Report Gene

7/23/20 14:57

Topic

Webinar ID

2020 Sagan

940 3997 6754

Question Det

Question

Answers

Mass precisions of 20% shouldnt be very difficult to get these days with EPRV. Am I missing a point here?

I guess it depends the mass of the planet you are targeting. Even in EPRV area 20% on 50 cm/s is not so easy to reach.

Exactly. Even with EPRV, 20% is not easy for small planets as the semi-amplitudes are not much larger than the instrumental error.

I totally agree. Also don’t forget that there might be stellar noise that might be correlated, so that the semi-amplitude (and mass) is degenerate to some extent with the noise power. For planets whose atmospheres can be characterised I guess this situation is quite rare though.

Why is the difference between an emission spectra and reflective spectra in the library of codes you showed ?

Emission spectra shows the flux that comes from the thermal emission from the planet itself, which is usually visible at longer wavelengths. Refelction spectra comes from computing the wavelength dependent albedo of the planet to determine how bright the planet is. The radiative transfer of those two problems are just slightly different so we have to have different routines. In picaso you can see the two different routines https://github.com/natashabatalha/picaso/blob/0737df3a7f7af930af0b94512c6869f94cfc07c9/picaso/fluxes.py

See “get_reflected_1d” versus “get_thermal_1d”

How do you find the planet’s equilibrium temperature?

You can calculate it using the luminosity of the star, the semi-major axis of the planet, and the bond albedo of the planet using the Stefan–Boltzmann law.

1xsolar what?

I think this referred to Solar metallicity and C/O ratio

How is metallicity derived?

Stellar metallicity is derived via stellar spectroscopy, using the absorption lines of various elements. If we get EPRVs, the stellar spectra are there so we usually use those.

Is there any relationship between planet and Stellar metallicity?

This maybe of interest: Do Metal-Rich Stars Make Metal-Rich Planets? New Insights on Giant Planet Formation from Host Star Abundances by Teske, et al., (2019) https://arxiv.org/abs/1912.00255

What would be the best way to deal with potentially changing atmospheres and/or clouds like has been seen in giant planet in our Solar System and brown dwarfs?

Instead of just setting a grey cloud model like I showed in the tutorial, you can use a cloud model like this one https://natashabatalha.github.io/virga/ which was grounded in observations of Jupiter and Brown Dwarfs. This model will give you full cloud profiles as a function of altitude instead of just setting a grey cross section.

Why wouldn't the atmospheric signal look like blackbody? I thought the features would be absorbed on a blackbody curve.

You would only expect a blackbody if you had perfectly isothermal profile and so no matter where in pressure you are optically thick, you are always getting the same flux. Thermal emission comes from the pressure level where you become optically thick, which depends on atmospheric absorption. The features that you see come from becoming optically thick at different temperatures. Usually there adiabatic structure to your PT profile. So in regions of high absorption, you optically thick higher up in altitude, colder, which results in less flux.

On what parameters or factors does critical metallicity depend?

For the planet atmosphere, its metallicity can depend on many things, from the material it accreted to the planet’s total mass.

What is common mode?

A common mode in this case is something that is present in all of the columns of the matrix, for example, the flux steadily decreases over time. You could take that trend out of the data and that would leave you with just noise and the planet spectrum (in a very idealised case).

How do you determine the optimal number of modes to subtract? I guess there is a trade-off between subtracting tellurics and self-subtracting planet lines. Along those lines, how does resolving power play into that trade-off?

I mentioned injections in answering live, but as for resolving power - you need enough to see the planet move across enough pixels on your detector to definititely separate it, so that then depends also on the orbital velocity of the planet itself and how long you can observe it for

The tellurics are multiplicative, but the PCA removes it as a subtraction? Can it be a problem for the planet portion of the signal?

In my understanding the PCA is done on the log flux, such that multiplications become sums. You can loook at Artigau et al. 2014 for details

https://ui.adsabs.harvard.edu/abs/2014SPIE.9149E..05A/abstract

Although I assume you need decently high SNR spectra? If your SNR per bin is close to 1, you'll have a lot of negative values.

