转载GMT课程lesson6

huolongguo

Lesson 6
Goals

Geological and geophysical analyses are often all about comparing disparate data sets. For example, a problem might involve understanding the relationships between volcano locations and faults, or might involve comparing variations in the gravity field and the distributions of faults and volcanoes. In potential field geophysics, we often compare the spatial distribution of anomalies with other mapped geologic or geophysical features. This kind of comparison helps sharpen interpretation of the potential field anomalies. An essential step in such comparisons is to make sure that the data are plotted together accurately. Obviously the quality of the interpretation depends on the quality of the mapping!

The goal of this lesson is to learn about plotting different types of geologic and geophysical together on the same map using GMT.

---

Seismic Tomographic Anomalies and Volcanoes

Over the last twenty years, geophysicsts have used seismic tomography to image variations in the Earth's crust and mantle in more detail than was previously possible. Seismic tomography works in a way similar to a CAT scan used to image the human body. Seismometers are deployed over a region of interest and these seismometers monitor either natural earthquakes or explosions (or other energy sources). Each seismometer senses the arrival times of seismic waves (for example the P (compressional) wave) associated with each seismic event. Using a simple model of the earth, one might predict the variation of P-wave arrival times among all the seismometers in the region (an exercise that is not too abstract when you realize that rate x time = distance, and that given the source and the seismometer location, it is possible to predict the travel path of the seismic wave). Of course, actual arrival times of P-waves at the seismometers will differ from arrival times predicted by a simple model.

In tomography, the earth is discretized into a series of cubic regions (voxels). The speed of P-waves within each vowel is assumed to be constant. Then using linear algebra techniques, the "best-fit" velocity distribution among all the voxels can be determined that best explains the variation in P-wave arrival times observed by the seismic network.

The variations in P-wave velocity are produced by changes in the elastic properties of rocks. One way to produce slow velocity regions is to develop partial melting along grain boundaries. Consequently, you might expect to see slow velocity regions correlating with evidence of active volcanism.

One of the best seismic tomographic images of the lower crust and mantle comes from northern Japan, where a dense seismic network is deployed. Hasegawa, Zhao, Nikajima, Tamura and others have been able to use these seismic tomographs to create realistic models of subduction zones. The following figure shows contours of P-wave velocity anomalies in northern Japan, using a tomographic model published by Zhao and others. The figure contours P-wave velocity variation along voxels located at a depth of approximately 40 km.