The impact of change in fluid seepage intensity on iron and phosphorus cycling in chimney-like seep carbonates
-
摘要:
冷泉碳酸盐岩形成过程中渗漏流体特征的变化可以改变参与的碳、硫等元素循环的生物地球化学过程,但对铁(Fe)和磷(P)元素循环及相关生物地球化学过程的影响目前还不清楚。本文对南海东沙海域一个冷泉碳酸盐岩烟囱横截面进行了矿物学、碳和硫地球化学及Fe和P组分进行了分析。结果显示,生物成因甲烷的厌氧氧化作用导致了极端13C亏损的自生碳酸盐的形成(δ13CVPDB<−55.5‰),并且形成过程中由于水合物分解,加入了富18O 流体,使得烟囱碳酸盐岩表现出异常高的氧同位素组成(5.5‰~5.8‰)。烟囱外层矿物组成以方解石为主,内层以文石为主,内层烟囱样品形成于更强的甲烷渗漏强度条件。样品中最主要的含磷物质是自生磷酸盐(PAuth),并且最高含量出现在烟囱内层。相比外层烟囱样品,内层铁氧化物含量更低,但是铁结合态磷(PFe)含量更高,推测可能是有蓝铁矿的形成。综合上述结果表明,在形成冷泉碳酸盐岩的强渗漏条件下,甲烷渗漏强度的增强会影响PAuth和PFe的丰度,可能更有利于蓝铁矿和自生磷灰石的形成,从而对冷泉环境中的磷和铁循环产生影响。
Abstract:During the formation of seep carbonates, variations in fluid seepage affect the biogeochemical processes that drive the cycling of various elements, such as carbon and sulfur. However, such an influence on iron (Fe) and phosphorus (P) cycling and related biogeochemical processes remains unclear. In this study, mineral compositions, carbon and sulfur geochemistry, Fe and P component were investigated on a cross section of a chimney-like seep carbonate sample collected from the Dongsha area of the South China Sea. Results reveal that the anaerobic oxidation of biogenic methane led to the formation of the extremely 13C-depleted authigenic carbonate rocks (δ13CVPDB<−55.5‰), and the addition of enriched 18O fluids from the dissociation of gas hydrates results in abnormally high oxygen isotope composition (5.5‰–5.8‰) observed in the chimney carbonates. Combined with the dominant minerals with calcite in the outer layer of the chimney carbonates and aragonite in the inner layer, we inferred that the carbonates in the inner layer were formed at the later stage under stronger methane seepage condition. The most significant P-bearing mineral in the samples is authigenic phosphate (PAuth), and the highest content is in the inner layer of the chimney carbonates. Compared with the outer chimney sample, the inner layer has lower Fe oxide content, but higher Fe-bound P (PFe) content, indicating the formation of vivianite. Therefore, we believe that under a stronger seepage condition during the formation of seep carbonates, an enhanced intensity of methane seepage can affect the abundance of PAuth and PFe, which may favor the formation of vivianite and authigenic phosphates, thereby influencing Fe and P cycling in cold-seep environments.
-
Keywords:
- methane seepage /
- phosphorus component /
- iron component /
- phosphorus cycle /
- seep carbonate
-
天然气水合物是一种似冰状固体化合物,由水分子和气体分子在低温高压条件下生成,具有可燃性,因此也被称为“可燃冰”[1]。水合物的燃烧效率极高,并且污染小,被视为是未来化石燃料的替代能源,美国、日本、韩国、印度等国家均已制定国家水合物勘探研究计划[2-4]。中国在近年来水合物勘探开发上也取得了较为显著的成效,2017年神狐海域的成功试采标志着中国成为全球第一个实现了在海域可燃冰开采中获得连续稳定产气的国家,同年,水合物被列为中国第173个矿种[5-9]。水合物在全球分布的范围十分广泛(图1),主要存在于海底沉积层和永久冻土层中。据估算,世界可燃冰总资源量大约相当于全球已知煤、石油和天然气总资源量的2倍,储量十分可观[10]。然而在复杂海底地质环境下勘探水合物具有相当大的困难,因此,准确勘探识别水合物和估算水合物储量一直以来都是国内外的研究热点问题。
