EGU21-9728
https://doi.org/10.5194/egusphere-egu21-9728
EGU General Assembly 2021
© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Palaeoclimatological interpretation of loess-paleosol sequences using log-ratio transformation of granulometric and geochemical data; Czech Republic

Daniel Šimíček1, Ondřej Bábek1, Karel Hron2, and Ivana Pavlů2
Daniel Šimíček et al.
  • 1Palacký University Olomouc, Faculty of Science, Department of Geology, Olomouc, Czech Republic
  • 2Palacký University Olomouc, Faculty of Science, Department of Mathematical Analysis and Applications of Mathematics, Olomouc, Czech Republic

The loess-paleosol sequences (LPS) are an important source of proxy data documenting climatic fluctuations in the Quaternary. Understanding the relationship between the grain size distribution and the geochemistry is crucial for the reliable palaeoclimatological and palaeoenvironmental interpretation of the LPS. For this purpose, the large granulometric (389 samples) and geochemical datasets (542 samples measured using EDXRF spectrometer) were acquired at four LPS in Moravia (eastern part of Czech Rep.). Czech loess represent an important part of the European loess belt, because they characterize region with alternating influence of the oceanic and continental macroclimate, modified by close contact to the front of both continental and alpine glaciation during glacial periods. Moreover, a very diverse geology is characteristic for the source area (the Bohemian Massif) of Czech loess. Therefore, distinguishing of provenance and transporting and post-deposition processes effects on the formation of the Czech LPS requires finding new approaches of statistical evaluation of datasets. Centered log-ratio (clr) transformation and scalar-on-function regression allow finding a relationship between the geochemical composition and the grain-size distribution of loess and soils. Centered log-ratio (clr) transformation was applied to the key elemental proxies of grain size, provenance and weathering and their spatial and stratigraphic distribution. Nearly all LPS samples are characteristic by bimodal grain-size distribution with the main modes corresponding to medium/coarse silt and clay fractions. The scalar-on-function regression shows that the grain-size control of the distribution of Al, Si, K, Ca, Fe, Rb, Sr and Zr is highly site-specific. The provenance signal is recorded especially in coarser-grained fractions transported for short distance. The content of the authigenic phyllosilicates in clay fraction, the alteration of feldspars and micas, low contents of K, and high values of Rb/Sr and Sr/Ca ratios reflect the intensity of weathering, which is highest in the weakly developed brown paleosols. The precipitation rates are suggested the most important microclimatic factor, which affect the intensity of loessification and pedogenic processes forming the Czech LPS.

This research was supported by the Czech Science Foundation (GAČR) research project 19-017685.

How to cite: Šimíček, D., Bábek, O., Hron, K., and Pavlů, I.: Palaeoclimatological interpretation of loess-paleosol sequences using log-ratio transformation of granulometric and geochemical data; Czech Republic, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9728, https://doi.org/10.5194/egusphere-egu21-9728, 2021.

Corresponding displays formerly uploaded have been withdrawn.