| Question Re | |||||||||||||||||||
| Report Gener | 7/20/20 15:03 | ||||||||||||||||||
| Topic | Webinar ID | ||||||||||||||||||
| 2020 Sagan S | 940 3997 6754 | ||||||||||||||||||
| Question Deta | |||||||||||||||||||
| # | Question | Answer(s) | |||||||||||||||||
| Monday 1 | There are usually multiple planets around any star. How you screen their effects on the star wobbling period? | The planets have different periods, so you can find them by doing an analysis that can find these periods - periodograms or Fourier analysis. | Is there any table for that like resolution of spectrograph and diameter of telescope? | ||||||||||||||||
| Monday 3 | How can we extract multiple planets and masses from one rv-phase curve? Is there a fourier/frequency analysis component to these pipelines? | Usually we compute the « posterior distribution » with what is called a Monte-Carlo Markov chain algorithm (MCMC). We will talk about that on Wednesday (hands on sessions). This analysis is done on the raw RVs, not on the phase-folded data | |||||||||||||||||
| Monday 4 | How to calculate the precision for specific telescope and spectrograph? | There are some answers to this question in my presentation on "Fundamentals of Instrumentation" (later today, and pre-recorded). | |||||||||||||||||
| Monday 7 | How do we estimate the distance of the exoplanet from its parent star? | You can use Kepler’s third law that relates the orbital period and the semimajor axis (period^2 is proportional to semimajor axis^3) | |||||||||||||||||
| Monday 8 | In the curve planet radius vs seperation, why RV observations were done mostly between 0.1 to 10 AU seperation? | The amplitude of an RV signal depends on the planet’s period — shorter period planets are easier to detect. Also, the longer the planet’s period, the longer we have to observe it to confirm the signal. So, RV observations have a bias against long period planets. | Very few planets exist within 0.1 au. Beyond 5 au the orbital periods are > 10 years, so only very long surveys can find them. Direct imaging can find planets beyond 10au. | ||||||||||||||||
| Monday 9 | Sad to see Latham's 1989 discovery become a brown dwarf due to astrometry followup. Would the 1988 Gamma Ceph count as the first exoplanet since it was subsequently confirmed in 2003? | Yes, I usually say so. Also Pollux b in 1993. | Latham 1989: "The unseen companion of HD114762: a probable brown dwarf". Nature. 339 (6219): 38–40. | ||||||||||||||||
| Monday 10 | Will RV space missions help to detect more planets and host just like Transits ?? | Make sure to check out the pre-recorded talk available from Professor Peter Plavchan, and to check out the Discussion Panel on Wednesday! | |||||||||||||||||
| Monday 11 | Can we detect asteroids or comets from other Sun planet systems? | Not by precise radial velocities, but we can detect asteriods being destroyed from the material they deposit in white dwarfs, and we can see very large comets when they transit other stars. | |||||||||||||||||
| Monday 12 | To know the mass of a planet from a RV analysis, we need to know the mass of the host star but how do we measure this mass in the first place? | In most cases the answer is that we have a pretty good idea about stellar masses from 150 years of astrophysics. It's better on the main sequence than for more evolved stars, however. | There are many ways to determine a star's mass. I for one use spectra for RV's that I can stack, and determine astrophysical parameters (logg, [M/H] and teff) and then use those parameters in isochronic models to determine a star's mass (and radius). | ||||||||||||||||
| Monday 13 | How is inclination found? | Usually, one needs to do this astrometrically by measuring the motion of the star in the plane of the sky, which is challenging but has been done for some long-period gas giants with HST FGS, for instance. In the special case where the planet transits, you can measure the inclination directly (to be near 90 degrees) | |||||||||||||||||
| Monday 14 | As the centre of mass lies almost inside the star, Still can we observe the Doppler effect? | Yes, the effect is small but detectable. You can find some numbers on slide 2 of my pre-recorded presentation on "RV in context"/Quirrenbach on the YouTube channel. | |||||||||||||||||
| Monday 16 | What sort of RV precision are we expecting in the era of ELTs | The goal is sub 10 cm/s, possibly down to 1-2 cm/s one day! | |||||||||||||||||
