Assessing sea level rise and extreme events along the China-Europe Sea Route

Lecci, Rita; Gwee, Robyn; Yan, Kun; Muis, Sanne; Pinardi, Nadia; She, Jun; Verlaan, Martin; Masina, Simona; Li, Wenshan; Wang, Hui; Causio, Salvatore; Novellino, Antonio; Alba, Marco; Kras, Etiënne; Gaytan Aguilar, Sandra; Calewaert, Jan-Bart

doi:10.5194/egusphere-2025-1763

Preprints

https://doi.org/10.5194/egusphere-2025-1763

Preprints

10 Jun 2025

| 10 Jun 2025

Assessing sea level rise and extreme events along the China-Europe Sea Route

Rita Lecci, Robyn Gwee, Kun Yan, Sanne Muis, Nadia Pinardi, Jun She, Martin Verlaan, Simona Masina, Wenshan Li, Hui Wang, Salvatore Causio, Antonio Novellino, Marco Alba, Etiënne Kras, Sandra Gaytan Aguilar, and Jan-Bart Calewaert

Abstract. The Intergovernmental Panel on Climate Change (IPCC) Sixth Assessment Report highlights the critical acceleration of global mean sea level (GMSL) rise, with trends surpassing historical rates observed over the past two millennia. The China-Europe Sea Route (CESR), a region of strategic importance for international trade, is particularly vulnerable to sea level changes and extreme events. This study integrates satellite altimetry, tide gauge records, and advanced hydrodynamic models to assess absolute and relative sea level variations, as well as extreme sea level events, across eight CESR sub-regions over the period 1993–2023.

Statistically significant mean sea level trends confirm consistent and systematic changes in sea level trends by decade and across regions. Notably the East China Sea, Yellow Sea and Bohai Sea show a decadal trend slowdown in the second (2003–2023) and third decade (2013–2023) with respect to the first one (1993–2003).

Accelerated regional mean SLA trends are observed in the North Indian Ocean, while Pacific sub-regions exhibited decadal variability. Discrepancies between tide gauge and satellite data in specific areas were attributed to land subsidence and inherent limitations of coastal altimetry.

Numerical modeling using the Global Tide and Surge Model (GTSM) provided estimates of return periods for extreme sea levels, identifying high-risk zones such as the Bay of Bengal and the South China Sea. However, challenges remain in capturing cyclone impacts, emphasizing the need for improved modeling frameworks.

By highlighting the importance of localized, data-driven approaches and continuous monitoring, the findings contribute to advancing climate resilience and informing risk mitigation strategies in this globally significant region.

Received: 13 Apr 2025 – Discussion started: 10 Jun 2025

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Status: final response (author comments only)

RC1: 'Comment on egusphere-2025-1763', Anonymous Referee #1, 07 Oct 2025

Line 85: Is there any further explanation about the multiple altimetry satellites mission used since it will affect the results such as in the Red Sea, the Persian Gulf, the Arabian Sea and the Bay of Bengal whose pattern of sea level rise was rather low in the first decade (1993-2003)?
Line 99: How do you handle exactly the coastal/shallow water problems in your altimetry satellite data?
Line 126: How do you define the mean anomaly sea level data for the area, whether you take it from one sample point or average of all sample points in the area?
The fonts in Fig.2 and Fig.6 are too small and they are hard to read

Citation: https://doi.org/10.5194/egusphere-2025-1763-RC1
RC2:
'Comment on egusphere-2025-1763', Anonymous Referee #2, 19 Nov 2025
Comment on “Assessing sea level rise and extreme events along the China-Europe Sea Route”
The work entitled “Assessing sea level rise and extreme events along the China-Europe Sea Route” evaluates the trends in absolute and relative mean sea level and return levels of extreme sea levels on the Europe China Sea Route covering 8 sea areas that include the Red Sea, Arabian Sea, Persian Gulf, Bay of Bengal, South China sea, East China Sea, Yellow Sea and Bohal Sea. The data for analysis is obtained from various sources such as satellite altimetry, tide gauge records and from hydrodynamic models. The results highlight the complex regional pattern in long-term sea-level trends. The calculated trends exhibit remarkable spatial and temporal variability, with increases 20–45% higher than the global mean and a clear eastward intensification. High spatial variability is also evident in the occurrence of extreme sea-level values. The research addresses a highly important topic, and the overall approach is clear with methods that follow widely accepted practices. However, the methodology for the extreme value analysis, along with its explanation, is not entirely clear. Several parts of the technical and methodological description would benefit from more consistent and coherent presentation. The manuscript could also be strengthened by emphasizing the innovative aspects of this study compared to the previous regional studies.
General comments:
The authors thoroughly discuss the global background of sea-level rise in the introduction. However, some important aspects appear to be missing. While the introduction effectively establishes the study’s overall motivation, it almost does not address extreme sea levels. Including a brief overview of regional long-term sea level trends and extreme events established in earlier studies, along with relevant background on their drivers, would strengthen this section. While several regional studies are discussed later in the results, presenting an overview in the introduction would provide context for the current work and help readers understand how this study builds on or complements the existing knowledge, rather than giving the impression that only global studies are available for this region.
The English language is generally clear. However, in the methods section the writing seems less polished, with occasional confusing phrasing, inconsistency in terminology. Organizing the section into subsections would help improve overall readability and help a reader to distinguish the exact workflow and purpose of each method. The methodology and also the results part would benefit from light language review.
Specific comments:
Please specify how was the mean monthly seasonal cycle (line 90) computed (method, period)?

