Kepler Mission - Hunting for Exoplanets
Kepler is part of NASA's Discovery Program designed to survey a portion of our region of the Milky Way to discover Earth-size exoplanets in or near habitable zones and estimate how many of the billions of stars in the Milky Way have such planets. The primary goal is to determine the frequency of Earth-size and larger planets in the HZ (Habitable Zone) of solar-like stars. The mission will monitor more than 100,000 stars for patterns of transits with a differential photometric precision of 20 ppm at V = 12 for a 6.5 hour transit. It will also provide asteroseismic results on several thousand dwarf stars. It is specifically designed to continuously observe a single FOV (Field of View) of > 100 deg 2 for 3.5 or more years. 1)
Finding extrasolar planets is extremely challenging and was not accomplished until 1995 when Mayor & Queloz, (1995) detected the first jovian-mass planet around normal stars. However, by making the observations from a spaceborne platform and using the transit method proposed by Borucki and Summers (1984), Earth-size planets, including those in the HZ, should be detected in substantial numbers.
The scientific objective of the Kepler Mission is to explore the structure and diversity of planetary systems. This is achieved by surveying a large sample of stars to:
• Determine the percentage of terrestrial and larger planets that are in or near the habitable zone of a wide variety of stars
• Determine the distribution of sizes and shapes of the orbits of these planets
• Estimate how many planets there are in multiple-star systems
• Determine the variety of orbit sizes and planet reflectivities, sizes, masses and densities of short-period giant planets
• Identify additional members of each discovered planetary system using other techniques
• Determine the properties of those stars that harbor planetary systems.
The Kepler Mission also supports the objectives of future NASA Origins theme missions SIM (Space Interferometry Mission) and TPF (Terrestrial Planet Finder),
• By identifying the common stellar characteristics of host stars for future planet searches
• By defining the volume of space needed for the search and
• By allowing SIM to target systems already known to have terrestrial planets.
Background of the Kepler Mission: 2)
The Kepler Mission developed over several decades as a way of answering the question: How frequent are other Earths in our galaxy? In particular, what is the frequency of Earth-size planets in the HZ (Habitable Zone) of solar-like stars? In the last half of the twentieth century, the astrometric and interferometric approaches to finding exoplanets were the favored methods. The surprising discovery by Wolszczan (1994) 3) based on timing of radio pulses from pulsars showed that a wider range of approaches should be considered. A paper by Rosenblatt (1971) 4) provided a quantitative discussion of another alternative method; searching for patterns of transits to get size and orbital period. To be successful, all three approaches depend upon adapting new technology; the underlying principles are well understood. A paper by Borucki and Summers (Ref. 3) corrected the detection probability in the paper and pointed out that ground-based observations of at least 13,000 stars simultaneously should be sufficient to detect Jovian-size planets, but that the detection of Earth-size planets would require space-based observations. Limitations to the detectability of planets by stellar variations was recognized (Borucki et al, 1985) 5) and discussed more fully by Jenkins (2002). 6)
To examine the technology needed to accomplish transit detection of exoplanets, NASA/ARC ( Ames Research Center) sponsored a workshop on high precision photometry in 1984 (Proceedings of the Workshop on Improvements to Photometry, 1984). The success of the first workshop encouraged a second workshop (Second Workshop on Improvements to Photometry, 1988) jointly sponsored by ARC and the NBS (National Bureau of Standards, now NIST) at Gaithersburg, Maryland in 1987. A wide range of subsystems was discussed including very stable band pass filters, 16-bit analog to digital converters, electronic amplifiers, and detectors. Ion-beam bombardment of band pass filters and the use of silicon diodes were recommended.
To further develop this approach NASA HQ funded the development and testing of proof-of-concept multichannel photometers based on silicon photodiodes. Tests conducted at the NBS and at Ames showed that the diodes had very high precision as expected, but that to reduce their thermal noise, they would need to be cooled to near liquid nitrogen temperatures. Two cooled multichannel photometers were built; the latter was based on an optical fiber feed to the cooled diodes. Erratic transmission of the multimode fibers doomed the latter (Borucki et al.1987 and Borucki et al.1988). 7) 8)
In 1992, NASA HQ proposed a new line of missions to address questions about the Solar System and that would also consider the search for exoplanets. Proposals for concept studies were invited and discussed at a workshop at San Juan Capistrano, CA. The proposals were for complete missions: science, technical, engineering, management, cost, and schedule were to be addressed. For this opportunity, a team was organized to propose a transit search for terrestrial planets. The proposed mission was called FRESIP (FRequency of Earth-Size Inner Planets) to describe its goal.
