Subsurface imaging of volcanoes
Subsurface imaging is a geophysical technique where data is turned into digitized images of the structures that lie beneath Earth's surface. These images are useful in helping to better understand our planet, help predict natural occurrences, obtain information about certain features, obtain the location of resources we want to acquire (for example, precious minerals or oil and gas), and a variety of other things.
The Earth can be broken down by it's four main layers; the inner core, the outer core, the mantle, and the crust. The inner core is the Earth's innermost layer. It is a solid sphere composed mainly of an iron-nickel allow, with a diameter of about 750 miles. The outer core is the next layer. Much like the inner core, it is composed of mainly iron and nickel. However, it is a fluid layer that is roughly 1400 miles thick. On top of the inner core, lies Earth's largest layer, the mantle. The mantle is roughly 1800 miles thick, making up 84% of Earth's total volume. It is composed mostly of silicate rocks that are rich in iron and magnesium. The crust, which is the outermost layer of the Earth, is relatively thin. Oceanic crust is only about 3 miles thick and continental crust varies from 6 to 47 miles thick. For a more detailed explanation of Earth's layers, refer to Layers of the Earth. Volcanoes can form at convergent plate boundaries, divergent plate boundaries, and at hotspots. The most widely accepted explanation for volcanism far away from plate boundaries (like in Hawaii or Yellowstone) is the mantle plume hypothesis. It states that under locations known as hotspots, magma is brought to the surface by a mantle plume, or a section of upwelling rock that is much hotter than the surrounding rock. Once the plume reaches the crust, it forms a magma chamber, spreading out in a cloud-like structure known as the "plume head" and provides lava for volcanoes.
Ways of Acquiring the Data
Subsurface images of volcanoes can be acquired in different ways. The technique used depends on the location and how deep you are trying to look. A few of the techniques that may be used include:
- Ground Penetrating Radar (GPR)
- Remote Sensing
- Seismic Attenuation Profile (SAP)
- Ambient Noise Tomography
Ground Penetrating Radar (GPR)
Ground-penetrating radar is a geophysical technique used to obtain images of the near-surface. It transmits electromagnetic energy into the ground, and measurements of the amplitude and travel time of the reflected waves are recorded. The contrasts of the electrical properties of the ground, defines the different boundaries.  Although GPR can see some, it is typically not the preferred method to acquire this type of data because of it's high frequencies limiting it's ability to reach greater depths without the cost of resolution. An example of GPR data for looking at magma chambers near the Great Rift Zone is provided (each arrow in the image is a different magma chamber).
Remote Sensing is the scanning of Earth via satellite in order to obtain information. Subsurface imaging can be done using radar interferometry. Radar interferometry is used to examine the topography of a volcano and to map surface changes like lava flows or deformations. Interferograms are the combination of two radar images of the same area, taken at different times. They can locate magma flows by showing surface deformations, which help in monitoring and analyzing subsurface activity. Radar is a widely used technique because there are few restrictions on the image taking process. Radar interferomtery provides complete spatial coverage from space, providing forecast and predictions of volcanic activity before the eruption occurs. 
Seismic Attenuation Profile (SAP)
Seismic Attenuation Profiles (SAP) is a method to image seismic attenuation structure from seismic reflection data. It has an advantage in application to geophysical imaging or rock property estimation in areas like volcanic areas or highly faulted areas, which are less reflective areas. This technique has this advantage since it does not require continuous reflections. An example of this advantage being displayed can be seen in the figure below. The downside of this method would be the resolution. The current SAP method, using spectral ratio, requires an averaging process to mitigate the influences from abnormal values caused by local noises. It is reliant on effective frequency preservation during data preprocessing, such as the normal moveout (NMO) correction and deconvolution.  Q is the attenuation quality factor. The larger Q is, the slower the energy attenuates. The smaller Q is, the faster the energy attenuates. The usual way of estimating Q from seismic reflection data is spectral ratios. Spectral ratios use the Fourier transform to compute the amplitude spectrum, so it has a time window problem. If this window is short, the spectrum is convolved with the transfer function of the window and frequency localization is lost. If this window is long, it is difficult to ascertain the spectral behavior of individual reflections. 
