Magnetic minerals in soils and paleosols as recorders of paleoclimate

Abstract

It is a fundamental challenge for geologists to create quantitative estimates of rainfall and temperature in past climates. Yet, records of past climates are integral for understanding the complexities of earth system dynamics. The research presented in this dissertation begins to establish a framework for reconstructing paleoclimates using the magnetic properties of fossilized soils. Magnetic minerals are ubiquitous in soils, and their composition, grain size, and concentration is often directly related to the ambient climatic conditions that were present during soil formation. Using rock magnetic methods, it is possible to sensitively characterize the magnetic mineral assemblages in natural materials - including soils and paleosols. The fundamentals of rock magnetism and many of the common methods used in rock magnetic applications are presented in chapter 2 and chapter 3, respectively. Chapter 4 reviews the physical, chemical, and biological factors that affect magnetic mineral assemblages in soils, the magnetic methods we use to characterize them, and the known relationships between magnetic minerals in soils and climate. A critical component to developing replicable tools for reconstructing paleoclimate is developing analytical and statistical tools that are accessible to the greater community. Chapter 5 introduces a new model, MAX UnMix, that was developed as an open-source, online tool for rock magnetic data processing that is designed to be user-friendly and accessible. Two case studies, on both fossil (Chapter 7) and modern (Chapter 6) soils, are presented and discuss many issues related to applying magnetic paleoprecipitation proxies in deep time. Chapter 7 discusses difficulties in disentangling the effects of pedogenesis, diagenesis, and recent surficial weathering in Paleocene-Eocene (56-55 Ma) paleosols. Chapter 6 explores the relative influence of soil forming factors (vegetation vs. climate) on controlling the pedogenic formation of magnetic minerals in soils developing across the forest-to-prairie ecotone in NW Minnesota. The body of research presented in this dissertation provides many challenges to future workers, while at the same time highlighting that rock magnetism should be a useful tool for researchers interested in deep time paleoclimates moving forward.