5.1 Distribution of GTO time among science cases
The Consortium has agreed to conduct the different GTO programs as a common effort, with the following distribution of GTO time among science cases:- 80% for low-mass extrasolar planets
- 10% for fundamental constants
- 10% for other, high-quality science to be determined at a later stage of the project
In the following sections it is assumed that the Consortium will receive about 160-240 VLT nights depending on the total investment (1 VLT night = 50 kEUR). This would represent about 1500 hours of observing time for extrasolar planets and 200 hours for fundamental constants. The use of this observing time is described in more details below for the two major science cases.
5.2 Extrasolar planets
5.2.1 Main scientific objectives
The exoplanet GTO program will consist of a survey of a carefully selected sample of nearby GKM dwarfs with the goal to detect a significant fraction of the planets orbiting them within ~1 AU down to minimum masses of ~1-2 Mearth. Some of these objects will be located in the habitable zone. Obviously, more massive planets at larger separations will be detected as well. This will allow us to accurately determine the overall frequency of such low-mass objects around solar-type stars and M dwarfs. The derived mass and semi-major axis distributions will provide new valuable inputs to improve our understanding of planet formation. The probably large number of multi-planet systems will open a new era for the study of planetary dynamics. Another expected field of investigation is the study of the correlations between the properties of low-mass objects and those of their parent star (e.g. chemical abundances).Finally, each short-period super-Earth or ice giant discovered with ESPRESSO will be considered as a potential transiting object, and follow-up photometric observations will be carried out to find the possible transit.
5.2.2 Number of potential targets
For GK dwarfs we aim at 10 cm/s accuracy with ESPRESSO in 1-UT mode. According to the expected instrumental performances, we will focus on bright stars with V < 8.5, for which a RV precision of ~5-10 cm/s can be obtained in 15-30 minutes of integration time. Such an exposure time is also needed to average out the stellar noise (p-mode oscillations and granulation).We will use the results from the HARPS GTO surveys and carry out dedicated programs with this instrument to establish the best possible list of targets. A search in the Hipparcos catalogue, selecting mid-G to mid-K dwarfs observable from Paranal with visual magnitudes brighter than 8.5, gives a total of more than 1000 stars. We then need to exclude spectroscopic binaries and active stars. In fact, we need the most inactive stars of the solar neighbourhood. These will be known from the HARPS GTO programs. Depending on the exact selection criteria, a list of at least 100-200 suitable G and K stars can be obtained.
Regarding the M targets, there are over 30 nearby M0.5-M4.5 stars brighter than V=12 mag and with declination < 10 deg, for which ESPRESSO can provide spectra with an accuracy better than 40 cm/s in one hour integration time or less. Note that this accuracy is enough to detect rocky planets in the habitable zones of Ms. These stars are single or very wide binaries, and most of them are also in the HARPS GTO sample. This will allow us to optimally select the best targets to follow with ESPRESSO.
5.2.3 Observing strategy
A trade-off has to be performed between the number of targets in the sample and the number of measurements on each target. The experience acquired with HARPS tends to stress the importance of accumulating a lot of measurements per star rather than observing a lot of stars, for two main reasons: 1) complex multi-planet systems are common and require many measurements to be resolved, and 2) the fraction of stars having low-mass planets seems to be high (at least 30% according to the most recent HARPS results), and therefore meaningful statistics can be obtained even with a relatively small sample.Although the best observing strategy will be studied in mode detail during later stages of the project, we can for the moment assume an average number of 40-50 measurements per star and 15-30 min of integration time per measurement. Obviously, the number of measurements per target will be adjusted according to the characteristics of the RV signals emerging from the data. With this strategy we can estimate how many stars can be observed during GTO time: 1500 hours / 15 hours per star = 100 stars. This number is well within the expected number of available targets. At first glance, 100 stars may appear somewhat low to obtain useful statistical properties of low-mass extrasolar planets. However, if we assume that 50% of the stars do have such planets, and that they always come in multi-planet systems with an average of 2-3 detectable planets per system, this would already give us about 100-150 planets. This is enough to reach the scientific goals mentioned above, although further refinements will require more observing time.
