Friday, June 19, 2015

(06-19-15) Camera out of focus = more valuable data from CCD

I find it funny that the most essential methodology for conducting my research at the CfA is to observe bright stars without focusing the charged-coupled device (CCD) detector.  This is necessary for the purpose of obtaining more photons during the observation run and reducing the scintillation noise.  However in order to understand that, I must put the incentive for my research project into context.

A planet orbiting their host star is an event that feels like winning the jackpot, for a scientist.  A plethora of fun facts and discoveries have been revealed in the past 2 decades due to these planets outside of our solar system that "transit", or pass in between the star and the line of sight of stargazers like us humans.  Because astronomers gather information by measuring the characteristics of the light that falls into our telescope, a planet transiting its host star is an exciting event because of the vast changes we see in the light from the bright host star that reaches our telescopes.  In terms of quantity, the light, or amount of photons, we measure in the telescope is less when a planet crosses past the face of a star.  This is illustrated lovely in the following "gif" image.

  Take note that the yellow line is the measurement of the amount of light that we would see from Earth.  Evidently, the measured brightness from the star decreases when a planet transits.  That yellow line being drawn on a Brightness vs. Time plot yields what is known as a light curve.
Also note that the larger (in radius) planet yields a deeper drop in the light curve.

Among the many uses for a light curve, you can determine the point in time when the planet stops moving toward the observer (Earthlings) and begins to move away, the ratio between the star's radius and the planet's radius, the amount of time is takes the planet to revolve thoroughly around its host star, how fast the planet is traveling through space, etc.

What I have previously explained is photometry, the measurements of an object's electromagnetic radiation.  Very soon, I will be measuring photons of bright stars via the MINERVA telescope array currently located on Mt. Hopkins.  I must observe these bright stars and simultaneously obtain enough photons to notice the drop in stellar brightness due to a transiting exoplanet while refraining from saturating the telescope camera's CCD.  This is why I will defocus my camera instead of focusing it on the stars.  Each pixel of the CCD has a limit as to how many photons it can sustain before its well becomes overwhelmed with photons that have been converted into electrons.  Once that threshold is reached, the photons that attempted to enter that overwhelmed pixel will be forced to move into the next nearest pixel.  It's like the photons are flooding the pixels and must find another pixel to fall into.
I can't have that happening during my observation.  That flooding or "bleeding" effect will mitigate the quality of the image and thereby damage the credibility of the desired data. Defocusing the telescope will decrease the chances of this bleeding to happen.  In doing so, I will have dispersed the star's photons from the start of the exposure and thereby decreased the likelihood that one of those pixels will be overwhelmed.  Also, the intentionally dispersed photons will be within a selected region of pixels that can partially be described by the known Point Spread Function.  Furthermore, defocusing the telescope decreases scintillation noise. Yay!

Scintillation noise is also known as observational scatter.  It is what causes stars to twinkle.  Electromagnetic radiation from stars reach Earth's atmosphere and is often absorbed by terrestrial particles, which then cause flourescent emission of light.  Unfortunately for astronomers taking photometry, this flourescent emission is not the same energy as the primary stellar radiation.  That conditions that govern the amount of unwanted scintillation are represented the equation
\[  \sigma_{scint} = \frac{0.004X^{7/4}e^{-h/H}}{D^{2/3}(2t_{exp})^{1/2}}  \]
where X is the airmass, h is the altitude of the telescope (in meters), D is the diameter of the telescope (in meters), \(t_{exp}\) is the exposure time of the CCD (in seconds) and H is 8000 meters representing the scaleheight of the atmosphere.
As the equation suggests, a longer exposure time results into less scintillation noise.  However, a longer exposure for bright stars such as mine will increase the likelihood of saturating the CCD.  This is why the defocusing technique has been significantly effective in the past.  It will allow me to observe bright stars at long exposure times (several minutes) without saturation.  I hope to achieve submillimagnitude (< 1 mmag) scintillation noise when I remotely observe on the MINERVA telescopes pretty soon.  Wish me luck!

No comments:

Post a Comment