The review panel found that the science value was very high and would have supported the concept had there been proof that detectors existed with sufficient precision and the requisite low noise to find Earth-size planets.
In 1994, the first flight opportunity for a Discovery-Class mission was announced. FRESIP proposed a 0.95 m aperture photometer to be placed in a Lagrange orbit. CCD detectors were substituted for the silicon detectors because of their capability of tracking many targets simultaneously and their ability to accept many different target patterns. The review panel considered the FRESIP photometer to be a telescope similar to the HST (Hubble Space Telescope) and thus far too expensive to qualify as a Discovery-class mission. The proposal was rejected.
Lab tests to prove that CCD detectors were suitable were funded by small grants from NASA HQ and ARC. The first paper presenting the results of lab tests demonstrating the CCD detectors had the requisite precision and low noise to detect transit patterns of Earth-size transits was published in 1995 (Robinson et al. 1995) 9). The experiment was carried out in the basement of Lick Observatory and used an old 512 x 512 Reticon front-side illuminated CCD. For many of the simulated stars a precision of 5 x 10 -6 was achieved. Back-side illuminated CCDs, where the light does not pass through the wire traces on its way to the active silicon were expected to have even higher precision. An accidental spilling of liquid nitrogen during the lab tests did not cause loss of precision because the records of centroid movement allowed the motions to be regressed out. In fact it was the mathematical identification and removal of the systematic noise that was the break through step that allowed the intrinsic precision of these detectors to be recognized.
In 1996, the second opportunity to propose for a flight mission was announced. Studies showed that mission costs could be reduced if photometer was placed in a solar orbit rather than a Lagrange orbit because of the reduction of space propulsion systems needed to stay in a Lagrange orbit. At the insistence of several members (Koch, Tarter, Sagan) of the team, the mission name was changed from FRESIP to "Kepler" to honor the German astronomer (Johannes Kepler, 1571-1630) who developed the laws of planetary motion and the principle needed to calculate optical prescriptions. Both are critical to the operation of the current mission. The mission cost was estimated in three different ways to show that the mission cost could be accomplished for the available budget. The proposal was rejected because no one had every demonstrated that the simultaneous, automated photometry of thousands of stars could be done. The review panel recommended that we build such a photometer to demonstrate the methods to be used. Funding was granted for such a demonstration from both NASA/HQ and NASA/ARC.
In 2000, the fourth opportunity to propose for a Discovery-class mission was announced and Kepler proposed for the fifth time. Kepler was one of three proposals selected from a total of 26 that was allowed to compete by writing a Concept Study Report and demonstrating readiness to proceed.
In December of 2001, Kepler was selected as Discovery Mission #10. Mission development started in 2002 by placing orders for the detectors.
During the years prior to selection, many events helped get the Mission concept accepted. Two major events were the discovery of extrasolar planets by Michel Mayor's team (Mayor and Queloz 1995)10) and Geoff Marcy's team (Marcy and Butler, 1996) 11) and success by several ground based transit search groups (Charbonneau et al. 2000). 12) Once the radial velocity technique had convincingly demonstrated that many exoplanets existed and NASA HQ recognized that the transit technique was proven and that the technology existed that could find Earth-size planets, both the development of the Kepler Mission and a vigorous ground based efforts were funded. In particular, the many years that the Kepler team devoted to convincing the science community, the technical review panels, and NASA HQ officials, helped promote the funding of ground-based transit surveys that are now so successful in finding and characterizing exoplanets. In turn the success of both the radial velocity and transit approaches helped the Kepler Mission to compete against the many excellent proposals received at every AO for a Discovery-class mission.
When a planet crosses in front of its star as viewed by an observer, the event is called a transit (Figure 1). Transits by terrestrial planets produce a small change in a star's brightness of about 1/10,000 (100 parts per million, ppm), lasting for 1 to 16 hours. This change must be periodic if it is caused by a planet. In addition, all transits produced by the same planet must be of the same change in brightness and last the same amount of time, thus providing a highly repeatable signal and robust detection method.
