Analysis Of Rainstorm Characteristics Based On Three-dimensional Copulas In Liaoning Province

Rainstorm,as a kind of extreme precipitation,has great precipitation intensity and can bring a lot of rainfall in a short time,and is the most important natural hazard.In the background of the monsoon,it often causes floods,debris flows,and other secondary disasters due to the rapid convergence.Urban waterlogging is caused due to infiltration,storage,drainage of precipitation is less than rain in the urban surface.The flooding flows into farmland and flooded crops,results in the decrease of production.At the same time,it affect the normal travel of human and other activities of society.Therefore,the analysis of regional rainstorm amount,intensity and frequency can be showing the inherent probability characteristics and contribute to a comprehensive understanding of rainstorm.Those are the basis of part of the risk analysis and research of rainstorm.Based on differences in regional climate and topography background,Liaoning province is divided into different partitions of wet and landscapes in this article.We selected representative stations of every district from country meteorological stations in Liaoning province by the Kendall correlation coefficient.And then,we are fitting probability distribution and calculating return period of the factors(The annual rainstorm day,rainstorm amount,annual average rainstorm intensity,and rainstorm rate of contribution)based on the daily precipitation from 1960 to 2012.Copula function is introduced based on the various precipitation factors,and it can build a multi-dimensional joint distribution.The optimal copula function used to fit extreme precipitation factors is determined by comparing the root mean squared error(RMSE)with Akaike information criterion(AIC).The return period characteristics of extreme precipitation and the probability can be analyzed contrasting with the different partitions of wet and landscapes after many factors combined.Research shows that:(1)A few factors which are the rainstorm amount of sub-moist mountainous area,annual average rainstorm intensity and rainstorm amount of sub-moist plain area,and annual rainstorm day,rainstorm amount,annual average rainstorm intensity and rainstorm rate of contribution of moist mountainous area obey generalized extreme value distribution.Some factors which are annual average rainstorm intensity and rainstorm rate of contribution of sub-moist mountainous area and rainstorm rate of contribution of sub-moist plain area obey extreme value distribution.The annual rainstorm day of moist and sub-moist mountainous area submit to poisson distribution in Liaoning province;it comply with normal distribution in sub-moist plain area.(2)The annual rainstorm day of sub-moist and moist mountainous area and sub-moist plain area,respectively,are evenly distributed 1-5 times,focusing on 0-2 times and focusing on 1-6 times.The rainstorm amount respectively focus to 50-350mm,50-300mm and 150-500mm,and the cumulative probability is up to 0.8,0.7,and 0.8.Annual average rainstorm intensity,respectively,are concentrated in the 60-100mm,52.5-150mm,60-120mm,and the cumulative probability is up to 0.8,0.84,0.85.The rainstorm rate of contribution are concentrated of 0.15-0.5,unconcentrated,focus on the range of 0.08-0.7.(3)The joint probability is respectively calculated by using AMH copula,Clayton copula and Frank copula in sub-moist mountainous and plain and moist mountainous area.In sub-moist mountainous area,the joint probability,respectively,show higher influence on rainstorm day and rainstorm amount,and the higher of rainstorm day and rainstorm rate of contribution,the lower influence on rainstorm day in the range 0-0.4,0.5-0.65 and 0.65-1 of the cumulative probability.In sub-moist plain area,the joint probability,respectively,show synchronization increases of rainstorm amount and rainstorm rate of contribution and the lag of rainstorm rate of contribution in interval 0-0.37,and the synchronization increases of rainstorm factors and the lag of cumulative probability.In moist mountainous area,the joint probability,respectively,show the lag of annual average rainstorm intensity in interval0-0.2,and the negative effect of annual rainstorm day and rainstorm amount.(4)The return period of annual rainstorm day(0-5times)is higher in sub-moist mountainous area and is greater than sub-moist plain area,and is lower in moist mountainous area.The return period of rainstorm amount and annual average rainstorm intensity is greater in moist mountainous area than sub-moist plain area,is lower in sub-moist mountainous area.The return period of the rainstorm rate of contribution(cumulative probability interval 0-0.4)is higher in moist mountainous area than sub-moist plain area,and is lower in sub-moist mountainous area.The upper limit of joint return period are respectively 100 years in sub-moist mountainous area,30 years in sub-moist plain area,and 50 years in moist mountainous area.It is significantly reducing,and means the probability increasing of rainstorm.The increase of the value of each rainstorm factors have significantly difference with the increase of return period.