Universität zu Köln

Seeger, Katharina ORCID: 0000-0003-0766-9818 (2025). Coasts at risk – Improving flood hazard and relative sea-level rise impact assessments in data-sparse coastal lowlands. PhD thesis, Universität zu Köln.

PDF Dissertation_Katharina_Seeger_published.pdf - Accepted Version Download (38MB)

Abstract

Coastal lowlands such as river deltas and low-lying coastal plains are highly vital environments, offering a wide range of ecosystem services that have attracted human settlement and economic development ever since, making them critical hubs of economic productivity and global trade. Today, almost 40 % of the world’s population live within 100 km distance from the coast and projections indicate global coastal population to increase by 60 % to 70 % by 2050 with highest growths being observed in Asian and African low-lying coastal areas of less than 10 m elevation above sea level – trends that are expected to continue in the future. With ongoing climate change, consequently rising sea level and shifting patterns in atmospheric circulation and extreme weather events, coastal lowlands face an increased exposure to flooding and sea-level rise. Densely populated river deltas, many of them located in tropical Southeast Asia, are particularly vulnerable as human activities have significantly altered deltaic dynamics and thereby increased the risk of flooding and permanent inundation. Comprehensive hazard exposure and impact assessments are needed to understand hazards and estimate the area, population and assets at risk, which will inform risk assessments and support the development of risk-sensitive coping strategies. Acknowledging the complex interactions of natural drivers and human pressures that shape the coastal-deltaic system helps to achieve a more holistic understanding and to design sustainable coastal management. For many coasts and deltas such as in Southeast Asia and Africa, however, the absence or unavailability of high-quality data often challenges to gain a complete understanding of processes, interactions and impacts in the context of flooding and to generate detailed and reliable information about flood hazards and exposure. Especially as both sea-level rise impact and flood inundation are closely related to land elevation, the quality of these assessments largely relies on vertical accuracy and proper datum referencing of the elevation data used. This thesis aims at improving the understanding of flood hazards and impacts and solving challenges in their assessments in the world’s data-sparse coastal-deltaic lowlands to guide disaster risk reduction and coping strategies in the view of future flood risk and relative sea-level rise. To achieve this overarching goal, the thesis addresses data-sparse coastal lowlands at various spatial scales, ranging from local (delta) scale by showcasing the Ayeyarwady Delta (Myanmar) and Mekong Delta (Vietnam), to regional (including five Asian megadeltas in South- and Southeast Asia), to global scale (the world’s coasts). The thesis entails four core sections that (i) assess the state-of-the-art of coastal hazards, including drivers and pressures as well as data and modelling requirements to improve the understanding; (ii) evaluate flood hazards in data-sparse coastal lowlands; (iii) highlight the need of proper handling of coastal land elevation and sea-level data and showing ways to reduce uncertainty in elevation assessments by elevation models; and (iv) unravel land elevation change and potential driving mechanisms and their relevance for future relative sea-level rise and flood exposure. Coastal flooding in the five Asian megadeltas of the Ganges-Brahmaputra-Meghna, Ayeyarwady, Chao Phraya, Mekong and Red River is controlled by geophysical drivers of monsoon and river discharge, tides, tropical cyclones, storm surges and wind-waves, as well as sea-level change, vertical land motion and land elevation dynamics, and sediment, shoreline and deltaic landform changes. Main human pressures include population growth and deltaic land-use land cover changes due to urbanisation, mangrove loss due to the expansion of agriculture and aquaculture, groundwater extraction, sand mining and upstream interventions (due to dam construction and impacts on river flow and sediment dynamics). Coastal flooding shows various interlinkages between direct and indirect societal, economic, technical, socio-technical, physical and ecological impacts as well as beneficial effects of ecosystem services but can often cause significant setbacks in socio-economic development and exacerbate poverty in these densely populated and highly resource-dependent lowlands. The Ayeyarwady and Red River deltas constitute the least studied deltas out of the five assessed while data collection and monitoring needs to be improved in all five megadeltas with regard to discharge and sediment gauges, sea- and water-level monitoring and tide gauges, near-shore and channel bathymetry measurements, subsidence monitoring as well as regular updates of elevation data. A standardised, integrative workflow was developed to perform first-order assessments of flood hazard exposure with respect to single flood types as well as multiple ones at low computational capacities. The highly flexible approach allows to evaluate flood-prone areas on regional and local scale, integrating local, high-quality data and, in its absence, regional or global substitutes, while it is independent of index values that are related to a specific unit of area or administrative boundaries. The workflow was demonstrated in the Ayeyarwady Delta to assess its exposure to monsoon and storm-surge flooding as well as relative sea-level rise and attribute flood components. In its design, the approach