| Compaction models, describing variations of velocity, density and porosity with depth, are important in basic and applied studies of the Earth. Many so-called global compaction trends have been proposed for different types of sedimentary rocks, but there is very little agreement among these published compaction trends. These differences are likely due partly to type of data analyses (e.g., cores vs. downhole logs) and partly to local variations in mineralogy and in the compaction and diagenesis processes. This study builds global compaction trends of siliciclastic sediments, based on applying a uniform technique to downhole logs from a global collection of sites; we exclude nonrepresentative environments such as accretionary prisms and overpressured sediments. Principal component analysis (PCA) was successfully applied on logs selected from the DSDP/ODP/IODP database to separate velocity and porosity data according to their lithologies. In contrast to earlier studies, we find global consistency in velocity and porosity compaction trends within each of these four lithologies: sands, silts, muds and clays.;Global compaction trends describe average behavior of deep-sea siliciclastic sediments. These trends can be used when no data exist to build local compaction trends. Regional compaction trends are always better predictors for the rate of physical properties change with depth and should be used if data are available. A major part of this study was to obtain reference compaction trends as well as time-to-depth conversion curves for IODP Leg 317 Canterbury Basin sites. The success of Canterbury Basin Expedition 317 depends on the accuracies of its compaction model and time-to-depth conversions. Despite poor core recovery and very limited well logging on Leg 317, our PCA-based method of determining lithology-dependent compaction trends was successful. |
