Font Size: a A A

Copula Based Analysis Of Extreme Weather

Posted on:2019-04-15Degree:MasterType:Thesis
Country:ChinaCandidate:D WangFull Text:PDF
GTID:2310330542954667Subject:Mathematics
Abstract/Summary:PDF Full Text Request
Extreme weather is becoming more frequent and stronger with the underground of global climate warming.There are many kinds of extreme weather,such as extreme high-temperature weather,extreme cold weather and heavy rainfall,which cause many natural disasters including flood and waterlogging,landside and mudslide.These disasters have led to massive economic loss and casualties.It's high time that we improve the ability of adapting to climate changes,forecasting extreme weather,which are precondition of prevent disasters.In addition,propagandizing extreme weather and its perniciousness should also be paid attention to.We choose temperature and wind speed as two main variables to describe the extreme weather due to the characteristics of weather in north of China.Dependence structure of these two variables are modeled to calculate the return period of extreme weather,which has significant meaning to predict extreme weather precisely,prevent disasters and reducing damages and improving the ability of adapting extreme weather.Daily weather data from year 1971 to year 2017 is used in our research.Joint distribution model of temperature and wind speed is constructed based on bivariate Copula model.Several methods of parametric estimation are used to estimate the parameters of Copula.Then we choose the best Copula model with the help of statistics.Finally,different types of return period are calculated,and their graphs are plotted by software,which is useful to analyze the risk of extreme weather in different conditions.We draw some conclusions about the Copula model.Firstly,generalized extreme distribution are more suitable to fit the data of temperature and wind speed,which corresponds to its advantages.Secondly,AMH Copula fit the joint distribution of two variables best,because AMH Copula is good at describe weak dependence.Thirdly,we find IFM method may be better than CML method according to the results of goodnessof-fit.Finally,by the graphs of return period,we find extreme weather is less likely to happen and this place is suitable to go skiing and launch sports events in winter.However,it is still important to make full preparation for extreme weather.
Keywords/Search Tags:Copula, Extreme Weather, Parameter Estimation, Return Period
PDF Full Text Request
Related items