Once detected, the planet's orbital size can be calculated from the period (how long it takes the planet to orbit once around the star) and the mass of the star using Kepler's Third Law of planetary motion. The size of the planet is found from the depth of the transit (how much the brightness of the star drops) and the size of the star. From the orbital size and the temperature of the star, the planet's characteristic temperature can be calculated. Knowing the temperature of a planet is key to whether or not the planet is habitable (not necessarily inhabited). Only planets with moderate temperatures are habitable for life similar to that found on Earth.
Target FOV (Field of View): Since transits only last a fraction of a day, all the stars must be monitored continuously, that is, their brightnesses must be measured at least once every few hours. The ability to continuously view the stars being monitored dictates that the FOV must never be blocked at any time during the year. Therefore, to avoid the Sun the FOV must be out of the ecliptic plane. The secondary requirement is that the FOV have the largest possible number of stars. This leads to the selection of a region in the Cygnus and Lyra constellations of our Galaxy as shown.
Figure 1: Illustration of a transit by a planet in front of its star as viewed by an observer (image credit: NASA/ARC)
Figure 2: Kepler's Field Of View In Targeted Star Field (image credit: NASA/ARC)
Kepler Science Team:
Hundreds of people across the country are involved in the Kepler Mission. NASA/JPL (Jet Propulsion Laboratory), Pasadena, Calif., managed the development of the project for NASA/ARC (Ames Research Center), Moffett Field, Calif., and is responsible for ensuring that Kepler's flight system performs successfully on orbit. NASA Ames managed the development of the ground system and will conduct scientific analysis for the mission. BATC (Ball Aerospace and Technologies Corporation) developed Kepler's flight system, including the spacecraft and the photometer, and is participating in mission operations. NASA Ames will manage flight operations after commissioning is completed (Ref.19) .
The Science Principal Investigator is William Borucki and the Deputy Principal Investigator is David Koch, both of NASA's Ames Research Center. Other members of Kepler's science team include Co-Investigators, a science working group and participating scientists.
The Co-Investigators include Gibor Basri, University of California at Berkeley, Berkeley, Calif.; Natalie Batalha, San Jose State University, San Jose, CA; Timothy Brown, LCOGT (Las Cumbres Observatory Global Telescope), Goleta, CA; Doug Caldwell, SETI Institute, Mountain View, CA; Jørgen Christensen-Dalsgaard, University of Aarhus, Denmark; William Cochran, McDonald Observatory, University of Texas at Austin; Edna DeVore, SETI Institute; Edward Dunham, Lowell Observatory, Flagstaff AZ; Nick Gautier, JPL, Pasadena, CA; John Geary, SAO (Smithsonian Astrophysical Observatory), Cambridge, MA; Ronald Gilliland, STScI (Space Telescope Science Institute), Baltimore, MD; Alan Gould, LHS (Lawrence Hall of Science), Berkeley, CA; Jon Jenkins, SETI Institute; Yoji Kondo, NASA/GSFC (Goddard Space Flight Center), Greenbelt, MD; David Latham, SAO; Jack Lissauer, NASA Ames; Geoff Marcy, University of California at Berkeley; David Monet, USNO (US Naval Observatory), Flagstaff Station, Flagstaff, AZ and Dimitar Sasselov, Harvard University, Cambridge, MA.
The Science Working Group is comprised of Alan Boss, Carnegie Institution of Washington, Washington D.C.; John J. Caldwell, York University, Canada; Andrea Dupree, SAO; Steve Howell, NOAO (National Optical Astronomy Observatory), Tucson, AZ; Hans Kjeldsen, University of Aarhus, Denmark; Soren Meibom, SAO; David Morrison, NASA Ames and Jill Tarter, SETI Institute.
Participating Scientists are Derek Buzasi, Eureka Scientific, Oakland, Calif.; Matt Holman, Harvard-Smithsonian CfA (Center for Astrophysics), Cambridge, MA; David Charbonneau, CfA; Sara Seager, Massachusetts Institute of Technology, Cambridge, MA; Laurance Doyle, SETI Institute; Jason Steffen, Fermi National Accelerator Laboratory, Batavia, Ill; Eric Ford, University of Florida, Gainsville; William Welsh, San Diego State University, San Diego, CA and Jonathan Fortney, University of California at Santa Cruz, Santa Cruz, CA.
