Interpretable Deep Learning Paradigm for Airborne Transient Electromagnetic Inversion

The extraction of geoelectric structural information from airborne transient electromagnetic (ATEM) data primarily involves data processing and inversion. Conventional methods rely on empirical parameter selection, making it difficult to process complex field data with high noise levels. Additionally, inversion computations are time-consuming and often suffer from multiple local minima. Existing deep learning-based approaches separate the data processing steps, where independently trained denoising networks struggle to ensure the reliability of subsequent inversions. Moreover, end-to-end networks lack interpretability. To address these issues, a unified and interpretable deep learning inversion paradigm based on disentangled representation learning is proposed. The network explicitly decomposes noisy data into noise and signal factors, completing the entire data processing workflow based on the signal factors, which makes the network more reliable and interpretable. Furthermore, physical constraints are incorporated into the learning process to enhance the physical consistency and reliability of the inversion results. The inversion results on field data demonstrate that the method can directly use noisy data to accurately reconstruct the subsurface electrical structure, thereby establishing a unified, interpretable, and physically constrained inversion paradigm for ATEM data processing.