Reconstructing Holocene (sub)tropical climate and cyclone variability using geochemical proxies

Abstract

Anthropogenic greenhouse gas emissions are responsible for a warming trend that cannot easily be reversed. This warming trend is expected to have a large impact on global weather patterns and local environmental conditions, for example by changing precipitation patterns, sea level rise and increasing tropical cyclone activity. Therefore, (sub)tropical coastal regions are expected to be heavily impacted by future climate change. To improve our understanding of the possible consequences of future climate change, paleo-archives can be used, which provide information on past climatic and environmental changes. The aim of the research presented in this thesis was to reconstruct Holocene environmental and climate conditions in (sub)tropical areas, including Florida (USA), the southwest of Japan and Queensland (Australia). In Florida Holocene environmental changes were studied in two estuaries: Tampa Bay and Charlotte Harbor. From mid to late Holocene, environmental conditions changed from fresh/brackish to more marine, which is a consequence of on-going sea-level rise and a relative decrease in terrestrial input. Storm deposits are more frequent in late Holocene sediments, which suggests that the tropical cyclone frequency increased. The impact of human activity in the area was studied in Rookery Bay, an estuary south of Charlotte Harbor. More dynamic and highly variable conditions during the 20th century are thought to reflect the effects of deforestation, canalizations and the construction of roads in the area. In Japan, Lake conditions in Lake Kaiike were reconstructed for the last 2000 years using biomarkers and compound specific hydrogen isotope ratios. Changes in precipitation and runoff, which are related to the intensity of the East Asian monsoon, strongly affected environmental conditions in the lake. The El Niño Southern Oscillation is thought to affect summer monsoon intensity in Asia. A peat core from Northern Queenslands, Australia, was used to test if hydrogen isotopes measured on long chain n-alkanes can be used for the reconstruction of tropical cyclones. For the past 50 years the signal closely follows the number of tropical cyclones that occurred within a 500 km radius. Further down the record however, the sample resolution becomes too low to capture the variability in cyclone-activity. Still, the absence of a clear trend in dD suggests that tropical cyclone frequency and/or intensity did not change appreciably over the last 250 years in Northern Queensland. The results presented in this thesis show that changes in sea level, precipitation and storm activity strongly have affected coastal environmental conditions in the (sub)tropics. If due to human activity coastal areas are altered, this may enhance the effect of natural climate change. Whether the activity of tropical cyclones will change in the future needs further evaluation. Additional paleo-records of storm activity may help to understand the processes involved, and perhaps new approaches like the use of dD to reconstruct precipitation patterns may provide new insights. Not only water temperature but also large-scale changes in the ocean-atmosphere system, will determine whether ongoing climate change will result in more and/or more intense tropical cyclones at specific locations.