I’m not sure to see what you mean, to me the flux is always positive and it’s not too much a problem. That said I answered your question based on a talk of Etienne Artigau I saw last year. I’m pretty sure he said the analysis is done in log flux but I’m not super sharp on the issue!

Yes, we work in log flux - the SNR of the stellar spectra are typicaly 100-250 (it’s only the planet spectrum itself that’s at SNR~1)

Will it be possible to investigate the atmospheres of rocky planets with this technique in the future?

Definitely, it’s mostly a case of getting a big enough light bucket for collecting photons. We may use transmission, emission, or even reflected light to do this, depending on the planetary system itself.

How do you generate the noise that we apply to the model spectra when cross-correlating models with the noisy planet data?

In this case, just using a random number generator to give a range of values from a Gaussian distribution, scaled for the signal-to-noise of the spectra.

What is the limit to detect molecules through this method?

There’s a number of things! I mentioned the line lists being an issue, but there is also just the number of photons you can receive. Also, the more lines you can detect the better, so it works better for molecules with many lines than singular lines. It also depends on the planet moving fast enough. If I get time, I’ll show you how this can work for very slow moving planets i.e. wide-separation directly imaged planets.

Is there any relation between the mass of the planet and the elemants found in the planets atmosphere ?

Yes, there is a large range. In Jupiter-mass planets we expect them to be H/He dominated, while for a rocky planet like Earth it is nitrogren dominated. The temperature plays a big role in the chemistry too.

Is there a possibility of detecting spectrum of planet in emission

Yep! So most of what we have detected so far is thermal emission coming from the planet’s dayside i.e. we look at it just before and after secondary eclipse (superior conjunction). As for actual emission lines themselves, this has been seen in the Nugroho papers for TiO and Fe II in the optical, and they are due to a temperature inversion in the planet atmosphere.

How can we find whether a planet is tidally locked or not ?

Hopefully I can show a slide on this, but if you look at Brogi et al. 2016 and Louden & Wheatley 2015 you’ll see an example - basically the width of the CCF tells you if the planet spectrum is rotationally broadened and you can compare this to the orbital period to see if that match or not for a tidally locked planet.

Do the spectra have enough resolution to detect phenomena like Zeeman splitting yet? If not, is there some other method used to detect a planet's magnetic field?

This is an excellent question and there is a very recent paper from Antonija Oklopčić (https://arxiv.org/abs/1910.02504) that looks at trying to measure magnetic fields using SPIRou.

What is the sweet spot in terms of wavelength regime for detecting planet atmospheres?

This depends entirely on what you’re trying to find - the planet’s brightness will be brightest near the the peak of its black body continuum typically, but if you wanted to find e.g. biosignatures, you’d have to specifically target the wavelengths where those species and spectrally active so e.g. 760nm for O2.

RV community is trying to push the detection of exoplanets to M-dwarf since planets are more easily detectable on low-mass stars. Is there some counter parts about planets habitality around such stars compared to Sun-like star (winds, X-ray emission, …) due to the proximity with the host star ?

I recommend taking a look to Rimmer et al. 2018 about RNA precursors on exoplanets.

While habtiable zone planets are easier to find around M-dwarfs compared to more massive stars, due to the increased activity (flares, CMEs etc.) of M-dwarfs and proximity of these planets to their host star could these planets actually be habitable?

There could indeed be a trade off. Rimmer et al.2018 (https://advances.sciencemag.org/content/4/8/eaar3302) shows that the amount of UV light alone can be an issue. Check their figure 4.

M dwarfs are being prioritized for observations because of how easily detectable the planets in their habitable zones are, but are there any other possible inhibitors (absence of volatiles, etc.) to these planets hosting life?

Check Rimmer et al.2018 which shows that more UV light would be needed for life as we know it around M dwarfs or cooler stars.