When fitting the linear trend, how were uncertainties computed, was autocorrelation checked in the residuals of the fitted regression to ensure the proper statistics (e.g standard errors, p-values)?

Although the overall trend is discussed in the text, you might consider adding to Fig. 2 or Fig. 3 as well. This would provide a visual reference alongside the decadal trends.

It remains unclear if the mean monthly seasonal cycle is removed also from the tide gauge data applied for comparison or complement the satellite altimetry grid points? Are the tide gauge data prepared similarly for both analysis of mean sea level trends and the extremes?

The description of the extreme-value methodology on the lines 135–139 seems ambiguous. The manuscript says return periods were derived “following the POT method by fitting a Gumbel distribution to the annual maxima,” while GPD fitting is applied to peaks above the 99th percentile. The annual-maxima (block maxima) and POT approaches are distinct, and Gumbel and GPD correspond to different frameworks of extreme value analysis. Also, the sentence “The GPD is parameterized by shape, location and scale parameters obtained following Maximum Likelihood Estimation (MLE).” on lines 137–138 appears to conflate the description of the Generalized Extreme Value (GEV) distribution, which is characterized by three parameters (Coles, 2004). While the GPD can technically be described as having three parameters if the threshold is considered a location parameter, in practice this threshold is chosen relying on statistics and is not estimated during the fitting procedure. Only the shape and scale parameters are fitted using MLE. The current wording is misleading. It would be helpful if the authors clarified exactly which method was used for each dataset and how the return levels were derived.

Additionally, it is unclear if GTSMip is used as dataset of precomputed extremes (lines 110–114) or whether authors perform their own EVA on it. Line 147 suggests that EVA with two methods is performed on GTSM data, which appears distinct from GTSMip, yet only one method is shown on Fig. 8. Based on Fig. 8, it seems that annual maxima with a Gumber fit were used for the observational data while a POT-GPD approach was applied to GTSMip. If so, the use of different EVA frameworks for different datasets raises concerns about comparability, since differences in return levels may stem from different in methodology or differences in the underlying data characteristics or both. Please clarify these points.

The calculation method for Gumbel distribution (MLE, method of moments, L-moments etc.) for parameter estimations should be also specified (for POT-Pareto MLE was mentioned). With the small dataset of annual maxima as 30-years MLE has shown unreliable results (Psutka and Psutka, 2015) while L-moments has proven to be more reliable (Ailliot et al., 2011).

Regarding EVA analysis, removing the annual mean from sea level records removes interannual mean level differences and normalizes each year, but it does not always fully remove the long-term effects. Were the vertical land motion and long-term sea-level rise also removed for EVA analysis? On some occasions even the annual maxima exhibit correlation, so both independence and stationarity should be tested. For example, the Mann-Kendall test provides a robust, nonparametric check for monotonic trends (either in maxima or in exceedances), while the Wald-Wolfowitz runs test (Yimaz and Perera, 2014) assesses the randomness and residual clustering, helping confirm whether block maxima or declustered peaks can be treated as independent and identically distributed.

From Fig. 8 it appears that, at some locations (e.g., Xiamen, Zhapo, Shanwei) the observed annual maxima fall outside the 90% confidence intervals of the fitted distributions. This may indicate that the selected distributions, or in case of POT-GPD approach the chosen threshold, are not adequately capturing the tail behaviour of the extremes. The issue may also relate to using different input datasets and methodologies for different EVA approaches, as noted in comment 6. For the annual maxima analysis, using Gumbel distribution may be restrictive, it is possible that in these locations a full GEV fit could be more appropriate as there seems to be a heavy tail (possible a positive shape parameter consistent with a Fréchet-type distribution). Overall, it would be worthwhile to assess whether the chosen distributions are indeed the most suitable for representing the extremes at these locations. Analysis or goodness of fit or likelihood-based comparisons could be helpful for assessing whether the chosen distribution adequately represents the tail.

The temporal resolution of satellite altimetry should be mentioned as well.

