Optimized algorithm recreates TianQin tunnel using muography

Muography is a powerful, non-invasive tool that can use cosmic-ray muons to image structures underground. Transmission muography can be used to map density distributions, but the accuracy of the reconstruction is algorithm-dependent, and some classical algorithms produce unsatisfactory results due to data smearing.

Qi et al. optimized a Metropolis-Hastings (MH) algorithm to reduce smearing and create muography reconstructions with more accurate density distributions.

An MH algorithm is a type of Markov chain Monte Carlo method, and the team’s MH algorithm “steers the sampling toward density-anomaly regions rather than perturbing the whole model uniformly,” said author Jian Tang.

The researchers validated the efficacy of their algorithm by testing it on muography data taken from the TianQin tunnel in Zhuhai, China.

To do this, the researchers measured the muons in the sky and inside the tunnel, creating a ratio to determine the cosmic ray muon survival rate and opacity measurements.

“Our reconstruction algorithm turns these opacity measurements into results: two-dimensional elevation maps of the air–rock interface, and a three-dimensional density distribution of the rock above the tunnel,” Tang said.

The agreement between the real air-rock interface and its reconstruction validated the reliability of the muography detector used. When compared to baseline algorithms, the MH algorithms exhibited a 72% precision improvement in high-density anomaly detection.

“In the future, we plan to apply our detector to more scenarios, such as karst caves and archaeological sites,” Tang said. “We will also continue to develop and refine our reconstruction algorithm in order to better understand the detector characteristics and expand its applications.”

Source: “Three-dimensional density and air-rock interface reconstruction with muography: Application to the TianQin tunnel,” by Songran Qi, Tao Yu, Shihan Zhao, Yunsong Ning, Aiyu Bai, Yu Chen, Yi Yuan, Mingchen Sun, Zhirui Liu, Liang Xian, Hengye Xu, Hao Jiang, Zhichao Wang, Shuhang Zhang, Su Zhan, and Jian Tang, Journal of Applied Physics (2026). The article can be accessed at https://doi.org/10.1063/5.0342694 .