An Updated Analysis of Long-Term Sea Level Change in China Seas and Their Adjacent Ocean with T/P: Jason-1/2/3 from 1993 to 2022

Abstract

This study analyzes sea level changes (SLCs) in China seas and their adjacent ocean (CSO) using data from the TOPEX/Poseidon and Jason-1/2/3 satellite altimetry missions from 1993 to 2022. A 30-year time series of sea level anomalies (SLAs) is established, with trends, spatial distribution, and periodicities analyzed through least squares linear fitting, Kriging interpolation, and wavelet analysis. The average yearly sea level rise in the CSO is 3.87 mm, with specific rates of 4.15 mm/yr in the Bohai Sea, 3.96 mm/yr in the Yellow Sea, 3.54 mm/yr in the East China Sea, and 4.09 mm/yr in the South China Sea. This study examines the spatiotemporal variations in SLAs and identifies an annual primary cycle, along with a new periodicity of 11 years. Utilizing 30 years of satellite observation data, particularly the newer Jason-3 satellite data, this reanalysis reveals new findings related to cycles. Overall, the research updates previous studies and provides valuable insights for further investigations into China’s marine environment.

1. Introduction

As global climate change persists, rising sea levels pose a significant threat to human habitats, potentially leading to catastrophic storm surges, floods, coastal erosion, and related hazards [1,2,3,4,5]. China, with its extensive 18,000 km of coastline and 14,000 km of island coastlines, is a crucial maritime nation. Many major cities and a significant portion of the population are concentrated in densely populated and economically developed eastern coastal regions. These areas are strategically important for Chinese economy, politics, military, and culture. Therefore, studying sea level rise along the Chinese coast and its adjacent areas is of utmost importance [6,7,8,9].

In recent years, the increase in and intensification of extreme weather events due to climate change have underscored the importance of capturing the latest trends and patterns in sea level change (SLC). Understanding recent SLC is essential for predicting future variations and assessing the impact of climate change on sea level. This information can help coastal cities and communities develop more effective adaptation and mitigation strategies, such as improving flood defenses, enhancing infrastructure resilience, and creating new land use plans. While extensive research has been conducted on global and regional sea level change, data are being updated rapidly. This study aims to re-estimate the changes in China’s sea areas using new data from 1993 to 2022.

Numerous domestic and international scholars have extensively researched global and regional SLC using satellite altimetry data [10,11]. A notable example is Jia et al., who used satellite altimetry data to examine global SLC from 2002 to 2020, finding an average global sea level rise of 3.3 mm/yr during this period [12]. Wang et al. derived a 24-year global ocean mass change using the Tongji-LEO2021 and Tongji-Grace 2018 monthly gravity field models from 1993 to 2016. Their results indicate that the global mean ocean mass change rate contributes approximately 54% to the global mean total sea level change rate of 2.85 ± 0.30 mm/yr [13]. Guo et al. investigated SLC in the China seas using data from three satellites—TOPEX/Poseidon (T/P), Jason-1, and Jason-2—and identified the patterns and spatial distribution characteristics of these changes [14]. Xiao et al. examined global SLC and that of the South China Sea from 2002 to 2020 using satellite altimetry data from four satellites (Jason-1, Jason-2, Jason-3, and HY-2B) and the SSALTO/DUACS sea surface height anomaly gridded product data [15].

Using satellite altimetry data to monitor SLC is one of the fundamental research methods in oceanography [16,17,18], with the T/P satellite series (TOPEX/Poseidon, Jason-1, Jason-2, and Jason-3) being among the most representative [19,20]. In this study, we use data from the TOPEX/Poseidon (T/P), Jason-1, Jason-2, and Jason-3 satellites to investigate SLC in the China Seas and their adjacent ocean (CSO) (100° E to 130° E, 0° N to 41° N) from 1993 to 2022. We establish a long-term time series of sea level anomalies (SLAs) spanning 30 years, incorporating the latest data from the Jason-3 satellite. Building on previous research, our goal is to identify new patterns and cycles of SLC, offering a more comprehensive and detailed assessment of sea level variability in the CSO. This study provides a scientific basis for improving coastal development strategies in China.

