Fine-scale (2m) wetland plant communities maps in Honghu Lake, China.

Wetland ecosystems have been subject to prolonged and severe ecological degradation. In recent years, wetland restoration has made significant contributions to addressing this issue. However, wetland plant communities are undergoing drastic changes during the restoration, and different plant communities are experiencing diverse restoration approaches. Existing technological methods struggle to achieve fine-scale monitoring of wetland plant restoration when access to high-cost remote sensing data such as hyperspectral imagery is limited. To address this challenge, this study proposes a fine-scale wetland plant communities mapping method that integrates image resolution enhancement and high-dimensional multi-index feature classification. The study employs a spatiotemporal-spectral fusion model, TemPanSharpening net, to improve the spatial resolution of long-term multispectral image sequences. Subsequently, multiple spectral features are selected and conveyed to a Transformer variant classification model. This approach is applied to map 2m resolution annual dynamics of wetland plant communities including <i>Phragmites australis</i>,<i> Zizania latifolia</i>, and<i> Nelumbo nucifera</i> in the Honghu Lake South, China. Compared to conventional remote sensing-based wetland plant communities classification approaches, this method significantly improves mapping granularity and achieves an overall accuracy of 88.21%. This research overcomes the limitations of fine-scale wetland plant communities mapping under constrained imaging conditions. It provides technical support for accurately monitoring the effectiveness of wetland plant restoration.

湿地生态系统长期面临持续且严重的生态退化问题。近年来，湿地修复工作为解决该问题作出了重要贡献。但在湿地修复进程中，湿地植物群落正发生剧烈变化，且不同植物群落所适配的修复方式亦存在差异。当前，在高光谱影像等高成本遥感数据获取受限的情况下，现有技术手段难以实现湿地植物修复的精细尺度监测。为应对这一挑战，本研究提出了一种融合图像分辨率增强与高维多指标特征分类的湿地植物群落精细制图方法。本研究采用时空光谱融合模型TemPanSharpening net，对长期多光谱影像序列进行空间分辨率提升；随后选取多组光谱特征，输入至Transformer变体分类模型中。该方法被应用于中国洪湖南岸区域2米分辨率的湿地植物群落年际动态制图，涵盖芦苇（*Phragmites australis*）、菰（*Zizania latifolia*）与莲（*Nelumbo nucifera*）三类植物群落。相较于传统基于遥感的湿地植物群落分类方法，本方法显著提升了制图精细度，总体精度达88.21%。本研究突破了成像条件受限场景下湿地植物群落精细制图的技术瓶颈，为精准监测湿地植物修复成效提供了技术支撑。