潮间带翅碱蓬（Chenopodiaceae）是滨海湿地的先锋植被，对维持潮间带物种多样性、 提升生态系统服务功能具有重要作用。近年来，不合理的人为活动严重威胁潮间带翅碱蓬 生态系统健康，因此，亟需开展翅碱蓬生态系统研究。传统的潮间带翅碱蓬监测方法以地 面调查和实验室分析为主，调查结果时空不连续、且耗费大量人力物力。遥感技术的迅速 发展为宏观研究翅碱蓬生态系统提供了重要数据基础，然而，目前仍缺少专门面向翅碱蓬 遥感识别与提取方法，也未形成系统、有效的翅碱蓬生态系统状况遥感评估方法，翅碱蓬 区域生长模拟更是空白。 针对上述问题，本文以辽东双台子河口湿地自然保护区的潮间带翅碱蓬生态系统为研 究对象，围绕翅碱蓬植被指数构建、翅碱蓬生态系统状况评估方法、翅碱蓬区域生长模拟 三个方面开展了系统性研究，主要结论与创新点如下： 

     （1）分析了翅碱蓬实测光谱特征，构建面向翅碱蓬遥感提取的植被指数。 从光谱匹配、特征参量提取、光谱敏感度分析、典型植被指数等多方面详细分析了翅 碱蓬地面实测光谱特征，揭示了翅碱蓬光谱特征规律，以翅碱蓬与绿色植被在绿波段与红 波段的反射率差异为基础，构建了专门面向翅碱蓬遥感提取的植被指数——Suaeda Salsa Vegetation Index（SSVI），并与常见植被指数提取结果进行对比验证及精度评估，结果表明， SSVI 翅碱蓬平均提取精度高达 90.7%，平均错分像元占研究区比例 1.3%，F1 为 0.8，提取 效果显著优于 NDVI、RVI、SAVI、MSAVI 等常见植被指数；同时，采用 SSVI 基于 Landsat 影像提取翅碱蓬，精度达到 89.3%，错分像元占研究区比例为 0.7%，F1 为 0.82，说明 SSVI 具有很好的适用性。 

     （2）重建翅碱蓬时空演变过程，构建了翅碱蓬生态系统状况评估方法。 基于研究区 1990 年、1995 年、2000 年、2005 年、2010 年和 2016 年的 Landsat 影像， 借助 SSVI 提取翅碱蓬空间分布信息，并采用监督分类方法提取了芦苇、光滩等其它 7 种 地物的空间分布数据，剖析了研究区翅碱蓬时空演变过程，结果表明：26 年间翅碱蓬面积 先增后降，于 2000 年达到峰值（19.56 km2），2016 年锐减至 7.83 km2，减少面积主要转化 为光滩，证明了翅碱蓬与光滩的演化关系。通过耦合生态系统格局、质量及服务功能，建 立了基于遥感技术的翅碱蓬生态系统状况评估方法（Suaeda Salsa Ecosystem state Assessment method, SEA），制定了 SEA 值分级标准，定量评估了翅碱蓬生态系统状况及变 化，结果表明，翅碱蓬生态系统状况以“差”和“中”等级为主，26 年间两者占研究区比例均 超过 90%，而研究区无“优”和“劣”等级；通过 Mann-Kendall 趋势分析法，厘定了翅碱蓬生 态系统 13.1%的区域呈显著减弱趋势，6.9%的区域呈显著增加趋势，表明生态系统状况整体上呈退化趋势。 

     （3）分析了翅碱蓬与光滩的演化关系，构建了基于水环境因素和生长过程的翅碱蓬 区域生长模拟模型。 分别基于 SSVI 和 DEM，验证了翅碱蓬-光滩间演化的多稳态双峰特征，结合多稳态理 论分析了翅碱蓬与光滩的演化关系。在此基础上，采用地理探测器，识别了 DEM 和全盐 为翅碱蓬主要生长驱动因素，为翅碱蓬区域生长模拟模型构建提供了依据。结合随机森林 模型，综合考虑水环境因素及翅碱蓬生长过程，耦合水动力参数和翅碱蓬生长过程信息， 建立了翅碱蓬区域生长模拟模型（Growth Region Simulation Modle of Sueada Salsa, GRSMS）；在 DEM、累计淹没时长和平均流速等多参数变化情景下，开展研究区翅碱蓬区 域生长模拟，并分析各情景下翅碱蓬生态系统状况变化。结果表明，DEM 升高将促进翅碱 蓬生长，增强翅碱蓬生态系统状况；累计淹没时长和平均流速的增加将有效抑制翅碱蓬生 长，减弱翅碱蓬生态系统状况。 本文构建的潮间带翅碱蓬植被指数（SSVI）、翅碱蓬生态系统状况评估方法（SEA） 和翅碱蓬区域生长模拟模型（GRSMS），有助于系统建立翅碱蓬遥感监测、分析与评估技 术体系，在推动翅碱蓬生态系统保护与管理方面具有重要的现实意义和应用前景。 

    Suaeda Salsa (Chenopodiaceae) is the pioneer vegetation of coastal wetlands, which plays an important role in maintaining intertidal species diversity and improving ecosystem services. In recent years, unreasonable human activities have seriously threatened the health of the Suaeda Salsa ecosystem in the intertidal zone, and it is urgent to conduct a comprehensive study on the Suaeda Salsa ecosystem. However, the traditional monitoring methods of Suaeda Salsa were based on ground survey and laboratory analysis, the survey results were not continuous in time and space and consumed a lot of human and material resources. The rapid development of remote sensing technology has provided an important data base for the macro-study of Suaeda Salsa ecosystem. However, methods for remote sensing identification and extraction of Suaeda Salsa are still lacking, and there is no systematic and effective comprehensive remote sensing evaluation method for Suaeda Salsa, even more the effective trend analysis is blank. 

