Advanced Journal of Environmental Science and Technology

ISSN 2756-3251

Advanced Journal of Environmental Science and Technology | Vol. 17 (8), pp. 308-322, August, 2026

Research Article

Hyperspectral Satellite Imaging Combined with Random Forest Modeling Maps Invasive Water Hyacinth Proliferation and Methane Fluxes in Tropical Reservoirs

Oluwaseun Adeyemi¹*, Priya Subramanian², Thiago Santos³

¹Department of Environmental Management, University of Ibadan, Ibadan, Nigeria
²Department of Civil Engineering, Indian Institute of Technology Madras, Chennai, India
³Department of Remote Sensing, National Institute for Space Research, Sao Jose dos Campos, Brazil

Corresponding Author: Oluwaseun Adeyemi — [email protected]

Abstract

Nutrient runoff drives explosive mats of invasive water hyacinth (Pontederia crassipes) across tropical hydroelectric reservoirs, blocking waterways and creating anoxic zones that emit potent methane gas. We fused PRISMA spaceborne hyperspectral imagery with floating-chamber greenhouse gas measurements inside a random forest machine-learning framework across three West African and South American reservoirs. The model classified floating macrophyte biomass density with 94.2% accuracy and predicted spatial ebullitive methane emissions with an R² of 0.88. Targeted mechanical harvesting guided by the satellite maps reduced reservoir-wide methane emissions by 41% while supplying feedstock for local biogas digesters.

Keywords: invasive aquatic weeds; water hyacinth; hyperspectral remote sensing; reservoir methane emissions; machine learning; ecosystem management

Manuscript Timeline: Received 13 July, 2026; Revised 10 August, 2026; Accepted 25 August, 2026; Published 30 August, 2026.

Citation: Adeyemi, O., Subramanian, P., and Santos, T. (2026). Hyperspectral Satellite Imaging Combined with Random Forest Modeling Maps Invasive Water Hyacinth Proliferation and Methane Fluxes in Tropical Reservoirs. Advanced Journal of Environmental Science and Technology, 17(8), pp. 308-322.