热带海洋学报

• • 上一篇    

基于多分辨率遥感影像的中国大陆海岸线提取与分类

邹朝乾1, 2,侯西勇1
  

  1. 1. 中国科学院烟台海岸带研究所, 山东 烟台 264003;

    2. 中国科学院大学, 北京 100049



  • 收稿日期:2025-11-14 修回日期:2026-01-21 接受日期:2026-01-26
  • 通讯作者: 侯西勇
  • 基金资助:

    国家自然科学基金项目(42176221)

Extraction and classification of China's mainland coastline based on multi-resolution remote sensing imagery

ZOU Zhaoqian1, 2, HOU Xiyong1   

  1. 1. Yantai Institute of Coastal Zone Research, Chinese Academy of Sciences,Yantai Shandong 264003, P.R. China;

    2. University of Chinese Academy of Sciences, Beijing 100049, China



  • Received:2025-11-14 Revised:2026-01-21 Accepted:2026-01-26
  • Supported by:

     National Natural Science Foundation of China (42176221).

摘要: 海岸线是陆海交界的重要生态地貌边界,精确提取其信息对海岸带资源管理、生态保护及空间规划具有重要意义。本文针对全国大陆海岸带集成30m、16m、10m分辨率遥感影像,针对典型区域集成4m、2m、1m分辨率卫星影像,基于多分辨率遥感影像系统分析大陆海岸线信息提取的尺度效应。结果表明:(1)中国大陆海岸线总长度随影像分辨率升高而增加,由30m的17567.55km增长至16m的18206.73km,以及10m的19152.17km,以自然岸线和养殖岸线为主,且养殖岸线比例随分辨率升高而增加,自然岸线比例则相应下降;(2)海区层面,各海区海岸线长度均随分辨率提高而显著增长,但增幅差异显著,北黄海相对增幅最大(16.6%),而南海增幅最小(5.3%);(3)省区层面差异显著,辽宁因基岩海岸分布广、形态曲折,海岸线长度随分辨率增幅最大,上海、天津因岸线高度人工化,增幅最小;(4)典型区域层面,伴随影像分辨率由30m提升至1m,海岸线长度显著增加,人工化或曲折岸段响应更强烈,且分辨率优于4m后长度增长趋缓;(5)在海岸线长度量算方面,4m分辨率能有效平衡精度与效率,是提取宏观形态特征的适宜尺度;而在岸线类型精确解译方面,宜进一步结合2m、1m等高分辨率影像,以实现对细碎岸段的可靠识别与分类。本研究揭示了海岸线长度及类型结构信息提取对影像分辨率的系统性响应特征,可为中国海岸带保护、岸线资源管理及尺度效应机制研究提供高精度的基础数据与方法支持。

关键词: 海岸线, 多尺度遥感, 尺度效应, 海岸带管理, 中国大陆

Abstract: The coastline, a critical eco-geomorphological boundary at the land-sea interface, is of significant importance for coastal zone resource management, ecological conservation, and spatial planning. Accurately extracting its information is therefore essential. This study, based on multi-resolution remote sensing imagery, systematically analyzes the scale effect of information extraction for China's mainland coastline. For the entire mainland coastal zone, remote sensing images at 30m, 16m, and 10m resolutions were integrated; for typical regions, satellite images at 4m, 2m, and 1m resolutions were utilized. Coastlines were extracted using an expert classification (visual interpretation) method, and the lengths and classified lengths derived from different resolutions were statistically analyzed at four hierarchical levels: the overall mainland, marine areas, provincial districts, and typical regions.The results indicate that:(1) The total length of China's mainland coastline increases as image resolution improves, growing from 17,567.55 km (30m) to 18,206.73 km (16m), and 19,152.17 km (10m). The coastline is dominated by natural and aquaculture shorelines; notably, the proportion of aquaculture shorelines increases with higher resolution, while the proportion of natural shorelines correspondingly decreases.(2) At the marine area level, coastline lengths in all regions significantly increase with improved resolution, though the magnitude of increase varies markedly. The North Yellow Sea exhibits the largest relative increase (16.6%), whereas the South China Sea exhibits the smallest (5.3%).(3) Significant differences are observed at the provincial level. Liaoning, owing to its intricate bedrock coasts, experiences the largest resolution-dependent increase in length. Conversely, Shanghai and Tianjin, characterized by highly artificialized shorelines, show the minimal increases.(4) At the typical region level, coastline length increases significantly as resolution improves from 30m to 1m. Artificialized or intricate coastal segments show a more pronounced response. Notably, the rate of increase in length tends to slow down once the resolution is finer than 4m.(5) Regarding coastline length measurement, 4m resolution effectively balances accuracy and efficiency, making it a suitable scale for extracting macro-scale morphological features. However, for the precise interpretation of coastline types, incorporating higher-resolution imagery (e.g., 2m, 1m) is advisable to achieve reliable identification and classification of fine-scale, fragmented segments. This study reveals the systematic response characteristics of coastline length and type-structure information extraction to image resolution, providing high-precision foundational data and methodological support for China's coastal zone protection, shoreline resource management, and further research into scale effect mechanisms.

Key words: coastline, multi-scale remote sensing, scale effect, coastal zone management, mainland China