Thickness Measurement for Mountain Glaciers With Water-Rich Ice Based on VHF GPR and PSO-CTM-CFAR Detector

• Published in: IEEE transactions on geoscience and remote sensing Vol. 63; pp. 1 - 13
• Main Authors: Wu, Yuxuan, Pang, Xiaoguang, Jiang, Liming
• Format: Journal Article
• Language: English
• Published: New York IEEE 2025; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Climate change; Constant false alarm rate; Constant false alarm rate (CFAR); Detection algorithms; Dispersion; glacier thickness; Glaciers; Granite; Ground penetrating radar; Ice; Ice caps; Ice cover; Ice formation; Ice thickness; mixed Weibull distribution; Moisture content; mountain glacier; Mountain regions; Mountainous areas; Particle swarm optimization; Signal to noise ratio; Stochastic processes; Thickness measurement; Time-domain analysis; Water; Water content; Water resources; water-rich ice; Weibull distribution; Asia
• ISSN: 0196-2892; 1558-0644
• DOI: 10.1109/TGRS.2025.3603270

Glaciers and ice caps in High Mountain Asia are changing rapidly in response to climate change, where ice thickness is a key parameter to describe glacier change. However, in high mountain areas of Asia, with the increasing liquid water content (LWC) of mountain glaciers and formation of water-rich ice layers, the dispersion caused by water-rich ice makes it difficult to directly identify the true location of ice-granite interface through ground penetrating radar (GPR) data. It is a core challenge for thickness measurement of mountain glaciers with water-rich ice to accurately detect the ice-granite interface. In this article, to extract ice-granite interface and measure the thickness of mountain glaciers with water-rich ice, we construct a stochastic dispersive medium model of mountain glaciers with water-rich ice based on FDTD algorithm, establish a mixed Weibull noise model to describe to clutters from liquid water in water-rich ice medium, and propose a corrected trimmed mean constant false alarm rate (CTM-CFAR) based on particle swarm optimization (PSO) to extract the characteristic of ice-granite interface. Compared with other CFAR, the method proposed in this article has higher precision and lower measurement error. In addition, with low signal-to-noise ratio (SNR) conditions, the proposed method also exhibits better performance than other CFAR-based methods.