地震勘探技术一直以来都是水合物识别与评价的重要地球物理方法[11]。早期通过识别地震剖面上的似海底反射层(BSR)来判断地层中的天然气水合物稳定带底界,BSR通常呈现出与海底平行、极性相反、高振幅强反射、与沉积地层相交等特征, 但地震剖面上的BSR与水合物分布并不是—对应的[12-14]。由于水合物的存在会导致地层的速度升高,所以有学者提出用速度异常的方法来识别水合物,但该方法仍有一定的问题,如有其他矿物充填对应的测井曲线上也有速度异常升高,因此速度异常的方法也存在指示不明的问题。后来,Liu等[15]提出利用纵横波速度增量比的方法来识别地层中的水合物,在布莱克海台ODP164航次的实际数据中取得较好的应用效果。Tian 和Liu[16]发现通过纵横波速度交汇图的方法也可以区分出含水合物地层,以及用纵横波速度增量交汇图的方法估算水合物饱和度范围,并在俄勒冈水合物海岭的实际数据中进行验证,结果与实际情况相一致。传统的水合物识别方法往往需要综合的判断分析,处理解释的时间长,水合物饱和度计算准确率低,亟待提出一套高效准确的水合物识别计算方法。随着人工智能的高速发展,机器学习作为人工智能领域的核心算法,分为监督学习、非监督学习和深度学习,无论是在理论研究还是实践应用方面都取得了巨大的突破[17-18]。目前通过机器学习的方法也解决了很多地质问题,例如,王迪等[19]利用深度学习方法定量地进行了烃类预测;Bai和Tan[20]采用机器学习方法进行储层参数预测并取得较好结果;陈钢花等[21]采用了双向长短时记忆神经网络有效解决了致密滩坝的储层岩性识别问题;Chen等[22]利用径向基函数神经网络对神狐海域水合物饱和度进行预测得到了较好的应用效果;Zhu等[23]基于机器学习方法对水合物的赋存类型进行了识别并估算了水合物饱和度。因此,本文考虑在中国南海海域借助人工智能的手段来进一步高效准确识别水合物资源。
1. 研究方法
本研究主要通过几种不同算法,包括AdaBoost算法、随机森林算法、Bagging算法、和最近邻(KNN)算法,对测井数据中的纵波速度和密度数据进行模型训练和测试,从而区分含水合物地层与非含水合物地层。
1.1 算法选择
1.1.1 AdaBoost算法
AdaBoost是基于Boosting的一种集成学习算法。该算法的核心思想是通过调整样本分布和弱分类器权值,将一系列弱分类器集成为一个分类精度高的强分类器,改变分类器在已有样本分布上的优势,提高泛化能力(模型在未曾见过的数据上的表现能力,即模型对于新的数据的适应能力)。AdaBoost算法通过改变训练集中各样本的权重实现迭代过程,根据每次训练集中各样本是否分类正确及上次总体分类的准确率,综合确定各样本的权重,把修改过权重的新数据集送给下层分类器进行训练,并将每次训练所得分类器融合起来,形成最终的决策分类器[24]。
1.1.2 随机森林算法
随机森林(Random Forest,简称RF)是一种集成多棵决策树的有监督学习算法。RF以决策树为基学习器构建Bagging集成的基础上,进一步在决策树的训练过程中引入了随机属性选择[25]。随机森林使用 booststrap 方法构建n个训练集,每个训练集对应生成一个决策树,总体就有n个决策树,因为每个决策树的数据集都不相同,所以每棵树又有少量区别。最后对所有的决策树的预测结果取平均减少预测的方差,提高在测试集上的性能表现。该算法具有简单、容易实现、计算开销小等特点,在很多现实任务中展现出强大的性能。
1.1.3 Bagging算法
Bagging算法是一种用来提高弱分类器准确率的集成方法,该算法对训练集实施有放回的随机样本重新采样,形成多个与训练集规模相近但各不相同的训练子集,在此基础上形成多个基分类器,在多个基分类器的基础上使用多数投票原则[26-27]。Bagging方法具有较大的随机性,但当采样次数足够多时会生成足够丰富的基本模型,可以得到十分精确的结果。通常在减小方差方面Bagging方法效果明显,因此Bagging方法对训练集微小变化敏感的模型可以提高其稳定性。
1.1.4 KNN算法
KNN算法是一种有监督的机器学习算法,通过遍历训练集样本,计算测试数据与训练样本点属性间的距离,找到与测试数据最近的K个样本点,采取少数服从多数的原则确定测试数据的类别[28]。KNN算法尤其适用于训练样本量较大且样本之间的差异性相对显著的海量数据分类,优势明显,准确度非常高。
1.2 评价指标
为了更好地确定水合物地层识别模型的预测精度和泛化能力,利用准确率、精确率、召回率、F1分数、ROC曲线来对分类模型进行性能评估。
识别水合物的过程是一个二分类问题,结果只有真和假两种。一般用T(True)表示预测结果正确(与实际相符),F(False)表示预测结果不正确(与实际不符),P(Positive)表示真实的正样本,N(Negative)表示真实的负样本,因此预测结果分为四种,如表1所示:
表 1 评价结果指标意义Table 1. Significance of evaluation results indicators预测为真 预测为负 实际为真 TP FN 实际为负 FP TN 准确率(Accuracy)用来衡量分类正确的比例,虽然准确率可以判断总的正确率,但在样本不平衡的情况下,并不能作为很好的指标来衡量分类结果:
$$ {\text{Accuracy = }}\frac{{{\text{TP}} + {\text{TN}}}}{{{\text{TP}} + {\text{TN}} + {\text{FP}} + {\text{FN}}}} $$ (1) 精确率(Precision)是指预测为正类样本数中实际为正类的比例:
$$ {\text{Precision}} = \frac{{{\text{TP}}}}{{{\text{TP}} + {\text{FP}}}} $$ (2) 召回率(Recall)是度量有多少个正类样本数被判定为正类:
