[ad_1]
A new artificial intelligence is transforming his brain into mapping seismic replicas. Scientists have formed an artificial neural network to study the spatial relationships between more than 130,000 major earthquakes and their aftershocks. When tested, the AI was much more effective in predicting the location of replicates than traditional methods used by many seismologists.
…
Although it is not possible to predict where and when an earthquake will occur, seismologists know some of them about replicas. "We have known for a long time that they will cluster spatially and will deteriorate over time," says geophysicist Susan Hough.
…
[F]Ault orientations in the basement can be as complicated as a mad three-dimensional quilt, and the stresses can push the flaws in many different directions at once. Imagine a book sitting on a table: the shear stress pushes the book to the side and can slide it to the left or right. The normal stress pushes on the book, perpendicular to the table, so that it does not move. A thorny IT problem can be tailored to a neural network.…
Another question is whether a prediction system using such an AI could intervene quickly after an earthquake so that its predictions of seismic replication are useful.
…
In the aftermath of an earthquake, these data would not be available for at least one day.
Read the full and original article: Artificial intelligence could improve predictions regarding earthquake replicas
Source link
