Six linear regression models for the magnetic susceptibility of sediment core GeoB18530-1

• Main Authors: von Dobeneck, Tilo, Bukar, Shettima
• Format: Dataset
• Language: English
• Published: PANGAEA
• Subjects: 509-1; ArcTrain; Calculated from linear regression; DEPTH; sediment/rock; GC; GeoB18530-1; Geochemistry; grain size distribution; Grand Banks Margin; Gravity corer; Heinrich event; Heinrich events; Ice-Rafted Detritus; Linear regression equation; Linear regression model; Magnetic susceptibility; Maria S. Merian; MSM39; Processes and impacts of climate change in the North Atlantic Ocean and the Canadian Arctic; Residual value; rock magnetism; SE Grand Banks Slope; Climate change; North Atlantic
• Online Access: https://doi.pangaea.de/10.1594/PANGAEA.971784

Six linear regression estimates of magnetic susceptibility were empirically tested for their ability to predict distinct HLs and ILs signal patterns. As predictors, we used MD magnetite content (IRM 20 mT), magnetic grain size (SIRM/X), dolomite and granite proxies (Ca/Sr and K/Fe), bulk porosity (%), dolomite and granite counts, and counts of the major 13 IRD species. As the obtained numbers and residues demonstrate, the rock magnetic proxies yield the best fits, followed by element ratios and porosity. The count-based predictors are incapable of estimating the older HLs 3-6 and ILs 4-6 correctly; their regression coefficients also do not conform to the measured susceptibilities of the IRD species.