Magnetic and mineralogical analyses of Ganzi and Luochuan sections

We present a first combined environmental magnetic and geochemical investigation of a loess-paleosol sequence (<55 ka) from the Chuanxi Plateau on the eastern margin of the Tibetan Plateau. Detailed comparison between the Ganzi section and the Luochuan section from the Chinese Loess Plateau (CLP) allows quantification of the effects of provenance and climate on pedogenic magnetic enhancement in Chinese loess. Rare earth element patterns and clay mineral compositions indicate that the Ganzi loess originates from the interior of the Tibetan Plateau. The different Ganzi and CLP loess provenances add complexity to interpretation of magnetic parameters in terms of the concentration and grain size of eolian magnetic minerals. Enhanced paleosol magnetism via pedogenic formation of ferrimagnetic nanoparticles is observed in both sections, but weaker ferrimagnetic contributions, finer superparamagnetic (SP) particles and stronger chemical weathering are found in the Ganzi loess, which indicates the action of multiple pedogenic processes that are dominated by the combined effects of mean annual precipitation (MAP), potential evapotranspiration (PET), organic matter and aluminium content. Under relatively high MAP and low PET conditions, high soil moisture favours transformation of ferrimagnetic minerals to hematite, which results in a relatively higher concentration of hematite but weaker ferrimagnetism of Ganzi loess. Initial growth of superparamagnetic (SP) particles is also documented in the incipient loess at Ganzi, which directly reflects the dynamic formation of nano-sized pedogenic ferrimagnets. A humid pedogenic environment with more organic matter and higher Al content also helps to form finer SP particles. We therefore propose that soil water balance, rather than solely rainfall, dominates the type, concentration and grain size of secondary ferrimagnetic minerals produced by pedogenesis.

(Table 1) Comparisons of dynamic and static Young's moduli of basalt at DSDP Holes 69-504B, 70-504B and 83-504B

Compressional- and shear-wave velocity logs (Vp and Vs, respectively) that were run to a sub-basement depth of 1013 m (1287.5 m sub-bottom) in Hole 504B suggest the presence of Layer 2A and document the presence of layers 2B and 2C on the Costa Rica Rift. Layer 2A extends from the mudline to 225 m sub-basement and is characterized by compressional-wave velocities of 4.0 km/s or less. Layer 2B extends from 225 to 900 m and may be divided into two intervals: an upper level from 225 to 600 m in which Vp decreases slowly from 5.0 to 4.8 km/s and a lower level from 600 to about 900 m in which Vp increases slowly to 6.0 km/s. In Layer 2C, which was logged for about 100 m to a depth of 1 km, Vp and Vs appear to be constant at 6.0 and 3.2 km/s, respectively. This velocity structure is consistent with, but more detailed than the structure determined by the oblique seismic experiment in the same hole. Since laboratory measurements of the compressional- and shear-wave velocity of samples from Hole 504B at Pconfining = Pdifferential average 6.0 and 3.2 km/s respectively, and show only slight increases with depth, we conclude that the velocity structure of Layer 2 is controlled almost entirely by variations in porosity and that the crack porosity of Layer 2C approaches zero. A comparison between the compressional-wave velocities determined by logging and the formation porosities calculated from the results of the large-scale resistivity experiment using Archie's Law suggest that the velocity- porosity relation derived by Hyndman et al. (1984) for laboratory samples serves as an upper bound for Vp, and the noninteractive relation derived by Toksöz et al. (1976) for cracks with an aspect ratio a = 1/32 serves as a lower bound.