$$ {\text{Recall}} = \frac{{{\text{TP}}}}{{{\text{TP}} + {\text{FN}}}} $$ (3) 精确率与召回率可能是相互矛盾的,因此需要对二者进行综合考虑,因此引入F分数,定义如下:
$$ {F_\beta } = \left( {1 + {\beta ^2}} \right) \cdot \rm\frac{{precision \cdot recall}}{{{\beta ^2} \cdot precision + recall}} $$ (4) 当式中β=1时,F分数为F1分数。F1是精确率和召回率的调和均值。当式中β>1时,召回率的权重高于精确率。相反,当β<1时精确率的权重高于召回率。
接受者操作特征(ROC)曲线是反映模型敏感性与特异性连续变化的综合指标。该曲线的横坐标为预测为正但实际为负的样本占所有负例样本的比例;曲线的纵坐标为预测为正且实际为正的样本占所有正例样本的比例。
$$\rm {{TPR}} = \frac{{TP}}{{TP + FN}} $$ (5) $$\rm {{FPR}} = \frac{{FP}}{{FP + TN}} $$ (6) ROC曲线下与坐标轴围成的面积(AUC)值为ROC曲线下与坐标轴围成的面积,是衡量算法优劣的一种性能指标,AUC值越接近1,则算法越优秀。
2. 实际应用
2.1 研究区概况
中国南海东沙海域位于南海东北部陆坡区(图2),该区域地理位置独特,海域沉积活动和构造运动相对复杂,具有被动大陆边缘特点,是天然气水合物生成的有利富集区,海底存在甲烷渗漏现象,冷泉发育活跃。2013年,广州海洋地质调查局开展了第二次天然气水合物勘查钻探项目(GMGS2),在该研究区进行了一系列的地质、地球物理、地球化学等方面的调查,并获取了大量的块状、脉状、层状、分散状及结核状等多种不同类型的水合物样品,孔隙型水合物和裂隙型水合物也被证实存在于研究区内,此次勘查在中国海域天然气水合物勘探中具有重要意义。在GMGS2-05、GMGS2-08和GMGS2-16钻井位置分别采集到了测井数据,因此本文采用这3个站位处的数据用来训练和测试水合物智能识别方法。
前期勘探已在05、08和16站位发现有天然气水合物,根据随钻测井获得了大量的测井数据,图3分别给出了05、08和16站位的纵波速度和密度测井曲线,蓝色区域为含水合物地层,本论文主要利用05、08和16站位测井数据作为学习数据集进行实验分析。在05站位,在海底以下约193~206.5 m 发现了大量的孔隙型水合物,并伴随有高速度和高电阻率异常特征;在08站位,发现有两层裂隙型水合物分别在海底以下约9~22和65~98 m;在16站位,裂隙型水合物出现在海底以下10~20 m,孔隙型水合物BSR在海底以下约189~198 m位置[27]。
2.2 模型调参和训练
在机器学习模型的训练过程中,为了使模型具有更好的泛化能力,将数据集分割为训练集和测试集。本实验选取05、08和16站位的测井数据,以深度位置计数分别取速度和密度点各4 247个建立水合物识别样本数据库(图3),其中抽取75%作为训练样本,然后将剩余25%数据点作为测试样本。训练集用于训练模型,测试集用于检验模型的泛化能力。从常规测井系列中优选对岩性变化敏感的纵波速度和密度作为水合物识别模型的输入特征,含水合物地层和非含水合物地层两种类别作为输出特征。
此外,合适的监督模型参数是提高分类器预测精度的关键。为了使上述机器学习监督模型在其它未知的例子上也表现出较好的分类性能,还需要调整模型在训练过程之前所设置的参数(即超参数)获取适当的参数值。但是,训练后的模型不能过于拟合训练集,否则会失去对测试数据的预测能力,导致出现过拟合。可以通过交叉验证结果生成的验证分数选择不引起模型过拟合或欠拟合的最优参数集。网格搜索是模型参数调试的常用方法,通过选择合理的参数范围网格搜索能够找出模型的最优参数组合。在本案例研究中,利用网格搜索对模型的超参数进行调试,试验不同的超参数范围,使用交叉验证来评估超参数值的所有可能组合,以获得泛化能力较好的改进模型。本文以准确度作为衡量分类性能的指标。对于每个模型,需要调优的参数见表2。
表 2 每个模型的超参数调整Table 2. Hyperparameter tuning for each algorithm监督模型 超参数 调整范围 最优参数 AdaBoost N estimators 5~25 15 Learning rate 0.1~0.9 0.5 RF max leaf nodes 5~45 10 n estimators 0~150 30 Bagging n estimators 10~500 100 Max samples 50~500 150 KNN N neighbors 5~200 10 weights uniform/distance uniform 3. 结果分析
应用不同机器学习模型对南海东沙海域的05、08和16站位的测井数据样本进行水合物识别分析,首先将数据样本中的已知为含水合物地层深度位置的数据标记为1,非含水合物地层深度位置的数据标记为0,识别结果如下:
(1) AdaBoost算法的分类效果
在训练AdaBoost分类器时,该算法首先训练一个基础分类器,然后进行集成,先选择决策树作为基础分类器,也就是弱学习器。其中基础分类器决策树算法的最大深度(max depth)为1,弱学习器的最大迭代次数(n estimators)为15,学习率为0.5,Scikit-Learn 默认使用SAMME.R,对样本集分类效果作为弱学习器权重,由于SAMME.R使用了概率度量的连续值,迭代一般比SAMME快。调整参数后的模型对测试数据进行分类,我们得到准确率、精确率、召回率和F1分数分别是
0.9171 、0.8625 、0.4726 和0.6106 。此外,AUC值为0.9146 。用AdaBoost 算法对05、08和16站位纵波速度和密度进行分类训练和测试,结果如图4,其中红色点表示含水合物地层纵波速度和密度交点,蓝色点表示非含水合物地层纵波速度和密度交点,中间的深绿色交线为算法的决策边界,决策边界左边为含水合物地层(绿色区域),决策边界右边为非含水合物地层(黄色区域)。该算法较好地将水合物与非水合物地层进行分类,算法稳定可靠。(2)随机森林算法分类效果
