A System Reanalysis of the Current Greenhouse Gases Budget of Terrestrial Ecosystems in Russia

作者：А. Shvidenko, Philippe Ciais, Prabir K. Patra, Ana Bastos, Shamil Maksyutov, Ronny Lauerwald, Benjamin Poulter, Dmitry Belikov, Naveen Chandra, М. В. Глаголев, Irina Terentieva, Д. В. Карелин, Juliya Kurbatova, I. N. Kurganova, A. A. Romanovskaya, В. Н. Коротков, Liudmila Mukhortova, Anatoly Prokushkin, Eric J. Gustafson, F. Kraxner, Vadim Mamkin, Н. В. Лукина, Andrey Krasovskiy, Еugene А. Vaganov, Dmitry Schepaschenko · 发表于：Global Biogeochemical Cycles · 年份：2025 · DOI：10.1029/2025gb008540 · 被引用次数：6 · 研究领域：Atmospheric and Environmental Gas Dynamics、Plant Water Relations and Carbon Dynamics、Science and Climate Studies

Abstract This study synthesizes the budgets of three greenhouse gases (GHG, namely CO 2 , CH 4 , N 2 O) for Russia over two decades (2000–2009 and 2010–2019) using bottom‐up and top‐down approaches, as part of the Regional Carbon Cycle Assessment and Processes, Phase 2 (RECCAP2). Published estimates of natural sources and sinks of these GHGs in Russia vary widely. Here, bottom‐up estimates are based on eddy covariance measurements, the Integrated Land Information System of Russia (ILIS‐LEA), field data, Dynamic Global Vegetation Models (DGVMs), and regional models. The bottom‐up approach estimated Net Ecosystem Exchange (NEE) at −0.64 ± 0.17 and −0.57 ± 0.14 Pg C yr −1 , for decades 2000–2009 and 2010–2019, respectively. Top‐down atmospheric inversions provide similar NEE carbon flux estimates with comparable uncertainties at −0.56 ± 0.26 and −0.73 ± 0.27 Pg C yr −1 for the two decades. Differences between these approaches arise from distinct flux components and structural assumptions. ILIS‐LEA indicates a slightly declining carbon sink in 2010–2019, driven by increased disturbances. In contrast, DGVMs suggest a stable carbon sink over both decades but they do not fully simulate the effects of disturbances and recovery. Top‐down inversions reveal an increasing CO 2 sink, suggesting with additional observed constraints on biomass carbon increment that soil and non‐forest biomes absorb more carbon than predicted by DGVMs and ILIS‐LEA models. A Bayesian averaging approach estima...