Rivers running green: water hyacinth invasion monitored from space

Summary

This study demonstrates the use of freely available Sentinel-2 satellite imagery to map water hyacinth invasions at high temporal and spatial resolution over a three-year period in Vietnam's Saigon River. Employing a Naïve Bayes classifier achieving 91% accuracy, the authors reveal pronounced seasonal patterns (peak in February during the dry season) and substantial interannual variability, with coverage driven by meteorological conditions. The work provides an openly available, scalable monitoring workflow applicable to freshwater systems globally, supporting the design of targeted mitigation strategies for this highly invasive species.

UK applicability

Water hyacinth is not an established invasive threat in UK freshwaters due to climate constraints, though the remote sensing methodology could be adapted for monitoring other invasive aquatic plants (e.g. floating pennywort, New Zealand pigmyweed) in UK rivers and wetlands. The automated workflow approach is directly transferable to UK monitoring infrastructure and policy applications.

Key measures

Water hyacinth spatial coverage (percentage of river area); seasonal dynamics; correlation with rainfall and relative humidity; classification accuracy of Naïve Bayes classifier; interannual variability in annual mean coverage

Outcomes reported

The study mapped water hyacinth coverage and seasonal dynamics in the Saigon River over three years (2018–2020) using Sentinel-2 satellite imagery with 91% classification accuracy. Peak hyacinth coverage reached 24% in the upstream section, with strong seasonal and interannual variability linked to rainfall and humidity patterns.

Topic tags