Holocene sea-level changes in Southeast Asia

Abstract

Sea-level changes are dynamic processes and changes in relative sea-level (RSL) are common occurrences in the history of Earth, with different rates over time. Since the beginning of the industrial revolution (ca. 1880 AD) greenhouse gas emissions increased and the related climate change accelerated the modern sea-level rise. This became an increasingly important threat within the last century, in particular for global societies where people are living on low-lying coastlines only a few meters above mean sea level (MSL). The study of Holocene (~11.5 ka BP) sea levels is an important tool to understand the natural processes affecting coastal regions, as paleo RSL positions can help assess long-term vertical land motion and , hence, the risk of local flooding and land loss. In the Holocene, sea-level rise was initiated by increasing temperatures of the Last Glacial Maximum (LGM) ~21 ka BP. Land-ice melt caused the eustatic sea level to rise, triggering processes related to glacial isostatic adjustments (GIA), such as, for example, ocean syphoning, land leveling and changes in the gravitational attraction and Earth rotation. Due to these processes, water masses migrated from the poles to equatorial regions, affecting many far-field areas that experienced a sea-level highstand. In fact, after ice-melting slowed down and stopped (~6 to 4 ka BP), water masses migrated back to the poles, generating RSL highstand patterns, which can be seen in the geological record above present sea level in many far-field (i.e., located away from the ice sheets) areas. At these places, different RSL indicators such as coral reefs, salt marshes or beachrocks, were deposited or developed above modern MSL. Today, they provide important benchmarks on local paleo sea-level histories, once post-depositional processes affecting their elevations are taking into account. In general, there are two types of RSL indicators: index points and limiting indicators. Index points are documenting the paleo sea-level position, and marine or terrestrial limiting points, were formed above or below the local tidal ranges. These latter indicate that the paleo RSL was surely above (marine limiting point) or surely below (terrestrial limiting points) the measured elevation of the indicator. According to well-established protocols, all RSL indicators need to have an age, obtained with radiometric dating, coordinates and elevation of the sample. RSL index points also need to have and associated indicative meaning (IM), composed of the indicative range (IR) and the reference water level (RWL). Additional to geological field data, Holocene sea-level studies must include the result of geophysical models that calculate the intensity and timing of GIA, taking into account different ice and earth model combinations. In this thesis, I worked to refine Holocene sea-level histories in Southeast Asia. First, I present a Holocene sea-level database, compiled from published literature. The database includes 546 data points of the broader region of South and Southeast Asia. Within different sample sets, we highlight that some problems such as data inconsistencies, general lack of data, or insufficient data information from literature complicate the analysis of the general sea-level history for the broader study region. Further, besides GIA, the main driver for the RSL highstand at many regions, we identify the need to better constrain syn- and post-depositional processes causing further departures from Eustasy. I then present the results of original fieldwork in the Spermonde Archipelago (SW Sulawesi). Here, we surveyed and dated 24 new RSL index points for the mid to Late-Holocene. Our data, based on highly accurate measurements of fossil microatolls, were supposed to help solving data inconsistencies within the rarely studied region. This study shows that the new data support previous results based on fossil microatolls but cannot completely solve local data inconsistencies derived from two older studies, comprising mainly marine and terrestrial limiting indicators. Using our data in combination with a large suite of GIA models, we discuss possible tectonic rates in our study area. In combination with previous studies, our data show the possibility that one populated island is subject to relevant subsidence. This should be evaluated by further studies, and opens up relevant questions regarding the fate of small, low-lying, heavily populated, tropical islands in face of climate change.

As understanding RSL histories is essential, another approach to gauge the relevance and importance of missing RSL information from our large database was developed within this thesis. We use a statistical analysis to better evaluate how GIA models fit geological field data. The results of this exercise can help to highlight if deviating RSL histories predicted by GIA models and observations are based on major data inconsistencies, a lack of data points or if post-depositional processes could be the reason for conflicts between RSL observations and model predictions in the related regions. Therefore, 16 study areas were compared to 54 GIA models providing positive or negative mismatches between model and data RSL predictions. The results are divided into two analyses, calculating the probability for the mismatch between the entire dataset and all GIA models and between each single index point and the models. The results show, that in three cases the results of both analyses differ significantly with each other, five regions indicate only low probability results implying that GIA is not the only driving factor and post-depositional processes cannot be excluded, and eight regions provide high probability results indicating GIA as the main driver for the paleo RSL positions.