Organisation of Ground-based Observation

Required telescope time for planet mass determinations

From the information on the planet yield, the cost of “equivalent precision” radial velocities, and the distribution of stellar activity in the sample, we can estimate the amount of telescope time required to measure the mass of PLATO candidates. These estimates are based on a detailed study assuming:

  • A distribution in planet masses and orbital separations based on Keplerresults (Fressin et al. 2013). We however do not consider hot and warm Jupiters, since they are expected to be already well studied in ground-based surveys and with TESS.
  • An ad-hoc number of required “phase points” for each considered planet mass to reach a given precision. This number is based on a very conservative experience gathered over the years in the framework of planet-search programmes at high precision.
  • An adequate match of the binning used and the period of the planet.

A realistic spread between different telescope sizes for a given planet category has been used. Observing facilities and observing strategy are selected in order to match required precision and photon-noise limits.


Estimates of ground-based telescope resources needed for follow up of planet candidates discovered during the PLATO Long-duration Observation Phase in both hemispheres.
Note 1:  The time spans of 7 years for the filtering observations and of 9 years for the radial velocity observations are assumptions based on estimates of available telescope resources.
Note 2: The numbers reported in the Table are global for northern and southern sky visibility.

It is worth noting that while we have been conservative in our analysis, there is much on-going theoretical and observational effort to understand the effects of activity signals on photometric and radial velocity measurements. There have already been some successes in correcting for activity albeit only in the case of the most massive planets (e.g. Hébrard et al. 2014; Bruno et al. 2016). The situation by 2025 is likely to be much improved.

Strategy for ground-based follow-up observations in view of available telescope resources

The PLATO mission is an effective planet survey, detecting hundreds of small planets and thousands of larger planets. Measuring planet masses for the full sample of expected PLATO planets around bright stars requires a significant amount of telescope time. A complete census of the sample will need a world-wide-effort. Here we discuss the strategy by which this goal can be achieved, taking into account a realistic estimate of available telescope resources.

To characterise all planets for their masses is a huge effort limited by several factors:

    1. Magnitude of the host star
      Based on experience of existing RV spectroscopy instruments, determining planet masses for small planets is limited to bright host stars (V≤ 11).
    2. Activity of the host star
      In recent years, RV spectroscopy data have shown that stellar activity is a limiting factor for planet mass measurements. In particular for the smallest planets only the quietest stars will allow us to determine accurate masses.
    3. Available telescope resources
      The required telescope resources require 1 – 8m class telescopes.
    4. Time period for which RV follow-up can be performed
      This is not a strict limit, because in principle scientifically interesting targets can and will be followed by the community over the time it needs to answer the science question investigated. For the sake of dimensioning the resources for the PLATO space mission, however, the RV follow-up is limited to 9 years (up to 5 years after end of observations). Targets not completed in that time interval will become part of the PLATO legacy.

The smallest planets will benefit from ground-based follow-up observations starting after one transit (this will be challenging). Obtaining spectroscopy while the spacecraft is obtaining photometry seems one of the best methods to mitigate stellar activity.

We assume that filtering observations are spread over 7 years. For the long-period planets (up to 1 year), additional telescope time is required to determine their masses. Mass determination therefore continues in years 8 and 9 after launch.

The number of nights given in Table 6.2are total estimates and need to be distributed among different observatories. The distribution of PLATO fields in the northern and southern hemisphere naturally leads to a differentiation on the required time.

In view of the large total telescope resources needed to determine planet masses for all host stars with V£11, a strategy is needed which keeps the effort within reasonable limits for existing and future observatories/instruments but still delivers the PLATO science. We plan proceeding in steps:

  1. Focus on a sample of PLATO’s unique planet parameter space, guaranteeing the maximum scientific return, taking into account the limiting factors a) and b) above.
    We point out that PLATO addresses a very broad range of science areas and “maximum science return” therefore cannot be obtained by considering a certain sample of planets defined by simple criteria such as host star brightness, orbital distance, planet size, etc. This means we expect that individual objects of high scientific interest will be investigated for all kinds of star/planet combinations. Here we simplify to a discussion of “samples” for the sake of dimensioning the follow-up resources.
  1. For a sub-sample of planets obtain PLATO telescope resources from major European observatories, namely ESO and La Palma.
    This sample shall be large enough to ensure the main science objectives can be reached at a minimum level.
  1. Enlarge the sample of planets with determined masses by obtaining time at smaller European observatories and with teams with guaranteed telescope time available.
  2. Enlarge the RV follow-up effort beyond Europe by including international observatories/teams.
  3. Remaining planets, not cleared in the 9 years PLATO observation+RV period, become part of the PLATO legacy and can continue to be studied by the international community in the years to come.

This strategy leads to essentially 80% of the best targets down to Super-Earths (~200planets) being characterised. Earth mass planets are by far the most difficult case (even a 2‒3MEplanet would be much easier to obtain radial velocity information on) and our observations are limited by the availability of an ESPRESSO like instrument on 8m class telescopes. None the less our success rate is already about 50% for longer period planets from ESO (keeping in mind that max ~60% can be reached from one hemisphere only). This improves to about 80% if a similar facility becomes available in the northern hemisphere. This should be seen against the backdrop that we do not expect to have any radial velocity information for any similar systems before the launch of PLATO.

If planets in the step-and-stare phases are also be considered, then additional resources need to be added. We note, however, that after NASA’s TESS mission and respective follow-up program a large fraction of short-period, hot, planets will already have RV data available at the time of PLATO launch. Adding the full sample may therefore significantly overestimate the needs for PLATO.

Note: We do not yet consider the P4 stellar sample of M dwarfs here but note that optimised facilities are available specifically for these objects (e.g. CARMENES), and should be continued until after PLATO launch.




PLATO – Revealing habitable worlds around solar-like stars
Definition Study Report, ESA-SCI(2017)1, April 2017