Kepler-68: Two transiting sub-Neptunes and a long-period Jupiter

Kepler-68: Three Planets, One With a Density Between That of Earth and Ice Giants
http://arxiv.org/abs/1302.2596

Quote :

NASA's Kepler Mission has revealed two transiting planets orbiting Kepler-68. Follow-up Doppler measurements have established the mass of the innermost planet and revealed a third jovian-mass planet orbiting beyond the two transiting planets. Kepler-68b, in a 5.4 day orbit has mass 8.3 +/- 2.3 Earth, radius 2.31 +/- 0.07 Earth radii, and a density of 3.32 +/- 0.92 (cgs), giving Kepler-68b a density intermediate between that of the ice giants and Earth. Kepler-68c is Earth-sized with a radius of 0.953 Earth and transits on a 9.6 day orbit; validation of Kepler-68c posed unique challenges. Kepler-68d has an orbital period of 580 +/- 15 days and minimum mass of Msin(i) = 0.947 Jupiter. Power spectra of the Kepler photometry at 1-minute cadence exhibit a rich and strong set of asteroseismic pulsation modes enabling detailed analysis of the stellar interior. Spectroscopy of the star coupled with asteroseismic modeling of the multiple pulsation modes yield precise measurements of stellar properties, notably Teff = 5793 +/- 74 K, M = 1.079 +/- 0.051 Msun, R = 1.243 +/- 0.019 Rsun, and density 0.7903 +/- 0.0054 (cgs), all measured with fractional uncertainties of only a few percent. Models of Kepler-68b suggest it is likely composed of rock and water, or has a H and He envelope to yield its density of about 3 (cgs).

A quite interesting system, worth of a long term follow-up campaign to better characterize the system architecture. In addition, with a quiet host star at Kp=10, Kepler-68c could become the first Earth-size exoplanet with a mass determination. A dedicated campaign with HARPS-North, similar to the one carried out for Corot-7b, has the potential to do that.

28 g/cm³? Not sure I believe that - as they say the significance of the RV variations for this planet is not very high. Very large error bars save the day!

The jovian is in the habitable zone with a fairly low-eccentricity orbit. So seems that such planets do not prohibit the formation of low-mass inner planets. Maybe this is a good sign for the prospects of captured habitable exomoons?

Possible transiting long-period sub-Neptune in a ~1000 day orbit.

Quote :

Independent transit searches of the Kepler photometry have been carried out by Ofir & Dreizler (2012) and Huang, Bakos & Hartman (2013) who also find evidence for Kepler-68b and Kepler-68c with the same period and transit depth within uncertainties. For Kepler-68c Ofir & Dreizler (2012) find a period of 9.60538 ± 0.00026 d and a planet radius of 0.86 Re. They also find a single transit from a possible third planet with a transit duration of ~8 hr implying a period of 970 ± 50 days, and planet size about 2.4 Re. Thus there is a possibility that a third planet transits Kepler-68. Kepler observations of Kepler-68 are ongoing, including acquisition of short cadence data. Inspection of data through Q13 has not shown further evidence for the several hundred day planet candidate. At a period of nearly 1000 days the next transit would not be expected until Q16 in early 2013

Also, where did the convention of numbering non-Kepler candidates in KOI systems starting from 20 come from?

Yeah I wondered that myself!

Lazarus wrote:

28 g/cm³? Not sure I believe that..

The "cgs" in the paper is not itself a unit but a measurement system, Centimetre–gram–second. The cgs unit of density is the g/cm³ (it is not SI). So they are saying 3.32 g/cm³.

I am well aware of what cgs is.

I was referring to the value given for planet c. (Table 4)

And yes I am aware that the value is in any case highly suspect because of the very low significance of the RV detection!

Galzi wrote:

A quite interesting system, worth of a long term follow-up campaign to better characterize the system architecture. In addition, with a quiet host star at Kp=10, Kepler-68c could become the first Earth-size exoplanet with a mass determination. A dedicated campaign with HARPS-North, similar to the one carried out for Corot-7b, has the potential to do that.

This star is included in the HARPS-N, Kepler list, already.

http://www.tng.iac.es/instruments/harps/data/Kepler_GTO_Targets.txt

@Lazarus Sorry, crossed wires, thought you were cubing a figure.

