DONATE

IAHR Document Library


« Back to Library Homepage « Proceedings of the 28th IAHR International Symposium on Ice ...

UAV-Based Mapping of Ice-Trapped Methane Bubbles Reveals Winter Ebullition Hotspots in a Cold-Region Reservoir

Author(s): Xiaoqiang Chen; Yiming Zhao; Wenfeng Huang; Zhijun Li; Hiroshi Cho

Linked Author(s): Yiming Zhao, Zhijun Li, Xiaoqiang Chen, Hiroshi Cho

Keywords: Ebullition; Methane; Ice-trapped bubbles; UAV; Cold-region reservoir

Abstract: Reservoirs are globally important sources of atmospheric methane (CH₄), yet ebullition remains the most spatially heterogeneous and uncertain emission pathway. Winter ice cover provides a unique opportunity for high-resolution assessment, as CH₄ bubbles released from sediments become trapped within or beneath ice and preserved as visible surface patterns. However, quantitative relationships linking remotely sensed bubble patterns to subsurface bubble structure and CH₄ content remain poorly constrained. Here, we developed an integrated UAV–object-based image analysis (OBIA)–X-ray computed tomography (CT) framework to identify winter CH₄ ebullition hotspots in an ice-covered reservoir in Northeast China. High-resolution UAV imagery was processed using OBIA to extract ice-trapped bubbles. Ice cores collected from hotspot and non-hotspot areas were analyzed using CT to quantify three-dimensional bubble structures, revealing bubble volume fractions of 0.3–2.4% and pronounced vertical stratification within ice layers. CH₄ concentrations in ice, trapped bubbles, and underlying water were measured using gas chromatography. Spatial clustering was evaluated using Moran’s I statistics. UAV-based bubble extraction achieved overall accuracies exceeding 0.80. Bubble distributions showed strong spatial autocorrelation (Moran’s I = 0.63, p < 0.001), with hotspot areas covering only 0.9% of the ice surface yet accounting for more than 95% of total bubble density. UAV-derived bubble area fraction was strongly correlated with CT-derived bubble volume (R² = 0.64) and ice CH₄ concentration (R² = 0.54). This study demonstrates that UAV-detected ice-trapped bubbles provide a reliable proxy for sediment methanogenesis and offers a transferable framework for reducing uncertainties in winter CH₄ assessments of cold-region inland waters.

DOI:

Year: 2026

Copyright © 2026 International Association for Hydro-Environment Engineering and Research. All rights reserved. | Terms and Conditions