随机森林算法将多个决策树组合在一起,随机选取不同的数据集,以便输出相似但不相同的模型结果,再将所有决策树结果整合在一起,作为输出结果,而这一训练方式,意味着很难过拟合,并且对噪音不敏感。通过网格搜索调整该模型超参数,确定决策树的个数(n estimators)为30,通过限制最大叶子节点数,防止过拟合,最大叶节点数设置为10。调整参数后的模型对测试数据进行分类,得到准确率、精确率、召回率和F1分数分别为
0.9388 ,0.8857 、0.6370 和0.7410 ,此外,AUC值为0.9278 。用随机森林算法对05、08和16站位纵波速度和密度进行分类训练和测试,结果如图5,其中红色点表示含水合物地层纵横波速度交点,蓝色点表示非含水合物地层纵横波速度交点,中间的深绿色交线为算法的决策边界,决策边界左边为含水合物地层(绿色区域),决策边界右边为非含水合物地层(黄色区域),该算法在水合物的识别问题上表现出很好的分类性能,算法稳定可靠,有很强的抗干扰能力、抗过拟合和泛化能力。(3)Bagging算法分类性能
在本算法中我们训练多个决策树去组成bagging集成算法,经网格搜索调参后,确定该算法包含100个决策树分类器(n estimators=100),每次从训练集中随机采样150个训练实例(max samples=150)进行训练,然后放回。一旦预测器训练完成,集成就可以通过简单地聚合所有预测器的预测来对新实例做出预测。最后通过该模型对测试数据的预测结果,得到准确率、精确率、召回率和F1分数分别为
0.9341 、0.8725 、0.6096 和0.7177 ,此外,AUC值为0.9449 。用Bagging算法对05、08和16站位纵波速度和密度进行分类训练和测试,结果如图6,其中红色点表示含水合物地层纵横波速度交点,蓝色点表示非含水合物地层纵横波速度交点,中间的深绿色交线为算法的决策边界,决策边界左边为含水合物地层(绿色区域),决策边界右边为非含水合物地层(黄色区域),该算法性能依赖于基分类器(本模型中使用决策树)的稳定性,能够较好地将水合物与非水合物地层进行区分。(4)KNN算法分类性能
在KNN算法中我们主要调整k值和选择距离的权重方式,k值越小越容易过拟合,当k=1时,这时只根据单个近邻进行预测,如果离目标点最近的一个点是噪声,就会出错,此时模型复杂度高,稳健性低,决策边界崎岖。但是如果k值取的过大,这时与目标点较远的样本点也会对预测起作用,就会导致欠拟合,此时模型变得简单,决策边界变平滑。在sklearn中k近邻算法还有一个weights参数,其默认值为uniform,此时是不考虑距离权重的,当weights=distance时,sklearn就会考虑距离的权重,权重和距离成反比,距离预测目标越近具有越高的权重。通过网格搜索最后确定模型的k值为200,weights=distance,用该模型对测试数据进行分类,我们得到准确率、精确率、召回率和F1分数分别为
0.9379 、0.8636 、0.6507 和0.7422 ,此外,AUC值为0.9580 用KNN算法对05、08和16站位纵波速度和密度进行分类训练和测试,结果如图7,其中红色点表示含水合物地层纵横波速度交点,蓝色点表示非含水合物地层纵横波速度交点,中间的深绿色交线为算法的决策边界,决策边界左边为含水合物地层(绿色区域),决策边界右边为非含水合物地层(黄色区域),该算法能够很好地将水合物与非水合物地层进行分类,准确率高,AUC值在这4种算法中最高(图8),因此该算法在该地区的水合物分类中所表现的性能最好。4. 结论与建议
本文通过4种机器学习算法对南海东沙海域GMGS2-05、GMGS2-08和GMGS2-16站位的纵波速度和密度数据进行训练,不断调整超参数选值确定最优模型来对水合物和非水合物点分类识别,引入准确率、精确率、召回率和F1分数分别是对识别结果进行评价(表3),此外用AUC值对模型性能进行更直观的评价。识别结果表明,① 与其他机器学习算法相比,随机森林算法在测试水合物分类过程中的准确率、精确率和最高,KNN算法的召回率和F1分数最高。② KNN算法的AUC值是最接近1的,也就是说该算法性能相对来说比较优秀。③ 这些算法都能较好地通过纵波速度和密度属性数据来识别水合物,为天然气水合物的高效准确识别提供了技术支撑。利用人工智能技术能够有效解决水合物研究中面临的难点问题,今后仍需在算法和输入地震属性特征方面进行更加深入的研究,不断优化模型算法,当前水合物在自然界中的赋存类型多样,后续值得应用人工智能手段进一步细化研究区分水合物的赋存状态的智能识别方法。
表 3 各算法评价指标计算结果Table 3. Calculation results of evaluation indicators of each algorithmSupervision algorithm Accuracy Precision Recall F1 值 AUC AdaBoost 0.9171 0.8625 0.4726 0.6106 0.9146 Random forest 0.9388 0.8857 0.6370 0.7410 0.9278 Bagging 0.9341 0.8725 0.6096 0.7177 0.9449 KNN 0.9379 0.8636 0.6507 0.7422 0.9580 -
图 1 采样位置和样品情况
a: 位于东沙海域的采样站位九龙甲烷礁site 3;b: 冷泉碳酸盐岩烟囱样品横切面和取样点S1到S10。冷泉碳酸盐岩烟囱的横截面直径约10 cm,用牙钻间隔约0.8 cm取一个点,共取10个点。
Figure 1. Sampling location and subsamples
a: The sampling site located in the Dongsha sea area, known as Methane Reef Site 3; b: the subsample of seep carbonates chimneys from S1to S10. The cross-section diameter of the seep carbonates chimneys is about 10 cm, and one point was taken about 0.8 cm apart for 10 points in total.