2. Satellite Altimetry Data and Processing

2.1. Sources of Study Areas and Altimetry Data

The study area encompasses the seas surrounding China: the Bohai Sea, Yellow Sea, East China Sea, and South China Sea, covering the region from longitude 100° E to 130° E and latitude 0° N to 41° N. The Bohai Sea, situated in the northernmost region of eastern China, is a nearly enclosed inland body of water, boasting an average depth of 18 m, with its deepest point plunging to 85 m (longitude 117.5° E–122° E, latitude 37° N–41° N). The Yellow Sea, a peripheral sea of the western Pacific Ocean, is positioned between the Chinese mainland and the Korean Peninsula. Its average depth measures 44 m, while its deepest point reaches 140 m (longitude 119° E–126° E, latitude 33° N–37° N). The East China Sea, situated between the South China Sea and the Yellow Sea, has an average depth exceeding 1000 m. It mainly consists of continental shelves with depths less than 200 m and reaches a maximum depth of approximately 2700 m (longitude 117° E–130° E, latitude 23° N–31° N). The South China Sea, a marginal sea in southern China, averages about 1212 m in depth and reaches a maximum depth of 5567 m in its central deep-sea plain (longitude 105° E–120° E, latitude 3° N–23° N) [21].

The T/P satellite, which was launched in 1992, is a key global ocean altimetry satellite tasked with measuring global ocean surface height and monitoring sea level changes. Subsequent missions encompass Jason-1 launched in 2001, Jason-2 in 2008, and Jason-3 in 2016, which build on the T/P satellite’s mission while enhancing measurement accuracy and coverage. The AVISO data center in France provided the multi-source satellite altimetry data used in this article, covering 30 years from 1 January 1993 to 31 December 2022. Data were selected from the L2P product for each satellite: T/P (January 1993 to August 2002, cycles 11 to 364), Jason-1 (August 2002 to January 2009, cycles 22 to 259), Jason-2 (January 2009 to October 2016, cycles 21 to 303), and Jason-3 (October 2016 to December 2022, cycles 24 to 143). The altimetry data utilized in this study were carefully chosen from the along-track Level-2p (L2P; version 02_00) products, which underwent prior processing to ensure quality. This preprocessing involved rigorous quality control measures and editing to filter out invalid ocean data, aiming to retain only valid measurements based on predefined criteria detailed in the L2P product handbook. These criteria encompass minimum and maximum thresholds for altimeter, radiometer, and geophysical parameters. Following this preprocessing, coastal data of poor quality, particularly those within a 10 km radius of the coastline, were excluded. The data were then computed using the CNES-CLS-2015 mean sea surface height model [22].

2.2. Collinear Processing

The four altimetry satellites discussed are repeat-track satellites with a cycle of approximately ten days, resulting in co-linear ground tracks. However, due to various factors such as instrument effects, gravitational forces, and radiation effects, the actual ground tracks of the satellites do not perfectly coincide and may contain errors. For the T/P series satellites, the offset between different tracks is about 1 km [23]. Therefore, co-location processing is necessary before conducting data analysis [24]. This process involves selecting a reference track based on specific rules and determining the longitudes and sea surface height anomalies of other tracks relative to those of this reference track at the same latitude points. In this study, the reference track is chosen as the one with the highest amount of T/P observational data, and observations from all four satellites are unified with this reference track [25]. The study area and reference trajectory are shown in Figure 1.

2.3. The Proper Weight

When calculating the SLC using data from the T/P series satellites, different weights must be assigned to varying latitudes to ensure the accuracy of the final results. During the calculation, spatial averaging of the SLAs for each cycle in a specific area produces the time series of SLAs [26]. The following weighted formula is used for the spatial averaging of points at latitude φ:

$$P_{\phi }=\sqrt{{sin}^{2}i-{sin}^2\phi }$$

(1)

In this context, ‘i’ denotes the orbital inclination of the satellite, with a specific value of 66° for the three altimeter satellites, as stated in [25]. When calculating the SLAs, observed values at each latitude are multiplied by their corresponding weights, and the average of these weighted values is taken to obtain a weighted average that reflects the influence of latitude. This approach allows for a more accurate representation of the contributions of different latitudes to the SLAs, leading to more reliable results.

2.4. T/P and Jason-1/2/3 Four Satellites to Establish a Continuous Time Series

The data used in this article are L2P data released by the AVISO center, which undergoes four steps: “Acquisition”, “Update Correction”, “Perform Input Checks and Quality Control”, and “Generate Products” [27]. L2P data refer to the dataset obtained after instrumental and geophysical corrections were applied to the original data [28]. Therefore, after co-linear processing, the weighted average of SLAs for each sea area and cycle can be obtained [29], establishing a long-term SLA time series in the CSO, as shown in Figure 2.

