Solar Tomography


Tomography has been used successfully in medicine for many years. Basically, a series of images (X-ray images for example) taken from many different angles can be used to reconstruct a 3D map of a patient's body. The same technique can be used on the solar corona (the Sun's outer atmosphere). Since the Sun, and the corona, rotates about once a month from the Earth's perspective, we can take a series of images of the corona during the rotation, and use the techniques commonly used in medicine to reconstruct maps of the 3D structure of the corona.

This is not a new idea. Since the 1970s solar scientists have used various techniques to attempt tomography on the solar corona. These techniques are named Solar Rotational Tomography (SRT). Compared to medical tomography, big problems arise in SRT. Two of the biggest problems are:

- The brightness of the corona drops very quickly with distance from the Sun. The brightness drops by a factor of over a thousand from near the Sun to far (from 2 to 6 solar radii from Sun center). This means that any small measurement error made NEAR the Sun is comparable to the actual measurement FAR from the Sun. This makes tomography unstable.

- We cannot see through the Sun. This means that there are large regions of the corona which are 'missing' in our data, which also makes tomography unstable.

For these two reasons, and others, solar science has not used the same excellent tomography techniques as used in medicine - until now that is. I have found that applying medical tomography techniques on flattened coronal images (flattened as in artificially removing the steep drop in brightness from coronal images) results in useful high-resolution maps of the coronal structure. Although this technique is in its infancy, the results show great potential. The new technique is called Qualitative Solar Rotational Tomography (QSRT).

Tests on model

Before applying the technique to real data, it is a good idea to test its accuracy on a model corona. This next figure shows the density in a model corona. This figure is a longitude-latitude map (like a world map), showing a 'shell' of the corona at a height of 4 solar radii. Red areas are high-density, and black areas low-density. Viewed in this way, this corona resembles the face of a six-eyed alien. The figure underneath shows the reconstruted corona using the QSRT technique. The reconstruction is not perfect, but it is pretty good.

It is interesting to see the development of the reconstructed map as we view the model corona from more and more viewing angles. This movie shows the QSRT technique in action.

Application to observations

The next figure shows the density distribution of the corona during March 2003, calculated from real observations made by the LASCO C2 coronagraph aboard SOHO using QSRT.

I can test the accuracy of the map by creating artificial or synthetic observations from the calculated density structure, and comparing them to the true observations. This comparison is shown for several different dates in the following figure. In this figure, the left column are the actual observations of the corona, and the right column the synthetic observations.

Click here to see animation

Summary and context

The QSRT technique gives a detailed map of the coronal density structure directly from coronal observations. The agreement between observation and the reconstruction is remarkable. These maps will be very useful for most areas of coronal and interplanetary solar wind research. Even though the technique is new (developed in the last year), it is proving extremely useful for the Institute of Maths and Physics (Aberystwyth University) Solar System group in interpreting their radio-wave observations of the solar wind. Research on Ultraviolet observations of the corona, conducted by myself and Shadia Habbal at the Institute for Astronomy, University of Hawaii are also benefitting greatly from these density maps.

These maps will also prove useful in the important field of space weather. Explosions at the Sun travel through space and often hit the Earth. These energetic magnetic clouds can disrupt communication, power supplies and be a major health hazard for astronauts and airline pilots. Understanding and predicting these storms is a major goal of solar science. The ability to map the whole 3D structure of the corona is a critical step towards achieving this goal.


© 2006 All rights reserved