Additional planetary companions in formerly 1-exoplanet systems

Forever alone? Testing single eccentric planetary systems for multiple companions
http://arxiv.org/abs/1307.0894

Study of single-planet eccentric systems, attempting to model them as 2-planet systems. The 2-planet fit produced plausible results for 10 systems, including HD 192310 (Gl 785) which has had a two-planet solution published already. The others are:

• HD 3651
  
• HD 7449
  
• HD 52265
  
• HD 65216
  
• HD 85390
  
• HD 89744
  
• HD 92788
  
• HD 117618
  
• GJ 649

Forever alone? Testing single eccentric planetary systems for multiple companions

Quote :

Determining the orbital eccentricity of an extrasolar planet is critically important for understanding the system's dynamical environment and history. However, eccentricity is often poorly determined or entirely mischaracterized due to poor observational sampling, low signal-to-noise, and/or degeneracies with other planetary signals. Some systems previously thought to contain a single, moderate-eccentricity planet have been shown, after further monitoring, to host two planets on nearly-circular orbits. We investigate published apparent single-planet systems to see if the available data can be better fit by two lower-eccentricity planets. We identify nine promising candidate systems and perform detailed dynamical tests to confirm the stability of the potential new multiple-planet systems. Finally, we compare the expected orbits of the single- and double-planet scenarios to better inform future observations of these interesting systems.

With some unexpected new entries (and updates) at EPE too:

http://exoplanet.eu/catalog/gl_649_c/ (Hot Super-Earth or Mini-Neptune)
http://exoplanet.eu/catalog/hd_117618_c/ (A "Venusian sub-Saturn")
http://exoplanet.eu/catalog/hd_3651_c/ (This is just weird...)
http://exoplanet.eu/catalog/hd_52265_c/ (...this one too...)
http://exoplanet.eu/catalog/hd_65216_c/ (Mass loss??)
http://exoplanet.eu/catalog/hd_85390_c/ (Saturn-kin)
http://exoplanet.eu/catalog/hd_89744_c/ (Another "weirdo")
http://exoplanet.eu/catalog/hd_92788_c/ (Resonance?)

Nice study... but this thread should be merged with this one.

Yeah, it should. But I think it should redirect here since there are exoplanets added to EPE catalog.

Merged.

Edasich wrote:

Forever alone? Testing single eccentric planetary systems for multiple companions

Quote :

Determining the orbital eccentricity of an extrasolar planet is critically important for understanding the system's dynamical environment and history. However, eccentricity is often poorly determined or entirely mischaracterized due to poor observational sampling, low signal-to-noise, and/or degeneracies with other planetary signals. Some systems previously thought to contain a single, moderate-eccentricity planet have been shown, after further monitoring, to host two planets on nearly-circular orbits. We investigate published apparent single-planet systems to see if the available data can be better fit by two lower-eccentricity planets. We identify nine promising candidate systems and perform detailed dynamical tests to confirm the stability of the potential new multiple-planet systems. Finally, we compare the expected orbits of the single- and double-planet scenarios to better inform future observations of these interesting systems.

With some unexpected new entries (and updates) at EPE too:

http://exoplanet.eu/catalog/gl_649_c/ (Hot Super-Earth or Mini-Neptune)
http://exoplanet.eu/catalog/hd_117618_c/ (A "Venusian sub-Saturn")
http://exoplanet.eu/catalog/hd_3651_c/ (This is just weird...)
http://exoplanet.eu/catalog/hd_52265_c/ (...this one too...)
http://exoplanet.eu/catalog/hd_65216_c/ (Mass loss??)
http://exoplanet.eu/catalog/hd_85390_c/ (Saturn-kin)
http://exoplanet.eu/catalog/hd_89744_c/ (Another "weirdo")
http://exoplanet.eu/catalog/hd_92788_c/ (Resonance?)

I have two questions.
-Why this planets are in the confirmed list? The autor say it's difficult to confirm this planets with available data!
-I dont understand your comments on each planet.

Answering Stalker:
 
1 - It's weird indeed. For this reason I've typed "unexpected" new entries. New EPE is showing such strange update forms (besides still missing significant exoplanets, both candidates and confirmed ones).
 
2 - Comments refer to strange orbital configurations assuming former 1-planet solution which imply a highly eccentric planet for each aforementioned cases.
 
However browsing Wittenmyer's paper the models assume 2-planets solutions with low or zero eccentricity for each planet in the system. "Mass loss" for HD 65216 (A)c alludes to its "shrunken" minimum mass estimate, from 2 Jupiter masses as in first report of HARPS team (I think it was HARPS) to 0.2 Jupiter masses, one tenth of former claim.

