Understanding the impacts of climate change and human activities on hydrological processes, especially the baseflow, is vital for sustainable water resource management. We analyzed the global changes in baseflow and baseflow index (BFI) for 2,374 global streamflow stations from 1970 to 2016 to examine their associations with precipitation, temperature, terrestrial water storage (TWS), normalized difference vegetation index (NDVI) representing the vegetation regimes, potential evapotranspiration (PET), and a humidity index (HI)‐precipitation/PET. Results showed that changes in both baseflow and BFI are significantly region dependent in continents, due to spatial differences of changes in climate and vegetation that determines the baseflow generation. Seasonal baseflow changed slightly, while seasonal BFI varied greatly, partly due to the shift from snowfall to rainfall and warming effects on glacial retreat and the timing of snowmelt. The multibasin baseflow elasticity analyses show that baseflow (BFI) was highly sensitive to changes in NDVI, precipitation, and temperature (temperature and PET). The multivariate regression analyses show that globally, changes in precipitation and TWS contributed to the majority (64.8% and 20.2%) of changes in baseflow, while changes in HI, PET, and precipitation contributed to the majority (30.3%, 28.9%, and 27.5%) of changes in BFI. The spatially varying interaction between climate, vegetation, and baseflow implies that regional adaptation of water resources utilization to climate change should consider the regional shifts in vegetation. This study quantified the elasticity and relative contribution of changes in geographic factors to changes in baseflow and BFI around the world and could inform practices for sustainable water management.

Floods in river deltas are driven by complex interactions between astronomical tides, sea levels, storm surges, wind waves, rainfall‐runoff, and river discharge. Given the anticipated increase in compound flood hazards in river deltas in a warming climate, climate‐informed regional to local extreme water levels (EWLs) is thus critical for decision‐makers to evaluate flood hazards and take adaptation measures. We develop a simple yet computationally efficient stress test framework, which combines historical and projected climatological information and a state‐of‐the‐art hydrodynamic model, to assess future compound coastal‐fluvial flood hazards in river deltas. Our framework is applied in the world's largest single urban area, China's Pearl River Delta (PRD), which is also characterized by densely crossed river network. We find that extreme sea level is the dominant driver causing the compound coastal‐fluvial flood in the PRD over the past 60 years. Meanwhile, there is large spatial heterogeneity of the individual and compound effects of the typhoon intensity, local sea‐level rise, and riverine inflow on coastal‐fluvial floods. In a plausible disruptive scenario (e.g., a 0.50 m sea‐level rise combined with a 9% increase in typhoon intensity in a 2°C warming), the EWL will increase by 0.76 m on average. An additional 1.54 and 0.56 m increase in EWL will occur in the river network and near the river mouth, respectively, if coastal floods coincide with the upstream mean annual flood. Findings from our modeling framework provide important insights to guide adaptation planning in river deltas to withstand future compound floods under climate change.

Analyzing the spatiotemporal characteristics and causes of landscape pattern changes in watersheds around big cities is essential for understanding the ecological consequence of urbanization and provides a basic reference for the watershed management. This study used a land-use transition matrix and landscape indices to explore the spatiotemporal change of land use and landscape pattern over Liuxihe River basin of Guangzhou in the southeast of China from 1980 to 2015 with multitemporal Landsat satellite data in response to the rapid urbanization process. Primary temporal and spatial influencing factors were first quantitatively identified through grey relation analysis (calculating correlation degree between land use changes and influencing factors) and Geodetector (detecting landscape spatial heterogeneity and its driving factors), respectively. Considerable spatial and temporal differences in land use and landscape pattern changes were observed herein, thus determining the influencing factors of these differences in the Liuxihe River basin. These changes were characterized by a large increase in construction land converted from cropland, particularly in the middle and lower reaches of the basin from 2000 to 2010, causing dramatic fragmentation and homogenization of the landscape pattern there. Meanwhile, the landscape pattern gradually transitioned from an agricultural land use dominant landscape to a construction land use dominant landscape in these regions. Furthermore, the rapid growth of a nonagricultural population and the transformation of industry primarily caused the temporal changes of landscape pattern, and the landscape spatial heterogeneity was mainly caused by the interaction of complicated geomorphology and anthropogenic activities in different spatial locations, particularly after 2000. This study not only provides an improved approach to quantifying the main spatiotemporal influencing factors of landscape pattern changes during different time periods, but also offers a reference for decision-makers to formulate optimal strategies on ecological protection and urban sustainable development of different regions in this study area.