24 July 2014

Oscillation analysis through scientific approach

Raw images of the small part of the solar surface. This is how it looks like
when Sun surface is observed in the real time.  Images are
 from DUNN telescope in NM, USA. I was observing.
This was unusually good data burst, without too much
artifacts.  
Recently I talked with a person who works in the industry. We discussed the techniques for oscillations testing. So I got the idea to describe how similar testing is usually done in science. I'll start from the top, covering all steps we go through before the actual oscillations are detected.
I used this approach in my own research.
The main goal for me was to establish the existence of different kinds of oscillations on a Sun and what they are doing there. Mind, the only data I was getting from the Sun, were images taken through various monochromatic filters. And I had to detect oscillations; acoustic, magnetics. I had to answer are those oscillations damped or not, how much energy they dissipate in their surrounding area, how much of energy is carried away, and what are the oscillations sources.  
The first step was going through all physics laws that involve behavior of the high-temperature plasma. This means going through the part of the physics that is described with MagnetoHydroDynamics (MHD), a study of the fluid behavior

A monochromatic image of the surface of the Sun.
This is an area where there is no strong magnetic fields.
I was observing using 1.5m solar telescope.  One cell you see
 here is approximately size of Spain.

in the presence of the magnetic field. This area has another layer of the complications added to the more broadly known parts of physics,  thermodynamic and fluid dynamics. Evaluating already known stuff from physics was done so that we could get some idea what we might see in the images of the Sun.
Then, models. Physics laws are modeled, covering the known conditions of the part of the Sun we plan to look at. Basically, this means, computer software is written to simulate a behavior of the high-temperature plasma under the same conditions that are already known to exist at that particular part of Sun we were studying. That model provides us with the starting point of how to see an acoustic oscillation in the monochromatic image of the Sun.
But that was just the start.
The collected images of the Sun had to be prepared for analysis, meaning we had to remove all influences that are not coming from the Sun. And trust me, there is all bunch of them.
Once, as a student, I managed to discover oscillations in our data that came from the construction site which was located two blocks away from the lab. Each time people at construction site turned on their machines power supply in the whole area reacted making my instrument register variation in voltage as an input data. Luckily that happened in Germany that has quite regimented working hours so it was easy to determine the cause of mysterious oscillations. (Workers in Germany tend not to change their working hours, or take unscheduled breaks. )
This is an example of the artifacts caused by malfunctioning
instrument that was later removed.  Example comes from data
I collected with 1.5m telescope. 
Of course, artifacts appear in Sun images. I've bumped into artifacts produced by cosmic radiation, increase of the temperature inside the instrument, effects of the other instrument on the one I'm using, and of course, distortions made by the Earth atmosphere. There were artifact caused even by lab partner walking across the lab when data were recorded. All artifacts caused by those influences need to be eliminated before the detection of the oscillation starts.
(This process of preparation is what develops the sceptic side in scientist. The first question we meet when we mention that we found something is, "Are you sure those are real and not some artifact?" And we have to prove to our strict competitive colleagues that we did our best to eliminate all artifacts before saying we discovered something.)
Then, we detect oscillations; in both, Sun data and data got from the model. Results are compared and if they are same, we get confirmation of our assumed conditions on the Sun.
If they are not, then, we tweak a model, using different conditions that  respect the physics laws, and start the process again.
This is an iterative method, the slow method. But the one that leads to the truth.

No comments:

Post a Comment