allows to complement index-based vulnerability and risk assessments while cloud-computed flood frequency maps serve as a fast-processed data product to validate existing global flood mappings and modelling studies. To investigate the impact of elevation data on sea-level rise impact assessments, the new, local AD-DEM was generated for the Ayeyarwady Delta and used to assess the performance of 10 commonly used global, space-borne elevation models after all datasets were referenced to local mean sea level. Local elevation data indicates the average elevation of the delta plain at ~1.3 m above local mean sea level which is overestimated by global elevation models by several metres. Best performing global elevation models differ by ~2 m RMSE from local data. Using latest IPCC sea-level rise scenarios, the consequences of inaccuracy of global elevation models were quantified in terms of area and population falling below sea level. Referring to best performing elevation models allowed to narrow down the range of area and population as was estimated by most inaccurate elevation models. As the proper alignment of coastal land elevation and sea-level data is critical for reliable sea-level rise and coastal hazard impact assessments, a systematic evaluation of the scientific literature was combined with quantifications of the most prominent errors on global and regional scale. >99 % of the evaluated assessments handled sea-level and land elevation data improperly and the most prominent error (25 % of the studies) was the neglection of datum conversion, followed by incomplete datum conversion (7 %). Consequently, actual sea-level height is underestimated by 0.2–0.5 m on average in coastal hazard assessments globally, regionally exceeding 1 m such as in Southeast Asia. The use of properly aligned sea-level and elevation data reveals worldwide 37 % more area and 67 % more people (i.e. increase of 88 million to 132 million people) to fall below sea-level following 1 m relative sea-level rise than estimated by existing scientific assessments. Works on the accuracy of the elevation data (case study on the Ayeyarwady Delta) and vertical datum offset (global assessment) were detailed by including also elevation change since data acquisition into account. As coastal-deltaic landscapes are often experiencing high rates of relative sea-level rise that sediment accretion cannot compensate for, thus resulting in elevation loss, elevation as indicated by old elevation models (some are 25 years old) may be completely off. A globally applicable approach was established and allows to quantify and attribute uncertainties in elevation assessment for data-sparse coastal lowlands using global elevation models to sources such as inaccuracy, vertical datum offset and elevation change since data acquisition. Applying the approach to the Mekong Delta not only reveals the best performing elevation model but also pinpoints errors in processing steps and indicates how much model performance (and its applicability) can be improved if certain processing steps are prioritised and properly applied. To move from (static) elevation to elevation dynamics to strengthen coastal hazard and impact assessments, first delta-wide InSAR-derived information on land elevation change in the Ayeyarwady Delta was estimated. Elevation change reveals a heterogenous pattern with moderate negative (e.g. elevation loss/subsidence) and positive elevation change (e.g. elevation gain/uplift) spread over the entire delta. Delta-wide average elevation change is -4 mm/yr, while areas of maximal elevation loss include the hotspot of Yangon City, where rates reach -100 mm/yr, as well as the northeastern delta and southern coastal range. Our findings are in line with previous studies for Yangon City, whereas the coastal pattern of elevation loss resembles that of other deltas in the world, suggesting potentially similar driving mechanisms. Limited by data scarcity and the final post-processing of the InSAR data still ongoing, relations between land elevation change and potential drivers were only tentatively determined. Implications on the delta’s exposure to relative sea-level rise and flooding highlighted the increased susceptibility especially in low-lying parts of the southern deltaic area, which already suffers from highest exposure and vulnerability. The works of this research constitute first steps in improving the understanding of flood hazards and impacts as well as solving challenges in their assessment in the world’s data-sparse coastal-deltaic lowlands. However, the outcomes of this reveal fundamental insights into the exposure of some of the least studied megadeltas, (i.e. the Ayeyarwady) while providing globally applicable workflows of flood hazard exposure and elevation uncertainty assessment that contribute to unravel the unknowns of flooding and relative sea-level rise in any data-sparse coastal lowland in the world.

Item Type:	Thesis (PhD thesis)
Creators:	Creators Email ORCID ORCID Put Code Seeger, Katharina k.seeger@uni-koeln.de https://orcid.org/0000-0003-0766-9818 UNSPECIFIED
URN:	urn:nbn:de:hbz:38-812284
Date:	2025
Language:	English
Faculty:	Faculty of Mathematics and Natural Sciences
Divisions:	Faculty of Mathematics and Natural Sciences > Department of Geosciences > Geographisches Institut
Subjects:	Natural sciences and mathematics Earth sciences Geography and travel
Uncontrolled Keywords:	Keywords Language Coastal hazards English Flood exposure English Relative sea-level rise impact English Coastal elevation English River Deltas English
Date of oral exam:	26 September 2025
Referee:	Name Academic Title Brill, Dominik PD Dr. Törnqvist, Torbjörn Prof. Dr.
Refereed:	Yes
URI:	http://kups.ub.uni-koeln.de/id/eprint/81228