The team members collaborate on various tasks within the project. For example:
• Scientists at SAO, USNO and LCOGT made the observations and interpreted the data used to build the Kepler Input Catalog.
• Scientists at SAO, Harvard, University of California at Berkeley, University of Texas at Austin, NOAO, Lowell Observatory and JPL will conduct the follow-up observing work to confirm discoveries, detect other planets in the systems and improve our understanding of the stellar properties.
• Educators at LHS and SETI Institute conduct the Education and Public Outreach program.
• Scientists at the University of Aarhus lead the Kepler Asteroseismic Science Consortium that determines stellar masses, sizes and ages from the Kepler data.
The Kepler Space Observatory, a PI (Principal Investigator) class mission, was competitively selected as NASA's tenth Discovery mission. NASA selected BATC (Ball Aerospace and Technologies Corporation) of Boulder, CO, as the prime contractor for both the photometer and spacecraft. The prime contractor is also responsible for operating the mission. This approach removes many contractual barriers to optimal mission design, efficiency, risk, and schedule for the flight hardware and software. Having a single contractor allows for a single systems engineering team and common subsystem engineering teams for software, thermal, integration and test, etc. for both the photometer and the spacecraft. This approach has allowed for the broadest possible trade space when conducting studies and further eliminates the need for defining many controlled interfaces to external entities, which may often be artificial. 15) 16) 17)
Systems engineering is an important discipline in the development and execution of space-astronomy missions. As observatories and instruments grow in size, complexity, and capability, we are forced to deal with new performance regimes – in many cases forcing us to find solutions to issues and error sources that could be safely ignored on past missions. Systems engineering, if applied rigorously and judiciously, can bring to bear a suite of processes and tools that can help balance risk, cost, and mission success. 18)
The Kepler mission has been optimized to search for Earth-size planets (0.5 to 10 earth masses) in the HZ (Habitable Zone) of solar-like stars. Given this design, the mission will be capable of not only detecting Earth analogs, but a wide range of planetary types and characteristics ranging from Mars-size objects and orbital periods of days to gas-giants and decade long orbits. The mission is designed to survey the full range of spectral-types of dwarf stars. Kepler utilizes photometry to detect planet's transiting their parent star. Three or more transits of a star with a statistically consistent period, brightness change and duration provide a rigorous method of detection. From the relative brightness change the planet size can be calculated. From the period the orbital size can be calculated and its location relative to the HZ determined.
The Kepler spacecraft (Figure 3) has significant heritage from Deep Impact and Orbital Express for many of its subsystems, particularly the avionics. The purpose of the spacecraft is to provide power, pointing and telemetry for the photometer. The three-axis-stabilized spacecraft is fully redundant and single-fault tolerant.
ADCS (Attitude Determination and Control Subsystem): Of primary concern for achieving the photometric precision is attitude stability. Image motion has an adverse affect on the photometric precision due to both the extended wings of the psf and the inter- and intra-pixel responsivity variations. The requirement is to keep the temporal frequency of anything that can affect the photometric precision well outside of the time domain for a transit. Transits can occur on time scales from an hour or so (a grazing transit of a planet with an orbit of a few days) up to 16 hours (a central transit of a planet with an orbit like Mars). To achieve the short term stability the ADCS needs to operate at about 10 Hz to keep jitter low. The specification is 0.1 arcsec (3σ) about each of three axes. To prevent long-term drifts, four fine guidance sensor CCDs are mounted to the scientific focal plane at the four corners. Note that in heliocentric orbit, the only external torque is solar radiation pressure (photons). Unlike Earth orbit, there is no gravity gradient, magnetic torquing or atmospheric drag. Control is provided by four reaction wheels, which are unloaded periodically by a twelve-thruster hydrazine reaction control system. There are ten coarse sun sensors, two star trackers, and two three-axes inertial measurement units for initial acquisition, roll maneuvers and safe-survival modes.