图 6 冷泉碳酸盐岩烟囱样品的晶质铁氧化物(FeOx2)和非晶质氧化铁含量(FeOx1)分别与铁结合态磷(PFe)的投点图
图中红色点为冷泉碳酸盐岩烟囱内层样品,黑色点为冷泉碳酸盐岩烟囱外层样品。
Figure 6. The cross-plot of crystalline iron oxides (FeOx2) and amorphous iron oxide content (FeOx1) from cold seep carbonate chimney samples vs iron-bound phosphorus (PFe), respectively
The red dots are seep carbonate chimney samples. The black dots are samples from the outer layers of seep carbonates.
-
[1] Peckmann J, Thiel V. Carbon cycling at ancient methane–seeps[J]. Chemical Geology, 2004, 205(3-4):443-467. doi: 10.1016/j.chemgeo.2003.12.025
[2] Feng D, Qiu J W, Hu Y, et al. Cold seep systems in the South China Sea: An overview[J]. Journal of Asian Earth Sciences, 2018, 168:3-16. doi: 10.1016/j.jseaes.2018.09.021
[3] Lu Y, Sun X M, Lin Z Y, et al. Cold seep status archived in authigenic carbonates: Mineralogical and isotopic evidence from Northern South China Sea[J]. Deep Sea Research Part II:Topical Studies in Oceanography, 2015, 122:95-105. doi: 10.1016/j.dsr2.2015.06.014
[4] Gong S G, Hu Y, Li N, et al. Environmental controls on sulfur isotopic compositions of sulfide minerals in seep carbonates from the South China Sea[J]. Journal of Asian Earth Sciences, 2018, 168:96-105. doi: 10.1016/j.jseaes.2018.04.037
[5] Feng D, Chen D F, Roberts H H. Petrographic and geochemical characterization of seep carbonate from Bush Hill (GC 185) gas vent and hydrate site of the Gulf of Mexico[J]. Marine and Petroleum Geology, 2009, 26(7):1190-1198. doi: 10.1016/j.marpetgeo.2008.07.001
[6] Mazzini A, Svensen H, Hovland M, et al. Comparison and implications from strikingly different authigenic carbonates in a Nyegga complex pockmark, G11, Norwegian Sea[J]. Marine Geology, 2006, 231(1-4):89-102. doi: 10.1016/j.margeo.2006.05.012
[7] Ritger S, Carson B, Suess E. Methane-derived authigenic carbonates formed by subduction-induced pore-water expulsion along the Oregon/Washington margin[J]. GSA Bulletin, 1987, 98(2):147-156. doi: 10.1130/0016-7606(1987)98<147:MACFBS>2.0.CO;2
[8] Roberts H H, Aharon P. Hydrocarbon-derived carbonate buildups of the northern Gulf of Mexico continental slope: A review of submersible investigations[J]. Geo-Marine Letters, 1994, 14(2-3):135-148. doi: 10.1007/BF01203725
[9] Campbell K A. Hydrocarbon seep and hydrothermal vent paleoenvironments and paleontology: past developments and future research directions[J]. Palaeogeography, Palaeoclimatology, Palaeoecology, 2006, 232(2-4):362-407. doi: 10.1016/j.palaeo.2005.06.018
[10] Paull C K, Chanton J P, Neumann A C, et al. Indicators of methane-derived carbonates and chemosynthetic organic carbon deposits: Examples from the Florida Escarpment[J]. Palaios, 1992, 7(4):361-375. doi: 10.2307/3514822
[11] Feng D, Chen D F. Authigenic carbonates from an active cold seep of the northern South China Sea: New insights into fluid sources and past seepage activity[J]. Deep Sea Research Part II:Topical Studies in Oceanography, 2015, 122:74-83. doi: 10.1016/j.dsr2.2015.02.003
[12] Wirsig C, Kowsmann R O, Miller D J, et al. U/Th-dating and post-depositional alteration of a cold seep carbonate chimney from the Campos Basin offshore Brazil[J]. Marine Geology, 2012, 329-331:24-33. doi: 10.1016/j.margeo.2012.10.001
[13] Peng X T, Guo Z X, Chen S, et al. Formation of carbonate pipes in the northern Okinawa Trough linked to strong sulfate exhaustion and iron supply[J]. Geochimica et Cosmochimica Acta, 2017, 205:1-13. doi: 10.1016/j.gca.2017.02.010
[14] Takeuchi R, Matsumoto R, Ogihara S, et al. Methane-induced dolomite “chimneys” on the Kuroshima Knoll, Ryukyu islands, Japan[J]. Journal of Geochemical Exploration, 2007, 95(1-3):16-28. doi: 10.1016/j.gexplo.2007.05.008
[15] Magalhães V H, Pinheiro L M, Ivanov M K, et al. Formation processes of methane-derived authigenic carbonates from the Gulf of Cadiz[J]. Sedimentary Geology, 2012, 243-244:155-168. doi: 10.1016/j.sedgeo.2011.10.013
[16] Peckmann J, Reimer A, Luth U, et al. Methane-derived carbonates and authigenic pyrite from the northwestern Black Sea[J]. Marine Geology, 2001, 177(1-2):129-150. doi: 10.1016/S0025-3227(01)00128-1
[17] Novikova S A, Shnyukov Y F, Sokol E V, et al. A methane-derived carbonate build-up at a cold seep on the Crimean slope, north-western Black Sea[J]. Marine Geology, 2015, 363:160-173. doi: 10.1016/j.margeo.2015.02.008
[18] 苏新, 陈芳, 陆红锋, 等. 南海北部深海甲烷冷泉自生碳酸盐岩显微结构特征与流体活动关系初探[J]. 现代地质, 2008, 22(3):376-381 SU Xin, CHEN Fang, LU Hongfeng, et al. Micro-textures of methane seep carbonates from the northern South China Sea in correlation with fluid flow[J]. Geoscience, 2008, 22(3):376-381.]