To Dr. Batalha - what is the y = mx+ b relationship you’re looking to reject when considering the feasibility of transmission spec. for a planet?

If you look at my 1x and 200x M/H plot in the tutorial with the data, you might imagine a case that is slightly more enhanced in metals that would result in a smaller spectral feature. If that were the case, the spectrum would fall “within” the error bars I am showing. Therefore, within the error bars you could simply draw that spectrum would be synonimous with a flat line. Only at higher precision would you be able to detect the water bump

Also to Dr. Batalha - why is there a degeneracy between mean molec weight, mu, and g?

Both increased gravity and decreased mean molecular weight have the effect of supressing atmospheric features. This paper https://ui.adsabs.harvard.edu/abs/2012ApJ...753..100B/abstract especially figure 3 talks about the challenges of getting the mean molecular weight right.

Can we expect silicon based life forms coould exist ?

Stay tuned for the talk by Jim Kasting!

What is the mass range where a planet can reasonably retain an Nitrogen Oxygen Atmosphere?

It depends on the stellar flux. Within the HZ, it's probably on the order of 0.5 Earth masses. Look for a paper by David Catling and Kevin Zahnle on 'the cosmic shoreline'.

Kaltenegger 2017 is a great review of HZ interpretations and biosignatures

https://ui.adsabs.harvard.edu/abs/2017ARA&A..55..433K/abstract

With the anticipation of life on Europa, has it been a study object for spectra study?

You can look at this paper to see what Europa would look like in reflected light compared to other solar system planets https://www.liebertpub.com/doi/10.1089/ast.2017.1763 unlike much of the discussion today about atmospheres Europa its spectrum is dominated by primary surface features!

When we can detect an exoplanet in the Andromeda galaxy in the future?

Check out some of the talk recordings from Tuesday. This was already addressed in the question/answer period there.

Can the RV method be used to measure the radius of exoplanets?

No, it can't. RVs effectively treat the orbiting bodies as point masses. So we need transits to provide radii.

For many of those plots, radius is inferred from mass and vice versa under an assumed density law, so that they can all be viewed on the same plot. You are right, only planets that transit have radii measured, and most masses come from radial velocity.

Can we use fourier analysis to separate multiple transists due to different panetary companions? what are the other methods to do so?

Transiting planets are often detected by constructing box-fitting least-squares periodograms. Check out the 2012 Sagan Summer Workshop for more details and hands-on practice: https://nexsci.caltech.edu/workshop/2012/agenda.shtml

We talk about a lot of planet statistics based on Kepler data which is a transit mission including some ~3000 planets studies only by transit (not confirmed by RV/TTV/astrometry). These could still be spurious detections. Aren't we wrong to make the claim of ~4000 planet detections here, some soleley based on the transit method?

No it is not. For a confirmation it has make three transits. It is unlikely that you will get a dubious 3 transit detection from a single target. Please correct me if I am wrong :)

But could false positive not be induced by stellar activity ?

You mean can we mistake a star spot with a transiting planet?? Well no!! . Both have differnent transiting geometry… U can Identify with that. :)

Can we get an idea of limb darkening from transit light curves

Yes. You can get the limb darkening from transit light curves.. :)

Why aren’t mass/radius compositon lines linear (for constant density)?

As you go to higher and higher masses, the material insid the planet is more compacted.

Somewhat linked to RM, can you talk what observations of gravity darkening in transits of planets around hotter stars can tell us about the system?

For gravity darkening, you could check out this article (https://www.nasa.gov/feature/goddard/2020/nasa-s-tess-delivers-new-insights-into-an-ultrahot-world) and ask Scott Gaudi about KELT-9b.

Also can you talk about the merits/drawbacks of photo-evaporation versus gas poor formation region models used to aim to explain the Fulton gap seen in equilibrium temperature versus planet radii plots?

I wanted to spend more time on the planet radius distribution, but that would have needed a second talk.

How does speckle imaging work?