Technical corrections:
Fig. 2. The text is difficult to read at normal size (100%). Increasing the font size would improve readability. This applies also for Fig. 3, Fig. 6 and Fig. 8. Please check for all figures if axis labels, legends and annotations are readable.
Line 168 “Figure 2 presents time series (1993–2023) of yearly mean sea level anomalies, averaged over the Global Ocean and the eight CESR sub-basins.” Fig. 2 description says, “Figure 2: Time series (1993-2023) of monthly means of sea level anomaly (black curve) from altimetry averaged over…”. Is it representing yearly or monthly average?
Figure 7. Overview of the tidal levels in the region over the period 1985-2014 derived from the GTSMip dataset. (Top) Mean Higher High Water (MHHW) (Bottom) Highest Astronomical Tide (HAT) Top could be a) and bottom could be b) as the letters are anyway marked on the figure.
Line 345 “Over the 1993–2023 period, mean sea level trends across CESR sub-regions were 20–45% higher than the global mean trend for the multidecadal period, 1993-2023, with considerable eastward increases.” Please adjust the sentence.
Line 346 “Notably decadal showdowns in the East China Sea, Yellow Sea and Bohai Bay mean sea level are present probably connected to the interplay of circulation changes affecting the mass transport across the open ocean boundaries, the water cycle and the steric sea level.” Please adjust for clarity.
Line 356 “This is like what we found in CESR sub-regions North Indian Ocean.” – Could be rephrased, for example: “This is consistent with our findings in the CESR sub-regions of the North Indian Ocean”.
Line 359 “Huang’s budget analysis finds that SLA rise of 1.20 ± 0.20 cm/decade in the upper 2,000 m can be attributed to the Ekman pumping and deep-water warming, which accounting 360 for 30% of the total sea level rise. Please adjust for clarity.
Several references in the reference list seem not to be cited in the main text. For example, Alam, (2014), Ali (1999), Anothy and Unnikrishnan (2013), Baburaj et al. (2020), Balaguru et al. (2018), and Bhatia et al. (2018) do not appear in the manuscript. This is only from the first page of reference list. Please check the entire reference list and ensure that all listed references are cited in text or remove those that are not used to.
Additionally, please check whether the publication years in the reference list match the in-text citations. For example, Muis et al., 2022 is listed on line 111, but in manuscript there are three references with Muis as first author, and none of these appear to be published in 2022.
References:
Coles, S. (2004). An Introduction to Statistical Modeling of Extreme Values (3rd ed.). Springer.
Ailliot, P., Thompson, C., & Thomson, P. (2011). Mixed methods for fitting the GEV distribution. Water Resources Research, 47, W05551. https://doi.org/10.1029/2010WR009417
Psutka, J. V., & Psutka, J. (2015). Sample size for maximum likelihood estimates of Gaussian model. In Computer Analysis of Images and Patterns (CAIP 2015).
Yilmaz, A., & Perera, B. (2014). Extreme rainfall nonstationarity investigation and intensity–frequency–duration relationship. Journal of Hydraulic Engineering, 19(6), 1160–1172. https://doi.org/10.1061/(ASCE)HE.1943-5584.00008
Citation: https://doi.org/10.5194/egusphere-2025-1763-RC2
RC3: 'Comment on egusphere-2025-1763', Anonymous Referee #3, 24 Nov 2025

