Confirmation of Earth Sized Candidates For The Kepler Project

Greetings To All,

As I understand it, before candidate exoplanets can be officially announced, they must first be confirmed through RV observations. Kepler is designed to detect Earth sized exoplanets through transit observations at distances of 1 AU for sun-like stars (G class). To my knowledge, the best RV equipment today does not have the sensitivity to detect such small mass bodies so far from it's host star. Therefore, how will these candidate exoplanets be verified?

http://obswww.unige.ch/Instruments/harps_north/science/White_Paper.txt
3.1 Main project goals:
Our main project has three broad scientific goals that can be achieved
by delivering planetary masses of certain precision:
(a) confirming an Earth-twin planet in the habitable zone of a G5V star
or later - 30% in mass;
(b) characterizing Earth-like planets of 2-5 M_Earth ("Super-Earths") in
different orbits: distinguishing between water-rich and dry planets - 10% in
mass;
(c) characterizing the transition between Super-Earths and Ice Giants (hot
Neptunes, for example) near 10 M_Earth - 5% in mass, or better.

For planet classes in goals (b) & (c), Kepler will determine planetary radii
of 5% or better, for stars in the "sweet spot" around V=12 mag. For the
Earth-like planets in goal (a), Kepler could provide 15% in radius precision.
All these estimates improve considerably for M-stars, but the probability
of discovery is assumed low. These values: 5% in radius, and (a),(b),& (c) in
mass, allow achieving our goals, as shown by the theoretical models
(Valencia,Sasselov,O'Connell 2006,2007; Fortney et al.2007). Note that
achieving goal (b) is essential, in a boot-strapping sense, to achieving
confidence in completing goal (a), and hence - of Kepler's legacy as whole.
HARPS-NEF in the Doppler domain is a true match to Kepler in the photometric
domain, thus providing vastly improved planetary mean densities for a
meaningful comparison to interior models.

3.2 Main project observing time requirements:
HARPS-NEF on a 4-m class telescope in the Northern hemisphere (e.g. WHT) can
achieve the above tolerances in mass determination with certain number of Doppler
measurements. That number of observations can be evaluated under the following
set of conditions. First, the Kepler light curve constrains the period and phase
of the planet's orbit. Second, we select targets of V=12 mag or brighter, for
which HARPS-NEF should achieve 1.0 m/s in 1 hour. This has been demonstrated with
the HARPS in Chile. We illustrate our estimate of required number of observations
to achieve 10% in planet mass at a 0.10 AU orbit in the table below:

Star/Planet 5 M_Earth 10 M_Earth
---------------------------------------------------
F0 (1.60 M_s) 158 40
G0 (1.05 M_s) 104 26
K0 (0.79 M_s) 78 20
M0 (0.51 M_s) 50 13
---------------------------------------------------

With such estimates in hand, we have determined the number of effective
clear observing nights (HARPS-NEF on WHT) required to accomplish the three goals
of our Main Project:
(a) for 2 planets (over 3 years) - 160 h = 16 nights per year;
(b) for 20 planets in different orbits - 250 h = 25 nights per year;
(c) for 20 planets in different orbits - 210 h = 20 nights per year.

3.3 Main project scheduling:
The 60 clear nights per year estimated above fall within the season of about
six months - April through October, when the Kepler field is accessible from
WHT. They do not have to be consecutive or time critical; queue scheduling would
work fine. We expect planets that orbit stars of different brightness and with
different periods, we can fill efficiently observing nights regardless of seeing
conditions. We anticipate having Kepler targets for goals (b) & (c) during year 1
and can begin work on them immediately - in summer 2009. For goal (a), we may have
to wait until year 3, i.e. begin in 2011, and continue until 2014.

The observing time estimates and scheduling make use of the expectation that
Kepler will provide a very large number of good targets in each of the categories
of planets. This will allow our team to select the best ones for follow-up, and
avoid parent stars that show high activity or with other problems. For example,
we estimate that about 50% of the initial targets will show intrinsic variations
which are too high for efficient HARPS-NEF follow-up.

If you were to observe a G-type star with an Earth-mass planet in a 1 AU orbit, you would have a very hard time finding that planet with today's technology. You don't even know the planet is there, what its orbital period is, or anything.

But if you detect transits, then you already know the orbital period and you know where to expect an RV signal. It no longer becomes blind searching. This allows you to place much stronger constraints on the mass of the planet than if you didn't even know a planet existed.

This is one of the interesting questions, especially when confronted with "noisy" stars. It took quite a bit of effort to extract the radial velocities of COROT-7's planets thanks to the activity of the star, and there you have the advantage that the orbital periods are very short.

My sincere thanks to all respondents.

The paper cited was written in Feb. 2007. Apparently HARPS-NEF, as a new instrument, had not yet been commissioned for science operations. Does anyone know for sure whether it is currently being used to verify Kepler candidates or is the HARPS instrument in Chile being used?

I believe I had read a while back that the Keck observatory would be used as well. Does anyone know whether it has similar capabilities and confirmation times to HARPS-NEF?