[19] 杨克红, 初凤友, 赵建如, 等. 南海北部冷泉碳酸盐岩矿物微形貌及其意义探讨[J]. 矿物学报, 2009, 29(3):345-352 YANG Kehong, CHU Fengyou, ZHAO Jianru, et al. Minerals’ Micro-Shape and Its Significance of Seep Carbonates in the North of the South China Sea[J]. Acta Mineralogica Sinica, 2009, 29(3):345-352.]
[20] Liang Q Y, Hu Y, Feng D, et al. Authigenic carbonates from newly discovered active cold seeps on the northwestern slope of the South China Sea: Constraints on fluid sources, formation environments, and seepage dynamics[J]. Deep Sea Research Part I:Oceanographic Research Papers, 2017, 124:31-41. doi: 10.1016/j.dsr.2017.04.015
[21] 邬黛黛, 吴能友, 叶瑛, 等. 南海北部陆坡九龙甲烷礁冷泉碳酸盐岩沉积岩石学特征[J]. 热带海洋学报, 2009, 28(3):74-81 WU Daidai, WU Nengyou, YE Ying, et al. Petrographic characteristics of authigenic carbonates from Jiulong methane reef of northern South China Sea[J]. Journal of Tropical Oceanography, 2009, 28(3):74-81.]
[22] Guan H X, Feng D, Wu N Y, et al. Methane seepage intensities traced by biomarker patterns in authigenic carbonates from the South China Sea[J]. Organic Geochemistry, 2016, 91:109-119. doi: 10.1016/j.orggeochem.2015.11.007
[23] Ge L, Jiang S Y, Yang T, et al. Glycerol ether biomarkers and their carbon isotopic compositions in a cold seep carbonate chimney from the Shenhu area, northern South China Sea[J]. Chinese Science Bulletin, 2011, 56(16):1700-1707. doi: 10.1007/s11434-011-4486-z
[24] 佟宏鹏, 冯东, 陈多福. 南海北部冷泉碳酸盐岩的矿物、岩石及地球化学研究进展[J]. 热带海洋学报, 2012, 31(5):45-56 TONG Hongpeng, FENG Dong, CHEN Duofu. Progresses on petrology, mineralogy and geochemistry of cold seep carbonates in the northern South China Sea[J]. Journal of Tropical Oceanography, 2012, 31(5):45-56.]
[25] Ge L, Jiang S Y, Swennen R, et al. Chemical environment of cold seep carbonate formation on the northern continental slope of South China Sea: Evidence from trace and rare earth element geochemistry[J]. Marine Geology, 2010, 277(1-4):21-30. doi: 10.1016/j.margeo.2010.08.008
[26] Han X Q, Suess E, Huang Y Y, et al. Jiulong methane reef: Microbial mediation of seep carbonates in the South China Sea[J]. Marine Geology, 2008, 249(3-4):243-256. doi: 10.1016/j.margeo.2007.11.012
[27] 陈多福, 黄永样, 冯东, 等. 南海北部冷泉碳酸盐岩和石化微生物细菌及地质意义[J]. 矿物岩石地球化学通报, 2005, 24(3):185-189 CHEN Duofu, HUANG Yongyang, FENG Dong, et al. Seep carbonate and preserved bacteria fossils in the northern of the South China Sea and their geological implications[J]. Bulletin of Mineralogy, Petrology and Geochemistry, 2005, 24(3):185-189.]
[28] 陈忠, 杨华平, 黄奇瑜, 等. 南海东沙西南海域冷泉碳酸盐岩特征及其意义[J]. 现代地质, 2008, 22(3):382-389 CHEN Zhong, YANG Huaping, Huang C Y, et al. Diagenetic environment and implication of seep carbonate precipitations from the southwestern Dongsha Area, South China Sea[J]. Geoscience, 2008, 22(3):382-389.]