| Monday 17 | Has Integral Field Spectrography been considered for exoplanetary detection? So as to probe many stars at the same time | To some extent, but the resolution of those spectrographs are far too low to reach the precision needed for RV exoplanet detection. | Yes, but not for EPRVs. IFS usually involves low resolution spectra, which are not amenable to precise Doppler work. IFS have been used to study the spectra and measure the orbits of directiy imaged giant planets . Other multiplexing techniques have been used, for instance the MARVELS survey which did simultaneous precise RVs on many stars. In general, though, EPRV work requires bright stars which are too far apart for practical multiplexing. | Interesting question - I think this would be a big challenge for an IFS to get the necessary spectral resoution. METIS on ELT will R=100,000 which is approaching what one needs for EPRV but the field of view would only really allow one star at a time in this case. It’s more designed for direct imaging and searching the close environs of a star for companion planets. I’ll talk a bit about this on Thursday. | |||||||||||||||
| Monday 18 | Is there an expected physical limitation to how precise RV measurements can be? Assume infinitely precise instrumentation - is there some physical limit we expect? | I’d recommend to also check out the Error Budgets talk by Sam Halverson. | |||||||||||||||||
| Monday 20 | How do you determine that a spectrograph with 0.8m/s RV precision misses 70% of planets in the green zone? Thank you! | We know the radius from the transit observations, and we can use models to estimate the corresponding mass (roughly). Then we know that we would need better precision to actually find the planets. | |||||||||||||||||
| Monday 21 | How do you correct for the motions of your telescope itself? I mean that the telescope movements and vibrations can introduce velocities of these orders right? | The telescope motions and vibrations are too small to matter. (You have to average their effect over the integration time, and vibrations have a zero mean.) But the motion of Earth around the Sun (and to a lesser extent the rotation of the Earth) is a very big effect that has to be corrected carefully. Fortunately this motion is known with extremely high precision. | |||||||||||||||||
| Monday 22 | Can Hi-Res multiplexed fiber instruments help with Telluric correction | Usually multiplexed instruments are not of sufficient spectral resolution. We need ultra-high resolution (R~100k or higher) to resolve individual telluric lines. | This has been done with the APOGEE spectrometer, which has 300 fibers over a 3 deg field. Typically about 20-30 across the field are on telluric standards. They use those standards to solve for a map of telluric absorption and apply it to the other science fibers. However, this only works to a degree. All sky imaging surveys (e.g. 2MASS) measured changes in the telluric absorption on spatial scales of ~10 arcsec, so you really need to have measurements on-sky very close to your star. | ||||||||||||||||
| Monday 24 | Davis et al 2017 PCA analysis (https://arxiv.org/abs/1708.00491) done in optical or also for IR? | Just optical, ~3925 to 6662A | |||||||||||||||||
| Monday 27 | is it possible that 2 planets are revolving around a star in such a way that one is moving along the other like radius of their orbit is different but thier angular position is exactly same? will it be possible to detect presence of both the planet? Or will it act like a dense or a planet with very high gravity as light will be bend more. | is it possible that 2 planets are revolving around a star in such a way that one is moving along the other like radius of their orbit is different but thier angular position is exactly same? will it be possible to detect presence of both the planet? Or will it act like a dense or a planet with very high gravity as light will be bend more. | I think you’re asking about co-orbital planets that share a common orbit. In general, planets in such orbits will interact with each other strongly. This usually means that if both planets are massive enough to be detetable, then they would not be stable and the system would not exist. There are specal configurations where they could be stable, but even then the interactions would be strong enough that we would probably be able to tell that the planets were not on strictly periodic orbits. But yes, in general the result would look roughly like a more or less massive planet (depending on their angular separation). | It is not possible for both planets to revolve exactly at the same angular position. However, in principle you can have two planets at the same periods. This is called the co-orbital motion. One of the possible co-orbital configurations is called tadpole, where the angle beetween the two planets is almost constant | |||||||||||||||