HD 92788 b's case is interesting since preliminary astrometric measurements gave true mass around 28 Jupiter masses. What if dealing with another substellar object? Another planet-brown dwarf system like HD 202206 but reverse?

Personally i wont add this planets in my catalog...

HD 92788 is indeed an interesting configuration.

Well obviously it's not a matter of dropping an additional planet into an existing system: they select single-planet systems with e > 0.3, and model it as two Keplerians with e < 0.2 (see section 2)

I.e. it is showing that an alternative model of the system can be used, it is not a discovery claim for additional planets. As the paper says, additional measurements of the system ought to resolve the degeneracy.

I don't get the reason of adding them as confirmed planets indeed. Moreover there was another similar paper (perhaps with the same author) analyzing more than 100 exoplanet systems and inferring several 2-planet systems for many "singleton" eccentric ones (with no discovery claim, of course).

However, concerning HD 3651 (=54 Psc) b's case, with putative orbital inclination of 83_-56⁺⁵⁶° (as in Simpson et al. 2010), it could be an interesting nearby target for transit search, couldn't it?

Also HD 3651 is one of the HARPS nearby bright stars GTO targets.

HARPS-N! HARPS-N!

…While I'm here, I should probably write out my thoughts from a week-and-a-half ago on this mess.

Alllright. So this idea for a paper is pretty good; eccentricities for RV planets are probably the most difficult parameter to constrain, and previous discoveries point to best-fit techniques overestimating the eccentricity of one or more planets if there is a currently undetected signal, well portrayed in the case of Mu Arae. Still, actually detecting these with limited datasets is going to be difficult and prone to false alarms.

For HD 3651, the most obvious issue is that the one-planet fit is better than the two-planet one. That's pretty sure indication that it is unjustified. Heck, we even have an explanation for the large eccentricity (0.6) in the presence of a brown dwarf companion at a few hundred AU, for which kozai oscillations are easily capable of pumping the planet's eccentricity and bringing it close to the star.

HD 7449 has missing data, though justifiably - the CORALIE data in Dumusque et al. is only presented in graphical form. Because of that about half of the dataset is missing, so it is rather premature to use this case; still, the CORALIE data can be somewhat crudely copied from the image, which I have done. Again, the one-planet fit is better than the two-planet one, but because they do not account for the trend caused by the second companion (note: though Dumusque et al. fit this as an eccentric ~5000-day planet, a parabolic trend provides pretty much the same fit and is much less dynamically questionable), both of their fits are rather incorrect.

HD 52265's case is more reasonable. While the two-planet fit is again slightly worse than the one-planet fit, the level of precision for the dominant CORALIE set makes this somewhat less significant. The biggest cause for concern is that the fit is effectively pivoted on the CORALIE data rather than the much higher precision Keck set because the former has much better sampling. It is too early to give a judgement here, but as b's eccentricity is none too unreasonable (0.35), this case is probably an overly eager one.

HD 65216 is definitely too eager. This paper shows no recognition for the second planet quietly announced by Mayor et al. (2011), which is probably to feature in an upcoming paper on new CORALIE multi-planet systems (based on the most recent CORALIE paper's use of some of those quietly announced planets). Their data is obviously not yet available, buuut there is some HARPS data on the archive that I keep rattling on about. Including those with the Mayor et al. (2004) CORALIE data are strongly against this paper's solution and in favour of the earlier one. Because the HARPS data is somewhat lacking the period and eccentricity of the outer planet have to be fixed, but the resultant mass is in good agreement (2.2 ± 0.3 M_J). To demonstrate that, look at the sweet, sweet velocity curve:



The outer planet is a somewhat high-mass Jupiter analogue.

For HD 85390… the writers of this paper evidently did not look at the paper where they got the data from. At all. Mordasini et al. plainly show that there is an upwards trend in the data, which is what is being detected here. While there is some evidence for curvature in the trend, particularly in the positive residuals in the first season, it is too early to discriminate between a trend or a periodicity.

I… don't think that much needs to be said for HD 89744. This paper shows a Chi² of 2.58 for a one-planet fit, and a Chi² of 62.54 for a two-planet fit. The latter is very, very unjustified. Again, there is a BD companion that explains b's large eccentricity.

For HD 92788, the two-planet fit is significantly worse than the one-planet fit (Chi²s of 5.70 vs. 2.29, rms' of 11.2 m/s vs. 8.6 m/s). Systemic will attempt to remove the inner planet if the eccentricity of b is let free in the two-planet fit, so this paper's solution is unjustified.

Concerning the claims of HD 92788 b being a brown dwarf, that comes from the same Simpson et al. two posts above. The value comes from finding the rotational inclination of the star and using the assumption that the planet orbits at the same inclination. For HD 92788 (and HD 69830), they use an extremely low rotational velocity value from Valenti & Fischer (2005), which leads to a very low rotational inclination. Values for rotational velocities generally show an instrument-to-instrument scatter of a km/s or two, so there is likely an overly large level of confidence in this value. Given the rarity of BDs in short orbits, HD 92788 b is probably a planet.

