Innovative Analysis of Large River Topography and Dynamics (Ref:4243)

Fluvial system sustainability is vital for a half billion riparian residents of global river floodplains and deltas, and it requires knowledge and understanding of the land surface changes through time in response to anthropogenic and climatic forcing. Germane to such understanding are high quality topographic data enabling the quantification of surface morphology and change, which are essential for understanding responses of lowland sedimentary systems to these forcings. However, global topographic data are of insufficient quality and resolution, creating major challenges for managing lowland river-floodplain complexes in large alluvial rivers. This scarcity of topographic data is especially applicable to low-gradient tropical rivers where 1) accurate topographic survey data are scarce, 2) floodplains have significant vegetation precluding satellite observation of surfaces, and 3) previous global data (eg., SRTM) are technically limited in terms of accuracy and resolution for geomorphic applications. The TanDEM-X high-resolution SAR mission (directory.eoportal.org/web/eoportal/satellite-missions/t/tandem-x) offers a way forwards, especially when paired with remote sensing data and machine learning statistical methods. Our exploratory work with the high-resolution (~12m) TDX product in Amazonia and SE Asia has identified means to remove trees using a combination of tomographic analysis, remote sensing, and machine learning using GPUs - producing bare earth DEMs for fluvial systems that can be verified for locations where we have field survey and Lidar data from prior and current NERC, NSF, and NASA-funded projects. The PhD student would focus on the refinement and application of these novel data and techniques towards quantifying the topography and evolution of project rivers - the lower Mekong River and large portions of the Amazon, major river systems that can be used to calibrate and test these approaches in coordination with ongoing NERC-funded research.