Nearshore sandbars crest position dynamics analysed based on Earth Observation data
2020
Abstract Understanding long-term underwater sandbars dynamics is a key element for optimal coastal zone management, but this endeavour is often held back by data availability. In order to increase the spatial and temporal resolution, availability and quality of sandbars observations, this study proposes and evaluates a new approach for automatic extraction of sandbars positions from freely available satellite imagery, using a case study of complex waters located on the Danube Delta coast, Black Sea. Validation of the satellite extracted sandbars locations against 4 in-situ measurements between 2016 and 2018 shows a very good correlation between the two datasets, demonstrating the high accuracy of satellite-derived data in revealing the underwater morphology of the upper shoreface along large areas and with high temporal resolution. The obtained sandbars positions were then integrated into an automatic workflow able to accurately represent the seasonal and multi-annual sandbars dynamics and also to give an insight about the sandbars alongshore behavioural changes. The results are similar to the ones previously suggested by other studies, which empowered different methods. These prove the high potential of the newly proposed algorithm to be used for the automatic extraction of the positions of underwater features in the surf zone on other sites worldwide, after proper tuning of parameters in accordance with local conditions. The main constraints of the proposed approach are represented by the omission of sandbars position that imprint a low signal on the satellite image due to increased water depth (especially for the outer bar), high turbidity (mostly for the inner bar) or any other noise generation situation (sun glint, small wind waves). The main advantages of the described algorithm are related to the simplicity of implementation and adaptability to other areas of interest and other types of satellite data.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
36
References
6
Citations
NaN
KQI