| Monday 18 | You emphasize the importance of the LFC for the calibration of EPRV instruments. What is the maximum frequency range that can be simultanously covered by a LFC? | 500 - 760 nm for EXPRES. Redder combs possible but not bluer combs (high energy melts the photonic crystal fiber) | |||||||||||||||||
| Monday 29 | What are typical exposure times for EXPRES to reach the expected errors? Is it for a certain magnitude of star? | depends on brightness - most stars are 10 minute exposures. but, we try to average over p-modes. | |||||||||||||||||
| Monday 30 | The radial velocity gives us the estimate of the minimum mass of the planet. How do we know whether it is a planet or a smaller star or a brown dwarf? | Andreas’ talk discusses mass vs. minimum mass and the typical mulitiplicative factor (I think it was 1.27?). | Yes, there are some comments on this question in my pre-recorded talk on "RVs in context" on the YouTube channel. The bottom line: for individual planets this may be difficult or even impossible, but for statistical analyses this can be taken into account rather easily. | ||||||||||||||||
| Monday 32 | What is a GP? | A “gaussian process”, which is a statistical noise model that accounts for correlated noise. We will hear more about this in later talks. | |||||||||||||||||
| Monday 33 | With your 30 cm/s precision, how does the mass uncertainty improve on Proxima b (or other generic planets)? | The ESPRESSO paper shows this (ESPRESSO is on a set of 8-m telescopes) - we would not get such good precision with a faint star. | |||||||||||||||||
| Monday 34 | How important is stellar magnitude in this whole process. I mean increasing spectral resolution would decrease the SNR right? So how do you decide a right balance between spectral resolution, SNR and exposure time required while designing the instrument? | There is some information on the spectral resolution that you would like to have in my presentation on "Fundamentals of Instrumentation" (next talk an on the YouTube channel). You normally try to expose long enough to get an SNR of about 100. | EXPRES was an experiment to see if the design choices would deliver high precision - we look at bright stars (V < 7) for this high precision. Can look at fainter stars, but lower SNR means lower RV precision. | ||||||||||||||||
| Monday 35 | Can we use the method to detect black holes at the galaxies centre? | HIRES is better for this (10-m telescope for the faint stars Andrea Ghez/UCLA is looking at) | |||||||||||||||||
| Monday 36 | Would it be possible to detect exomoons with this high precision? | Its difficult becuase they are low mass, but there are indirect methods. For example, we can infer the precence an exomoon in variations of the transit times (assuming the host planet is transiting). Then if we measure the mass of the host planet with RVs we know the mass of the moon. | |||||||||||||||||
| Monday 37 | How does magnetic field contribute to noise? | To know about the contribution of magnetic fields, check out the talks/panel discussion on Wednesday, as well as their pre-recored counterparts that are available now. | |||||||||||||||||
| Monday 38 | Are there any more candidates for benchmark EPRV stars to use as standards? | Great question - we hope everyone will think about this. Our best cases were highly eccentric known planets. There probably are no stars w/o planets! | The list, of 8 stars, for the NEID spectrograph can be found here: https://neid.psu.edu/observers-and-proposers-faq/ | ||||||||||||||||
| Monday 39 | If we use whatever calibration technique (Th-Ar lamps, hollow cathods, combs, etc..) to probe the stellar wavelength shifts we introduce some error due to the uncertanties correct? For example, the lamp degrades over time, there is physical uncertainty in dt x de >= h, so why not do RVs in pixel domain? We do not need absolute RVs for detection correct? | Thanks for your question. You are right that absolute RVs are not, in general, needed for planet detection. But, the detector pixels themselves can physically move over time due to thermal fluctuations of the instrument, etc. So, in pixel space, the stellar Doppler shift is degenerate with instrumental shifts. | |||||||||||||||||
