Isolation of paleofield information from the natural remanent magnetization of marine sediments using stratigraphic networks

Observations of the Earth's magnetic field discovered dramatic changes of its intensity. In the last ca. 100 years the dipole moment decreased by about 10 %.Such decreases of the geomagnetic field were observed already in the history of the Earth's magnetic field and even times were the polarity completely reverses. To understandmore of the dynamics of such variations in intensity of the Earth's magnetic field high-resolution data from the ancient magnetic field are needed. Sedimentary sequenceshave become more and more interesting for paleointensity investigations because they give continuous records of the past variations of the Earth's magnetic field and have areasonable global distribution. The mechanism of acquisition of natural remanent magnetization in sediments is strongly influenced by a variety of factors and the current state of knowledge about them is poor.The aim of this study is to enhance the understanding of remanence acquisition in sediments. The complete study includes four publications, which are published or have beensubmitted for publication. The basis for this study are eight sediment series, recovered on a North-South profile across the Subtropical Front in the subtropical and subantarctic South Atlantic. Noticeable lithologic variations are observed between the different coresof the transect. Different water-masses strongly affect sediment accumulation and composition. Sedimentation rates decrease from South to North from about 4cm/kyr to lessthan 1cm/kyr. High-resolution measurements of magnetic and compositional data are performed for each core to form a multi-parameter data set. Newly adopted mathematicaltechniques are developed and by using multi-parameter signal correlations it is possible to define a high-resolution timescale for the eight sediment series. Taking into account correlations between all pairs of multi-parameter data sets, and correlation errors between core pairs (A,B) and (B,C) are controlled by comparison with (A,C). The final age model is called a stratigraphic network. The stratigraphic network provides to study thevariations of paleomagnetic, paleoclimatic and oceanographic conditions during time of all sediment series. In addition, the results enable to trace temporal shifts of the Subtropical Front.Based on this stratigraphic network it is possible to investigate high-resolution measurements of rock magnetic properties, the natural remanent magnetization (NRM) andartificial remanences like anhysteretic and isothermal remanent magnetization (ARM and IRM). Principal component analysis (PCA), which is a standard statistical tool to extract relevant information from a complex data set, is applied to isolate the predominant independent environmental signal components from the rock-magnetic parameters. This method reduces lithologic and climatic influence upon the relative paleointensity record. For thenetwork cores a homogeneity interval, where remanences are only affected by a single environmental signal, was found. Relative paleointensity was determined in thishomogeneity interval and the results were used to construct a first relative paleointensity stack for the central South Atlantic covering the last 300 ka. This stackconnects southern and northern hemisphere relative paleointensity data.The final aim of this study is to analyze the influence of lithology upon relative paleointensity, and to assess lithologic influences based on a comparison between manydifferent recordings of the same field history. Due to the regional proximity of the records, they have experienced approximately the same magnetic field history andvariations between the single relative paleointensity signals are caused by their different lithologic composition. The performed studies upon rock-magnetic andcompositional data as well as investigations of the relative paleointensity of the network cores provide a more detailed analysis of parameters, which influence the naturalremanence of sediments. Therefore, a linear extension of the standard relative paleointensity determination is developed which enables to quantitatively test and comparethe influence of different sediment properties and finally correct for such influences.The new relative paleointensity determination is used to construct a new corrected paleointensity stack for the central South Atlantic. This corrected paleointensity stackcan be compared to other local and global paleointensity data. Another finding of this study is that the ratio between corrected and uncorrected relative paleointensity stack reveals a hidden climate signal, which indicates that climatic variations in sediment composition are present in non-ideal sediment sequences. The results contribute to theoverall understanding of remanence acquisition in sediments and enhance the possibility to use sedimentary records as recorders of the ancient magnetic field