HD 117618 is the only case here where the new planet can have any significant confidence, though that is because this is the only one where this paper supplies new data. In an e = 0 1-planet fit, the second signal has a periodogram FAP of 0.07, which is rather low but is likely significant as it is well clear of the periodogram noise. The second planet is probably real, but requires further data to be more certain.

Gliese 649 is a tentative case. Though it has a periodogram FAP of ~0.1, unlike the above it lies at effectively the same level as the periodogram noise (which is itself caused by the significant change in error bars halfway through the dataset). Though the second signal harbours a significant improvement to the fit, the current dataset is too lacking to make any definitve judgement.

So, that's that. These should really not be on the EPE at all, except for perhaps the last two as candidates.

In other news, new AAT data! omnomnomnom. I'll give a rundown of what's new:

• HD 2039 b has a somewhat improved constraint of its parameters, though coverage of its ascending phase is still largely missing. The orbital period is now well constrained, but the minimum mass and eccentricity are as loose as they were in Butler et al. (2006), because the RV maximum has not yet been observed.
  
• HD 20782 b finally has better periastron coverage! Now there are 6 datapoints around it, improving on 2 from O'Toole et al. (2008). The planet's parameters are now well constrained. The orbital period is considerably longer than the two previous values - 585 & 590 days to 597.1 days - and the minimum mass is a little lower too. With an eccentricity of about 0.950 ± 0.009, HD 20782 b is confidently the most eccentric planet over HD 80606 b. It has an a priori transit probability of about 4%, which will be interesting to follow up on.
  
• HD 23127 b is about the same as it was six years ago, though with a slightly lower eccentricity. The rms of the residuals are still high, which seems to be instrumental.
  
• HD 38283 b was announced only a couple of years ago, so there is not much new data. The best-fit eccentricity has become very high (0.64), but that is caused by one outlying point and excluding that the eccentricity is negligible. The tentative 120-day signal seems to have dissappeared.
  
• HD 39091 (Pi Mensae) b now has two orbits observed, double that of Butler et al. (2006). The one-planet fit is good and is not much different to the previous orbit, but the residuals clearly show a slight upwards trend. There is no known stellar companion, so given the small size of the trend it may be a distant planet. Given the unusual parameters of b (high mass, high eccentricity, moderately long period), planet-planet interactions may have been important in the system's past.
  
• The chances of HD 102365 b being real continue decreasing; the periodogram FAP has dropped from 0.00002 to 0.17 with only one additional season of observations. The newest season has many observations at about 10 m/s above previous data, for some reason.
  
• HD 108147 has only 3 more observations than the Butler et al. (2006) dataset. The residuals show a promising 62.6-day peak.
  
• This is the first update in a long time on HD 142415, and it's a bit weird. b's parameters are better constrained, but the residuals are very high; these tentatively follow a ~3000-day sinusoid, which may be planetary. The residuals are still high, which can be pointed to stellar activity (Log R_hk = -4.5, a value similar to a Hyades FG-dwarf).
  
• HD 187085 b has similar parameters to the announcement paper, except with much lower eccentricity; however, the best fit for that dataset is also of lower eccentricity than was announced, so the parameters remain effectively unchanged.
  
• HD 213240 b remains pretty much the same as the Butler et al. (2006) parameters.
  
• Finally, HD 216437 (Rho Indi) b also remains mostly unchanged, though the residuals show a tentative downwards trend.
  

Somewhat weird they include the 89744 as one of the "plausible" ones really.

Haven't checked whether there are any significant differences between v1 and v2...

Shellface wrote:

In other news, new AAT data!

Shellface, if you don't mind me asking, where are you finding the new RV data for these systems?

They're at bottom of this topic's paper, after the tables and figures. No embargoargo here!

Lazarus wrote:

Forever alone? Testing single eccentric planetary systems for multiple companions
http://arxiv.org/abs/1307.0894

Study of single-planet eccentric systems, attempting to model them as 2-planet systems. The 2-planet fit produced plausible results for 10 systems, including HD 192310 (Gl 785) which has had a two-planet solution published already. The others are:

...

• HD 7499
  ...
  

Shellface wrote:

HARPS-N! HARPS-N!

HD 7449 has missing data, though justifiably - the CORALIE data in Dumusque et al. is only presented in graphical form. Because of that about half of the dataset is missing, so it is rather premature to use this case; still, the CORALIE data can be somewhat crudely copied from the image, which I have done. Again, the one-planet fit is better than the two-planet one, but because they do not account for the trend caused by the second companion (note: though Dumusque et al. fit this as an eccentric ~5000-day planet, a parabolic trend provides pretty much the same fit and is much less dynamically questionable), both of their fits are rather incorrect.