| Monday 41 | Which part of spectrum is preferred for analysis and why? | Typically the visible has been the most fruitful, from ~450 - 700 nm. For M-dwarfs, we are pushing this out to as far as 1.7 microns. | It depends on the stellar type. You want high SNR, which favors the spectral range in which the star is brightest. You also want many lines in the part of the spectrum you observe. So for Solar-type stars the blue part of the visible spectrum is preferrred, for M dwarfs the red part is very good. | ||||||||||||||||
| Monday 42 | When the spectrum is sliced, should each slice be considered has independent or could the slices be merged later without any danger ? | Both have been tried, and both have tradeoffs. If you treat them independently, you can handle differences in the RV information across the fiber face (which happens for complicated reasons related to how fibers propagate light). But doing so requires more detector pixels, and detector pixels are the most expensive thing in your instrument. It also places difficult constraints on the cross-disperser. Merging them together solves both of these, but you have to be very certain that the doppler shift is constant across your fiber. | |||||||||||||||||
| Monday 44 | Won’t the image slicers introduce nonunifomtiy at the spectrograph input due to thermal, compared to a direct fiber output fed to spectrograph? | This is why we dump the 2 half moons into a rectangular fiber - gives a smooth scrambled image into the instrument | |||||||||||||||||
| Monday 47 | What does the quoted precision associated to RV instruments correspond to (we say m/s or cm/s)? I assume it depends on exposure time? stellar magnitude? stellar spin? are exposures combined? etc. | Sam Halverson’s talk goes into this! Check out his prerecorded talk. | As I showed in my presentation, the precision depends on the SNR, and the SNR is proportional to the square root of the number of the photons. That is where the stellar brightness and exposure time come in. The stellar rotation enters through the "Q factor", because more rapid rotation smears out the information in the spectrum and thuse reduces Q. | Thanks! Exactly, so my question was what are the assumptions when people quote a typical number for the precision? | |||||||||||||||
| Monday 49 | How do we calculate the quality factor precisely? It relates to the amount of lines etc.. but is there an annalytical equation or a way to compute it? | Bouchy, et
al, Fundamental photon noise limit to radial velocity measurements
(A&A Volume 374, Number 2, August I 2001) https://www.aanda.org/articles/aa/abs/2001/29/aa1316/aa1316.html |
There is no simple formula to do that, because it depends on the spectrum, so it is different from star to star. It also depends on the resolution of the spectrograph. In practice you write a small computer program to calculate Q from these inputs. The math is shown in a bit more detail in the paper cited in my talk. | ||||||||||||||||
| Monday 50 | Quality factor of Bouchy from optical wavelengths, how does it translates to IR ? | The prinicple is exactly the same. One just has to use a different part of the spectrum to compute the Q factor. | |||||||||||||||||
| Monday 51 | I mean to say that in CARMENES, starimage is divided into two parts, in that case doesn’t it affact of stellar spectra? | It is possible to design the optics and cameras such that there are no missing parts of the spectrum. | |||||||||||||||||
| Monday 53 | How do you account for losses when optical fibres are used? | We can measure the losses, but they are generally slowly varying in wavelength and so do not have a large impact on RV measurement except that they rob you of some signal. This is annoying, but their benefits outweigh these losses. | |||||||||||||||||
| Monday 54 | Can you do anything other than slicing to make the spectrograph illumination slit smaller? | You can use adaptive optics. You can also slice the *pupil*, for instance by having many small telescopes each contribute a spot. This is the principle behind, for instance, the PRV instrument MINERVA. | |||||||||||||||||
| Monday 55 | Would a telescope with smaller focal length help reduce the spectrograph size? And in a diffraction limited case would the spectrograph dependence on telescope aperture disappear? | 1) No, because you have to presever étendu (A x Omega), and both A and Omega are fixed by the size of the telescope and the angular size of the star. 2) Yes! Diffration limited intsruments are a promising way forward but *extremely* difficult to build. This is the principle behind Justin Crepp’s iLocator project at Notre Dame (https://ilocater.nd.edu/). | Yep - we're building iLocater right now. Should have the spectrograph in the lab right about the end of 2020! We use small single-mode optical fibers (about 6 microns diameter rather than 50/100 micron used in seeing-limited instruments) to illuminate the instrument. The challenge is getting enough light in to the fiber initially to minimize photon noise and that's linked to adaptive optics performance. But the nice thing is we get a very high resolution in a compact size which helps with many of the systematic effects we've heard about. And we don't have any modal noise! | ||||||||||||||||