Figure 3: The Kepler spacecraft with the HGA (High Gain Antenna) deployed and the photometer (image credit: NASA, KeplerTeam)
EPS (Electrical Power Subsystem): The EPS is based on a direct-energy transfer architecture. The solar array is designed to produce at least 615 W at 29±4 V at the end of mission in the nominal observing attitude. Solar-array strings are switched as required to provide power to flight segment loads. The spacecraft is rotated 90º every three months to maintain the Sun on the solar array. The solar array is thermally isolated from the spacecraft and photometer. A Li-ion battery is provided to support launch and emergency modes, but is not needed for the observing mode.
The solar array is rigidly mounted to the spacecraft's upper deck. As such, it pulls double-duty on this mission, providing power, as well as shielding the photometer from direct solar heating. The solar array is on four non-coplanar panels and totals 10.2 m2 of triple-junction photovoltaic cells. It contains 130 strings each composed of 22 cells. The solar array is expected to generate up to 1,100 W of electrical power. Unlike most spacecraft solar arrays that are deployed or articulated, Kepler's solar array is fixed.
TCS (Thermal Control Subsystem): TCS is responsible for maintaining spacecraft component temperatures within operational limits. The solar array and thermal blankets shield the photometer from direct solar heating. The solar panels themselves are made out of a special material to minimize heat flow to the photometer, and their finishes also help regulate panel temperature. Kepler is also protected by an "active" thermal control system that consists of heat pipes, thermally conductive adhesives, heaters and temperature sensors. Propane and ammonia flowing through pipes embedded in the spacecraft's exterior panels cool the focal plane. Various parts of the spacecraft that need to be heated in order to operate are equipped with controlled heaters but insulated to avoid heating the photometer.
Avionics: The spacecraft avionics are derived from the design used for the Orbital Express mission. They are fully redundant and can be cross switched between the A and B sides. The processors are the same as for the photometer, radiation hardened PowerPC 750s built by BAE. The avionics provide command and telemetry processing, formatting and storage of spacecraft housekeeping data, thermal control processing, ADCS processing, a mission unique board for items like the cover release, and network interfaces between all of the subsystems and with the photometer. Redundant crystal oscillators are used for on-board time keeping with drift rates of less than 5 x 10-11.
RF Telecommunications: Telemetry for the stored data will be transmitted to the ground using a Ka-band (32 GHz) high-gain antenna (HGA) with a diameter of 0.8 m. Data rates range up to 2.88 Mbps and use a 35 W TWTA (Traveling Wave Tube Amplifier). The command uplink and realtime engineering data downlink will use an omni-directional X-band (8 GHz) antenna system and a 25 W TWTA. A 34 m BWG (Beam Wave Guide) antenna is baselined for the uplink transmitter. The one-time release HGA boom and the redundant two-axes gimbal are the only mechanisms on the spacecraft. The command contacts and data downlinks should not interrupt the precision or recording of the scientific data.
Table 1: Key spacecraft parameters 19)
Spacecraft structures and mechanisms: The majority of Kepler's systems and subsystems are mounted on a low-profile hexagonal box which is wrapped around the base of the photometer. The hexagonal box structure consists of six shear panels, a top deck, bottom deck, reaction control system deck, and the launch vehicle adapter ring. Construction of the shear panels, decks, and solar array substrates, consists of sandwiched aluminum face-sheets on an aluminum honeycomb core. The six shear panels provide structure to accommodate mounting of the spacecraft electronics, portions of the photometer electronics, battery, star trackers, reaction wheels, inertial measurement units, radio equipment, and high- and low-gain antennas.
The top deck shear panel provides the mounting surface for the solar array panels. The bottom deck provides the interface to the photometer and also supports the thrusters, associated propellant lines, and launch vehicle umbilical connectors. The reaction control system deck is attached to the inside of the launch vehicle adapter ring, and provides a mounting surface for the tank, pressure transducer, latch valves, and propellant lines. The base of the photometer is mounted to the lower deck.
