| 引用本文: | 王怀军,曹 蕾,俞嘉悦,等.基于EOF分析和GAMLSS模型的淮河流域极端气候事件非平稳特征[J].灌溉排水学报,2021,(5):125-134. |
| WANG Huaijun,CAO Lei,YU Jiayue,et al.基于EOF分析和GAMLSS模型的淮河流域极端气候事件非平稳特征[J].灌溉排水学报,2021,(5):125-134. |
|
| 摘要: |
| 【目的】将经验正交函数分析（EOF）时间系数作为广义可加模型（GAMLSS）的分析对象，揭示流域尺度极端气候事件的非平稳性特征。【方法】以EOF和GAMLSS为基础，将时间、气候指数和CO2量作为协变量，分析淮河流域极端气候事件的时空分布特征和非一致性规律。【结果】降水极值为平稳时间序列，无显著变化。气温极值显著变化，暖极值以1985年为分界，先下降后上升，冷极值中冷夜日数和霜冻日数呈持续下降趋势。气温极值为非平稳性变化，且CO2变化是导致极值非平稳的原因。气温极值的非平稳变化导致不同重现期下重现水平发生显著变化，暖昼日数（Tx90p）第一主成分20年一遇重现水平由1965年的40 d增加到2015年的60 d；冷夜日数（Tn10p）20年一遇重现水平由1965年的90 d下降到-25 d。【结论】GAMLSS模型可以模拟气候极值序列位置、尺度参数的平稳和非平稳性变化，与EOF结合，可以减少站点尺度非平稳分析的不确定性。 |
| 关键词: 极端气候事件；EOF；GAMLSS；淮河流域 |
| DOI：10.13522/j.cnki.ggps.2020299 |
| 分类号: |
| 基金项目: |
|
| Using EOF and GAMLSS to Analyze Nonstationary Extreme Climate Events in Huai River Basin |
|
WANG Huaijun, CAO Lei, YU Jiayue, LU Yuanyuan, FENG Ru, YANG Yaxue, YE Zhengwei, SUN Xiaohui
|
|
1.School of Urban and Environmental Sciences, Huaiyin Normal University, Huai’an 223300, China;2. State Key Laboratory of Hydrology-water Resources and Hydraulic Engineering, Nanjing Hydraulic Research Institute,
Nanjing 210029, China; 3. Research Center for Climate Change, Ministry of Water Resources, Nanjing 210029, China; 4. Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China;5. Water Conservancy Bureau of Dezhou City, Dezhou 253000, China
|
| Abstract: |
| 【Objective】Global climate change results in nonstationary extreme climate events which cannot be analyzed using traditional frequency analysis method. The objective of this paper is to test the feasibility of time coefficient of empirical orthogonal function method (EOF) and the generalized additive model (GAMLSS) to analyze non-stationary extreme climate events.【Method】We took data measured from the Huai river basin in 1965-2015 as an example, analyzing the nonstationary extreme climate events in it using the EOF and GAMLSS, with climate indices and atmospheric CO2 concentration taken as covariates.【Result】The first loading of EOF1 of the climate extremes reflects the change in extreme climate events in the basin. The first corresponding principal component (PC1) of the extreme precipitations did not show significant trend, but of the extreme temperature did with the warm extremes decreasing first followed by an increase from 1965—2015. The cold night extremes and the number of frost days had both been in decrease, and the correlation coefficient between extreme precipitations and their contributing factors was low. In contrast, extreme temperature was nonstationary and correlated to CO2 concentration at significant level, largely because the change in CO2 was nonstationary. The GAMLSS model can describe both stationary and nonstationary climate extremes. Overall, the temporal change in extreme precipitation was approximately stationary and can be described by a logistic distribution. The number of warm extremes had been in decline until 1985, but has seen an increase since then. The nonstationary change in extreme temperature led to a significant change in the period of return, with the 20-year return in the first principal component of warm days increasing from 40 d in 1965 to 60 d in 2015, while the 20-year return in the first principal component of the cold night decreasing from 90 d in 1965 to 25 d in 2015.【Conclusion】GAMLSS can describe both stationary and nonstationary climate extremes; combining it with EOF can reduce analysis uncertainty. |
| Key words: extreme climate event; empirical orthogonal function method; GAMLSS; Huai river basin |