In any event, it looks like confirmation times will be extensive. It will be interesting to see how many exoplanets are announced in January as part of the first wave. I'm guessing less than 50 with the vast majority being hot Jupiters (no surprise there). I wish the NASA Kepler website provided some general statistics on the number of candidates found each month so at least we had some idea what the potential catch might be.

AlSchmitt wrote:

The paper cited was written in Feb. 2007. Apparently HARPS-NEF, as a new instrument, had not yet been commissioned for science operations. Does anyone know for sure whether it is currently being used to verify Kepler candidates or is the HARPS instrument in Chile being used?

I believe I had read a while back that the Keck observatory would be used as well. Does anyone know whether it has similar capabilities and confirmation times to HARPS-NEF?

Not earlier than February 2010
See the schedule for the WHT telescope
http://catserver.ing.iac.es/schedules/schedules_view.php

About telescope Keck see my post
https://solar-flux.forumotion.com/extrasolar-news-and-discoveries-f2/kepler-results-t282-75.htm#2699

HARPS in ESO the field Kepler not available because of curvature of the globe.

Yeah, he was asking about the HARPS-N instrument, which will be in the northern hemisphere. I've read somewhere that a lot of its time will be used for Kepler candidates.

Sirius_Alpha wrote:

Yeah, he was asking about the HARPS-N instrument, which will be in the northern hemisphere. I've read somewhere that a lot of its time will be used for Kepler candidates.

http://obswww.unige.ch/Instruments/harps_north/science/White_Paper.txt
3.3 Main project scheduling:
The 60 clear nights per year estimated above fall within the season of about
six months - April through October, when the Kepler field is accessible from
WHT. They do not have to be consecutive or time critical; queue scheduling would
work fine. We expect planets that orbit stars of different brightness and with
different periods, we can fill efficiently observing nights regardless of seeing
conditions. We anticipate having Kepler targets for goals (b) & (c) during year 1
and can begin work on them immediately - in summer 2009. For goal (a), we may have
to wait until year 3, i.e. begin in 2011, and continue until 2014.

Appeared timetable of observant time for first semester 2010


http://nexsci.caltech.edu/missions/KeckSolicitation/proposal-current.shtml

Quote :

2010A</A>

W. Borucki

NASA Ames

Key

Follow-up Observations of Kepler Targets

x

HIRESr

http://www2.keck.hawaii.edu/observing/schedule/index.php
Borucki users HIRES 27.04.2010 (1/2 nights), 01.05.2010 (1/2 nights), 04.05.2010(1/2 nights), 25-27.05.2010 (1/2 nights), 28-30.06.2010, 01.07.2010, 24-26.07.2010.

In general approximately 10 observant nights

http://catserver.ing.iac.es/schedules/schedules_view.php
Not HARPS-NEF in semestr 2010A (February - July 2010) in WHT

Interesting HARPS-NEF begin work in 2010? Window for observations of the field Keplera is closed at October 2010 and is opened afterwards at April 2011 only.

Fear that delays with HARPS-NEF vastly reduce amount of the small planets Keplera in 2010.

Borislav wrote:

Interesting HARPS-NEF begin work in 2010? Window for observations of the field Keplera is closed at October 2010 and is opened afterwards at April 2011 only.

Fear that delays with HARPS-NEF vastly reduce amount of the small planets Keplera in 2010.

http://sci.esa.int/science-e/www/object/index.cfm?fobjectid=46020
on 54 pages

Quote :

While radial velocity follow-up of giant planets detected by Kepler can be performed with available facilities, such as the HIRES spectrograph on the Keck telescope, for small planet candidates, Kepler will have to rely on HARPS-North, which will not be on the sky before 2011, and even then, will provide radial velocity follow-up capabilities for small planets only for the few brightest stars in the Kepler sample.

This isn't really surprising, in my opinion. The RV confirmation stage is always the bottleneck of any transit based exoplanet survey. Recall the SWEEPS survey that provided several planet candidates, of which only two were confirmed, because the other stars were too dim for RV follow-up.

Kepler will find lots of planet candidates around stars that will be too dim to be accessible to today's telescopes. We may have to wait for the next generation of telescopes to really confirm all of the Kepler planets.

Quote :

We may have to wait for the next generation of telescopes to really confirm all of the Kepler planets.

That makes me wonder why Kepler wasn't originally designed to survey the nearest Earth mass exoplanets in a wide field instead of exoplanets in a narrow and deep field. More like the canceled TESS project or the ongoing MEarth project. Having nearby candidates would make the verification process much easier using existing equipment.

Furthermore, once the verification process was complete, such nearby worlds could be characterized in follow-up studies. This would have provided the greatest bang for the buck in overall scientific gain.

Even though Kepler will return some great science, perhaps the project should have tackled the nearest exoplanets first.

They wanted a spacecraft that would help guide the SIM mission. For that, you need a census of planets around a variety of stars. The transit method is far more favorable of distant stars than near ones anyhow.

SIM (or an astrometry mission like it) will tackle the nearest exoplanets.