| Monday 57 | Can you briefly explain the value of signal scrambling in fiber optic cables again? | Coherent light through a fiber produces speckles. as the fiber moves, the speckles shift around, changing the input illumination of the spectrograph. scrambling (inverting near / far fields, agitating fibers) swishes the speckles around with high frequency, averaging over the speckles. | |||||||||||||||||
| Monday 58 | What is scrambling and why is it important? | there are interference modes in fibers (worse for lasers) and when the fiber moves, the speckles shift around. So, scramble to mix the speckles. | "Scrambling" means that the output is much more uniform than the input. That is important because non-uniform fiber output is a source of radial velocity errors. | ||||||||||||||||
| Monday 60 | Diffraction order m and wavelength realtion depend on width between the mirrors in fabry perot. Fabry perot are not really stable, sometimes even in controleed conditions. How do we deal with that? | The FPs are in vacuum with thermal control. Slow drifts calibrated out. | |||||||||||||||||
| Monday 61 | How can using iodine cell be better in some ways? and in what ways? | I abandoned I2 cells - I believe that the forest of I2 lines mask our ability to extract photospheric velocities (line profile variations). I2 is great for an unstable spectrograph and you can expect a precision floor of about 2 m/s. | Iodine gives you a way to simultaneously calibrate the wavelength scale and point spread function of the spectrograph for light traveling exactly the same path as the starlight. The downsides are that is contaminates the spectrum, eats into your preceous star light, and makes the reduction pipeline more complex. However, it is much easier/cheaper to design a spectrograph using iodine to get to that 2 m/s precision floor compared to a stablized spectrograph that is aiming for that same precision. | ||||||||||||||||
| Monday 62 | If you use calibration through your science fiber, how do you account for persistence or is that not a problem? | This has not been a problem with our silicon STA CCD | |||||||||||||||||
| Monday 63 | A EPRV instrument mounted on Namyth platform , viable? | Like HIRES. There are just engineering issues (vibration, temperature harder to control). | |||||||||||||||||
| Monday 64 | For Prof. Fisher, in your slide re. K>3m/s detectable with ~2m/s RV precision + many observations, does the planet ratio imply that ~53% of jupiter sized planets will also be missed? Since the ratios in both that slide and the previous one are scaled around the neptune sized planets? | I think this is “small” numbers for hot Jupiters. RVs do a better job at finding jupiters in wider separations. | Why do RVs do a better job at finding jupiters further out? I would think that Hot Jupiters would lead to larger semi amplitudes since P is smaller. | ||||||||||||||||
| Monday 65 | Two questions about rotation and PRVs: 1) How does vsini relate to rv precision? I have seen plots showing how RV precision degrades with increasing vsini - how are these curves calculated? Can this be determined analytically? 2) what possible future pathways to being able to improve RV precision on non-slow rotating stars (assuming you can already deal with the activity that more rotation brings)? | Since this comes up in Sam’s talk, I think I will save this question for the next panel. | |||||||||||||||||
| Monday 66 | Can you please re-explain the first scaling of transit depth 100-10-1? | The depth of the transit is given by the fraction of the star covered by the planet, which is the area of the planet disk divided by the area of the stellar disk. So it is the square of the radius ratio. | |||||||||||||||||
| Monday 71 | For Andreas...for your two fibers...what is done to ensure the manufacture of the fibers is producing consistent fibers, so the physical properties of the fibers do not contribute errors? | It is usually ok to buy one long fiber (you can get km lengths) and cut it into shorter pieces. The pieces are very similar to each other. | |||||||||||||||||