The paper addresses an important subject; namely rising sea levels in regions which are both highly populated and significant in terms of international trade. The paper does not provide particularly ground-breaking results, insight or methodology, however it does make an important contribution by highlighting trends and variability in the areas of interest. The Appendix, describing the validation of GTSM and Coast-RP is also a valuable contribution and will be useful for other researchers in the field.
As a whole, the paper is well structured but within some sections it can be confusing for the reader, because of the mix of datasets, regions and analysis. I think it would be very worthwhile if thought could be put into improving the clarity within sections, particularly the Datasets, Methodology and Results. There are three very different sources of data - satellite altimetry, tide gauges and global model reanalysis - and, for the most part, they require different methodologies and provide different kinds of results and insights. It would be good to make this clear right at the start in the introduction, and then clearly separate them in subsequent sections using subheadings.
It should also be made clear that extreme water levels are considered only for present day conditions and that no attempt has been made to assess the impact of climate change on extreme water levels - only on mean sea level
Note that the references have not been properly curated. I found two citations in the text which did not have references and 27 references that didn't have a citation!
Specific comments
Lines 46-47: this statement should be supported by a reference
Section 2.1 Datasets: Please consider using sub-headings to clearly distinguish between types of datasets
Lines 88-89 and 99-100: These two sentences, mentioning the challenge of satellite altimetry near the coast, appear to be contradictory. The first says that the problem is addressed by wet tropospheric corrections (I have no idea what this is - reference required) while the second says that this method becomes inaccurate near the coast. It should be clearly stated how coastal areas have been considered. This could be by omitting - the subsequent analysis are not dependent on coastal areas.
Line 92: The RLR dataset are described as being relative to 7000mm below mean sea level. Is there an epoch at which this mean sea level is determined?
Line 111: I can't find a reference for the 'Climate Change Initiative Coastal Sea Level Team 2020' citation
Line 106 & 130-132: I think that the authors have not undertaken correction to local land subsidence but it is considered a source of uncertainty when considering tide gauge data and comparison with other sources. Please can the authors make this clear.
Section 2.2 Methodology: Please consider using sub-headings to clearly distinguish between types of datasets
Line 135: This sentence contains a contradiction - annual maxima are not part of the peaks-over-threshold method.
Line 140-145: Please note that the extreme value analysis is relevant and presented only for a present day scenario not considering climate change
Line 145: I don't understand what using Weibull plotting formulations means - please clarify
Line 168-192: Please make clear what dataset was used here
Line 210: I think saying 'areas like the Arabian Sea and South China Sea' is too generic for this paper. Be specific - there are limited number of regions of interest and it would not harm the paper to list all where these monsoonal processes have an impact.
Line 213: As above, I see know harm in being specific - are there other relevant currents/regions?
Line 213-214: 'SLA trends exceed 4 mm/yr along the coasts of the Yellow and East China Seas'. Does the problems with SLA in coastal zones affect these areas?
Line 214-215: 'The South China Sea displays maximum SLA trends in its open ocean regions, influenced by stronger westward currents through the Luzon Strait compared to southeastern warm currents.' This is a throwaway comment - are the currents through the Luzon Strait the cause? If so then there would need to be justification. Otherwise, please change the wording.
Line 217: Looks more like north and northwest Bay of Bengal
Line 225: This area includes several seasonally varying currents - are these annual average?
Line 236: Were the land effects on satellite altimetry that was previously mentioned an issue here?
Line 252: Figure 6: Look like time series of annual means to me not monthly
Line 287: Reference to 'Park et al 2021' is missing
Line 322: This problem of small sample size of historical cyclones is well known, but I think the results here are a good empirical illustration of it and worth highlighting.
Line 349: 'significant positive acceleration in North Indian Ocean'. I'm not sure the evidence is strong enough for this assertion - there is a clearly a lot of interdecadal variability that makes acceleration on a regional scale difficult to indentify.
Line 354 onwards: The context suggests that the abbreviation 'SLA' as used in the conclusions section would make more sense if it meant Sea Level Anomaly (which is a common meaning). However, on line 67 it was defined as Satellite Altimetry. Please clarify or correct.
Line 373-376: Extreme water level - particularly when cause by tropical cyclones - are very sensitive to local geography (depth and extent of coastal shallows, shape of coastline with bays and inlets etc). Care should be taken when applying extreme water levels at one location to others, even if nearby.
Figure A1: Check whether the reproduction of this figure is good enough to see the grid - when I printed it out, I couldn't and it is not that great in the PDF.
Line 392-393: The sentence 'If we use all stations that have more than 10 years of data, this increases to 36 stations' appears superfluous as these were not used. Please either remove or clarify why this might be useful to know.
Line 412: I think this should be return levels rather than 'return periods'
Page 22 to 27: Many of these references are not cited in the text and 2 in the text are not in the references. Some references are incomplete e.g. include et al when authors should be written in full.

Citation: https://doi.org/10.5194/egusphere-2025-1763-RC3

Viewed

Total article views: 912 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
764	130	18	912	25	36

HTML: 764
PDF: 130
XML: 18
Total: 912
BibTeX: 25
EndNote: 36

Views and downloads (calculated since 10 Jun 2025)

Month	HTML	PDF	XML	Total
Jun 2025	129	23	7	159
Jul 2025	38	32	0	70
Aug 2025	113	25	1	139
Sep 2025	396	13	2	411
Oct 2025	50	22	4	76
Nov 2025	38	15	4	57

Cumulative views and downloads (calculated since 10 Jun 2025)

Month	HTML	PDF	XML	Total
Jun 2025	129	23	7	159
Jul 2025	38	32	0	70
Aug 2025	113	25	1	139
Sep 2025	396	13	2	411
Oct 2025	50	22	4	76
Nov 2025	38	15	4	57

Viewed (geographical distribution)

Total article views: 900 (including HTML, PDF, and XML) Thereof 900 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 26 Nov 2025

Short summary

This study explored how sea level is changing along the China-Europe Sea Route. By combining satellite and in-situ observations with advanced modeling, the research identified ongoing sea level rise and an increasing frequency of extreme water level events in some regions. These findings underscore the importance of continued monitoring and provide useful knowledge to support long-term planning, coastal resilience, and informed decision-making.


Total:	0
HTML:	0
PDF:	0
XML:	0