Today' s astroseismology paper gives a radius for planet b of 2.53 times that of earth, so it could be more like an ice giant (very hot) than a benchmark planet, with lower density than published in the discovery paper

Neverthelles there seems to be 4 planets b, c, d and e with period around 970 days and radius 2.4 earth

pochimax wrote:

Today' s astroseismology paper gives a radius for planet b of 2.53 times that of earth, so it could be more like an ice giant (very hot) than a benchmark planet, with lower density than published in the discovery paper

Stellar parameters in the two papers appear to be the same:

Fundamental Properties of Kepler Planet-Candidate Host Stars using Asteroseismology
http://arxiv.org/abs/1302.2624

Table 2

T_eff = 5793 ± 74 K
[Fe/H] = +0.12 ± 0.07 dex
ρ = 0.7903 ± 0.0054 g cm^-3
R = 1.243 ± 0.019 R_⊙
M = 1.079 ± 0.051 M_⊙

Kepler-68: Three Planets, One With a Density Between That of Earth and Ice Giants
http://arxiv.org/abs/1302.2596

Table 4

T_eff = 5793 ± 74 K
[Fe/H] = +0.12 ± 0.074 dex
ρ = 0.7903 ± 0.0054 g cm^-3
R = 1.243 ± 0.019 R_⊙
M = 1.079 ± 0.051 M_⊙

Difference in planetary parameters seems to be that the latter paper re-analyses the Kepler photometry, also simultaneously taking into account the radial velocities, while the asteroseismology paper uses the transit parameters from the KOI catalogue.

So radius correction from 2.52 to 2.31 not otherwise

Quote :

So radius correction from 2.52 to 2.31 not otherwise

Yes, i agree. Thanks.

I wonder but we do not know impact parameter for 4th putative planet at tentative 970 days orbit.Would be good to know possible the closest and the longest distance from host as Jupiter mass planet or 2.4 r planet somehow does not suit there or I am wrong?

Call me crazy… but maybe we should read the entire text…?

Quote :

In Figure 4 we present vetting evidence to further investigate the veracity of the single transit of presumed long period reported in Ofir & Dreizler (2012) and shown in their Fig. 3. The detrending of raw Kepler data adopted by Ofir & Dreizler (2012) results in a ∼8 hour wide intensity drop that is adequately fit with a standard transit model. In the raw LC (simple aperture photometry – no detrending) Kepler data their event is easily seen at BJD - 2455000 = 403.2. However, the raw LC data shows a drop of intensity across this event of about 0.0001 which may be easily seen by drawing linear fits to the data before or after the event. Such behavior is commonly seen with sudden pixel sensitivity losses associated with radiation damage to a single pixel (see, e.g. discussion from discovery in HST ACS data by Gilliland & Bohlin (2007) and Christiansen et al. (2012a) for presence in Kepler data) – and ultimate recovery of most, but not all of the depressed sensitivity. Such sensitivity drops can be particularly difficult to tell from transits in LC data which blur both sudden (spurious) drops, and short ingress/egress events. The spurious signature may be more easily seen in short cadence data as shown in the bottom panel of Figure 4 – here the event looks more like a sudden sensitivity loss and recovery, than a transit. Note that the two occurrences of Kepler-68b transits about one day from both ends show symmetric responses about the transit centers, while the single-event behavior is quite asymmetric with shape characteristic of a pixel sensitivity loss. The middle panel shows that the pipeline PDC-MAP treatment completely removes any suggestion of the single-event transit while nicely preserving the real events from Kepler-68b. These considerations do not prove the single-event transit as an instrument systematic false positive, but this interpretation is favored by us. Definitive proof would come if a difference image analysis across transit could isolate the intensity drop to one pixel, while the stellar signal spans several pixels. We tested for this, but were only able to show that the source is within the set of saturated pixels for Kepler-68, leaving open an astrophysical source.

Quote :

Fig. 4.— The upper panel shows raw Kepler long cadence data spanning six days centered on the “monotransit” event shown in Fig. 3 of Ofir & Dreizler (2012). The middle panel shows the same data after pipeline processing with PDC-MAP. The lower panel shows the same time period with short cadence raw data.

If it were a real transit, the bottom of the "trough" would show a downwards gradient, considering that the intensity went down by about 0.0002% before and after the shape. It does not, so the shape is more likely a 0.0004% decrease in intensity for about 1/4 of a day, which then increased a little after.