Figure 4: Kepler spacecraft integrated with Photometer (image credit: NASA, Kepler Team)
Figure 5: The Kepler spacecraft in Astrotech's Hazardous Processing Facility in Titusville, FL in February 2009 (image credit: NASA)
Launch: The Kepler Observatory was launched on March 7, 2009 (03:49:57 UTC) on a ULA (United Launch Alliance) Delta-II 7925-10L vehicle from Space Launch Complex 17B at the Cape Canaveral Air Force Station, FL. 20)
Orbit: The continuous viewing needed for a high detection efficiency for planetary transits requires that theFOV (Field of View) of the photometer be out of the ecliptic plane so as not to be blocked periodically by the Sun or the Moon. A star field near the galactic plane that meets these viewing constraints and has a sufficiently high star density has been selected. 21)
An Earth-trailing heliocentric orbit with a period of 372.5 days provides the optimum approach to meeting of the combined Sun-Earth-Moon avoidance criteria within the launch vehicle capability. In this orbit the spacecraft slowly drifts away from the Earth and is at a distance of 0.5 AU (worst case) at the end of four years. Telecommunications and navigation for the mission are provided by NASA's DSN (Deep Space Network).
Figure 6: The spacecraft must execute a 90 degree roll every 3 months to reposition the solar panels to face the Sun while keeping the instrument aimed at the target field of view (image credit: NASA)
Figure 7: Illustration of the Kepler spacecraft in orbit (image credit: NASA)
Key Mission Requirements:
Key considerations when looking for planetary transits are the probability for the orbital plane to be aligned along the line of sight and the number of stars to monitor. The probability of orbital alignment is simply the ratio of the stellar diameter to the orbital diameter. For the Sun-Earth analogy the probability is 0.5%. Hence, one needs to monitor many thousands of stars before one can arrive at a statistically meaningful result, null or otherwise.
In addition, a sequence of transits with a consistent period, depth and duration must be detected to be confident and to confirm the existence of a planet. A Sun-Earth-like transit produces an apparent change in brightness of the star of 84 ppm (parts per million) with a duration of 13 hours, if it crosses near the center of the star. For a statistically significant detection, the minimum single transit Signal to Noise Ratio (SNR) requirement is taken to be 4σ, leading to a combined average significance of 8σ for 4 such transits. The detection threshold is set at 7σ, yielding a detection rate of 84% while controlling the total number of expected false alarms to no more than one for the entire experiment. The total system noise, defined to be the CDPP (Combined Differential Photometric Precision), must be less than 21 ppm in 6.5 hours (half of a central transit duration).
The resulting driving requirements for the Kepler Mission are:
1) A CDPP of 20 ppm in 6.5 hrs and the ability to detect a single Earth-like transit with an SNR>4
2) The capability to monitor >100,000 stars simultaneously (>170,000 stars in the first year)
3) A mission duration of at least four years.
Sensor complement: (Photometer)
Table 2: Main parameters of the photometer
The instrument has the sensitivity to detect an Earth-size transit of an mv=12 G2V (solar-like) star at 4 σ in 6.5 hours of integration. The instrument has a spectral bandpass from 400 nm to 850 nm. Data from the individual pixels that make up each star of the 100,000 main-sequence stars brighter than mv=14 are recorded continuously and simultaneously. The data are stored on the spacecraft and transmitted to the ground about once a month. 22)
The sole instrument aboard Kepler is a photometer (or light meter), an instrument that measures the brightness variations of stars. The photometer consists of the telescope, the focal plane array and the local detector electronics.
Telescope: Kepler has a very large field of view — approximately 100 square degrees — for an astronomical telescope. The photometer optics are a modification of the classic Schmidt telescope design. They include a 0.95 m aperture fused-silica Schmidt corrector plate and a 1.4 m diameter 85% light weighted ultra-low expansion-glass primary mirror. The mirror has an enhanced silver coating. The optical design results in 95% of the energy from a star being distributed over an area at the focal plane of approximately seven pixels in diameter. The primary mirror is mounted onto three focus mechanisms, which may be used in flight to make fine focus adjustments. The focus mechanisms can adjust the mirror's piston, tip and tilt. While electrical power is required to move the focus mechanisms, they are designed to hold the position of the primary mirror without continuous power. A sunshade is mounted at the front of the telescope to prevent sunlight from entering the photometer. Kepler is the ninth largest Schmidt telescope ever built and the largest telescope ever to be launched beyond Earth orbit.
