Monitoring Water Hyacinth Growth Stages Using Machine Learning Techniques in Sentinel-2 Time Series

Citation

Pirbasti MA & Akbari V (2024) Monitoring Water Hyacinth Growth Stages Using Machine Learning Techniques in Sentinel-2 Time Series. IGARSS 2024 - 2024 IEEE International Geoscience and Remote Sensing Symposium, Athens, Greece, 07.07.2024-12.07.2024. https://doi.org/10.1109/igarss53475.2024.10641219

Abstract
Water hyacinth (Pontederia crassipes) is recognised as the most notorious invasive species worldwide. Although its threats and effects are fully documented, its spatial distribution is still poorly understood, especially in complex environments such as wetland systems. This study aimed mapping the spatiotemporal distribution of invasive water hyacinth (WH) in Anzali International Wetland (AIW), whose habitat is endangered by the presence of WH; it was conducted using Sentinel-2 Multi-Spectral Instrument (MSI) 2022 data. Specifically, this study sought to identify multispectral remote sensing variables and in-situ field data using machine learning (ML) methods to detect and map WH growth stages. We used four images dominated by four growth stages: early, mid, high, and decaying stages to train our ML classifier. We used Random Forest (RF) algorithm for training our training samples achieving an overall classification accuracy (OA) of over 98%. These findings were further supported by statistical analysis, such as F1 (above 96%) and Intersection over Union (IoU) (above 92%), indicating the high performance quality of the used algorithm. Our study provides valuable insights into using ML algorithms for mapping WH growth stages, which can significantly contribute to can help decision-makers to take necessary measures to manage the spread of water hyacinth with multiple growth stages in the same region.