Uncertainty in Modeled Eco-hydrologic Responses to Climate Change in Korea Mountain Wetlands

Mountain wetlands have small catchment areas and limited storage capacity, making them sensitive to changes in climate forcing. However, previous studies have mainly evaluated mean future changes, limiting the joint interpretation of the likelihood of exceeding specific thresholds and the origins of associated uncertainty. This study applied an eco-hydrologic model to Janggun Wetland and Hwaeom Marsh in southeastern Korea to project future changes in wetland water level, inundation days, and leaf area index (LAI) under multiple future climate forcings. Exceedance probability, a hierarchical Bayesian model, and Shapley decomposition were used to evaluate the likelihood that future changes would exceed specific criteria and to assess the relative contributions of future climate uncertainty and model parameter uncertainty. The results showed that Janggun Wetland exhibited a wetting tendency, with increases in wetland water level and inundation days, whereas Hwaeom Marsh showed mixed hydrologic responses and a decreasing tendency in LAI. In Janggun Wetland, the probability of inundation days increasing relative to the present was approximately 50% in the near future and increased to more than 90% in the far future. In Hwaeom Marsh, the probability of LAI decreasing relative to the present exceeded 70% in the far future. Future climate uncertainty generally had a greater influence than model parameter uncertainty, although their relative contributions varied by wetland, variable, threshold, and future period. This study provides scientific evidence for wetland monitoring, model improvement, and climate change adaptation management.