Analysis Of The Variation Characteristics Of Extreme Precipitation And Its Attribution During Rainy Season In The Southeast Coast Of China Under The Non-stationary Condition

Global warming has caused the changes in hydrologic circulation process, which induced many extreme precipitation events in global. Then, the extreme precipitation event has caused great economic loss and has aroused wide attention of a group of international and domestic scholars. The southeast coastal region of China covers Shanghai, Zhejiang, Fujian, Guangdong and Hainan, which is located between the largest continents (landmass of Eurasia) and the largest ocean (the Pacific), suffering the most serious extreme precipitation disaster. Since adopted the reform and opening up policy, the economy of the southeast coastal area developed rapidly with the high concentration of population and economy. However, all kinds of natural disaster happened frequently, especially the secondary disaster consist of flood, landslide and mud-rock flow which brought about by extreme precipitation events and constituted a serious treat on development of regional economy and society. Therefore, it is a fundamental work to analyze change characteristics of the extreme precipitation series and explore the reason of this change comprehensively.This study focuses on non-stationary characteristics of precipitation series in a changing environment. Mean and variance of precipitation extremes series were chose to represent the non-stationary characteristics. Using Empirical mode decomposition (EMD) and Mann-Kendall test to analyze non-stationary deterministic trend of research area in pre- and later flood period from 1960-2012. Contribution analysis also was used to investigate the contribution of mean and variance trends to the precipitation extreme increments. The impact of the non-stationary characteristics of precipitation extremes on the risk of extreme precipitation was explored by time-varing model. Selecting large-scale climate factors for pre and later flood period and Green house emissions data to further explore the cause of change characteristics of extreme precipitation events. Created the comprehensive climatic index and human activities index and simulated the 2-year precipitation event and the hundred -year precipitation event. The contribution analysis method and GAMLSS model was be used comprehensively to explore the impact of global climate change and human activities on the risk of extreme precipitation and the changes of precipitation in different return level. Result indicates that:1) Precipitation extremes series (Dmax) in southeastern region shows significant non-stationary characteristics in both pre-and later flood season. In pre-flood period, the deterministic trend has increased in inland area of Zhejiang, Fujian and Guangdong province, also in the western coast of Guangdong province. In later-flood period, the deterministic trend of Dmax characterized by a tendency of increasing in southeast and reducing in northwest. There are steady trend on the Dmax of Fujian. The Dmax has a increasing trend in the coastal areas of Guangdong and Hainan, whereas a reducing trend in central region. 2) 2) In pre-flood period, the changes of mean can mainly account for the deterministic trend of Dmax series except in Fujian. In later-flood period, the change of variance is main influential factor in Shanghai, Zhejiang, Guangdong and Hainan. The trend of variance led the overall trend of Dmax series importantly in Fujian.3) Time-varing model can be applied to accurately simulate the trend of extreme precipitation events under the non-stationary condition. The change of variance in precipitation series influences the extreme precipitation and the risk of extreme precipitation.4) In pre-flood period, the contribution of human activity to precipitation extreme in the stage of rapid social-economic development (1986-2012) is much greater than the stage of slow social-economic development (1960-1985), especially in Zhejiang and Guangdong Province. In later flood period, large scale climate changes has main influence on the trend of extreme precipitation, but human activities may have led the increase of extreme precipitation year by year by explore the whole time series.5) The central and southeast coast of Guangdong province is the region with the maximum precipitation in southeastern China during the pre-flood period. In addition, precipitation with various return levels has increasing in northeastern Zhejiang, the most area of Fujian, west coast and north central part of Guangdong. In later flood period, the precipitation amount of 2-year and 100-year precipitation event is fewer in inland and rich in coast area. Shanghai, the central and eastern Zhejiang and the northeast and east coast of Fujian suffering a more intense 2-year precipitation event clearly. The 100-year extreme precipitation event has enhanced remarkably in the northeast coast and northwest inland of Fujian and the east area of Guangdong.