Environmental coherence framework for multi-sensor remote sensing: water hyacinth assessment in Lake Tana

Area of Science:

• Environmental Science
• Remote Sensing
• Ecology

Background:

• Invasive water hyacinth poses a significant threat to Lake Tana, the source of the Blue Nile.
• Effective monitoring is crucial for managing this aquatic invasion, especially in data-scarce regions.

Purpose of the Study:

• To conduct a comprehensive 11-year remote sensing assessment of water hyacinth invasion dynamics in Lake Tana.
• To introduce and validate a novel environmental coherence framework for assessing remote sensing algorithms.
• To provide data-driven insights for targeted environmental management of aquatic invasions.

Main Methods:

• Utilized an 11-year (2013-2024) remote sensing dataset from Landsat 8/9, Sentinel-1, and Sentinel-2.
• Developed and applied a novel environmental coherence framework for indirect validation of multisensor algorithms against hydro-meteorological drivers.
• Employed Google Earth Engine for a replicable workflow in monitoring aquatic invasions.

Main Results:

• Sentinel-2 NDVI and NDVI+FAI were identified as the most effective indicators for monitoring water hyacinth, showing high environmental coherence.
• The invasion exhibited a complex 'boom-bust' cycle, with a peak in 2018-2019, followed by a decline and recent resurgence.
• The environmental coherence framework provided ecologically grounded validation, suitable for areas lacking systematic field surveys.

Conclusions:

• Established a robust, replicable Google Earth Engine workflow for monitoring aquatic invasions in data-scarce tropical regions.
• The study provides critical insights into the invasion dynamics of water hyacinth in Lake Tana, informing management strategies.
• The environmental coherence framework offers a valuable alternative for validating remote sensing data in ecological studies.