Figure 2 shows that the SLC in the CSO exhibits an overall rising trend with an evident annual periodicity. In general, the sea level has undergone several phases of abnormal changes. From 1993 to 1996, the SLAs in the study area continuously increased. Between 1997 and 1999, a notable, abnormal SLC occurred, potentially linked to the El Niño event in 1998, resulting in a sharp rise in sea level during this period. Following a period of relative stability from 1999 to 2001, the sea level declined from 2002 to 2005. From 2006 to 2014, the sea level rose sharply, with a slower increase from 2014 to 2015, which may be attributed to the El Niño event that commenced in 2014. In 2020, a notable variation in sea level was noted, potentially linked to the La Niña phenomenon that emerged during the late 2020 and initial 2021 period.

Utilizing a consistent methodology, an analysis was carried out on datasets sourced from the Bohai Sea, Yellow Sea, East China Sea, and South China Sea, to establish a comprehensive long-term (spanning 30 years) SLA sequence that incorporates information from four altimetry satellites, as depicted in Figure 3.

3. Results and Analysis

3.1. Study on the Interannual SLC in the CSO

The p-value for the fitted result, calculated using the t-test, verifies the reliability of the linear trend. For both Figure 4 and Figure 5, the p-values are all less than 0.01, suggesting strong statistical significance [30]. These rates are shown in Figure 4, which illustrates an overall sea level rise rate (SLR) of 3.87 mm/yr in the CSO.

Using the same method, the SLA time series from Figure 3 was linearly fitted. As illustrated in Figure 5, the annual rise rates for the Bohai Sea, Yellow Sea, East China Sea, and South China Sea are 4.15 mm/yr, 3.96 mm/yr, 3.54 mm/yr, and 4.09 mm/yr, respectively.

The results presented here align well with those reported by Wang et al., who indicated that from 1993 to 2021, the SLR in China’s coastal areas was 4.0 mm/yr. Specifically, the Bohai Sea and Yellow Sea experienced a rise of 3.5 mm/yr, the East China Sea experienced one of 3.3 mm/yr, and the South China Sea experienced a rise of 3.6 mm/yr [31]. The observations by Wang et al., which indicated an annual rise of approximately 4.42 mm for the South China Sea from 1993 to 2016 [32], are broadly consistent with our findings. These comparisons suggest that the SLA rise rates reported in this study are reliable. Compared to the previous study, our research uses data over a longer period, establishing long-term sea level changes and reflecting the most recent rate of change in the CSO.

Using the aforementioned method, we calculated the rate of SLAs for each trajectory point on the reference orbit in a similar manner. The rate of sea level change varies spatially, with different marine regions exhibiting distinct average rates of change [14,33]. We used Kriging interpolation to grid the trajectory data, creating a 60′ × 60′ grid. The distance from trajectory points to grid points was used as weights, and we calculated the sea level change rate at each grid point with a search radius of 60′. The statistical analysis depicted in Figure 6 demonstrates that approximately 95.2% of the reference points (7168 out of 7530 analyzed in the study area) exhibit an upward sea level trend, whereas only 4.8% (362 reference points) show a decreasing trend. This finding is consistent with Mu et al.’s assertion that coastal sea levels in China have significantly risen over the past four decades (1980–2020) [34].

Figure 6 illustrates an overall rising trend in sea level across the China Seas and adjacent oceans. It shows that, from north to south, the average SLR is higher in the Bohai Sea compared to the Yellow Sea and East China Sea, while it is relatively similar to that of the South China Sea. The northern part of the South China Sea experiences a slower rate of sea level rise compared to the southern part, with a more pronounced rise observed near the southeastern part of Hainan Province. The East China Sea displays an overall distribution pattern of higher elevations in the west and lower elevations in the east, with a more rapid sea level increase observed near the Yangtze River’s mouth. Conversely, the South China Sea exhibits a pattern of higher elevations in the southeast and lower elevations in the northwest, with the northern vicinity of the Philippines experiencing the highest rate of sea level rise within the studied area.

3.2. Study on the Intra-Annual Change in Sea Level in the CSO

The average SLA for each season is calculated based on the time series from various points on the reference orbit. This enables us to determine the average rate of sea level rise for each point across the four seasons. Using Kriging interpolation, we generate a map showing sea level rise rates for each season, as shown in Figure 7.

Figure 7 shows that in spring (March, April, May), the overall SLR in the CSO region is relatively high. The fastest rising area is situated in the southern waters of the Philippines, while the rise rate is relatively lower in the southeastern part of Taiwan.

In summer (June, July, August), compared to spring, the SLR increases in the Yellow Sea and the South China Sea. The highest rise rate is observed in the southwestern part of the Philippine Islands, while the lowest rate is found in the East China Sea.