Ambient Noise Tomography
The Earth is constantly vibrating, normally imperceptibly, and these vibrations are known as ambient noise. The ocean waves and wind are what mainly cause this noise, but vehicles and machinery are also responsible. Ambient noise tomography requires pairs of seismic stations, so that directions can be determined. It proceeds in three main steps. The first is the cross-correlation of continuous records of up to months of seismic noise for each pair of seismic stations in the area. This happens because the noise is mostly surface waves traveling in random directions, and this step brings out the waves that pass both stations. The second step is estimating the velocity of the surface wave for all pairs of stations. The velocity is simply a function of the frequency of the wave (a phenomenon known as dispersion) and is most sensitive to the S-wave velocity structure. For the final step, the results of the dispersion for all the station pairs are used to construct a three dimensional image of the S-wave velocity structure. An example of a couple of slices from a three dimensional S-wave velocity model, acquired at the Katmai Volcanoes, is shown below. The warmer colors depict areas with relatively low seismic wave velocity, while the colder colors indicate areas with relatively high seismic wave velocities. Anomalous can be identified by deviations from the general pattern that seismic wave velocities normally increase with depth (mainly due to the pressure increasing). 
Having seismic data of volcanoes helps with predicting a volcanic eruption, the amount of lava the eruption will have, the type of magma being erupted, and the potential hazards associated with the eruption. Using seismometers, scientists will be able to tell when there is an unusual increase in earthquake activity. From this, they will then begin monitoring other data such as gas, ground deformation, and satellite imagery. Comparing the satellite images to past satellite images, they can determine if magma is rising towards the surface. Looking at the size of a mantle plume will give an idea of how much lava will be coming out of the volcano when it erupts. Using the P and S-wave velocities, we can determine the type of magma the volcano has. Different magmas have different properties, with viscosity being one of the most important when talking about how the magma will act in an eruption. Magmas with a felsic composition (such as basalt) have a relatively low viscosity. Therefore, once the pressure is released, the lava will be more explosive as compared to a magma with a more mafic composition (rhyolite) which has a higher viscosity. The potential hazards associated with an eruption are a combination of all of these factors. If we know roughly when a volcano will erupt, the amount of lava that will be released, and the type of eruption, people in the surrounding areas can be warned and evacuated if needed. Having this data helps us better prepare for an eruption and ultimately safe lives.
This page is currently being authored by a student at the University of Houston. This page will be complete by March 20, 2018.
- Rowell, C. R., Pidlisecky, A., Irving, J. D., & Ferguson, R. J. (n.d.). Characterization of lava tubes using ground penetrating radar at Craters of the Moon National Monument, USA. Retrieved March 20, 2018, from Characterization of lava tubes using ground penetrating radar at Craters of the Moon National Monument, USA.
- Ground‐penetrating radar sounding in mafic lava flows: Assessing attenuation and scattering losses in Mars‐analog volcanic terrains. (2006, June 29). Retrieved March 20, 2018, from https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2005JE002589
- Haas, C. (n.d.). Remote Sensing Used in Monitoring Active Volcanoes. Retrieved March 21, 2018, from http://academic.emporia.edu/aberjame/student/haas1/volcanoes.htm
- Tsuru, T., No, T., & Fujie, G. (2017, January 03). Geophysical imaging of subsurface structures in volcanic area by seismic attenuation profiling. Retrieved March 20, 2018, from https://earth-planets-space.springeropen.com/articles/10.1186/s40623-016-0592-0
- Hong-Bing, LI, et al. “Characteristics of Seismic Attenuation in Gas Reservoirs in the Domain of Wavelet Scales.” Chinese Journal of Geophysics, Wiley-Blackwell, 31 May 2013, agupubs.onlinelibrary.wiley.com/doi/pdf/10.1002/cjg2.579.
- Thurber, C. (n.d.). Earthquake Studies Reveal the Magmatic Plumbing System of the Katmai Volcanoes. Retrieved March 20, 2018, from https://www.nps.gov/articles/aps-v11-i1-c6.htm