{"id":44039,"date":"2026-10-27T15:55:32","date_gmt":"2026-10-27T19:55:32","guid":{"rendered":"https:\/\/www.ecoportal.net\/en\/?p=44039"},"modified":"2026-10-27T15:55:32","modified_gmt":"2026-10-27T19:55:32","slug":"3d-seismic-scans-machine-learning","status":"publish","type":"post","link":"https:\/\/www.ecoportal.net\/en\/3d-seismic-scans-machine-learning\/44039\/","title":{"rendered":"3D seismic scans and machine learning traced hidden branches through five major faults beneath China\u2019s Tarim Basin"},"content":{"rendered":"

Geoscientists focusing on China used a combination of high-resolution 3D seismic imaging and machine learning algorithms<\/strong>. <\/span><\/p>\n

The result was mapping of multi-scale strike-slip fault networks in the deep Tarim Basin of China. <\/span><\/p>\n

Traditional deep subsurface exploration in remote continental basins depended on seismic data obtained from grids with large spacing<\/strong>.<\/span><\/p>\n

These could not detect secondary fracture lines. <\/span><\/p>\n

Combining pattern recognition algorithms with vast geophysical datasets made it possible for researchers to map patterns of structural deformations along five major faults<\/strong>.<\/p>\n

Mapping deep subsurface fault networks using 3D seismic imaging<\/h2>\n

To map subsurface structures in a large sedimentary zone of the Tarim Basin, super precise geophysical surveys are needed.<\/p>\n

Two-dimensional line surveys of seismic lines cannot show the three-dimensional structure<\/strong> of deep strike-slip fault zones.<\/p>\n

Geophysicists conducted many surveys through the use of 3D reflection seismic studies in the Shunbei central region.<\/p>\n

Seismic waves<\/strong> were detected.<\/span><\/p>\n

The high-resolution reflection seismic surveys reveal minor vertical displacements that are thousands of feet below extensive layers of desert sand.<\/span><\/p>\n

3D reflection surveys provide an accurate structural model<\/strong> needed for tracing extensive fault systems.<\/span><\/p>\n

Utilizing machine-learning algorithms to identify fracture patterns<\/h2>\n

The manual analysis of huge volumes of seismographic data takes months of hard work and often fails to detect the faintest low-order fault structures.<\/span><\/p>\n

Engineers applied machine learning classifiers<\/strong> to detect low-order structural anomalies from millions of seismographic data points.<\/span><\/p>\n

Many additional low-order branch connections between major fault fragments were found automatically using such algorithms.<\/span><\/p>\n

Pattern recognition helped speed up the process of fault mapping<\/strong> and detection while eliminating possible human-induced biases.<\/span><\/p>\n

Map tracing helped detect micro-fractures that went unnoticed by the standard manual approach, and the picture of what is happening below the surface is now much clearer.<\/span><\/p>\n

Tectonic stress simulation by physical sandbox modeling<\/h2>\n

Numerical modeling is not enough to model the physics of the processes involved in strike-slip fault development<\/strong> on a multi-scale basis over millions of years.<\/span><\/p>\n

Physical sandbox experiments were conducted using sand and silicone materials to model the process of deep basement faulting.<\/span><\/p>\n

Compression in the physical apparatus modeled the structural geometry seen in seismic studies<\/a> of the region, and the potential for deformation became visible<\/strong>.<\/span><\/p>\n

The experiments showed that the large regional faults accommodate the strain from the tectonic plate movement in the early stages.<\/span><\/p>\n

By absorbing the primary strain, it prevents any geometric change or reactivation of the secondary fault structures<\/strong>.<\/span><\/p>\n

Mechanical strain distribution in deep fault systems<\/h2>\n

Tectonic deformation over continental interiors does not occur uniformly through all possible fractures in a specified area.<\/span><\/p>\n

The main fault systems carry major shearing stresses<\/strong>, and thus create stress shadow zones around the secondary rock bodies surrounding them.<\/span><\/p>\n

There is strict hierarchical control of how multi-faults accommodate the regional tectonic stresses on their scales, according to AGU \u2014 Tectonics<\/a><\/span>.<\/span><\/p>\n

Subordinate secondary faults remain inactive<\/strong> once the main faults have adjusted themselves with respect to the primary strain mechanism.<\/span><\/p>\n

Knowledge about strain partitioning behavior will help to estimate the pathways of fluids stored inside deep carbonate reservoirs<\/a> below the Earth’s surface.<\/span><\/p>\n

Assessing the significance for deep basin resources prospecting and tectonics<\/h2>\n

It is vital to comprehend the network of the branch faults<\/strong> in the Tarim Basin to have a clear understanding of continent deformation and the geological history of the region.<\/p>\n

Strong structural models improve the production process involved in extracting oil and geothermal energy in challenging desert environments.<\/p>\n

Low-order fault branch complexities should be understood to prevent dangerous drilling problems<\/strong> in deep excavation.<\/p>\n

Revised fault maps at different scales will help to bridge the gap between plate tectonic theory and reservoir engineering.<\/p>\n

Integration of geophysics will make sure that future drilling projects<\/a> will be accurately applied to deep continental fault systems<\/strong>.<\/p>\n","protected":false},"excerpt":{"rendered":"

Geoscientists focusing on China used a combination of high-resolution 3D seismic imaging and machine learning algorithms. The result was mapping of multi-scale strike-slip fault networks in the deep Tarim Basin of China.  Traditional deep subsurface exploration in remote continental basins depended on seismic data obtained from grids with large spacing. These could not detect secondary …