…”