Description of work The work needed to be done to reach the objectives of this work-package can be broken down in three tasks: Development on a new data reduction methods, applying the new data calibration methods to archival Spitzer and HST observations of exoplanet atmospheres, applying the new data calibration methods to Early Release Science observations with JWST of exoplanet atmospheres. The work within this work-package will be shared by MPIA and SRON with a contribution by CEA. For task 1 MPIA will be the main responsible, while for task 2 and 3 the work will be shared between MPIA, SRON and CEA. For task 1 we estimated to need the equivalent of 24 months full time position., while for tasks 2 and 3 we estimated we would need the equivalent of a 9 month full time position for each of the two tasks. Task 1: Development of a new data reduction method. At present several thousand transiting exoplanet systems have been discovered. For relatively few systems, however, a spectro-photometric characterization of the planetary atmospheres could be performed due to the tiny photometric signatures of the atmospheres and the large systematic noise introduced by the used instruments. Several methods have been developed to deal with instrument and atmospheric noise. These methods include high precision calibration and modeling of the instruments, modeling of the noise using methods like principle component analysis or Gaussian processes and the simultaneous observations of many reference stars. Though significant progress has been made, most of these methods have drawbacks as they either have to make too many assumptions or do not fully utilize all information available in the data to negate the noise terms. The main objective of this work program is the development and implementation of novel data calibration and spectral extraction methods based on causal connections within the data. This new calibration approach was successfully used for photometric observations with the Kepler observatory and we intend to expand the application of this method to spectroscopic observations. The new calibration method requires no assumptions on the functional form of any systematic noise and can substantially improve the precision of the calibration compared to currently implemented schemes. We will intend to make a thorough comparison of the new calibration method we intend to implement and current schemes by blind tests on simulated data with various types of systematic noise. This will ensure that we can demonstrate the reliability of the applied methods and advantages compared to other schemes in an unbiased way. Task 1 is also closely related to Work Package 3. Until now, most transit spectroscopy studies have used a sequential approach of first deriving the instrument calibration, then extracting the spectra including noise estimates and finally analyzing the spectra, retrieving the properties of the exoplanet atmosphere. The ultimate goal of this program is to combine spectral retrieval models with the data calibration methods, to simultaneously model both signal and noise. An important aspect of this task, therefore, is to implement the calibration methods in such a way that they can be combined with the methods developed Work Package 3. Task 2: Applying the selected data reduction method to archive Spitzer and HST data. We intend to apply the method developed in task 1 to archival Spitzer and HST observations of transiting exoplanets. The goal of this task is two fold, first it will allow us to thoroughly test the in task 1 developed methods on data observed with a wide variety of instruments all having their own type of systematic noise sources. This will enable us to demonstrate the accuracy and applicability of our novel calibration methods. Second, at present the archival data is calibrated and analyzed using a range of methods, all having their own assumptions and limits. This makes it hard to make an unbiased comparison of the entire available data set and limits the scientific interpretations. By making a homogeneous analysis of the available data we intend to ensure that no such biases are present. Also, preliminary tests of the data reduction concept have shown that significant signal-to-noise increase of the planetary spectra and lightcurves is possible, resulting in a far better constraint on the properties of the planetary atmospheres. We intend to produce a homogeneous catalog of planetary spectra at the beginning of the JWST science operations. Having reliable estiates of the target properties would provide a substantial support for and the scientific interpretation of and the planning for new observations with JWST as it will enable us to choose the optimal observing strategy and target sample. Task 3: Applying the selected data reduction method to ERS JWST data From spring 2019 onward, a number of exoplanet systems will be observed within the Early Release Science (ERS) program of the JWST observatory. These data will become immediately public, as one of the main goals of this program is to demonstrate the scientific capabilities of JWST and to improve the observational strategies of scientific programs targeting exoplanets. It is, therefore, of the utmost importance that a timely and consistent calibration of the data is undertaken. In this task we will apply our calibration methods to the ERS data exoplanet spectra to produce a homogeneous set of exoplanet atmosphere spectra. Such a data set will allow for an unbiased comparison of the results and the spectra produced in task 2. To assist the implementation of the calibration methods for JWST data, we will use the existing instrument simulators developed by the JWST instrument teams and the Space Telescope Science Institute (STScI) to create simulated data sets on which we can test our codes. We will also use team data from detector tests we have access to as participant of the MIRI EC consortium. This will ensure that we will have a fully working version of the calibration software package when the first data of JWST will be observed. A large number of exoplanet systems will be observed after the end of this program. Our goal is that the methods created and delivered by us will form the basis of a standard calibration and analysis approach for the JWST data such that over time a homogeneous catalog can be created. |