EGU23-11414
https://doi.org/10.5194/egusphere-egu23-11414
EGU General Assembly 2023
© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.

Temporal variations of United Arab Emirates coastline from 1991 to 2021

Justine Sarrau1 and Abdelgadir Abuelgasim2
Justine Sarrau and Abdelgadir Abuelgasim
  • 1United Arab Emirates University, Geography and urban sustainability, United Arab Emirates (justine.sarrau.pro@gmail.com)
  • 2United Arab Emirates University, Geography and urban sustainability, United Arab Emirates (a.abuelgasim@uaeu.ac.ae)

In the context of global sea level rising, coasts are directly impacted. The retreat to coastlines and submersion of anthropic installations are among the major impacts. It is thus imperative to continuously monitor the coastlines status and devise the means and techniques to effectively assess their status. The United Arab Emirates (UAE) for example is a country which has a long sandy coastline. In this research, an algorithm was developed that makes use of remote sensing temporal data to assess the variability of the coastline in the UAE. The algorithm is used to automatically extract the whole coastline between 1991 and 2021 from Landsat 5 and 8 satellite images. They were selected for 1991, 2001, 2013 and 2021 because of the availability of data, and the significant changes that have been done in coastal areas due to urban development during this period.

Only the Landsat spectral bands of green and near-infrared were utilized to calculate the spectral index of detection of the coast DDWI (Direct Difference Water Index). It is the first step of the algorithm developed. Then is used an automatic threshold Otsu to differentiate the land from water. The result is filled to remove the main artifacts and a canny edge detector is used to detect the coastline. At the end of the algorithm, the result is georeferenced because it lost it during the process. The georeferenced layer is polygonised so that the remaining artifacts are easier to remove. Then, a mask layer was created including boats, clouds, etc… and it is removed from the polygonised layer to get the final extracted coastline.

The preliminary findings of this study show that the sandbanks have increased during the period of the study along the Arabian Gulf waters, suggesting that the coastline is retreating. The results showed a development of the sandbanks towards the Arabian Gulf in several places along the northern coastline but also their general retreat on the north-western one. This can be explained by the sediment settlement or the backfills that have been done to create new islands especially around Abu Dhabi city and Dubai. The creation of mangroves plantations or port infrastructures in the same place has completely changed the coastline layout of the UAE.

On the other side of the UAE, along the sea of Oman, the sandbanks have retreated, suggesting either soil erosion by water currents or advancement of the coastline. The results show no significant change at all and no sandbanks. The only changes observed are linked to the anthropic modification of the coast. While the coastline did not change, the developed algorithm detected scattered sandbanks as the coastline. This confusion likely comes from the similar reflectance of sandbanks in shallow water with the sand of the coast. A further improvement for the developed algorithm will be pursued in the future to reduce such confusions.
This methodology is applicable worldwide, but it is necessary to monitor the results for sandy areas such as the Middle East.

How to cite: Sarrau, J. and Abuelgasim, A.: Temporal variations of United Arab Emirates coastline from 1991 to 2021, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-11414, https://doi.org/10.5194/egusphere-egu23-11414, 2023.