Search

• Geographically isolated wetlands (GIWs) are vulnerable to climate change. • A hybrid Machine Learning (ML) – hydrologic modelling – RS approach was used. • Multiple indices and variables derived to assess water storage volume (WSV) of GIWs. • Trend and rate of changes in WSV showed different patterns across spatial scales. • Future changes in WSV are driven by hydrologic, climatic, and geographic features. Geographically Isolated Wetlands (GIWs) are essential ecological and ecosystem entities that are vulnerable to climate change. The water storage volume (WSV) of GIWs is an important indicator to monitor their health and ecosystem services. Therefore, the projection of changes in future WSV of GIWs is crucial for conservation policies that are typically set at the regional scale. A limiting factor for the extended spatial projection of WSV of GIWs is the lack of data on their geometry, bathymetry, and the complexity of hydrologic relation to the surrounding drainage basin and the climate at a regional scale. In this study, we developed a hybrid Machine Learning (ML) – hydrologic modelling approach, from which a number of indices and variables derived to assess WSV of GIWs. We projected future changes in WSV of about a hundred thousand GIWs across 17 large river basins in western Canada, using a nine-model ensemble climate simulations under two greenhouse gas emissions scenarios (i.e., RCP 2.6 and RCP 8.5). We estimated the baseline WSV by developing ML models using water depth measurements and geographical information features from 234 surveyed GIWs across the study area. Future WSV was quantified using a three-step procedure, where water balance simulations were facilitated through the Soil and Water Assessment Tool (SWAT) for the 2018–2034 period, and GIW specific mass-balance components were adapted using remote sensing data. The spatiotemporal analysis of the WSV offers new insights, and reveals (1) a baseline provincial WSV of 550 million m3, with an average of 5,508 m3 per GIW, (2) a decreasing future WSV at annual rates of −2.58 and −2.62 million m3 for RCP 2.6 and RCP 8.5, respectively, and (3) the minimum and maximum WSV loss rates occurring during October and February, respectively, and (4) a diverse trend and magnitude of changes in WSV across 17 river basins. Finally, we state that while a group of hydrologic, climatic, and geographic processes and factors drives WSV of GIWs, the dominating driving force in WSV trends might be spatially and temporally different depending on the environmental conditions and characteristics of river basins and their associated GIWs. This study facilitates a better understanding of the connections between the WSV of GIWs and watershed-scale hydrology and climate change across multiple spatial scales. [ABSTRACT FROM AUTHOR]