Abstract

Understanding the relationship between hydrological connectivity (HC) and water level (WL) is crucial for effective water resource management and wetland restoration. However, current knowledge regarding this relationship is limited. This study proposed an integrated nonstationary and uncertain analysis framework (INUAF) to investigate the HC-WL relationship with reference to the Baiyangdian wetland, which is a fragmented wetland in North China. With the INUAF, the interannual and intra-annual variations of both HC and WL were examined, together with the wavelet coherence and lag effects between the two variables at multiple scales. The results highlighted marked nonstationarity in HC, WL, and the relationship between them. Scale-dependent lag effects revealed that HC lags WL by 37 days (131 days) at the 1 a scale (4 a scale), and leads WL by 190 days at the 8 a scale, indicating a complex coupled relationship between HC and WL. Additionally, the INUAF was applied to evaluating the uncertainty in the response of lagged HC to varied WL. Results indicated that every 0.2-m increase in WL led to a 2.2%-2.4% higher probability of maintaining high HC for WL between 6.0 and 8.0 m, but a 10%-11% higher probability for WL between 8.0 and 9.0 m. We suggest that a WL of > 8.4 m would produce a probability of > 50% for achieving high HC. These findings provide valuable insights into the HC-WL relationship and could contribute to wetland restoration efforts.