Here are a few links about speckle imaging: - https://www.nasa.gov/feature/speckle-instrument-brings-astronomical-objects-into-focus

How can we confirme about the transiting exoplanets whether they have massive core or not?

It’s tricky because of the degeneracies. For solar system objects, we measure the higher moments of the gravitational field. For transiting exoplanets, we can’t really peek beneath the surface, but we could measure the bulk density and place a lower limit on the amount of material likely to be in the core (e.g., iron). We can also determine the compositions of disintegrating planets currently being eaten by white dwarfs. It’s difficult to tell exactly which part of the planet is being injested, but we can assess whether the material is more core-like or mantle-like.

For a particular planetary system, are all the planets formed from the same protoplanetary disc? If yes then why do planets differ in their orientation?

Dynamical interactions among planets and close fly-bys with nearby stars in crowded regions can alter orbits, but that’s an active area of research. :)

Correct me if I am wrong but there are different theories of planet formation like core collapse and disk instability :) It can be determined from Mass of the planet and host star metallicity :)

Yes, they form from the same disk. In general, they orbit together in a common orbital plane, but sometimes planet-planet interactions or the passage of nearby stars can knock them into highly inclined orbits.

Fairly often, I think we hear that our solar system is unusual/other systems don't look like ours. How probable is it that maybe it's not that strange, but limitations to our observations obscure the features that set the systems we look at and our own apart?

Good point. We know that many systems are different from our own, but we don’t yet have the sensitivity to say exactly how many systems look like our own. We do know that most of the planets we’ve detected are unlike those in our own solar system. For instance, the Kepler-11 planets are closer in and the HR 8799 planets are farther out.

Why is early Mars and recent Venus higher up on y-axis for a brighter star, hotter than the current Sun? If the Sun is getting brighter and hotter w/ time, shouldn’t those limits be below 5800K ?

Sorry, I skipped all the details on the HZ. Earth-like planets orbiting redder stars have lower albedos because Rayleigh scattering is lower and absorption of stellar near-UV radiation is larger. Just the opposite for planets orbiting bright blue stars.

I think our moon also played an important role in making our planet habitable. Adding to the habitable zones, greenhouse effect, etc. Would a planet also need a moon for maintaing its habitability over time?

Although this does not directly answer your question, I suggest this paper which talks about the influence of Earth's moon on its habitability, and the impact (or potential lack of an effect) we would expect on planets orbiting single or multiple star systems. it heavily focuses on the implications of obliquity on the evolution of life: https://iopscience.iop.org/article/10.3847/1538-4357/ab46b5

For the habitability, would a large gas planet near the terrestial planet have an important role, just as Jupiter would block out dangerous celestial bodies coming to Earth using its gravitational force?

Interesting paper related to this here: https://ui.adsabs.harvard.edu/abs/2020AJ....159...10H/abstract Also I’ve seen talks by the first author about how Jupiter throws just as many roacks towards Earth as it does away! Can’t find the related paper though

I am also curious about the effect of the exomoons as well on exoplanets. I've read that lunar tides give allow a large biodiversity for our Earth!

What is the main technological reason to have a TESS-like mission (looking at relatively bright stars across the sky) come later and not before a Kepler-like mission, which focused on relatively fainter candidates at one particular region of the sky. Was it in some way easier to design a planet hunting machine like Kepler and not TESS a decade back?

Before Kepler, we didn’t know that there were so many small planets that TESS could detect. Kepler could have discovered that small planets with high transit probabilities were rare. The more complicated answer is that planning for TESS began before the Kepler results were known, but it is true that a TESS-like mission isn’t as compelling if transiting planets are rare.

Kepler was designed to measure eta_Earth with transits. This requires nearly continuous monitoring of a larger number of stars for over three years. This means that it wasn’t able to look at multiple fields; it had to focus on one field for nearly four years. TESS only looks at most regions of the sky for ~27 days, not nearly long enough to detect transits of Earthike planets oribiting sunlike stars. TESS was designed to find transiting planet around bright stars, which are located all of over the sky, so it has to moving to new fields to monitor nearly the entire sky.