There are different inherent biases in radial velocity and transit surveys. If a system is significantly inclined from edge-on, the radial velocity will report lower masses for the planets, while a transit survey won't spot it at all. In addition, the transit method only detects the planet at one point in its orbit, at other times you have no idea the planet is there*, while for radial velocity measurements you can detect variations in RV right through the orbit.

Hence good strategies for transit surveys differ from those for radial velocity surveys. In all likelihood we'll get a few solid detections plus some statistical properties of the terrestrial planet distribution, which will be useful for follow-up missions.

* Ok, you might also have sensitivity to secondary transits, and for close-in planets you may be able to get phase curves, but for the most part it is the case that the planet cannot be detected for the majority of its orbit.

Sirius_Alpha,

Quote :

The transit method is far more favorable of distant stars than near ones anyhow.

Aren't exoplanets around nearby stars easier to detect than exoplanets around distant stars, all other factors considered equal? Or did I misunderstand your comment?

While it would be easier to detect a planet transiting a nearby star than it would be to detect the same planet transiting a distant star, there are simply much more distant stars than there are near ones.

The fraction of planets that transit will be far smaller than the fraction of planets that do not. So not too many of the nearby stars will host transiting planets.

The easy way to find transiting planets is to stare at a field of stars and watch their brightness, looking for transit events. Since the vast majority of stars in such a field will be distant, most of the transit events detected will for stars that are rather far away.

Sirius_Alpha,

Thank you for the explanation. I now understand your original comment.

I see more clearly than ever the brilliance of the MEarth project. By carefully selecting the most favorable nearby stars for study and by conducting the survey in an highly efficient (automated) manner, it mamimizes the chance for discovery of Earth-like worlds. Rather than take a shotgun approach, it takes a targeted analytic approach. And it allows for easy verification and follow-up characterization. It will not detect as many exoplanets as Kepler, but the ones it finds should be of extremely high quality.

AlSchmitt wrote:

That makes me wonder why Kepler wasn't originally designed to survey the nearest Earth mass exoplanets in a wide field instead of exoplanets in a narrow and deep field. More like the canceled TESS project or the ongoing MEarth project. Having nearby candidates would make the verification process much easier using existing equipment.

TESS is very different animal.

First its aim was to scan the Entire Sky for the Transiting planets with periods of 60 days or less around the closest 1 Million or so stars with V12 or better brightness. Only for Neptune size, maybe the odd Super -terrestrial type. More like a COROT mission dedicated to planets type.

Kepler aims for complete statistics for 4+ year period orbits, for typically V12 or less bright stars (Bright stars would swamp the long 6.5 hour iterations) in a particular region. Also down to Mars size, though with ~ 1 year orbits. I can't stress enough how hard it is to get down to this size. Kepler mission is in a class by it self, nothing planned or in orbit comes remotely close.

We need both.

TESS planets would have seen a lot more interest, for the dream of space travel. Kepler field too far away.

As a side note it is becoming more doubtful if kepler targets would be useful for providing JWST targets. TESS would have been the better bet.

There is ASTrO (All Sky Transit Observer).
http://nexsci.caltech.edu/workshop/2009/beichman_transits.pdf

I believe ASTRO was down selected when TESS was chosen.

Is it still being funded?

TheoA wrote

First its aim was to scan the Entire Sky for the Transiting planets with periods of 60 days or less around the closest 1 Million or so stars with V12 or better brightness. Only for Neptune size, maybe the odd Super -terrestrial type. More like a COROT mission dedicated to planets type.

far as i know TESS would detect earth-size planets around our neighborhood stars

unfortunality NASA don't know do bussines, and mission wasn't chosen...

Daniel wrote:

unfortunality NASA don't know do bussines, and mission wasn't chosen...

I find it hard to fault them for this too much. Extrasolar planets are fascinating, but they're but a small part of the astronomy field (even if they are exciting!). There's a lot of other interesting science to be done that doesn't directly involve extrasolar planets.

yeah you're right Sirius_Alpha.

but i think that planet around Red dwarf in habitable zone of our neighborhood stars i should be high priority by NASA,

as i see on the last 30,40 years, since that NASA send the last mans on moon,the NASA start with a new plan and change over again,again and again...like walk around the cicle,as we see now on the new vision of space exploration,to send the man back to moon...

i'm not against human exploration of space,but if start the goal,must reach the goal...and with this NASA burget, NASA don't go to reach nothing!

for the last years i lose my faith in how USA govenament deal with NASA, speciality on science projects...

that is the fact of USA space program unfortunately...

i put my faith on the european space agency ESA.

it's hard to see ESA science projects be cancel,like NASA does.

for example:

JIMO jupiter ice moons orbiter cancel,

TPF on hold,

return the man to the moon,constelation project,tight burget(this way in my opinion NASA never go to build ARES V the gargo rocket,only ares I,to reach the low orbit,like the shuttle does),this way probabily China go There first,and build a moon base,beusace don't change they space plan easily.

SIM planet quest, now it's Lite SIM less powerful,and delay,no lauch date,

etc,etc....

oh yes delay the kepler for 2 year nearly cancel

Dawn nearly cancel

that is what i remenber now...