[29] 陈忠, 黄奇瑜, 颜文, 等. 南海西沙海槽的碳酸盐结壳及其对甲烷冷泉活动的指示意义[J]. 热带海洋学报, 2007, 26(2):26-33 CHEN Zhong, Huang C Y, YAN Wen, et al. Authigenic carbonates as evidence for seeping fluids in Xisha Trough of South China Sea[J]. Journal of Tropical Oceanography, 2007, 26(2):26-33.]
[30] 葛璐, 蒋少涌, 杨涛, 等. 南海北部神狐海区冷泉碳酸盐岩的地球化学特征[J]. 矿物学报, 2009, 29(S1):370 GE Lu, JIANG Shaoyong, YANG Tao, et al. Geochemical characteristics of seep carbonates from Shenhu area, northern South China Sea[J]. Acta Mineral Sinica, 2009, 29(S1):370.]
[31] 刘关勇, 王旭东, 黄慧文, 等. 南海北部烟囱状碳酸盐岩记录的冷泉流体活动演化特征研究[J]. 地球化学, 2017, 46(6):567-579 LIU Guanyong, WANG Xudong, HUANG Huiwen, et al. Variations in fluid sources and seepages archived in carbonate chimneys from the northern South China Sea[J]. Geochimica, 2017, 46(6):567-579.]
[32] 韩喜球, 杨克红, 黄永样. 南海东沙东北冷泉流体的来源和性质: 来自烟囱状冷泉碳酸盐岩的证据[J]. 科学通报, 2013, 58(19): 1865-1873 HAN Xiqiu, YANG Kehong, HUANG Yongyang. Origin and nature of cold seep in northeastern Dongsha area, South China Sea: Evidence from chimney-like seep carbonates[J]. Chinese Science Bulletin, 2013, 58(30): 3689-3697.]
[33] 陈选博, 韩喜球. 南海东北陆坡烟囱状冷泉碳酸盐岩生长剖面的碳、氧同位素特征与生长模式[J]. 沉积学报, 2013, 31(1):50-55 CHEN Xuanbo, HAN Xiqiu. Carbon and oxygen isotope characteristics of the growth profile of a seep carbonate chimney from the northeastern slope of the south China sea and its formation model[J]. Acta Sedimentologica Sinica, 2013, 31(1):50-55.]
[34] 杨克红, 初凤友, 赵建如, 等. 南海北部烟囱状冷泉碳酸盐岩的沉积环境分析[J]. 海洋学报, 2013, 35(2):82-89 YANG Kehong, CHU Fengyou, ZHAO Jianru, et al. Sedimentary environment of the cold-seep carbonate chimneys, North of the South China Sea[J]. Acta Oceanologica Sinica, 2013, 35(2):82-89.]
[35] Sun Y D, Peckmann J, Hu Y, et al. Formation of tubular carbonates within the seabed of the northern south China sea[J]. Minerals, 2020, 10(9):768. doi: 10.3390/min10090768
[36] Jin M, Feng D, Huang K J, et al. Behavior of Mg isotopes during precipitation of methane-derived carbonate: Evidence from tubular seep carbonates from the South China Sea[J]. Chemical Geology, 2021, 567:120101. doi: 10.1016/j.chemgeo.2021.120101
[37] Hsu T W, Jiang W T, Wang Y. Authigenesis of vivianite as influenced by methane-induced sulfidization in cold-seep sediments off southwestern Taiwan[J]. Journal of Asian Earth Sciences, 2014, 89:88-97. doi: 10.1016/j.jseaes.2014.03.027
[38] Egger M, Rasigraf O, Sapart C J, et al. Iron-mediated anaerobic oxidation of methane in brackish coastal sediments[J]. Environmental Science & Technology, 2015, 49(1):277-283.
[39] Egger M, Jilbert T, Behrends T, et al. Vivianite is a major sink for phosphorus in methanogenic coastal surface sediments[J]. Geochimica et Cosmochimica Acta, 2015, 169:217-235. doi: 10.1016/j.gca.2015.09.012
[40] Liu J R, Izon G, Wang J S, et al. Vivianite formation in methane-rich deep-sea sediments from the South China Sea[J]. Biogeosciences, 2018, 15(20):6329-6348. doi: 10.5194/bg-15-6329-2018
[41] Wu D D, Xie R, Liu J, et al. Zone of metal-driven anaerobic oxidation of methane is an important sink for phosphorus in the Taixinan Basin, South China Sea[J]. Marine Geology, 2020, 427:106268. doi: 10.1016/j.margeo.2020.106268
[42] Zhou J L, Du M R, Li J W, et al. Phosphorus species in deep-sea carbonate deposits: implications for phosphorus cycling in cold seep environments[J]. Minerals, 2020, 10(7):645. doi: 10.3390/min10070645
[43] 胡钰, 刘春阳, 陈庆王, 等. 南海东沙海域甲烷渗漏活动对不同Fe组分和P形态组成的影响[J]. 地球化学, 2023, 52(4):439-447 HU Yu, LIU Chunyang, CHEN Qingwang, et al. Methane seepage impacts the compositions of different iron speciations and phosphorus fractions in the Dongsha area of the South China Sea[J]. Geochimica, 2023, 52(4):439-447.]