MagAO Imaging of Long-period Objects (MILO). I. A Benchmark M Dwarf Companion Exciting a Massive Planet around the Sun-like Star HD 7449

Quote :

We present high-contrast Magellan adaptive optics (MagAO) images of HD 7449, a Sun-like star with one planet and a long-term radial velocity (RV) trend. We unambiguously detect the source of the long-term trend from 0.6-2.15 \microns ~at a separation of \about 0\fasec 54. We use the object's colors and spectral energy distribution to show that it is most likely an M4-M5 dwarf (mass \about 0.1-0.2 \msun) at the same distance as the primary and is therefore likely bound. We also present new RVs measured with the Magellan/MIKE and PFS spectrometers and compile these with archival data from CORALIE and HARPS. We use a new Markov chain Monte Carlo procedure to constrain both the mass (>0.17 \msun ~at 99% confidence) and semimajor axis (\about 18 AU) of the M dwarf companion (HD 7449B). We also refine the parameters of the known massive planet (HD 7449Ab), finding that its minimum mass is 7.8+3.7−1.35 \mj, its semimajor axis is 2.33+0.01−0.02 AU, and its eccentricity is 0.8+0.08−0.06. We use N-body simulations to constrain the eccentricity of HD 7449B to ≲ 0.5. The M dwarf may be inducing Kozai oscillations on the planet, explaining its high eccentricity. If this is the case and its orbit was initially circular, the mass of the planet would need to be ≲ 10.8 \mj. This demonstrates that strong constraints on known planets can be made using direct observations of otherwise undetectable long-period companions.

So HD 7449 (A)c has to be considered spurious because of presence of HD 7449 B. Along with Kepler-444 A HD 7449 A turns out as another interesting case of exoplanetary system in binary/multiple stellar system (sorry for redundance). There are also interesting constraints about orbital inclination of the planet (>46,4°) and true mass (<10.8 Mj). Could it be worth a long-period transit search?

Findings like these ones encourage me to keep on trusting in the announcement, someday, of an interesting piece of news about nearby 36 Ophiuchi AB system, despite past disproving of substellar/planetary companion around both the stars...

Hey, I remember writing all that! That was a long time ago, huh.

This is about what I expected for HD 7449. The RV trend is rather large, but has already experienced a turnover, so a binary period around a century looked about right. Still, that's a very tight binary for a giant planet host! Seeing as the system is metal-poor, this system looks rather awkward to explain with core accretion.

Considering the fairly moderate binary orbital period, a constrained orbit should be possible within a few decades.

Quote :

Findings like these ones encourage me to keep on trusting in the announcement, someday, of an interesting piece of news about nearby 36 Ophiuchi AB system, despite past disproving of substellar/planetary companion around both the stars...

There is definitely some analogy between the stellar orbits of Kepler-444 and 36 Oph (AB), so planet formation may very well have worked in the latter. Though, 36 Oph is rather young and active, so planet detection may be difficult.

I agree, Shellface. Radial velocities would be biased by stellar activity and a transit search would be as much difficult due to reduced distance and putative starspots too. But who knows? I keep fingers crossed

Uh, same for another nearby binary: 70 Ophiuchi AB.

http://www.spacedaily.com/reports/Monster_planet_is_dancing_with_the_stars_999.html

Monster planet is 'dancing with the stars'

Missing orbital phase information for HD 7449 Ab in the paper, ah well. Also, that's quite an increase in mass for the planet!

Quote :

Also, that's quite an increase in mass for the planet!

Looking again, that doesn't seem right - b's orbital parameters are little changed from Dumusque et al. (2011)'s fit (I'm pretty sure the semi-amplitude is smaller, though this paper doesn't state their value), yet their msin(i) is seven times higher? And why are its errors larger by a similar value? I suspect there is some sort of calculation error here…

(This paper doesn't appear to state their value for the (primary) stellar mass, but if I understand right, their value would have to be ~20 Msol to explain the difference alone, which is clearly incorrect.)

From the Rodigas et al. paper:

Quote :

To constrain the planet’s mass (m), we used the distributions of K_b , P_b , and e_b, drew random Gaussian-distributed values for the stellar mass M_∗ having mean = 1.05 M_⊙ and standard deviation = 0.09 M_⊙, assumed M_∗ ≫ m, and then solved for m sin i_b using the well-known relation...

Which is consistent with the stellar mass from Dumusque et al. (2011)

Judging by figure 2 of Rodigas et al., the semi-amplitude appears to be somewhere around 20 to 25 m/s, versus the 41.59^+10.79_-19.52 from Dumusque et al. (2011).