In autumn (September, October, November), the South China Sea shows a notable overall increase in sea level, while other regions remain largely consistent with summer conditions. However, a rare decrease in the SLR is observed in the East China Sea.

In winter (December, January, February), the SLR in the South China Sea continues to increase. In contrast, the rise rates in the Yellow Sea and Bohai Sea slow down compared to those in summer and autumn. The lowest rise rate is observed in the East China Sea. The South China Sea exhibits a faster rise rate compared to that of the other three sea regions.

Overall, most areas of the South China Sea experience a relatively rapid increase in sea level. The rates of sea level rise in the Bohai Sea, Yellow Sea, and East China Sea show significant seasonal variations. Notably, the rare downward trend observed in the East China Sea is particularly significant.

To accurately present the average rate of SLC for each season over the past 30 years in the CSO, the data in Figure 7 were analyzed point by point, resulting in

Table 1. Average sea level rise rates (mm/yr) for each season.

Region/Rate/ Season	Spring	Summer	Autumn	Winter
Bohai Sea	3.6	3.8	4.4	3.4
Yellow Sea	3.5	4.1	3.9	3.3
East China Sea	3.1	3.3	2.4	3.2
South China Sea	3.6	3.7	4.0	4.2
China seas and their adjacent ocean	3.6	3.7	3.8	4.1

.

From the data in Table 1, it is evident that the Bohai Sea undergoes a notably faster sea level increase during the first three seasons, with a slight slowdown in winter. In the Yellow Sea, the SLR in spring is comparable to that of the Bohai Sea. It reaches its highest point in summer and slows down in autumn and winter. In the East China Sea, the SLR is the slowest in the CSO region, especially in autumn, with a rise rate of only 2.4 mm/yr. The South China Sea maintains a consistently rapid SLR throughout the year, with the highest rate observed in winter at 4.2 mm/yr. Overall, the CSO region exhibits a relatively stable SLR, maintaining a rapid upward trend throughout the year.

3.3. Spatial Characteristics of the SLC in the CSO

Shown in Figure 8 is the specific spatial pattern of the yearly average SLA across the CSO from 1993 to 2022, derived through calculating the temporal average of trajectory point data utilizing the Kriging interpolation technique.

The SLA in the CSO is generally positive. However, there are specific regions where the average SLA is negative. Analysis of 9571 trajectory points reveals that 8512 of these points have a positive yearly average SLA, accounting for 88.9% of the total. Negative yearly average sea level rise values are observed in the western part of the East China Sea, the southwestern part of the Philippines, and the western part of the South China Sea.

As shown in Figure 9, average SLA distribution maps for each of the four seasons were produced to investigate the distribution characteristics of the yearly average SLA throughout the year.

In spring, a large area of negative SLA values is observed in the CSO region, especially in the Yellow Sea and Bohai Sea, where the SLA values are predominantly negative. In the East China Sea, the western part has lower SLA values while the eastern part exhibits higher SLA values. The South China Sea also shows widespread negative SLA values, but positive maximum SLA values are observed near the Xisha Islands.

In summer, compared to spring, the overall SLA values in the CSO region are positive. The most notable change occurs in the Bohai Sea and Yellow Sea, where SLA values shift from almost entirely negative to almost entirely positive, with negative SLA values occurring only near the Qingdao area. In the East China Sea, SLA values are generally positive, particularly in the eastern region, where they increase more rapidly. In the South China Sea, while the overall SLA values are positive, significant negative SLA values persist in the Gulf of Tonkin and the southwestern South China Sea.

In autumn, the SLA values across the CSO region reach their peak for the year, all remaining consistently positive. Notably, the SLA values in the Bohai Sea and Yellow Sea experience significant increases, with the highest values recorded in the eastern Yellow Sea. In contrast to those in spring and summer, the East China Sea displays higher SLA values in its eastern waters than in the west. As for the South China Sea, its SLA values remain relatively stable, hovering around 10 cm.

In winter, the CSO region exhibits vast areas with negative SLA values, particularly in the Bohai Sea, Yellow Sea, and East China Sea. Compared to those in autumn, the overall SLA values in the South China Sea decrease, but an unusual increase in SLA values is observed in the Beibu Gulf region.

3.4. Characteristics of the SLC Cycles in the CSO

To investigate the periodic characteristics of SLC in the CSO, we analyzed a 30-year SLA time series dataset, as shown in Figure 2, using wavelet analysis in PyCharm (2023.2.1). The resulting wavelet power spectrum and global wavelet spectrum are presented in Figure 10. The analysis reveals that the CSO region exhibits two primary periodic components of 1 year and 11 years, along with additional components of 2.5 years and 5 years.

