Selection of target stars and field of view

Continuously monitoring approximately 100,000 quiet, late-type target stars will provide a statistically meaningful estimate of the frequency of terrestrial planets in the HZ of solar-like stars. Centered on a galactic longitude of 70° and latitude of +6°, the FOV satisfies both the constraint of a 55° sun-avoidance angle and provides a very rich star field. This FOV falls within the Cygnus-Lyra constellations and results in looking in a tangential direction from the galactic center. In the 100 square degree Kepler FOV, there are approximately 450,000 stars brighter than 15th magnitude. A ground-based observation program led by David Latham (SAO) and Tim Brown (HAO) is underway to observe 2 x 106 stars in the FOV brighter than 17th magnitude. A unique color-filter system, based on the Sloan system and augmented with special filters, is used to identify both the luminosity class and spectral type of each star. Ancillary information from the 2 Mass catalog is also used. The resulting catalog allows the Kepler mission to choose only F through M dwarfs and to exclude giants and early spectral types from the target list. By classifying the stars for which we have complete photometry and all three 2 Mass bands available down to K = 14.5th magnitude, several thousand M dwarfs can be found and put on the target list. Because of their small diameter, these stars will provide sufficient signal-to-noise ratio (SNR) for detection of terrestrial-size planets even though they are dimmer than the majority of other targets. The Kepler results for the frequency of terrestrial planets orbiting M dwarfs are important because most nearby stars are M dwarfs.

Stellar variability sets the limit to the minimum size of planet that can be detected. It reduces the signal detectability in two important ways:

Figure 3. Power spectra of solar variability at solar maximum and minimum. Also shown are energy spectra of 8-hr and 10-hr transits (from Jenkins 2002.)

• The variability introduces noise into the detection passband and thereby reduces the SNR, and thus the statistical significance, of transits.

• Because the flux of every target star is ratioed to the fluxes of several surrounding stars to reject common-mode instrument noise, variability of the stars used in the normalization introduces noise into the target-star signal.

The second concern can be alleviated by measuring the variability of each star relative to an ensemble of others and then iteratively removing the noisiest from the list of comparison standards. To mitigate the effects of the first concern, stars must be chosen that have low variability.

Power spectra for the Sun at solar maximum and minimum are shown in Figure 3.

Also shown are the energy spectra for transits with 8- and 10-hour durations. It's clear that most of the solar variability is at periods substantially longer than those associated with planetary transits. In particular, the Sun's variability for samples with duration similar to that for transits is about 10 ppm. For stars rotating more rapidly than the Sun, the power spectrum will increase in amplitude and move to shorter periods, thus increasing the noise in the detection passband.

Stellar variability in late-type main-sequence stars is usually associated with the interplay of the convective layer and the internal magnetic field. Because the depth of the convective layer is a function of the spectral type of the star, and because the activity level is higher when the star is rotating rapidly, the variability of solar-like main-sequence stars is related to both their spectral type and rotation rate. Further, because the rotation rate decreases with age, the age of a star is an important variable. Thus, we expect that the factors that influence the variability of target stars are age and spectral type.

The age and rotation rate of the Sun are approximately 5 Gyr and 27 days, respectively. The age of the Galaxy is about 13.7 Gyr; about two-thirds of the stars are older than the Sun and are expected to be at least as quiet as the Sun. That extrapolation cannot be verified by examining the actual photometric variability of solar-like stars, because no

Figure 4. Activity indicator RhK versus stellar spectral type (Henry et al. 1966). The Sun lies in the middle of the inactive region. Its value of RhK varies throughout the bin during the solar activity cycle. Note that most stars are less active than the Sun.

star other than the Sun has been measured to the requisite precision. However, the RHK index is believed to be well correlated to stellar variability. It is based on the spectral line profile of the Calcium ii H and K lines and is readily measured with ground-based telescopes.

Figure 4 shows measurements of the RHK index for a variety of spectral types. As can be seen from this figure, about 70% of the stars are found to have an index at least as low as that of the Sun. Hence we plan to choose approximately 150,000 late-type dwarfs to monitor during the first year of observations, and then gradually eliminate those that are too variable, to find Earth-size planets. This action will limit the time needed by the Deep Space Network to receive telemetry from the spacecraft as it recedes from Earth.

Although most solar-like stars are expected to have stellar activity levels no higher than that of the Sun, planets can still be found around stars with higher activity levels if the size of the planets are somewhat larger than Earth, or if:

• They are found around later spectral types;

• The stars are brighter (less shot noise);

• They are closer to the central star (more transits);

• The transit is closer to a central transit than assumed here (grazing transit). Planets in the HZ of K dwarfs have orbital periods of a few months, and therefore would show about 16 transits during a four-year mission. Figure 5 shows the minimum size a planet would require to produce an 8a detection versus the amplitude of the stellar variability, assuming that the frequency distribution of the stellar noise is the same as that of the Sun. The upper curve shows that the amplitude of the stellar noise would need to be at least eight times that of the Sun before it would prevent planets slightly larger than twice the radius of the Earth from being detected. For short-period planets showing 16 transits, planets as small as 1.4 times the radius of the Earth would still be

Stellar Variability, per

Figure 5. Effect of increased stellar variability on the minimum-sized planet that can be detected with 8a.

Stellar Variability, per

Figure 5. Effect of increased stellar variability on the minimum-sized planet that can be detected with 8a.

detectable for stars, even if the star had eight times the amplitude variability of the Sun (see also Jenkins 2002).

Was this article helpful?

0 0

Post a comment