Journal
QUATERNARY INTERNATIONAL
Volume 514, Issue -, Pages 119-129Publisher
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.quaint.2018.11.037
Keywords
Paleoclimatology; Pleistocene; Last Glacial Maximum; Eastern Mediterranean; FastICA
Ask authors/readers for more resources
In this study, we present the results of independent component analysis (ICA) of previously published Lake Van data covering the last 250 kyr BP, to shed light on the precipitation and temperature regime in eastern Anatolia. The data processed were the element intensities of Ca, Fe, K, Mn, and Si analyzed by XRF core scanner; concentrations of TOC and CaCO3 content; and B* (color reflectance) from the Ahlat Ridge sediment record. Our analysis is based on application of ICA on the data by changing the initial random unit vector several times and clustering possible independent components through average-link agglomeration. As components extracted by ICA do not have a hierarchy, mutual information, which is a measure of information content between two random variables, is used as a measure of similarity by which to select candidate components. As a result, we argue, the independent component (Van-IC8), which shows the highest similarity to the Greenland delta O-18 record and visually similar to other regional temperature indicating data, can be read as a proxy for temperature variability. We also assert that the independent component (Van-IC7) which has the highest similarity to B* and visually similar to other regional precipitation proxies, and with the lake level reconstruction from another sediment profile from Lake Van is a proxy of precipitation variability across the region. Our results show that the region's temperature approximately maps onto global records, i.e. warm during interglacials and cold during stadials. However, the precipitation proxy reveals that the region was not dry, or at least as wet as it is today, during the end of the MIS 6 and the LGM. The MIS 5e/c and Holocene were characterized by a wet period followed by dry intervals and Dansgaard/Oeschger events are characterized as being warm and wet.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available