Hydroclimatic vulnerability of forest carbon stocks: a global assessment of stress and precipitationsheds

Forests play a major role in the global carbon cycle, but their capacity to store carbon and maintain carbon uptake is closely tied to water availability. This thesis examines the hydroclimatic vulnerability of forest carbon stocks by linking locally observed stress within forests with the wideratmospheric moisture-source regions that supply part of their rainfall. It develops an exploratory diagnostic framework that first identifies major aboveground forest carbon stocks, then maps where these stocks show significant hydroclimatic drying or weakening net carbon uptake, and finally characterizes the precipitationsheds of four major forest carbon systems: the Amazon, Congo, Southeast Asia, and the boreal forest belt. The results show that major aboveground forest carbon stocks occupy only about 6% of global land grid cells, but around 40% of these cells show either significant drying or weakening net carbon uptake. Backward moisture tracking with WAM2layers shows that these forest carbon systems are situated within large and regionally distinct precipitationsheds. Mean moisture transport distances range from roughly 3,000 km in the Congo to 4,000 km in the Amazon, while more than 70% of tracked moisture originates outside the corresponding regional forest sink in all four systems. Upwind source regions also differ strongly in country distribution and land cover. Southeast Asia and the Amazon are predominantly ocean-supplied, the Congo is nearly balanced between oceanic and terrestrial sources, and the boreal system is mainly land-supplied. The comparison between major stock systems and their stress-affected subsets shows that stress-affected areas do not have fundamentally separate precipitationsheds. Instead, they remain embedded within the broader source-region systems of the major carbon stocks, with only modest shifts in relative source contributions. The thesis therefore argues that hydroclimatic forest carbon vulnerability can be understood through both local stress and wider precipitationsheds. This framing helps identify the spatially extended systems in which questions regarding carbon persistence in forests and governance require further investigation.