[44] Chen Q W, Hu Y, Peckmann J, et al. The formation of authigenic phosphorus minerals in cold-seep sediments from the South China Sea: Implications for carbon cycling below the sulfate-methane transition[J]. Marine and Petroleum Geology, 2023, 155:106425. doi: 10.1016/j.marpetgeo.2023.106425
[45] Poulton S W, Canfield D E. Development of a sequential extraction procedure for iron: implications for iron partitioning in continentally derived particulates[J]. Chemical Geology, 2005, 214(3-4):209-221. doi: 10.1016/j.chemgeo.2004.09.003
[46] Henkel S, Kasten S, Poulton S W, et al. Determination of the stable iron isotopic composition of sequentially leached iron phases in marine sediments[J]. Chemical Geology, 2016, 421:93-102. doi: 10.1016/j.chemgeo.2015.12.003
[47] Alcott L J, Krause A J, Hammarlund E U, et al. Development of iron speciation reference materials for palaeoredox analysis[J]. Geostandards and Geoanalytical Research, 2020, 44(3):581-591. doi: 10.1111/ggr.12342
[48] Ruttenberg K C. Development of a sequential extraction method for different forms of phosphorus in marine sediments[J]. Limnology and Oceanography, 1992, 37(7):1460-1482. doi: 10.4319/lo.1992.37.7.1460
[49] Huerta-Diaz M A, Tovar-Sánchez A, Filippelli G, et al. A combined CDB-MAGIC method for the determination of phosphorus associated with sedimentary iron oxyhydroxides[J]. Applied Geochemistry, 2005, 20(11):2108-2115. doi: 10.1016/j.apgeochem.2005.07.009
[50] Chen F, Hu Y, Feng D, et al. Evidence of intense methane seepages from molybdenum enrichments in gas hydrate-bearing sediments of the northern South China Sea[J]. Chemical Geology, 2016, 443:173-181. doi: 10.1016/j.chemgeo.2016.09.029
[51] Hu Y, Feng D, Peckmann J, et al. The crucial role of deep-sourced methane in maintaining the subseafloor sulfate budget[J]. Geoscience Frontiers, 2023, 14(3):101530. doi: 10.1016/j.gsf.2022.101530
[52] Hu Y. Pore water geochemistry and quantification of methane cycling[M]//Chen D F, Feng D. South China Sea Seeps. Singapore: Springer, 2023: 129-148.
[53] Hu Y, Luo M, Peckmann J, et al. Quantifying the extent of authigenic carbonate formation in shallow marine sediments through a correlation between carbonate precipitation rate and sulfate flux[J]. Geophysical Research Letters, 2023, 50(19):e2023GL104296. doi: 10.1029/2023GL104296
[54] Bohrmann G, Greinert J, Suess E, et al. Authigenic carbonates from the Cascadia subduction zone and their relation to gas hydrate stability[J]. Geology, 1998, 26(7):647-650. doi: 10.1130/0091-7613(1998)026<0647:ACFTCS>2.3.CO;2
[55] 杨克红, 于晓果, 初凤友, 等. 南海北部甲烷渗漏系统环境变化的碳、氧同位素记录[J]. 地球科学, 2016, 41(7):1206-1215 YANG Kehong, YU Xiaoguo, CHU Fengyou, et al. Environmental changes in methane seeps recorded by carbon and oxygen isotopes in the northern South China Sea[J]. Earth Science, 2016, 41(7):1206-1215.]
[56] Ruttenberg K C. The global phosphorus cycle[J]. Treatise on Geochemistry, 2014, 10:499-558.
[57] Egger M, Hagens M, Sapart C J, et al. Iron oxide reduction in methane-rich deep Baltic Sea sediments[J]. Geochimica et Cosmochimica Acta, 2017, 207:256-276. doi: 10.1016/j.gca.2017.03.019
[58] Slomp C P, Mort H P, Jilbert T, et al. Coupled dynamics of iron and phosphorus in sediments of an oligotrophic coastal basin and the impact of anaerobic oxidation of methane[J]. PLoS One, 2013, 8(4):e62386. doi: 10.1371/journal.pone.0062386
[59] Egger M, Kraal P, Jilbert T, et al. Anaerobic oxidation of methane alters sediment records of sulfur, iron and phosphorus in the Black Sea[J]. Biogeosciences, 2016, 13(18):5333-5355. doi: 10.5194/bg-13-5333-2016
[60] Kraal P, Burton E D, Rose A L, et al. Sedimentary iron–phosphorus cycling under contrasting redox conditions in a eutrophic estuary[J]. Chemical Geology, 2015, 392:19-31. doi: 10.1016/j.chemgeo.2014.11.006
[61] 张劼, 雷怀彦, 杨鸣, 等. 南海北部陆坡沉积物中P-S-Fe的相互作用及其对划分硫酸盐–甲烷转换带的指示意义[J]. 地学前缘, 2018, 25(3):285-293 ZHANG Jie, LEI Huaiyan, YANG Ming, et al. The interactions of P-S-Fe in sediment from the continental slope of northern South China Sea and their implication for the sulfate-methane transition zone[J]. Earth Science Frontiers, 2018, 25(3):285-293.]
[62] Filippelli G M. The global phosphorus cycle: past, present, and future[J]. Elements, 2008, 4(2):89-95. doi: 10.2113/GSELEMENTS.4.2.89
[63] März C, Riedinger N, Sena C, et al. Phosphorus dynamics around the sulphate-methane transition in continental margin sediments: Authigenic apatite and Fe(II) phosphates[J]. Marine Geology, 2018, 404:84-96. doi: 10.1016/j.margeo.2018.07.010