Results

In March of 2000 we used 5 nights at the Kitt Peak National Observatory 36-inch (0.9m) telescope and wide field MOSAIC CCD camera to survey two areas of the sky for extra-solar planet transits. The observing time allowed us to test and refine the same techniques that we will employ in our future projects with the RCT. While we didn't find hard evidence for a transiting planet, we did find some intriguing light curves and met all of our expectations in terms of data quality.

The Fields
Each field that we surveyed was a square region of sky 59 arcminutes on a side. For comparison, the diameter of the moon spans roughly 30 arcminutes on the sky. These are very wide fields for a professional telescope. The richer of the two fields contained 11,500 detectable stars. Over the course of the 5 nights, we obtained 200 high quality exposures of this field. We show sample data here to explain how our project works.

Stellar Colors

A 3-color image taken from part of a transit survey field observed in March 2000 at the Kitt Peak 36-inch telescope. This image was constructed by adding three images together, each of which was observed through a different color filter (U, V, and I). The three individual images were assigned blue, green, and red colors according to the wavelength of light transmitted by their corresponding filters. The combined image reveals the different colors (or surface temperatures) of the stars present in the field.

Most of our exposures were taken through a medium-wavelength filter, called V, that transmits green and yellow light. We used these data to construct our light curves. In addition, we observed the fields once through other filters to observe the brightness of each star at other wavelengths, both longer and shorter than the wavelengths to which our eyes are sensitive. By noting the brightness of a certain star in one filter relative to another, we can quantify its color or spectral class (surface temperature). This is one way to determine the size of each star we are observing. By knowing the size of a star that shows a transit in its light curve, we can determine the size of the object making the transit. This is an important step towards confirming or rejecting it as a planet.

Light Curves

Light curves of three stars observed at the Kitt Peak 36-inch telescope. The top panel shows the light curve of a non-variable star. Its brightness stays at a constant level apart from random noise in each measurement. The middle panel shows the light curve of an eclipsing binary star. The combined light of the two stars diminishes periodically as one passes in front of the other. The bottom panel shows another variable of unknown type.

Our reduced light curves span about 4.5 hours each night for 5 consecutive nights. Except for a few gaps, the data points in the light curve are spaced 5.7 minutes apart, which is the time between exposures. Each individual exposure is 3 minutes long. Our photometric precision depends on the brightness of the star being measured and the observing conditions at the time the exposure is taken. The best precisions we reached are about 0.002 magnitudes, which means that we measure a star's intensity to 2 parts in 1000. Depending on the type (size) of the star being observed, this precision is sufficient to detect transits by Jupiter-sized, Neptune-sized, or even Earth-sized planets (see the the Photometry Details page).

Transits
A few stars show events in their light curves that could potentially be planetary transits. In the figure below, we show a transit-like event that occured in one star. For approximately 1.5-hours on this night, the star dimmed by 1.4 percent. This signature is like that expected if a giant planet transited this star. The colors of this star show that it is like the Sun. This means that a 1.4 percent diminution could be produced by the transit of a planet with a radius 1.2 times that of Jupiter. However, there are other possible explanations for this event: a grazing, or partial eclipse by a stellar companion or another unrecognized type of stellar variability. More observations are needed to determine what is happening to this star.

The light curve of a star on one night that shows a transit-like event. We observed that this star dimmed by 1.4 percent for a duration of 1.5 hours on one night. Such an event could be caused by the transit of a giant planet, but more observations of this star are needed to determine its cause.

It is unfortunate that we were not able to observe this field for more than 5 nights. We didn't observe, and were unlikely to observe, a star showing more than one of these transit-like events in its light curve. Multiple transits by the same planet are what we need to confirm it as a planet. To observe multiple transits by one planet we need a longer timeline of observations. A transitting planet with a short orbital period would dim the star repeatedly, once per orbit by exactly the same amount.

What would the light curve of a transitting planet look like? The appearance would vary considerably depending on the size of the planet, its orbital period and inclination, type of star, and brightness of the star involved. We can take one case as an example. Consider a Jupiter-sized planet orbiting a Sun-like (G2 Main Sequence) star in a 3.42-day period orbit with an orbit that takes it directly into our line-of-sight towards the star. The transits by this planet would last for 3.9 hours and would dim the star by 1 percent. Below, we plot a simulated light curve for the star observed each night for 4 weeks.

A 4 week-long simulated light curve of a star with 5 whole or partial transits occuring on nights #1, 8, 11, 18, and 25 when the star's brightness drops by 1 percent. During the other nights, this star does not vary, but stays at a constant brightness. To produce this artificial light curve, we assume a Jupiter-sized planet transitting a G2 dwarf star (like the Sun) with an orbital period of 3.42 days. The data points in this light curve each have an uncertainty of 0.2 percent, producing a random scatter about the true brightness of the star. These light curve data are like those we expect to obtain in future observations.

We are embarking on a new telescope project to gather observations very similar to these simulations. Searches for extra-solar planets using the 50-inch Robotically-Controlled Telescope at Kitt Peak are scheduled to start during 2001.

On to the next section.
Back to the ESP main page.