Various factors, such as climate characteristics, seafloor topography, and geographical location, affect the variability of SLC periodic characteristics across different maritime regions. To examine the periodic characteristics of SLC in various maritime regions of China, we conducted wavelet analysis on the average SLA time series from the four regions shown in Figure 3, obtaining their respective power spectra. Figure 11 displays the wavelet power spectrum and global wavelet spectrum for these four maritime regions.

Figure 3 illustrates that the SLA in the Bohai Sea peaked at 39 cm in 2021 and dipped to its lowest point of −22 cm in 1996. The amplitude of SLA in the Bohai Sea is relatively large, indicating significant annual variation in sea level. This variation is mainly due to the Bohai Sea’s high latitude and considerable temperature changes between spring and autumn. Overall, the sea level in the Bohai Sea exhibits a steadily rising trend, with relatively minor variations during the periods of 1997–1999 and 2014–2015. Figure 11 illustrates that the Bohai Sea experiences its most prominent sea level fluctuations over a one-year period.

The SLA in the Yellow Sea reached its maximum value of 27 cm in 2022 and its minimum of −14 cm in 1996. Similar to that in the Bohai Sea, the SLA amplitude in the Yellow Sea was relatively modest, with no significant anomalies during the periods of 1997–1999 and 2014–2015. This suggests a minor overall impact, although both seas could have been influenced by El Niño and La Niña events. Overall, sea level in the Yellow Sea showed an increasing trend from 1993 to 2022. Like that in the Bohai Sea, the primary periodic variation in sea level in the Yellow Sea is a one-year cycle.

The SLA in the East China Sea peaked at 22 cm in 2020 and reached a minimum of −14 cm in 1996. Over the past 30 years, sea levels in the East China Sea have shown a steady increase, with an overall upward trend from 1993 to 2022. Wavelet analysis indicates that the primary periodic variation in sea level in the East China Sea is a one-year cycle, with other periodic cycles being less pronounced.

In the South China Sea, there are two primary periodic components: one year and eleven years. This aligns with the periodic components observed in the CSO.

Upon comparison, it is evident that the annual cycle is a prominent periodic component in each sea area, with notable 11-year cycles also observed in the CSO and South China Sea regions. This article identifies new periodic characteristics of SLC through the study of long-term SLA time series, offering a more comprehensive and in-depth assessment of SLC in the CSO. This establishes a foundational scientific rationale for the enhanced development of coastal regions.

4. Discussion

This study utilizes data from the T/P, Jason-1, Jason-2, and Jason-3 altimetry missions to investigate SLC in the CSO from 1993 to 2022. It establishes a 30-year time series of SLA and employs methods such as least squares linear fitting, Kriging interpolation, and wavelet analysis to examine trends, spatial distribution, and periodicities of SLC.

Key findings include the following: (1) The average SLR in the CSO from 1993 to 2022 is 3.87 mm/yr. Specifically, the rates are 4.15 mm/yr for the Bohai Sea, 3.96 mm/yr for the Yellow Sea, 3.54 mm/yr for the East China Sea, and 4.09 mm/yr for the South China Sea. (2) Analyzing the spatiotemporal variations in SLAs in the CSO reveals significant insights. This study examines the annual and seasonal spatiotemporal changes in sea level across these regions, identifying spatial characteristics in SLA variations throughout the seasons, as well as the spatial distribution of annual average SLA and seasonal fluctuations in SLA. (3) The identified periodic characteristics of SLC include a dominant one-year cycle across all studied regions, with notable 11-year cycles also observed in the CSO and South China Sea regions.

This study utilizes satellite altimetry data spanning 30 years to update previous research and establish a longer time series of SLA. By leveraging the relatively new and underexplored Jason-3 data, it reanalyzes SLC in the CSO. This study indicates that the annual cycle has significant periodicity for the region, consistent with the findings of Guo et al. [14]. Additionally, it reveals previously unreported 11-year periodic sea level variations in the region, which are consistent with the well-known Schwabe cycle of approximately 11 years of solar activity [35], providing valuable reference for future studies on Chinese maritime environments. These findings highlight the significance of extended observational periods and advanced satellite technologies in enhancing our understanding of sea level dynamics.

One limitation of the study lies in the inadequate exploration of the influence of the El Niño phenomenon on sea level variations. El Niño is a quasi-periodic climate event that may have a significant effect on the marine environment, including SLC. The lack of comprehensive analysis in this area may prevent a complete understanding of SLC. Future research can consider incorporating the El Niño factor into the analysis and evaluating its potential impact on sea level in the CSO.