Could you elaborate on why stitching transmission spectra together is challenging?

There are a number of reasons. If the stars are variable, and that variability changes with wavelength, and changes over time then it is likely one can “offsets” in your transmissio spectra in two different bandpasses taken at two different times.

It’s really because a transit depth is a relative measurement, and you can’t control what your star is doing. I’m only a theorist, I’m sure an observer could give a better and more specific answer. ;)

What enables the ~ 4x improvement in noise floor for OST when compared to JWST? Is it just the colder temperatures?

It would be a new kind of detector — so, technology development. There are a few conference proceedings and papers in the literature on the onging tech development.

Here is a paper: https://www.spiedigitallibrary.org/conference-proceedings-of-spie/10698/1069844/A-highly-stable-spectrophotometric-capability-for-the-Origins-Space-Telescope/10.1117/12.2311896.short?SSO=1

Would HabEx be able to perform other science observations while the starshade is moving for two weeks?

Yes. HabEx could either search for potentially habitable planets using the coronagraph, or use the other two instruments to do general astrophysics. It’s also possible to use multiple instruments simultaneously.

What is the difference on the spectrum could obtain Habex or Ariel mission?

Habex would do optical reflected light spectrum of planets (even potentially Earth-like planets) around Sunlike stars. ARIEL would do emission and transmission spectra of *transiting planets,* in the infrared, of planets mostly larger or hotter than the Earth. So, different planet populations and different kind of spectra.

For groundbased telescopes in the range of 1M-2M diameter (those ones with relatively small diameters), how would planet confirmation be held with photometry with transit graph plottings?

Small ground-based telescopes can collect photometry to detect brightness decreases due to eclipsing binaries and identify false positives. In the most favorable circumstances (i.e., large planet/star radius ratios, clear skies), they can also observe deep transits. If the transit time and depth matches expectations from TESS observations, then they increase the likelihood that the transit-like event is caused by a real transiting planet.

How well would portions or instruments of each of the different decadal survey proposals work together? In other words, after all this work is done, is it likely that at the end instead of choosing one final design, they will all be stitched together to make a single mission?

The LUVOIR and HabEx architectures and requirements are sufficiently similar that one could imagine designing a mission that was some sort of hybrid mission (e.g., with an intermemediate size aperture.) Origins and Lynx are suffiently different that this wouldn’t be possible.

Does Luvoir bring exo-moon detection into the picture?

No, you can't simultaneously separate the planet from the star and the moon from the planet. At best, you would get a time-averaged spectrum of the combined planet-moon system.

Check out the appendix to the LUVOIR Final Report for more info about detecting exomoons with LUVOIR. It’s in section A.19. https://asd.gsfc.nasa.gov/luvoir/reports/LUVOIR_FinalReportAppendices_2019-08-26.pdf

I really like the idea that the Earth transmission spectra has considerably changed during the lifetime of the Earth. It remind me a reflection of Michel Mayor in a book written in 2002. Michel highlighted the fact that when life just emerged on Earth, no bio-signatures would have been detected. On the opposite, some geological process can sometimes mimic biological signatures. Its conclusion was that it would therefore be really hard to claim spectroscopic detection of life outside the solar system. According to you, is this statement still valid 20 years later ?

The simultaneous presence of significant amounts of O2 and CH4 remains a pretty reliable biosignature.

Thank you for the nice introductions in those three competing concepts. However, on the first glimps in this short amount of time they sound pretty similar. Could you discuss in the panel the differences between the concepts? (It would "spice up" the discussion to hear the pro's for each instrument from the representatives of the competing missions)

Part of the answer to this is the overall target sample size that can be achieved, and the sensitivities of the different missions. Each has a substantial concept study report that was posted a few months ago. I encourage you to check those out for all the fine details!

Could James Kastings tell about which paper he was referring to (by Sara Seager)?

Sara Seager, Exoplanet Habitability, Science (2013)