Figure 9: Inspection of the 1.4 meter primary mirror honeycomb structure. The mirror has been 86% light weighted, and only weighs 14% of a solid mirror of the same dimensions (image credit: NASA, Kepler Team)
FPA (Focal Plane Array): At the heart of the photometer is the Focal Plane Array. This consists of a set of CCDs (Charged Coupled Devices), sapphire field flattening lenses, an invar substrate, heat pipes and radiator.
The CCDs are the silicon light-sensitive chips that are used in today's TV cameras, camcorders and digital cameras. The CCDs aboard Kepler are not used to take pictures in the conventional sense. Kepler's wide-field optics reflect light from the star field onto the array of 42 CCDs. Each of the 42 CCDs are 59 x 28 mm in size and contain 2,200 by 1024 pixels, that is, individual picture elements, for a total of 95 Mpixels. The CCDs are four-phase, thinned, back-illuminated and anti-reflection coated devices. Each device has two outputs, resulting in a total of 84 data channels. The CCDs are mounted in pairs and have a single sapphire field-flattening lens over each pair. The optics spread the light of the stars over several pixels within an individual CCD to improve differential photometry thus making the system less sensitive to inter-pixel response variations and pointing jitter.
The focal plane is cooled to about -85º Celsius by heat pipes that carry the heat to an external radiator. Data from the CCDs are extracted every six seconds to limit saturation and added on board to form a 30-minute sum for each pixel. The array is supported midway between the Schmidt corrector and the primary mirror.
Figure 10: Completed flight focal plane array with the 42 science CCDs and four fine guidance CCDs in the corners (image credit: NASA, Kepler Team)
Local detector electronics: A local detector electronics box communicates with the 84 data channels and converts the CCD output analog signals into digital data. The electronics box is located directly behind the focal-plane array in the center of the photometer structure. It has more than 22,000 electronic components tightly packed into a volume measuring slightly more than one cubic foot. Careful thermal engineering was required in order to isolate the cold detectors from the heat of the detector electronics. The data are stored in the spacecraft's solid-state recorder and transmitted to the ground approximately once a month.
Data handling: Since the entire 95 Mpixels of data cannot be stored continuously for 30 days, the science team has pre-selected the pixels of interest associated with each star of interest. This amounts to about 5 %of the pixels. These data are then requantized, compressed and stored. The on-board photometer flight software gathers the science and ancillary pixel data and stores them in a 16 GB solid-state recorder. Data are required to be stored and downlinked for science stars, p-mode stars, smear, black level, background and full FOV images.
The Kepler focal plane is approximately 30 x 30 cm in size. It is composed of 25 individually mounted modules. The 4 corner modules are used for fine guiding and the other 21 modules are used for science observing. Attached are some pictures that show a single science module and the assembled focal plane with all 25 modules installed.
Note that the fine guidance modules in the corners of the focal plane are very much smaller CCDs than the science modules. On the left, a single science module with two CCDs and a single field flattening lens mounted onto an Invar carrier. On the right of Figure 11, a focal plane assembly with all 21 science modules and four fine-guidance sensors, one in each corner, installed. Under normal operations, each module and its electronics convert light into digital numbers. For the darkest parts of the image between stars, we expect these numbers to be very small (but not zero). Correspondingly, for the brightest stars in the image, much larger numbers are expected creating an image of each observed star and its background neighborhood. 23)
Figure 11: Left: Single 2-CCD module, Right: Full focal plane CCD assembly (image credit: NASA, Kepler Team)
Selecting the Kepler Star Field: The star field for the Kepler Mission was selected based on the following constraints:
1) The field must be continuously viewable throughout the mission.
2) The field needs to be rich in stars similar to our sun because Kepler needs to observe more than 100,000 stars simultaneously.
3) The spacecraft and photometer, with its sunshade, must fit inside a standard Delta II launch vehicle.
The size of the optics and the space available for the sunshield require the center of the star field to be more than 55º above or below the path of the sun as the spacecraft orbits the sun each year trailing behind the Earth.
This left two portions of the sky to view, one each in the northern and southern sky. The Cygnus-Lyra region in the northern sky was chosen for its rich field of stars somewhat richer than a southern field. Consistent with this decision, all of the ground-based telescopes that support the Kepler team's follow-up observation work are located at northern latitudes.
