Class accuracy statistics.

Shore area wetlands (lacustrine fringe) play a critical role as ecotones that support biodiversity, provide habitats for spawning and refuge, and exhibit high levels of primary productivity. They facilitate significant exchanges of materials between aquatic and terrestrial ecosystems. To effectively manage and preserve these important resources, it is essential to understand their distribution, size, and dynamic changes. This study aimed to create an accurate map of shoreline wetlands using multi-temporal and multi-source data, wetland indicators such as wetland hydrology (WH), hydrophytic vegetation (HV), hydric soil (HS), and radar imagery from Sentinel-1A, employing Geomatica software. Additionally, ArcGIS software was used to map the topographic position (TP), Lake Bathymetry (LB), and HS indicators for wetlands. The analytical hierarchy process and weighted overlay methods were also applied in the mapping process for integrating all the indicators to obtain the final extent of shoreline wetlands. The TP wetland indicator map covered about 55,364 ha, while HS covered around 55,151 ha within a 3 km buffer from Lake Tana. The map of WH indicator for wetlands revealed that permanently inundated areas accounted for roughly 591,312 ha, and when temporarily inundated areas were included, the total coverage increased to 607,053 ha. HV, including invasive water hyacinth, covered over 74,772 ha. Overall, shoreline wetlands were predominantly located within three kilometers of the terrestrial area from Lake Tana, totaling 26,664 ha. The overall accuracy of land use and cover classification was recorded at 79%, with a Kappa statistic of 0.70, indicating that the resulting map is of acceptable quality. The integration of multi-temporal and multi-source data, along with wetland indicators and radar imagery from Sentinel-1A using Geomatica software, has provided valuable insights into the spatial distribution of shoreline wetlands in Lake Tana. The findings from this study will serve as an important reference for future research aimed at effectively managing and conserving these vital resources.

滨岸湿地（湖岸边缘带，lacustrine fringe）作为生态交错带，在支撑生物多样性、提供产卵与庇护栖息地以及维持高水平初级生产力方面发挥着关键作用。它们助力水生与陆地生态系统之间开展大规模的物质交换。为有效管理与保护这类重要资源，明晰其分布范围、面积大小及动态变化至关重要。本研究旨在利用多时序多源数据、湿地指标（包括湿地水文（wetland hydrology, WH）、水生植物（hydrophytic vegetation, HV）、湿土（hydric soil, HS））以及Sentinel-1A雷达影像，借助Geomatica软件精准绘制滨岸湿地分布图。此外，本研究还使用ArcGIS软件绘制湿地的地形位置（topographic position, TP）、湖泊水深（Lake Bathymetry, LB）以及HS指标分布图。制图过程中，还应用了层次分析法与加权叠加法整合所有指标，以最终确定滨岸湿地的分布范围。在塔纳湖（Lake Tana）周边3公里缓冲区内，TP湿地指标图覆盖面积约为55364公顷，HS湿地指标图覆盖面积约为55151公顷。湿地WH指标图显示，永久淹没区域总面积约为591312公顷，若纳入临时淹没区域，总覆盖面积增至607053公顷。包括入侵物种水葫芦在内的HV覆盖面积超过74772公顷。总体而言，滨岸湿地主要分布在距塔纳湖陆地边界3公里范围内，总面积达26664公顷。本次土地利用与土地覆盖分类的总体精度为79%，Kappa系数为0.70，表明所生成的地图质量符合可接受标准。本研究整合多时序多源数据、湿地各类指标以及Sentinel-1A雷达影像，并借助Geomatica软件，为解析塔纳湖滨岸湿地的空间分布提供了宝贵视角。本研究成果可为未来开展此类重要资源的有效管理与保护相关研究提供重要参考。