基于空间贝叶斯层次模型的淮河流域气候极值特征分析

王怀军; 潘莹萍; 冯 如; 肖明贤

引用本文:	王怀军,潘莹萍,冯如,等.基于空间贝叶斯层次模型的淮河流域气候极值特征分析[J].灌溉排水学报,2020,39(5):102-110.
	WANG Huaijun PAN Yingping, FENG Ru, XIAO Mingxian,WANG Huaijun PAN Yingping, FENG Ru, XIAO Mingxian,WANG Huaijun PAN Yingping, FENG Ru, XIAO Mingxian,et al.基于空间贝叶斯层次模型的淮河流域气候极值特征分析[J].灌溉排水学报,2020,39(5):102-110.

【打印本页】【下载PDF全文】【查看/发表评论】【EndNote】【RefMan】【BibTex】

←前一篇|后一篇→

过刊浏览高级检索

本文已被：浏览 1282次下载 1169次	码上扫一扫！
分享到：微信更多字体:加大+\|默认\|缩小-
基于空间贝叶斯层次模型的淮河流域气候极值特征分析
王怀军，潘莹萍，冯如，肖明贤
1.淮阴师范学院城市与环境学院，江苏淮安223300；2.南京水利科学研究院水文水资源与水利工程科学国家重点实验室，南京210029；3.南京水利科学研究院水利部应对气候变化研究中心，南京210029； 4.北京师范大学地理科学学部，北京100875

摘要:

【目的】验证空间贝叶斯层次模型在极端气候事件当中建模的适用性，探明淮河流域极端气候事件的空间分布规律。【方法】基于空间贝叶斯层次模型，将经度、纬度与海拔作为模型协变量捕捉气候极值的空间变化特征。在建模过程中，将广义极值函数（GEV）作为其边际分布，采用马尔可夫链蒙特卡罗算法（MCMC）确定空间极值模型所需的参数值。选用淮河流域1960—2015年日最大降水量（RX1day）、日最高气温（TXx）作为极端气候变量进行建模，将模型结果按站点提取，并与基于GEV的站点结果进行对比。【结果】空间贝叶斯层次模型能够很好地模拟淮河流域气候极值，模型结果的参数及各重现水平与直接基于站点数据的GEV分析相近。RX1day不同重现水平从流域西北向东南增加；TXx不同重现水平具有典型的经向地带性，从流域东部往西部增加。【结论】研究建立的空间极值模型可以获得没有观测台站所在位置的极端气候重现水平，该结果拓展了淮河流域极端气候事件时空规律研究。

关键词: 极端气候；贝叶斯层次模型；空间极值；淮河流域

DOI：10.13522/j.cnki.ggps.2019215

分类号:

基金项目:

Using Spatial Bayesian Hierarchical Model to Analyze Extreme Climate Indexes in Huai River Basin

WANG Huaijun PAN Yingping, FENG Ru, XIAO Mingxian,WANG Huaijun PAN Yingping, FENG Ru, XIAO Mingxian,WANG Huaijun PAN Yingping, FENG Ru, XIAO Mingxian,et al

1. School of Urban and Environmental Sciences, Huaiyin Normal University, Huaian 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

Abstract:

【Objective】The objective of this paper is to test the feasibility of using spatial Bayesian hierarchical model to estimate the spatiotemporal change in extreme climate events in the Huai river Basin.【Method】The hierarchical Bayesian spatial model, HKEVP, was used to capture the spatial variation in extreme climates indexes, with the longitude and transverse coordinates and its associated latitude as independent variables and the generalized extreme value distribution (GEV) as the marginal distribution. We first calculated the parameters in the model using the Markov Chain Monte Carlo method (MCMC) and then applied the model to analyze the maximum daily precipitation (RX1day) and highest daily temperature (TXx) in each month from 1960—2015. The GEV parameters and the return level derived from the model were compared to those calculated from the maximum likelihood estimates.【Result】The hierarchical Bayesian model was adequate for estimating the GEV parameters for locations without observation, and the RX1day return level at different return periods increased from northwest to southeast in the basin. The return level of TXx was longitudinally zonal and its value was lower in the east than in the west.【Conclusion】The proposed spatiotemporal model for estimating extreme climate indexes is able to obtain the return levels of the extreme values at ungauged station, and the results improved our understanding of the spatiotemporal change in extreme climate indexes in Huai River Basin.

Key words: extreme climate indexes; Bayesian hierarchical model; spatial extremes; Huai river basin