Distances to the Kepler Stars: Kepler will be looking along the Orion spiral arm of our galaxy. The distance to most of the stars for which Earth-size planets can be detected by Kepler is from 600 to 3,000 light years. Less than 1% of the stars that Kepler will be looking at are closer than 600 light years. Stars farther than 3,000 light years are too faint for Kepler to observe the transits needed to detect Earth-size planets.
Figure 12: The Kepler Field of View (image credit: NASA, Kepler Team)
The ground segment facilities, shown in Figure 13, are used to operate the Flight Segment and analyze the data. Overall mission direction will be provided from the Mission Management and Science Offices hosted by the SOC ( Science Operations Center) at NASA/ARC ( Ames Research Center) in Mountain View, California. Strategic mission planning and target selection is done at the SOC. Target selection will utilize an input catalog especially generated by a team of Co-Is (Co-Investigators) led by the SAO (Smithsonian Astrophysical Observatory) that will provide the means to discriminate between dwarf and giant stars. 24)
Scientific data analysis, direction for the FOP (Follow-up Observing Program) implemented by Co-Is, and final interpretation will also be performed at NASA Ames. Flight Segment operations management, tactical mission planning, sequence validation, and engineering trend analysis will be directed by a FPC (Flight Planning Center) at BATC (Ball Aerospace Technologies Corporation) in Boulder, Colorado. Command and data processing, Flight Segment health & status monitoring, DSN scheduling, and spacecraft ephemeris propagation is the responsibility of the MOC (Mission Operations Center) at HTSI (Honeywell Technology Solutions Inc.) facility in Columbia, Maryland. Uplink and downlink telecommunications will use the NASA/JPL DSMS (Deep Space Mission System ), i.e., the DSN 34 m antennas located around the world.
The DMC (Data Management Center) at the STScI (Space Telescope Science Institute) in Baltimore, Maryland receives the "raw" Level 1 data and performs pixel-level calibration. The resulting Level 2 data set is archived by the DMC and forwarded to the SOC for further processing. The SOC processing includes generation of calibrated photometric light curves (returned to the DMC as a Level 3 data set for inclusion in the Kepler public archive) and transit detection. STScI also provides p-mode analysis. After an extensive data validation process, follow-up observations on each planetary candidate will be performed. The FOP is necessary to eliminate intrinsic false positives due to grazing-eclipsing binaries and extrinsic false positives due to background eclipsing binaries or discriminate between terrestrial transits of the target star and giant planet transits of a background star.
Figure 13: Overview of the ground segment (image credit: NASA, Kepler Team)
Systems Engineering Organization
While there are 6 major partners involved in the Kepler mission, the bulk of the systems engineering work at the Mission/Project System-level and Segment-levels is performed via collaborative effort between JPL, ARC, and BATC. The distribution of effort can be split into 4 major tasks: Science Systems Engineering, Mission/Project Systems Engineering, Flight Segment Systems Engineering, and Ground Segment Systems Engineering. Given the Kepler team structure, these 4 tasks are covered in a distributed fashion. For example, the Science Office at ARC leads the "Science Systems Engineering" effort – which primarily involves science requirements synthesis and follow-up observing program planning - and receives significant support from the Ground Segment Systems Engineer at ARC and the Project System Engineer and Mission Scientist at JPL in the areas of requirements sub-allocation and validation. Likewise, the Project System Engineer leads mission-level engineering efforts but receives substantial support from the Flight Segment System Engineer at BATC in the areas of end-to-end performance modeling, mission-level technical performance metric tracking, launch vehicle interface definition, and mission planning and trajectory design.
Given the potential for confusion and/or gaps in the lines of roles and responsibilities (an issue which seriously impacted some past missions), the Kepler project was careful to establish the diagram in Figure 14 which clarifies those relationships. It's worth highlighting the need for tight coupling of science and engineering on space missions. Care must be taken to avoid misunderstandings and gaps between science needs and engineering implementations. On Kepler, we have established a very tightly knit systems engineering team, which includes representation from the Science Team in the form of the Deputy PI and Mission Scientist – together with the Project System Engineer, they have "one foot rooted in each camp".
Figure 14: System Engineering Roles & Responsibilities (image credit: NASA, Kepler Team)