Moho Depth Disposition of the Contiguous United States: A Multi‐Modal Data Driven Approach

• Published in: Journal of geophysical research. Machine learning and computation Vol. 2; no. 3
• Main Authors: Mir, Ramees R., Makhdoomi, Sameem, Huang, Hui, Ma, Zhitu, Mousavi, S. Mostafa
• Format: Journal Article
• Language: English
• Published: 01.09.2025
• ISSN: 2993-5210; 2993-5210
• DOI: 10.1029/2025JH000604

Precise determination of Moho depth is essential for understanding lithospheric deformation, lower crustal rheology, and crust‐mantle interactions. This study employs a Random Forest Regressor (RFR) model to predict Moho depth across the contiguous United States using a variety of geophysical data, including gravity anomalies, topography, heat flow, magnetic anomalies, sediment thickness, and electrical conductivity. The RFR outperforms the Airy isostatic model and other Machine Learning (ML) approaches, achieving a Coefficient of determination ( R 2 ) of ∼0.75, Root Mean Square Error of ∼3.3 km, and a standard deviation of ±10 km for ∼80% of predictions. Permutation importance identifies elevation and Bouguer anomaly as primary predictors, with their independent contributions confirmed despite correlation. Sensitivity analyses validate robustness, though the model is not transferable to oceanic crust. An innovative error estimation methodology integrates correlated and uncorrelated noise with variable standard deviations to capture geophysical data uncertainties and enhance model reliability. This study highlights the potential of accessible ML methods for first‐order Moho depth estimation in data‐scarce regions. We created a map of the depth to the Mohorovičić discontinuity (Moho) across the United States. The Moho is the boundary between the Earth's crust and the mantle, thus an important parameter to understand the Earth's structure. Instead of relying on traditional methods, we trained a computer program to recognize patterns in different types of data, such as gravity, elevation, and heat flow. Our method to obtain the map can help scientists to design and apply similar methods to data scarce regions such as central and northern Africa and parts of Asia. Our study also shows how machine learning can be used to make sense of complex data in geology, and how these techniques can be easily taught and understood, even by those new to this technology. Developed a Random Forest Regression model predicting Moho depth of US with R 2  = ∼0.75 and Root Mean Square Error = ∼3.3 km Introduced an error propagation framework accounting for epistemic and aleatoric uncertainties This study highlights the potential of accessible Machine Learning methods for first‐order Moho depth estimation in data‐scarce regions