     In response to the above problems, we took the intertidal Suaeda Salsa ecosystem in the Shuangtaizi Estuary Wetland Nature Reserve in eastern Liaoning Province as the study area, carried out research on the construction of the Suaeda Salsa vegetation index, the assessment method of Suaeda Salsa ecosystem state, and the growth region simulation of the Suaeda Salsa ecosystem. The main conclusions and innovations are as follows: 

    (1) Analyzed the spectral characteristics of Suaeda Salsa, and constructed the remote sensing vegetation index for Suaeda Salsa. 

    This paper analyzed in detail the ground-measured spectral characteristics of Suaeda Salsa from various aspects such as spectral matching, feature parameter extraction, spectral sensitivity analysis and typical vegetation index, revealed the ground-measured spectrum characteristics of Suaeda Salsa. Based on the difference of red and green band between Suaeda Salsa and green vegetation, a new vegetation index specifically for Suaeda Salsa Vegetation Index (SSVI) was constructed. Comparative analysis showed that the average extraction accuracy of SSVI was 90.7%, the average error fraction in the study area was 1.3% and F1 index was 0.8. As a result, the SSVI extraction was ignificantly better than that of common vegetation indexes such as NDVI, RVI, SAVI and MSAVI. Besides, we made use of SSVI to extract Sueada Salsa based on Landsat image with the accuracy of 89.3%, the average error fraction in the study area was 0.7% and F1 index was 0.8, which indicated that the SSVI had the good applicability. 

    (2) Reconstructed the spatiotemporal evolution process of the Suaeda Salsa ecosystem and constructed an assessment method for the Suaeda Salsa ecosystem state. 

    Made use of the Suaeda Salsa Vegetation Index (SSVI), the spatial distribution of Suaeda Salsa in the study area in 1990, 1995, 2000, 2005, 2010 and 2016 was extracted using Landsat images. Based on the supervised classification method, the distribution of other 7 land covers such as reed and bare beach were extracted. Then, we described the spatiotemporal evolution process of Suaeda Salsa ecosystem during the 26 years, the results showed that Sueada Salsa area increased before reducing trend, the maximum area appeard in 2000 (19.56 km2 ), then decreased to 7.83 km2 in 2016, the dreased Sueda Salsa mainly converted to bare beach, which indicated the evolution relationship between Sueda Salsa and bare beach. Coupled with the ecosystem pattern, quality and service functions, we established the Suaeda Salsa Ecosystem State Assessment method (SEA) based on remote sensing technology, developed the SEA grading standards, and quantitatively assessed the state and changes of the Suaeda Salsa ecosystem, the results showed that the general and weak grading was the main state in the Sueada Salsa ecosystem, which covered more than 90% of the study area, on the contrast, there was no good and bad grading. Meanwhile, took the advantage of Mann-Kendall trend analysis method, we found the 13.1% of the Sueada Salsa ecosystem showed decreasing trend while 6.9% showed increasing trend, which indicated that the overall Suaeda Salsa ecosystem showed decreasing trend. 

    (3) Analyzed the evolutionary relationship between Suaeda Salsa and bare beach, and constructed a growth region simulation model of Suaeda Salsa on the basis of water environment factors and growth process. 

    We verified respectively the multi-stable and bimodal characteristics of the Suaeda Salsa evolution to bare beach the based on SSVI and DEM, and analyzed the evolution relationship between Suaeda Salsa and bare beach combing with the multi-stable theory. Besides, the geographic detector was used to analyze and identify the growth driving factors of Suaeda Salsa, which provided the data basis for the model construction of growth dynamic simulation and analysis. Fully considering the water environment factors and the growth process of Suaeda Salsa, coupled with the hydrodynamic parameters and the growth state information of Suaeda Salsa, a growth region simulation model of Suaeda Salsa was established combined with the random forest model (Growth Region Simulation Modle of Sueada Salsa, GRSMS). Under the multi-parameter change scenarios such as DEM, cumulative submergence time and average flow rate, the simulation and analysis of the growth state of Suaeda salsa in the study area have been carried out. The results showed that the increase of DEM will promote the growth of Suaeda salsa, enhanced the Sueada Salsa ecosystem state, the increase of cumulative submergence time and the average flow rate would effectively inhibit the growth of Suaeda Salsa and weaken the Sueada Salsa ecosystem state. 

    The Suaeda Salsa vegetation index (SSVI), Suaeda Salsa Ecosystem state Assessment method (SEA) and Growth Region Simulation Model of Sueada Salsa (GRSMS) constructed in this paper were helpful to establish a systematic technology system for Suaeda Salsa remote sensing monitoring, analysis and assessment, which has important practical significance and application prospects in promoting the protection and management of the Suaeda Salsa ecosystem.