A hierarchical Bayesian approach to the modified Bartlett-Lewis rectangular pulse model for a joint estimation of model parameters across stations

详细信息    查看全文

• 作者：Jang-Gyeong Kim^a ; Hyun-Han Kwon^a ; ^{hkwon@jbnu.ac.kr} ; Dongkyun Kim^b
• 关键词：Hierarchical Bayesian model ; Modified Bartlett-Lewis rectangular pulse model ; Uncertainty ; Rainfall simulation
• 刊名：Journal of Hydrology
• 出版年：2017
• 出版时间：January 2017
• 年：2017
• 卷：544
• 期：Complete
• 页码：210-223
• 全文大小：2142 K
• 卷排序：544

文摘

A hierarchical Bayesian model is proposed to jointly estimate parameters. Parameter uncertainty is formally taken into account through a Bayesian framework. Parameters with better reproduction of rainfall statistics are reliably estimated. Seasonal patterns of the estimated parameters are correctly identified. Spatial variability of the parameters is substantially reduced.