Carbon Sinks in Northern China During 1982–2023: Pattern, Evolution and Driving Forces

The ecological environment in Northern China is fragile, its overall carbon sink capacity has long been underestimated, and its spatial differentiation is significant. Based on the Carnegie-Ames-Stanford Approach (CASA) model and the Thornthwaite memorial model, this study analyzed the spatial-temporal characteristics and driving mechanisms of actual net ecosystem productivity ( NEP_A ) in Northern China from 1982 to 2023. The results show that: 1) NEP_A fluctuated between 139.48 and 174.74 g C/(m²·yr), with the year 2000 serving as a pivotal transition point after which ecological restoration projects reversed the preceding declining trend. Approximately 52.07% of the study area exhibited carbon sink improvement, concentrated in the eastern plains such as the Songnen Plain, Liaohe Plain, and Haihe Plain, while 47.93% exhibited degradation, primarily in the western arid region. 2) The relative contributions of climate change and human activities to carbon sink changes were quantitatively separated. In carbon sink-improved areas, human activities dominated, contributing 61.51% versus 38.49% from climate change. In carbon sink–degraded areas, climate change and human activities contributed nearly equally (51.94% and 48.06%, respectively). Among climatic factors, precipitation exerted the most extensive influence, with NEP_A showing a significant positive correlation (P < 0.05) in 31.93% of the study area, particularly in the Hulunbuir Sandy Land and the northern Haihe Plain. These findings elucidate the spatially differentiated driving mechanisms of vegetation carbon sinks under climate and human impacts, providing a scientific basis for region-specific ecosystem management and the achievement of carbon neutrality goals. Future research incorporating higher-resolution data and advanced attribution frameworks is needed to reduce uncertainties associated with mixed-pixel effects in heterogeneous landscapes.