台风影响下西沙灰沙岛的时空变化特征
|
肖海婷(2000—), 女, 江西省赣州市人, 硕士研究生, 从事珊瑚礁遥感研究。email: xht_gis@163.com |
Copy editor: 殷波 , YIN Bo
收稿日期: 2024-04-30
修回日期: 2024-06-12
网络出版日期: 2024-06-20
基金资助
国家自然科学基金项目(4192200195)
国家自然科学基金项目(42030502)
国家自然科学基金项目(42090041)
Spatiotemporal changes of lime-sand islands in the Xisha Islands under the impacts of typhoons
Copy editor: YIN Bo
Received date: 2024-04-30
Revised date: 2024-06-12
Online published: 2024-06-20
Supported by
National Natural Science Foundation of China(4192200195)
National Natural Science Foundation of China(42030502)
National Natural Science Foundation of China(42090041)
文章基于哨兵二号(Sentinel-2)遥感影像, 使用梯度向量流-主动轮廓模型(gradient vector flow-snake, GVF-Snake)提取南海西沙北岛、中岛和南岛的海滩基部线(toe of beach, ToB)和植被边界线, 获取了灰沙岛及其海滩面积。采用突变检验法, 判别出2016—2022年造成研究区岛屿面积发生突变的台风事件。通过统计台风事件中灰沙岛归一化植被指数(normalized difference vegetation index, NDVI)和海滩面积变化, 以及ToB线和植被边界线的移动距离, 分析了西沙灰沙岛受台风影响后的时空变化特征及其影响因素。结果显示: 1) 台风导致灰沙岛植被以轻度受损为主, 中度及以上受损植被分布在岛屿沿岸, 导致了植被边界线的后退, 而沉积物随受损植被剥脱, 并随风浪冲刷至海滩, 导致了ToB线的向海推进。受台风路径方向影响, 迎风面沿岸首先遭受冲击, 灰沙岛迎风面沿岸植被重度受损, 因而导致ToB线和植被边界线在迎风面具有更大的移动距离; 2) 在台风影响末期, 由于灰沙岛沿岸植被受损, 植被边界线的后退和ToB线的向海推进, 使得海滩面积扩大, 而在台风过后, 因失去植被的固定, 在海滩堆积的松散沉积物在风浪作用下从岛屿运离至礁坪, 使得海滩面积缩减。因此, 具有固定与缓冲作用的沿岸植被在维持灰沙岛的稳定性方面具有重要作用; 3) 台风风速增大和台风路径与灰沙岛之间距离的缩短, 以及台风期间海洋表层流速的增加, 易使得灰沙岛遭受极端波浪和风暴流的影响, 将增大灰沙岛严重受损植被的面积, 进而造成更大的岸线移动的距离和海滩面积的变化。此外, 台风可引起沉积物在灰沙岛及其周围环境中的运移, 进而直接改变灰沙岛海滩面积及其形态。台风过后灰沙岛上大量松散沉积物的流失, 导致了灰沙岛面积突变减小, 加之全球变暖造成生物碳酸盐产率下降, 以及来自周围环境的沉积物供应不足, 可能是南海西沙灰沙岛面积长期呈下降趋势的原因。
肖海婷 , 黄荣永 , 刘羿 , 余克服 . 台风影响下西沙灰沙岛的时空变化特征[J]. 热带海洋学报, 2025 , 44(2) : 157 -177 . DOI: 10.11978/2024099
In this study, we explored the changing characteristics of spatial patterns of lime-sand islands in the Xisha Islands under the impacts of typhoons and their driving factors, which are of great significance to the ecological protection and sustainable development of lime-sand islands in the South China Sea (SCS). First, to determine the area of the BeiDao, ZhongDao and Nandao islands of the Xisha Islands and their beaches, the toe of beach line (ToB) and the boundary of vegetation were extracted using the gradient vector flow-snake (GVF-snake) model based on Sentinel-2 remote sensing images. The monthly average area of the study islands from 2016 to 2022 was examined using the change point detection method, and the typhoon events that had caused significant changes in the size of the study islands were identified. Finally, the spatial and temporal characteristics of the islands affected by typhoons were analyzed based on the changes in lime-sand islands area, vegetation area, beach area and normalized difference vegetation index (NDVI) during typhoon events. The results are shown as follows. 