Geomagnetic field for 0-3 ka: 1. New data sets for global modeling
Geomagnetic field for 0-3 ka: 1. New data sets for global modeling
复制标题
DOI：
10.1029/2008gc002295
复制
发表时间：
2009-06-13
影响因子：
3.5
通讯作者：
Constable, C. G.
中科院分区：
文献类型：
--
作者：
Donadini, F.;Korte, M.;Constable, C. G.
古地磁和古地磁记录被用于区域和全球的地球磁场研究。我们对五个新编译的数据集进行了描述和评估，这些数据集也被Korte等人（2009）用于生成过去3000年来地磁场的一系列时变球谐模型。数据来自我们汇编的湖泊沉积物记录和在线数据库GEOMAGIA50v2。这五个选集可从EarthRef数字档案馆（http://earthref.org/cgi-bin/erda.cgi?n=944）获得。数据根据材料来源分组，我们对考古磁文物和熔岩流（ARCH3k_dat数据集）和沉积物（SED3k_dat数据集）进行了单独的可靠性评估。数据的总数比以前的编译多55%。每个组使用不同的标准选择约束数据集。考古数据的筛选是基于原始数据提供者提供的不确定性。湖泊沉积物数据的评估依赖于预先设定的年龄不确定性和以下一项或多项：与同一地区的考古磁数据的比较，几个湖泊之间的区域一致性，以及与全球考古磁模型的一致性。我们讨论了较大的无约束数据集和较小的（可能）更可靠的数据集的相对优点。在每种情况下，受约束的数据集消除了高达35%的先验可用数据，并依赖于对数据质量的潜在主观评估。鉴于现有数据有限，我们的分析表明，在全球野外建模过程中，迭代拒绝少量（1-1.5%）的离群数据是一种较好的方法。模型和数据之间的特定区域比较支持Korte等人的无异常值CALS3k。基于沉积物和考古文物的所有可用测量数据的3模型目前提供了0-3 ka场的最佳全球代表；ARCH3k。1模型能更好地拟合较密集的欧洲古地磁数据，在该地区可能效果更好。
Paleomagnetic and archeomagnetic records are used in both regional and global studies of Earth's magnetic field. We present a description and assessment of five newly compiled data sets, also used in the companion paper by Korte et al. (2009) to produce a series of time-varying spherical harmonic models of the geomagnetic field for the last 3000 years. Data are drawn from our compilation of lake sediment records and from the online database, GEOMAGIA50v2. The five selections are available from the EarthRef Digital Archive at http://earthref.org/cgi-bin/erda.cgi?n=944. Data are grouped according to the source of material, and we conducted separate assessments of reliability for archeomagnetic artifacts and lava flows (the ARCH3k_dat data set) and for sediments (SED3k_dat). The overall number of data is 55% greater than in previous compilations. Constrained data sets were selected using different criteria for each group. Winnowing of archeological data was based on uncertainties supplied by the original data providers. The lake sediment data assessment relied on preassigned age uncertainties and one or more of the following: comparisons with archeomagnetic data from the same region, regional consistency among several lakes, and consistency with global archeomagnetic models. We discuss relative merits of a larger unconstrained data set or a smaller (possibly) more reliable one. The constrained data sets eliminate a priori up to 35% of the available data in each case and rely on potentially subjective assessments of data quality. Given the limited data available our analyses indicate that iterative rejection of a small number (1-1.5%) of outlying data during global field modeling is a preferable approach. Specific regional comparisons among the models and data support the conclusion that Korte et al.'s outlier-free CALS3k.3 model based on all available measurements from sediments and archeological artifacts currently provides the best global representation of the 0-3 ka field; the ARCH3k.1 model provides a better fit to the denser European archeomagnetic data and may be better in that region.