Sentinel-2 Time Series and Classifier Fusion to Map an Aquatic Invasive Plant Species along a River—The Case of Water-Hyacinth

Freshwater ecosystems host high levels of biodiversity but are also highly vulnerable to biological invasions. Aquatic Invasive Alien Plant Species (aIAPS) can cause detrimental effects on freshwater and their services society, raising challenges decision-makers regarding correct management. Spatially temporally explicit information the occurrence aIAPS in dynamic systems is essential implement efficient regional local action plans. The use unmanned aerial vehicle imagery synchronized with free Sentinel-2 multispectral data allied classifier fusion techniques may support more monitoring actions for non-stationary aIAPS. Here, we explore advantages such a novel approach mapping invasive water-hyacinth (Eichhornia crassipes) Cávado River (northern Portugal). Invaded non-invaded areas were used evolution spectral attributes Eichhornia crassipes through time series (processed by super-resolution algorithm) that covers March 2021 February 2022 build an dataset (presence or absence). Analysis behavior throughout year allowed detection regions greater capacity distinguish target plant from surrounding environment. Classifier implemented biomod2 predictive modelling package fed selected firstly extract signature day secondly identify pixels similar reflectance values over time. Predictions statistical machine-learning algorithms ensembled map invaded spaces across whole study area during all seasons classifications attaining accuracy (True Skill Statistic, TSS: 0.932; Area Under Receiver Operating Curve, ROC: 0.992; Kappa: 0.826). Our results provide evidence potential our invaders time, applicable assessment success control as well implementation long-term strategic monitoring.