In a study published in National Science Review, a research team led by Prof. ZHOU Yongqiang from the Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, revealed that the age of dissolved organic carbon (DOC) in global rivers is fundamentally controlled by how long carbon resides in soils before entering aquatic systems.

Researchers constructed the first high-resolution global maps of riverine DOC concentration, alongside its radiocarbon (Δ¹⁴C) and stable carbon isotope (δ¹³C) signatures. This was achieved by integrating a comprehensive global database with machine learning approaches. The study revealed the sources, spatial distribution, and age characteristics of riverine DOC, quantifying contributions from different endmembers.

The results show that riverine DOC spans an age range from modern carbon to material exceeding 29,000 years, with a mean radiocarbon age of approximately 221 years. Nearly 60% of DOC has ¹⁴C ages younger than 100 years. However, aged carbon persists in high-latitude and high-elevation regions, particularly where permafrost thaw and glacial processes mobilize long-stored carbon pools.

Using a four-endmember isotope mixing model, the researchers found that fossil or petrogenic carbon contributes only 6.7% to the global DOC pool, while modern terrestrial organic carbon and riverine autochthonous production dominate, contributing about 38% and 44%, respectively. Holocene-aged sediment-derived DOC accounts for approximately 10.7%, especially in high-latitude floodplains.

The study highlights that soil carbon residence time, shaped by climate and hydrological processes, is a key regulator of riverine DOC age. Notably, warming-induced permafrost thaw is accelerating the release of ancient "old carbon" into river systems, which can then be transported downstream and participate in aquatic biogeochemical processes, potentially enhancing carbon-cycle feedbacks to the climate system.

These findings fill a gap in global-scale understanding of riverine carbon cycling, linking terrestrial carbon storage, mobilization, and aquatic processing into a unified framework. They provide a mechanistic basis for predicting how ongoing climate change may reshape carbon cycling across the land–water continuum.