The exoplanet GTO program can be carried out over a time span of about 3 years, but a longer duration would be also acceptable. Distributed over 3 years, it would represent about 25 nights per semester.
5.3 Fundamental constants
5.3.1 Main scientific objectives
The main objective of the GTO programs dedicated to the search for fundamental constant variability will consist of establishing whether present hints of variability for α and μ are real or not. The goal will be achieved by HR- USS measurements with 1 ppm accuracy, or better, for both the constants. The new observations will either confirm the claimed variability or will place more stringent bounds at lower level. A confirmation of variability with high statistical significance is of crucial importance. It will open a window to new physics beyond the Standard Model and possibly shading new light on the nature of Dark Energy. On the other hand, more stringent bounds are important to constrain various theoretical models which leads to the variability of fundamental constants.5.3.2 Number of potential targets
The whole GTO program will be dedicated to deep spectroscopic observations of few carefully selected targets which offer the best characteristics for achieving the most precise measurements of the two constants. Observations are rather time-demanding since high S/N is required. We may expect to be able to observe somewhat like 4 targets for α and 4 targets for μ probing the constants at the level of 1 ppm or better. Present hint for α variability is of Dα/α = (-5.7 ± 1.1) ppm from averaging 143 systems and for μ, Dμ/μ = (20 ± 6) ppm for two systems. With ESPRESSO we expect to measure both constants at the level of 1 ppm or better in individual systems thus clarifying the reality, once for ever, of the claimed variability.The targets will be selected from the whole sample of absorbing systems. For α the extant sample comprises the 143 systems studied by Murphy et al, of which several are reachable from Paranal; the 23 systems of the Chand et al sample drawn from the ESO Large Program (PI J. Bergeron) of bright QSOs; the two bright QSOs HE 0515-4414, Q 1101-264 with V=15 and 16 mag respectively. Since we are dealing with relatively bright QSOs, it is unlikely that we will have many new potential targets in the next six years. The criteria for target selection are: i) the brightness of the QSO, ii) the number of components forming the absorption system, and iii) the intrinsic width of the metal lines. In addition important things to be considered are also the strength and location of the more crucial lines for the analysis and the redshift of the system. The latter is important since theories predict larger variations for higher redshift. Primary targets are the brightest QSO HE 0515-4414 and QSO 1101-264, with redshift 1.15 and 1.84, respectively. HE 0001-2340 drawn from the Chand et al sample, which shows two extremely simple absorption systems at z=2.185 and 2.187 is an additional target matching all the above mentioned criteria.
The number of extant targets for Dμ/μ is much more restricted since the number of high redshift systems showing H2 is presently rather limited. However, there are several ongoing surveys searching H2 absorption systems which may produce new high redshift systems suitable for μ observations. Both photon noise and instrumental systematics contributes to the error budget in the measurements in QSO 0347-3819 and QSO 0405-443. Thus ESPRESSO observations of these two systems will provide significantly higher accuracy are crucial to establish the μ variability.
An additional very interesting target is provided by the high redshift H2 system at z=4.224 towards PSS J1443+2724 (Ledoux et al 2006). This target has V=19.3 and cannot be observed with UVES with the required signal-to-noise to probe Dμ/μ. Considering the possible time evolution for the constants this is an extremely interesting target to be followed with ESPRESSO observations in the 4 UTs mode.
5.3.3 Observing strategy
The approximately 20 GTO nights preliminary allocated by the consortium to this research could be split into two parts equally distributed for α and μ. For both constants we might exploit both the 1UT and 4UT modes. The 1 UT will be used to achieve the highest accuracy measurements by using the brightest suitable sources while leaving the 4UT mode to perform a somewhat less accurate measurement but at slightly higher redshift. In this case the decrease in the precision is compensated by a much larger time leverage.5.4 Additional science cases
Approximately 20 GTO nights for other, high-quality science will be determined at a later stage of the project among the science described in chapter 4.