| Monday 74 | What do you expect to be the greatest non-instumental “noise-maker”? Is there a possibility that all these incredible increases in instrumental precision would be made unreachable by external sources of error? | You can’t know until you build an instrument that has precision less than those other terms. This allows you to “see” the other error sources (photospheric velocities) | |||||||||||||||||
| Monday 75 | How important would you say outlier identification in the 1D spectra is after the various standard corrections have been applied (i.e. to account for calibration unc., tellurics, etc.)? Would you say problems with outliers or single data point errors are larger contributors to overall line RV noise? | We don’t have outlier rejection, but take great care with our analysis and have a very stable, well-calibrated system. | |||||||||||||||||
| Monday 76 | Can photonic laterns helpful at the input end? | We thought a lot about this. But, it seemed that there are other challenges that outweigh the benefits (from what we could tell) | |||||||||||||||||
| Monday 77 | Sometimes when observing one has to choose between either a simultaneous sky reference vs. e.g. Fabry-Perot reference for the second fiber. Is there any benefit in being able to do both at the same time? Why are instruments limited to two fibers? | There is a tradeoff between how much cross-dispersion you can get (which is necessary to separate the orders so they do not overlap) and how big your detector is (so you can capture all of the orders you want). If you add in a 3rd fiber, like we did with NEID and HPF, then you need more cross dispersion, so you loose some orders. | |||||||||||||||||
| Monday 78 | If I am not mistaken HARPS was mounted on a single telescope and ESPRESSO has 4 telescopes combined. Why do we go for larger telescopes if we could have smaller ones combined creating a larger "effective" primary mirror? how do we gain with one larger telescope compared to 'summing' smaller ones? | This might be a practical consideration. The VLT (4 telescopes) existed, so the challenge was how to combine the fibers from all 4 telescopes. | Yes, this is a very good question. At the moment we are "opportunistic", in that we build spectrographs for existing telescopes. We take what we can get. If we can find the money for a dedicated RV facility, building an array of smaller telescopes instead of a single bigger one would be a very serious alternative. | ||||||||||||||||
| Monday 80 | Considering the upcoming powerful imaging mission you mentioned and EPRV as precursor, how do you think EPRV surveys should balance between coverage (more stars) and depth (less stars more observations) ? | Great question! I think this is addressed by the Thursday talks, although Chad should feel free to answer it. :) Jenn Burt and I led part of the “NASA-NSF EPRV Initiative” to investigate what signals we could detect on stars of different magnitudes assuming different instrument/telescope performance. The final report from the effort should be coming out in the next few months. | It turns out that most of the space mission proposals are severely target limited. In their nominal 2-3 year mission life that is devoted to planet characterization, they can only characterize 50 to a few hundred (at very most) planets. That number is actually well matched to what EPRV efforts can also reasonably survey. If you require a large number of RV observations per year on each of your targets stars (say 100 visits), then you pretty quickly become limited by the amount of telescope time available. That is difficult to expand except by adding more facilities into the survey, which is something we looked at closely in the working group study. | ||||||||||||||||
| Monday 83 | To what extent or to what accuracy can we correct for contribution of our (Earth) velocity to the radial velocities? | 1 or 2 cm/s | For the details, see the paper by Jason Wright and Jason Eastman from 2014: Barycentric Corrections at 1 cm/s for precise Doppler velocities (https://arxiv.org/abs/1409.4774). | ||||||||||||||||
| Monday 94 | What is the required precision needed on the timestamp of a RV measurement - exposure | Wright & Eastman 2014 is a great reference | |||||||||||||||||
| Monday 84 | Mosaic echelle will it be a problem at all? | Mosaic echelles (as the one I showed in my talk) are difficult to manufacture, and therefore expensive. But if you get a good one, there are no technical problems with it, other that you lose some light in the gap between the mosaic sections. | However, you can also get a "bad" one where the phasing between the two mosaics is not done very well. Then your resulting pupil has a lot of problems. | ||||||||||||||||
