Systematic assessment of the uncertainty in integrated surface water-groundwater modeling based on the probabilistic collocation method

• Published in: Water resources research 2014-07, Vol.50 (7), p.5848-5865
• Main Authors: Wu, Bin, Zheng, Yi, Tian, Yong, Wu, Xin, Yao, Yingying, Han, Feng, Liu, Jie, Zheng, Chunmiao
• Format: Article
• Language: English
• Subjects: Data collection; Freshwater; Groundwater; GSFLOW; Heihe River Basin; integrated modeling; probabilistic collocation method; River basins; Rivers; Surface water; surface water-groundwater interaction; Surface-groundwater relations; uncertainty analysis
• ISSN: 0043-1397; 1944-7973
• DOI: 10.1002/2014WR015366

Systematic uncertainty analysis (UA) has rarely been conducted for integrated modeling of surface water‐groundwater (SW‐GW) systems, which is subject to significant uncertainty, especially at a large basin scale. The main objective of this study was to explore an innovative framework in which a systematic UA can be effectively and efficiently performed for integrated SW‐GW models of large river basins and to illuminate how process understanding, model calibration, data collection, and management can benefit from such a systematic UA. The framework is based on the computationally efficient Probabilistic Collocation Method (PCM) linked with a complex simulation model. The applicability and advantages of the framework were evaluated and validated through an integrated SW‐GW model for the Zhangye Basin in the middle Heihe River Basin, northwest China. The framework for systematic UA allows for a holistic assessment of the modeling uncertainty, yielding valuable insights into the hydrological processes, model structure, data deficit, and potential effectiveness of management. The study shows that, under the complex SW‐GW interactions, the modeling uncertainty has great spatial and temporal variabilities and is highly output‐dependent. Overall, this study confirms that a systematic UA should play a critical role in integrated SW‐GW modeling of large river basins, and the PCM‐based approach is a promising option to fulfill this role. Key Points Systematic uncertainty analysis for integrated surface water-groundwater models A holistic view of the modeling uncertainty achieved with a low computing cost Insights into process understanding, model calibration, and data collection