the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 10 Nov 2023
Changing Sea Level, Changing Shorelines: Comparison of Remote Sensing Observations at the Terschelling Barrier Island
Abstract. Sea level rise is associated with increased coastal erosion and inundation. However, the effects of sea level change on the shoreline can be enhanced or counteracted by vertical land motion and morphological processes. Therefore, knowledge about the individual contributions of sea level change, vertical land motion and morphodynamics on shoreline changes is necessary to make informed choices when applying coastal defence measures. Here, we assess the potential of remote sensing techniques to detect a geometrical relationship between sea level rise and shoreline retreat for a case study at the Terschelling barrier island at the Northern Dutch coast. First, we find that sea level observations from satellite radar altimetry retracked with ALES can represent sea level variations between 2002 and 2022 at the shoreline when the region to extract altimetry timeseries is chosen carefully. Second, results for cross-shore timeseries of satellite-derived shorelines extracted from optical remote sensing images can change considerably dependent on choices made for tidal correction and parameter settings during the computation of timeseries. While absolute shoreline positions can differ on average by more than 200 m, the average trend differences are below 1 m yr-1. Third, by intersecting the 1992 land elevation with time variable sea level, we find that inundation through sea level rise caused on average -0.3 m yr-1 of shoreline retreat between 1992 and 2022. The actual shoreline movement in this period was on average between -2.8 m yr-1 and -3.2 m yr-1, leading to the interpretation that the larger part of shoreline changes at Terschelling is driven by morphodynamics. We conclude that the combination of sea level from radar altimetry, satellite derived shorelines and land elevation provides valuable information about the influence of sea level rise, vertical land motion and morphodynamics on shoreline movements.
- Preprint
(9655 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (extended)
-
RC1: 'Comment on egusphere-2023-2320', Anonymous Referee #1, 03 Dec 2023
reply
This paper is about exploiting satellite-based observations to observe changes at land-sea boundary. The authors propose a case study at the Terschelling barrier island located at the Northern Dutch coast. Different satellite data sets are used: radar altimetry to measure absolute sea level changes; optical imagery (Landsat) to detect shoreline changes. Independent data sets are then used to validate the results. In particular the authors use tide gauges to measure in situ relative sea level changes; GNSS stations to measure vertical land motion; LIDAR observations to get land elevations and ocean depths.
The paper is certainly of scientific interest and deserves publication. It is well written with well stated scientific objectives and in depth data analyses; thorough methodology; description of results and discussion. The idea of a complete remote sensing approach to study changes at land-sea interface is actual and especially necessary for coastal areas where no other data are available. Of course, the land-sea changes over time are governed by a number of processes and the monitoring requires fine resolution and careful attention in using satellite observations. Satellites are not designed for these purposes, but with customized processing it is possible their usage, knowing the limitations. The authors in the study area have valuable ground truth and auxiliary information to validate results and assess uncertainty. The paper is in the direction of a community paper just published that they should cite in their manuscript:
Laignel, B., Vignudelli, S., Almar, R., Becker, M., Bentamy, A., Benveniste, J., ... & Verpoorter, C. (2023). Observation of the coastal areas, estuaries and deltas from space. Surveys in Geophysics, 1-48.
Hereinafter I have some recommendations to the authors:
- The authors use satellite radar altimetry to measure sea level. We know that this technique needs some specialized processing to retrieve data near coast. The authors use a coastal product that implements a retracker called ALES during two decades since 2002. However, the investigation of climate-related signals (e.g., trends) requires careful attention in ensuring homogeneity of processing between the various mission, removal of possible drifting in corrections, etc. For this reason ESA launched the Sea Level Climate change initiative to produce a validated gridded product that now is available through Copernicus (https://cds.climate.copernicus.eu/portfolio/dataset/satellite-sea-level-global). Moreover virtual stations are also provided here https://climate.esa.int/en/projects/sea-level/data/ . The authors should use these products as benchmark to assess that their processing chain is consistent.
- The authors mentions various papers related to the synergy of altimetry, tide gauge and GNSS data. I suggest to integrate with recent papers that provide updated inverse methods aiming at a better characterization of the errors in estimating the sea level trends
De Biasio F., Vignudelli S., Sea Level Change in the Mediterranean Sea from Satellite Altimetry and Tide Gauge. In Proceedings of Oceans from Space Conference (Editors: V. Barale, J.F.R. Gower, L. Alberotanza), 24-28 October 2022, Venice, Italy, 152-153, doi:10.57648/OceansFromSpaceV-2022-PROCEEDINGS.
De Biasio F., Vignudelli S., Baldin G.: Revisiting Vertical Land Motion and Sea Level Trends in the Northeastern Adriatic Sea Using Satellite Altimetry and Tide Gauge Data, Journal of Marine Science and Technology, 8(11), 949, doi:10.3390/jmse8110949, 2020.
- The GNSS time series contains discontinuities from antenna and receiver changes. It should be recalled that GNSS is a point, sometime not co-located with tide gauge. Estimation of the VLM depends on how the station is managed and how logs a re updated. We have seen differences between the various services around the world. Our feeling is that only local people can assess well the significance of VLM trends and errors. Sometime using InSAR can help, but I don’t want ask authors to add these data if they are not expert with this technique. I just like authors inform readers about caveats when using GNSS stations. In Table 1, please add error to your VLM estimation (versions 1,2,3)
- Detection of shoreline. Usage of state-of-the-art products is fine. However, Sentinel-2 would provide more revisiting and better resolution. Landsat, like most other imaging satellites, is multispectral. The bands are Blue, Green, Red, near IR, and short wave IR, all with 30 m resolution There are one or two (depending on the satellite) thermal IR bands with 60 m resolution There is a panchromatic image with 15 m resolution. In principle all spectral bands can contribute towards land-water discrimination, but in practice only a few bands provide robust and substantial leverage on classification land-water. Blue and green have the least contrast due to a combination of low and variable land albedo and possible strong and variable reflection from below water substrate. IR bands are generally better for water discrimination because there is a sharp increase in land albedo and increased absorption in water, leading to greater land-water contrast. There are two methods for improving the resolution: panchromatic sharpening and spectral un-mixing. The latest can improve detection from 30 m to 5 meters (see http://meetingorganizer.copernicus.org/EGU2013/EGU2013-9681.pdf ). I suggest authors to discuss a bit the various methods and highlights pros and cons about using a customized processing or using global products.
- Comparison of altimetry with TG to estimate accuracy. I don’t understand well how the tow measuring systems are made homogeneous. Comparison should be instantaneous. DAC and tides (if relevant) need to be removed as the two systems do not measure the same place. Some earth tides are seen partially by the TG. The recipe needs to be reported in appendix of the paper
- Table 3: errors in trends need to be provided. Also significance of the trend should be checked (e.g. using the Mann–Kendall test).
A key point I would like mentioning is the replication of the approach to other sites and hopefully globally, following the promising validation in the study site. It is important to understand if a full remote sensing global application is feasible and if not the authors should explain how to fill the gaps. The authors highlight the need of land elevation data in high spatial and temporal resolution with high accuracy. Can SAR Interferometry fille the gap to measure land changes ? more and more SAR small satellites are going to be launched.
Having said that I warmly authors to consider my constructive comments and I will be available to read the revised version. I have not found any typo in the text and figures. I much appreciate the careful attention in writing and explain the scientific story telling.
Citation: https://doi.org/10.5194/egusphere-2023-2320-RC1
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 179 | 80 | 17 | 276 | 11 | 11 |
- HTML: 179
- PDF: 80
- XML: 17
- Total: 276
- BibTeX: 11
- EndNote: 11
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1