Invasive alien plants (IAPs) threaten biodiversity and critical ecosystem services worldwide. There is, therefore, an urgent need to develop intervention measures to control the spread of IAPs. Efforts to control and monitor the spread of IAPs would require their current and detailed distribution over a large geographic area. Recently launched multispectral instrument on-board Sentinel-2 provides free data with good spatiotemporal and spectral resolution, compared to Landsat datasets. The Sentinel-2 dataset, therefore, can be a useful source of the IAPs spatial information required for detection and monitoring purposes. We combined Sentinel-2 data with a radiative transfer model to discriminate IAPs (Acacia mearnsii and Acacia dealbata) from surrounding native tree species in Van Reenen, KwaZulu-Natal, South Africa. The forward mode of combined PROSPECT leaf optical properties model and SAIL canopy bidirectional reflectance model, also referred to as PROSAIL was used to simulate reflectance corresponding to bands of Sentinel-MSI, while the PROSAIL model inversion retrieved leaf area index (LAI) and canopy chlorophyll contents (CCC) of the IAPs and native species. Both reflectance and retrieved properties were used to map the distribution of the species within the study area. Our results showed that A. mearnsii and A. dealbata could be accurately discriminated from the surrounding native trees using integrated PROSAIL Sentinel-2 based model. We found that CCC– and LAI-based (% accuracy = 92.8%, 91.4% for CCC and LAI, respectively) modelling produced a higher classification accuracy than field sampling-based modelling (Accuracy = 90.2% (IAP), 82.2% (NAT) and kappa coefficient = 0.84 (IAP), 0.78 (NAT)). Simulated bands corresponding to Sentinel-2 data, on the other hand, produced species maps comparable to field sampling-based maps. Overall, the integrated PROSAIL Sentinel-2 inversion approach proved suitable for detecting and mapping IAPs over a large area. Due to the high spatiotemporal coverage of Sentinel-2, satellite images, the model developed showed the potential to contribute to the IAPs monitoring systems.

外来入侵植物（IAP）威胁着世界范围内的生物多样性和关键的生态系统服务。因此，迫切需要制定干预措施来控制IAP的传播。要控制和监视IAP的传播，就需要它们在广泛的地理区域中的当前详细分布。与Landsat数据集相比，Sentinel-2板上最新推出的多光谱仪器可提供具有良好时空和光谱分辨率的免费数据。因此，Sentinel-2数据集可以成为检测和监视所需的IAP空间信息的有用来源。我们将Sentinel-2数据与辐射传输模型结合起来，以区分IAP（相思树和相思树））来自南非夸祖鲁-纳塔尔省Van Reenen的周围原生树种。使用PROSPECT叶片光学特性模型和SAIL冠层双向反射模型的组合正向模式（也称为PROSAIL）来模拟与Sentinel-MSI波段相对应的反射率，而PROSAIL模型反演则获取叶面积指数（LAI）和冠层叶绿素IAP和本地物种的含量（CCC）。反射率和检索到的属性均用于绘制研究区域内物种的分布图。我们的结果表明，A。mearnsii和A. Dealbata可以使用基于PROSAIL Sentinel-2的集成模型与周围的本地树准确地区分开。我们发现，基于CCC和LAI的模型（对于CCC和LAI的准确度分别为92.8％，91.4％）比基于现场采样的模型产生的分类准确度更高（准确度为90.2％（IAP），82.2％（NAT ），kappa系数= 0.84（IAP），0.78（NAT））。另一方面，与Sentinel-2数据相对应的模拟带产生的物种图可与基于野外采样的图相比。总体而言，事实证明，集成的PROSAIL Sentinel-2反演方法适用于在大范围内检测和映射IAP。由于Sentinel-2卫星图像的时空覆盖范围很大，因此开发的模型显示出有可能为IAP监控系统做出贡献。