Please use this identifier to cite or link to this item: http://hdl.handle.net/1946/4456
Time series analysis based on paleoclimatic proxies from lake sediments in Iceland
High-resolution proxy data that can be used to reconstruct temperature variability through time is essential in order to estimate the amplification of present warming compared with previous times. This thesis presents results from statistical analyses of paleoclimate records from sedimentary cores recovered from two different lakes in Iceland. The aim of the research was to determine if these proxy records preserve regular climate periodicities and whether the periodicities can be explained by external forces that have influenced past environmental changes in the North Atlantic region.
The first part discusses proxy data from lake sediments spanning the past 10,000 years recovered from Haukadalsvatn, a lake in northwestern Iceland. The spectral analyses of the proxies, biogenic silica (BSi) and total organic carbon (TOC), indicate significant periodicities through the last 10 ka. In the BSi record ~1200, 130, 90 and 70 year cycles are most apparent, whereas ~220, 47 and 40 year cycle are predominant in the TOC record. Differences in the periodicities of the two proxies are explained by differences in their primary controls. The periodicities in the BSi show considerably more stability through time than in the TOC record, where irregularity is due to competition between within-lake productivity and terrestrial soil erosion as the controlling variables on TOC. The climate proxies preserved in Haukadalsvatn‘s sediment is thought to respond to known climate variability in the North Atlantic tied to changes in both oceanic and atmospheric variability.
In the second part, a 3000-year varve-thickness record from Hvítárvatn, a glacial lake in central Iceland was used. The first-order low-frequency trend of the varve thickness record reflects increased erosion through the Late Holocene, reaching a peak during the Little Ice Age (LIA). Superimposed on this trend are large inter-annual to decadal fluctuations in varve thickness. The spectral analyses show that dominant variations in the varve thickness record, after removing the non-linear low-frequency variability, are 100 to 85, 35, 13, 5 and 2 to 4 year cycles. Some of these cycles show similar variability to the North Atlantic Oscillation (NAO) and Atlantic Multidecadal Oscillation (AMO). That relationship is supported by a significant correlation between varve thickness and summer NAO index as well as summer AMO index in the time domain.