Lake sediments are key long-term organic carbon sinks, but spatiotemporal heterogeneity, sources, and shifting controls of burial in large arid-zone lakes remain poorly quantified. This study presents a systematic multicore, multiproxy investigation of Bosten Lake (arid NW China), integrating Bayesian chronology, grain size, stable isotopes (δ¹³Corg), C/N ratio, and MixSIAR to reconstruct ~150-year spatiotemporal organic carbon burial dynamics. Temporal phases were identified via BEAST and CONISS, while Bayesian regression disentangled the relative importance and uncertainty of climatic, hydrological, and anthropogenic drivers. Results reveal strong spatial heterogeneity in organic carbon burial rates (OCBR). The outlet exhibited exceptionally high and variable burial (70.18–198.35 g m⁻² yr⁻¹), whereas profundal sites showed lower, more stable values (25.56–31.38 g m⁻² yr⁻¹). Despite its arid setting, Bosten Lake demonstrated high burial efficiency, exceeding previously reported global averages for lakes. MixSIAR identified two dominant compositional sources: nearshore sediments were dominated by topsoil inputs (>60%), while deep-water sediments were dominated by autochthonous phytoplankton (>60%). Temporally, whole-lake OCBR evolved in three phases: low before the 1980s, moderate increase during 1980–2000, and sharp acceleration after ~2000 CE. Bayesian regression identified that pre-2000 variability was dominated by natural climate forcing, with temperature as a primary positive driver in most cores. Our findings highlight that reliable carbon-budget assessments in large arid lakes require multicore designs to resolve internal heterogeneity. Furthermore, recent anthropogenic factors and hydrological alterations can rapidly amplify lacustrine carbon sequestration, with important implications for regional carbon cycling and management under ongoing warming and land-use change.
Rapid urbanization is increasingly recognized as an important influence on regional climate change, yet its effects remain difficult to quantify. Using multitemporal land-use data and daily climate records over past decades, we developed a grid-to-grid nonstationary framework to quantify the effects of urbanization on regional climate extremes in the Minjiang River Watershed (MRW), the largest watershed in Southeast China. The results indicated that urbanization in the MRW has accelerated since 2000, modifying both precipitation and temperature extremes, with the strongest changes occurring in the highly urbanized southeast. The variability of 3-day precipitation (Rx3day) exhibited an inverse trend, characterised by a significant increase from 1980 to 2000 and a subsequent slight decrease from 2001 to 2020, within specific areas. Long-duration temperature indices, especially Tx90day, changed from decreasing or weak trends during 1980–2000 to pronounced increases in the highly urbanized area during 2001–2020. Precipitation responses remained spatially heterogeneous, but long-duration indices showed larger changes in highly urbanized areas. Notably, the 90-day precipitation (Rx90day) has increased by more than 25% in highly urbanized areas.The proposed method quantified the impacts of urbanization on local or regional climate and would be helpful for understanding the relationship between human activities and climate change.
Human activities are reshaping the background state of the climate system, with potential consequences for the spatial organization of atmospheric variability across distant regions and different vertical levels. Atmospheric teleconnections constitute an important component of this interconnected Earth system, yet their three-dimensional organization and recent evolution remain insufficiently understood. To address this issue, we constructed multi-layer climate networks, including both within-layer and cross-layer networks, using atmospheric temperature anomalies at multiple pressure levels from ERA5 reanalysis data. By analyzing lagged correlations, we characterized horizontal teleconnections and cross-layer statistical coupling across different atmospheric layers, with particular attention to the lower troposphere (850 hPa) and upper troposphere (200 hPa). The results show scale-dependent spatial patterns consistent with planetary-scale wave-like variability and reveal a systematic redistribution of positive and negative connectivity scales between the lower and upper troposphere. Sliding-window analyses further indicate an increase in temperature-based statistical connectivity within and between atmospheric layers over recent decades. These findings provide a network-based perspective on the evolving three-dimensional statistical organization of atmospheric connectivity during recent decades of pronounced anthropogenic climate change. The observed changes in atmospheric connectivity provide complementary information for understanding how large-scale atmospheric variability is evolving in a changing climate.
One defining feature of the Anthropocene is the unprecedented entanglement of global food systems, in which international agricultural trade acts as a primary anthropogenic force reshaping land, water, and biogeochemical cycles across continents. The China-Brazil agricultural nexus exemplifies this tele-coupled human-Earth interaction, yet the environmental costs, particularly nitrogen (N) pollution and water scarcity, are often asymmetrically distributed between importing and exporting regions. Here, we investigate how human-driven trade restructuring, combined with technological and spatial interventions, can modulate these pressures within planetary boundaries. By integrating substance flow analysis with a constrained optimization framework, we quantify the N burdens embedded in Brazil to China feed trade and evaluate a counterfactual shift toward direct animal-product trade. We find that current feed trade supports Chinese livestock production at the cost of 0.74 Tg N lost to aquatic systems in China, primarily due to inefficient human-managed manure cycles. Transitioning to animal-product trade reduces bilateral N losses by 33%, but transfers production-side environmental pressures to Brazil, which represents a classic Anthropocene pollution displacement dilemma. However, through deliberate human interventions, including advanced manure recycling, high-efficiency crops, and spatial planning that aligns livestock with regional water availability, Brazil could reduce its cropland use by 28% and N losses by 47%, cutting synthetic fertilizer demand by up to 82%. Our findings demonstrate that human agency, expressed through coordinated trade policy and production optimization, can substantively mitigate the environmental externalities of globalized food systems. Nevertheless, these outcomes represent theoretical mitigation potentials under modeled constraints, and realizing them requires multilevel governance to address transboundary environmental justice and avoid unintended leakage in multilateral markets.