| Monday 88 | What are the limitations of EPRV technique? | This is a SUPER important question but also kind of a moving target. I would say many of the speakers will touch on this theme, but Sam’s prerecorded talk is a good place to get an idea of many of the instrumentation-related limitations. But as several speakers have said, the stars are a “tall tent pole”, holding back progress. Some of the Wednesday talks are focused on stars, so check those out, too. | |||||||||||||||||
| Monday 92 | what is the proper way to assess the scatter of the RV time-series? standard deviation, mean absolute deviation, standard error (of the mean), RMS? for example when observing RV standards. Usually they differ significantly from error from single measurement. Essentially, I want to know if from those values I am extracting the most out of the doppler information contained in the spectra. | I think this is a great question and will be covered in depth on Tuesday. | |||||||||||||||||
| Monday 95 | If the planet orbiting the star is orbiting in a plane perpendicular to us, then how do we measure it's doppler shift? | We can only measure the radial component of velocity, so we would not see any RV from perpendicular orbits. | |||||||||||||||||
| Monday 96 | How would RV precision be affected in optimal vs non-optimal extraction of data? | If done correctly, optimal extraction (see Petersburg et al. 2020 https://iopscience.iop.org/article/10.3847/1538-3881/ab7e31/meta) gives higher fidelity spectra and this translates to better precision. The RV impact is at least 0.5 m/s. | |||||||||||||||||
| Monday 99 | I’ve never heard of etalon, what is it? | An etalon in its simplest form is a pair of plane-parallel plates between which light can bounce many times. You can also search for "Fabry Perot" (see: https://bit.ly/3jkmWb8). | |||||||||||||||||
| Monday 101 | Where does DACE grab all the data from for each target? From the exoplanet archive? Or there’s someone that does the search for all possible data? Or simply just data that the european team has access to? Is it possible to load your own data? | Yes, you can certainly load your own data into DACE. If you do that I reccomend that you make an account so that you can save your work (but thats not strictly required). | It should be .rdb files and when you click on the import button, you will have the mandatory and optional columns that your rdb files need | It looks like space-delimited text, with certain required columns. | |||||||||||||||
| Monday 105 | Can public work on the data? Are there data available for public to work on? | All of the data that is available publicly in the platform (e.g., not private data that you upload) can be accessed by the public and you! | |||||||||||||||||
| Monday 102 | Sorry, but I thought that RV offset were calculated joinly with Keplerian model, in order that several set of data from different observatories reproduce the signal observed. I guess that your given offsets are some kind of initial solutions to start the minimization. Am I right? | Correct. The offset can also be fit as a free parameter. | |||||||||||||||||
| Monday 106 | What does this correlation represent and why it is significant? | We're going to get into that in the talks Tuesday. | |||||||||||||||||
| Monday 108 | How is the periodogram calculated? Lomb Scargle? or some modified/improved version of it to fit keplerian orbits? | Nathan pointed out that this is included in more detail in Part 4 of the tutorial on the Sagan webpage | |||||||||||||||||
| Monday 109 | When you clicked on keplarian, Your plot only took points from kepler data set? | That shows the “Keplerian” orbital model. All datasets are included. Kepler data is not actually radial velocities so it will not usually be shown in this part of DACE. | |||||||||||||||||
| Monday 110 | Why is the peak at 1d due to instrumental problems be exactly there, why not a peak at longer period (e.g. 10d) or even shorted (eg 0.1d) ? | There are many things that happen at exactly 1 day, including daily thermal cycles, daily maintenance on the telescope, but it’s primarily the windowing function (when exactly you observe). There is a degeneracy because of aliasing between a long-period signal and one that occurs dirunally. | |||||||||||||||||
| Monday 111 | Which equation is used in define Keplerian? | See the slides from my talk this morning. This will also be covered tomorrow morning. |