1) The vegetation of the islands was mainly slightly stressed by the typhoons, and the severely stressed areas were mainly located along the shorelines. The typhoons caused a significant retreat of the vegetation boundary of the islands, and the sediments were stripped off with the damaged vegetation and washed to the beach with the wind and waves, resulting in the seaward advancement of the ToB line. The windward coasts were the first to be impacted by the direction of the typhoon path, and the vegetation along the windward side of the islands was heavily damaged, resulting in more significant changes in vegetation, vegetation boundary, and the ToB line on the windward side. 2) As the vegetation along the lime-sand islands was damaged, the retreat of the vegetation boundary line and the seaward advancement of the ToB line resulted in the expansion of the beach area, whereas after the typhoon, due to the loss of vegetation fixation, the loose sediment accumulated on the beach was transported away from the islands to the reef flats by the wind and wave action, resulting in the reduction of the beach area. Therefore, coastal vegetation that buffers and fixes the sand is critical to the stability of lime-sand islands. 3) Increased typhoon wind speeds and shorter distances between typhoon paths and the islands, as well as increased surface currents during the typhoon, made the islands exposed to extreme wave and storm currents, which led to an increase in the area of severely damaged vegetation on the islands, resulting in greater shoreline movement and changes in beach area. In addition, typhoons can cause sediment transport between the lime-sand islands and their surroundings, which can directly change the beach area and morphology of the lime-sand islands. The loss of large amounts of unconsolidated sediments on the lime-sand islands after the typhoon, which led to an abrupt decrease in the size of the islands, together with the decrease in the rate of biogenic carbonate production due to global warming and the insufficient supply of sediments from the surrounding environment, may be the reasons for the long-term decreasing trend in the size of the lime-sand islands in the Xisha Islands.
Key words: lime-sand island; typhoon; remote sensing; Xisha Islands
图1 研究区位置示意图a. 西沙群岛位置及台风路径示意图。基于自然资源部标准地图服务网站下载的审图号为GS(2023)2762号标准地图制作; b. 七连屿Sentinel-2卫星数据(成像日期: 2021年4月9日); c. 北岛、中岛和南岛及其分区示意图 Fig. 1 Location of study area. (a) Location of Xisha Islands and typhoon tracks; (b) Sentinel-2 satellite data for Qilianyu (imaging date: April 9, 2021); (c) Map of the BeiDao, ZhongDao and NanDao islands and the islands sections |
表1 对研究区造成影响的台风信息Tab. 1 The list of typhoons information affecting study islands |
| 序号 | 编号 | 名称 | 日期 | 强度等级 |
|---|---|---|---|---|
| 1 | 1603 | 银河 | 20160725—20160728 | 强热带风暴 |
| 2 | 1621 | 莎莉嘉 | 20161012—20161020 | 超强台风 |
| 3 | 1704 | 塔拉斯 | 20170710—20170717 | 强热带风暴 |
| 4 | 1710 | 海棠 | 20170727—20170803 | 热带风暴 |
| 5 | 1719 | 杜苏芮 | 20170911—20170916 | 强台风 |
| 6 | 1720 | 卡努 | 20171011—20171016 | 强台风 |
| 7 | 1809 | 山神 | 20180716—20180724 | 热带风暴 |
| 8 | 1822 | 山竹 | 20180907—20180917 | 强台风 |
| 9 | 1826 | 玉兔 | 20181021—20181103 | 强台风 |
| 10 | 1904 | 木恩 | 20190702—20190704 | 热带风暴 |
| 11 | 1907 | 韦帕 | 20190730—20190803 | 热带风暴 |
| 12 | 1912 | 杨柳 | 20190825—20190830 | 热带风暴 |
| 13 | 1928 | 北冕 | 20191125—20191206 | 强台风 |
