Frog: A Global Machine-Learning Temperature Calibration For Branched Gdgts In Soils And Peats
Frog: A Global Machine-Learning Temperature Calibration For Branched Gdgts In Soils And Peats
复制标题
DOI：
10.1016/j.gca.2021.12.007
复制
发表时间：
2021-12
影响因子：
5
通讯作者：
P. Véquaud;A. Thibault;S. Derenne;C. Anquetil;S. Collin;S. Contreras;A. Nottingham;Pierre Sabatier;J. Werne;A. Huguet
中科院分区：
文献类型：
--
作者：
P. Véquaud;A. Thibault;S. Derenne;C. Anquetil;S. Collin;S. Contreras;A. Nottingham;Pierre Sabatier;J. Werne;A. Huguet
支链甘油二烷基甘油四醚 (brGDGT) 是一类细菌脂质，随着时间的推移，它们已成为大陆环境中稳定的温度和 pH 值古代理。然而，之前的研究表明，除了温度和 pH 值之外的其他参数，例如土壤湿度、热状况或植被也会影响 brGDGT 在土壤和泥炭中的相对分布。这可以解释在这些设置中使用年平均气温 (MAAT) 和 pH 进行全球 brGDGT 校准中的大部分残余散射。尽管 brGDGT 分析方法有所改进并开发了精细模型，但与土壤和泥炭中 brGDGT 分布和 MAAT 之间的全局校准相关的均方根误差 (RMSE) 仍然很高（~5°C）。本研究的目的是使用机器学习算法根据全球扩展数据集（即 775 个土壤和泥炭样本，即在之前可用的全球校准中添加 112 个样本）开发新的全球陆地 brGDGT 温度校准。统计分析强调了除 MAAT 之外的潜在混杂因素对 brGDGT 相对丰度具有不同影响的五个簇。结果还揭示了使用单一指数和简单线性回归模型来捕获 brGDGT 对温度变化的响应的局限性。因此，提出了一种基于随机森林算法的新改进校准，即所谓的使用 brGDGT 的 PaleOMAAT 的随机森林回归（FROG）。这种多因素和非参数模型可以克服单一指数的使用，并通过考虑 MAAT 与各个 brGDGT 的相对丰度之间的非线性关系来更能代表环境的复杂性。 FROG 模型代表了针对土壤和泥炭的精确 brGDGT 温度校准（R2= 0.8；RMSE = 4.01 °C），比之前的全球土壤校准更稳健、更准确，同时在扩展数据集上提出。这种新颖的改进校准在分别涵盖过去 110 公里和上新世的两个古档案中得到了进一步应用和验证。
Branched glycerol dialkyl glycerol tetraethers (brGDGTs) are a family of bacterial lipids which have emerged over time as robust temperature and pH paleoproxies in continental settings. Nevertheless, it was previously shown that other parameters than temperature and pH, such as soil moisture, thermal regime or vegetation can also influence the relative distribution of brGDGTs in soils and peats. This can explain a large part of the residual scatter in the global brGDGT calibrations with mean annual air temperature (MAAT) and pH in these settings. Despite improvements in brGDGT analytical methods and development of refined models, the root-mean-square error (RMSE) associated with global calibrations between brGDGT distribution and MAAT in soils and peats remains high (∼5 °C). The aim of the present study was to develop a new global terrestrial brGDGT temperature calibration from a worldwide extended dataset (i.e. 775 soil and peat samples, i.e. 112 samples added to the previously available global calibration) using a machine learning algorithm. Statistical analyses highlighted five clusters with different effects of potential confounding factors in addition to MAAT on the relative abundances of brGDGTs. The results also revealed the limitations of using a single index and a simple linear regression model to capture the response of brGDGTs to temperature changes. A new improved calibration based on a random forest algorithm was thus proposed, the so-called randomForestRegression for PaleOMAAT using brGDGTs (FROG).This multi-factorial and non-parametric model allows to overcome the use of a single index, and to be more representative of the environmental complexity by taking into account the non-linear relationships between MAAT and the relative abundances of the individual brGDGTs. The FROG model represents a refined brGDGT temperature calibration (R2= 0.8; RMSE = 4.01 °C) for soils and peats, more robust and accurate than previous global soil calibrations while being proposed on an extended dataset. This novel improved calibration was further applied and validated on two paleo archives covering the last 110 kyr and the Pliocene, respectively.