Zhuhai Branch of State Key Laboratory of Earth Surface Process and Resource Ecology, Beijing Normal University, Zhuhai 519087, China.
State Key Laboratory of Cryospheric Science, Northwest Institute of Eco-Environment and Resources, Chinese Academy of Sciences, Lanzhou 730000, China.
Sci Total Environ. 2022 Jul 20;831:154761. doi: 10.1016/j.scitotenv.2022.154761. Epub 2022 Mar 24.
Sedimentary soil organic carbon (SOC) stored in thermokarst lakes and ponds (hereafter referred to as thaw lakes) across high-latitude/altitude permafrost areas is of global significance due to increasing thaw lake numbers and their high C vulnerability under climate warming. However, to date, little is known about the SOC storage in these lakes, which limits our better understanding of the fate of these active carbon in a warming future. Here, by combining large-scale field observation data and published deep (e.g., 0-300 cm) permafrost SOC data with a random forest (RF) machine learning technique, we provided the first comprehensive estimation of thaw lake SOC stocks to 3 m depth on the Tibetan Plateau. This study demonstrated that combining multiple environmental factors with the RF model could effectively predict the spatial distributions of the thaw lake SOC density values (SOCDs). The model results revealed that the soil respiration, normalized difference vegetation index (NDVI), and mean annual precipitation (MAP) were the most influential factors for predicting thaw lake SOCDs. In total, the sedimentary SOC stocks in the thaw lakes were approximately 52.62 Tg in the top 3 m, with 53% of the SOC stored in the upper layers (0-100 cm). The SOCDs generally exhibited high values in eastern Tibetan Plateau, and low values in mid- and western Tibetan Plateau, which were similar to the patterns of the land cover types that affected the SOCDs. We further found that the SOCDs of thaw lakes were generally higher than those of their surrounding permafrost soils at different layer depths, which could be ascribed to the erosion of soil particles or leaching solution from the thawing permafrost soils to lakes and/or enhanced vegetation growth at the lake bottom. This research highlights the necessity of explicitly considering the thaw lake SOC stocks in Earth system models for more comprehensive future projections of the carbon dynamics on the plateau.
高纬度/高海拔多年冻土区热喀斯特湖和池塘(以下简称融湖)中储存的沉积土壤有机碳(SOC)由于融湖数量的增加及其在气候变暖下的高碳脆弱性而具有全球意义。然而,迄今为止,人们对这些湖泊中的 SOC 储量知之甚少,这限制了我们更好地了解这些活跃碳在未来变暖中的命运。在这里,我们通过结合大规模野外观测数据和已发表的深层(例如,0-300cm)多年冻土 SOC 数据以及随机森林(RF)机器学习技术,首次对青藏高原融湖 SOC 储量进行了至 3m 深度的综合估计。本研究表明,结合多个环境因素和 RF 模型可以有效地预测融湖 SOC 密度值(SOCD)的空间分布。模型结果表明,土壤呼吸、归一化差异植被指数(NDVI)和年平均降水量(MAP)是预测融湖 SOCD 的最主要影响因素。总的来说,在 3m 深度内,融湖中沉积的 SOC 储量约为 52.62Tg,其中 53%的 SOC 储存在上层(0-100cm)。SOCD 一般在青藏高原东部较高,在中、西部较低,这与影响 SOCD 的土地覆盖类型模式相似。我们进一步发现,在不同层深处,融湖的 SOCD 一般高于其周围多年冻土土壤的 SOCD,这可能归因于融冻土壤颗粒的侵蚀或淋溶溶液向湖泊的迁移以及/或湖底植被生长的增强。本研究强调,在地球系统模型中明确考虑融湖 SOC 储量对于更全面地预测高原上的碳动态是必要的。