| 14 | 2003 | 森拉克 | 20200731—20200803 | 热带风暴 |
| 15 | 2011 | 红霞 | 20200914—20200919 | 强热带风暴 |
| 16 | 2022 | 环高 | 20201108—20201116 | 强台风 |
| 17 | 2104 | 小熊 | 20210611—20210613 | 热带风暴 |
| 18 | 2113 | 康森 | 20210905—20210913 | 强热带风暴 |
| 19 | 2118 | 圆规 | 20211008—20211014 | 台风 |
| 20 | 2122 | 雷伊 | 20211212—20211221 | 超强台风 |
| 21 | 2203 | 暹芭 | 20220629—20220707 | 台风 |
| 22 | 2207 | 木兰 | 20220808—20220811 | 热带风暴 |
| 23 | 2209 | 马鞍 | 20220821—20220826 | 台风 |
| 24 | 2216 | 奥鹿 | 20220921—20220929 | 超强台风 |
图3 Sentinel-2和GF-2影像提取结果对比a. 遥感数据及成像时间; b. 图a中黄色方框所指区域中ToB线提取结果对比; c. 植被边界线提取结果对比。图b和图c中红色虚线表示Sentinel-2影像提取结果; 绿色曲线表示GF-2影像提取结果 Fig. 3 Comparison of Sentinel-2 and GF-2 image extraction results. (a) Remote sensing data and time of imaging; (b) Comparison of ToB line extraction results in the region indicated by the yellow box in (a); (c) Comparison of vegetation boundary extraction results. The red dashed lines in Figures b and c represent the Sentinel-2 image extraction results; The green curve represents the GF-2 image extraction result |
图4 2016—2022年研究区岛屿的面积时间序列及其线性趋势a. 北岛; b. 中岛; c. 南岛。图中方程式表示灰沙岛面积的一元线性拟合方程 Fig. 4 Time series of the area of study islands and their linear trends (dashed line) from 2016 to 2022. (a) BeiDao island; (b) ZhongDao island; (c) NanDao island. The equation in the figure represents a univariate linear fitting equation for the area of Grey Sand Island |
图5 研究区岛屿面积的Mann-Kendall检验(a、c、e)和滑动t检验(b、d、f)a, b: 北岛; c, d: 中岛; e, f: 南岛。图中灰色竖纹表示台风影响时段 Fig. 5 Mann-Kendall test (a, c, e) and moving t-test (b, d, f) for the area of study islands. (a, b) BeiDao island; (c, d) ZhongDao island; (e, f) NanDao island. The gray vertical lines in the figure indicate the period of typhoon influence |
表2 台风“杜苏芮”、“环高”和“奥鹿”影响前后Sentinel-2影像选取Tab. 2 Selection of Sentinel-2 images before and after the impacts of Typhoons Doksuri, Vamco and Noru |
| 名称 | 台风路径至岛屿 最短距离/km | 距岛最近时风速/(m·s-1) | 影像日期 | ||
|---|---|---|---|---|---|
| 台风前 | 台风影响末期 | 台风后 | |||
| 杜苏芮 | 88.91 | 48 | 2017/08/13 | 2017/09/17 | 2017/10/27 |
| 环高 | 75.95 | 33 | 2020/10/31 | 2020/11/25 | 2020/12/15 |
| 奥鹿 | 147.56 | 52 | 2022/09/16 | 2022/10/01 | 2022/11/05 |
图6 研究区岛屿受台风“杜苏芮”(a~c)、“环高”(d~f)和“奥鹿”(g~i)影响后植被受损格局及不同受损情况像元占比a, d, g: 北岛; b, e, h: 中岛; c, f, i: 南岛。各分图中的饼状图为不同损伤程度的像元占比(单位: %) Fig. 6 Damage pattern and proportion of stressed pixels of vegetation in study islands after the impact of Typhoons Doksuri (a~c), Vamco (d~f) and Noru (g~i). (a, d, g) BeiDao island; (b, e, h) ZhongDao island; (c, f, i) NanDao island |
图7 研究区岛屿受不同台风影响后植被边界线净移动量(a~c)和ToB线净移动量(d~f)Fig. 7 Net movements of vegetation boundary (a~c) and ToB line (d~f) in study islands after the impact of different typhoons |
表3 研究区岛屿受台风“杜苏芮”影响后ToB线和植被边界线移动情况Tab. 3 The NSM of the ToB line and vegetation boundary of study islands after Typhoon Doksuri |
| 岛屿 | 分区 | 海滩基部线 | 植被边界线 | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 平均/m | 最大/m | 最小/m | 向陆/% | 向海/% | 稳定/% | 平均/m | 最大/m | 最小/m | 后退/% | 扩张/% | 稳定/% | |||
| 背风面 | ||||||||||||||
| 北岛 | ||||||||||||||