Source-linked AI summary
PandExo: A Community Tool for Transiting Exoplanet Science with JWST & HST
Natasha E. Batalha, Avi Mandell, Klaus Pontoppidan, Kevin B. Stevenson, Nikole K. Lewis, Jason Kalirai, Thomas Greene, Loïc Albert, Louise D. Nielsen, Nick Earl
TL;DR
Existing JWST studies modeled theoretical instrument noise to estimate performance and detectable atmospheric features, but the community needed accessible simulations for observing planning. PandExo provides open-source JWST and HST time-series spectroscopy simulations using Pandeia components and multiple noise formulations, with calculations agreeing well with instrument-team results. It also supports realistic HST uncertainty estimation and observation planning, while some regime-specific approximations remain.
Problem
JWST studies needed to characterize instrument performance and theoretically detectable atmospheric features, while the community needed accessible simulations to integrate these results into observing plans.
Method
PandExo is an open-source Python package and online tool that uses Pandeia components and computes noise with LMF and MULTIACCUM formulations for JWST and HST spectroscopy.
Results
Instrument-team calculations are in good agreement with PandExo’s LMF noise calculations across the reported JWST benchmarking.
Takeaways & Limitations
PandExo provides realistic HST/WFC3 uncertainties, supports instrument-setup optimization and scheduling requirements, and is available with documentation, tutorials, and an online interface.
Takeaways & Limitations
For small group counts, PandExo may produce slightly conservative MIRI values because its near-IR detector formula does not capture MIRI’s more efficient read-reset scheme.
Abstract
from arXiv · showhide
As we approach the James Webb Space Telescope (JWST) era, several studies have emerged that aim to: 1) characterize how the instruments will perform and 2) determine what atmospheric spectral features could theoretically be detected using transmission and emission spectroscopy. To some degree, all these studies have relied on modeling of JWST's theoretical instrument noise. With under two years left until launch, it is imperative that the exoplanet community begins to digest and integrate these studies into their observing plans, as well as think about how to leverage the Hubble Space Telescope (HST) to optimize JWST observations. In order to encourage this and to allow all members of the community access to JWST & HST noise simulations, we present here an open-source Python package and online interface for creating observation simulations of all observatory-supported time-series spectroscopy modes. This noise simulator, called PandExo, relies on some aspects of Space Telescope Science Institute's Exposure Time Calculator, Pandeia. We describe PandExo and the formalism for computing noise sources for JWST. Then, we benchmark PandExo's performance against each instrument team's independently written noise simulator for JWST, and previous observations for HST. We find that \texttt{PandExo} is within 10% agreement for HST/WFC3 and for all JWST instruments.
1. INTRODUCTION
Before JWST launch, researchers developed independent models to predict instrument performance and recoverable atmospheric properties, creating a need to integrate these studies into observing plans. PandExo addresses this need with an open-source simulator for JWST and HST time-series spectroscopy modes.
- JWST’s four instruments span 0.6–28 µm and offer low- and medium-resolution modes for exoplanet science.
- Early studies estimated JWST performance and predicted attainable signal-to-noise ratios for NIRCam transit and eclipse spectra.For 1000 seconds, predicted NIRCam SNRs ranged from approximately 5 for faint to approximately 90 for bright G2V stars.
- Later sensitivity models examined atmospheric constraints for habitable planets and high-SNR spectroscopy of small planets around nearby M dwarfs.One study predicted high-SNR NIRSpec spectra for 1–10 M⊕, 400–1000 K planets within approximately 50 pc after co-adding 25 transits.
- Independent simulations incorporated spacecraft jitter, drift, flat-field errors, background noise, and time-varying astrophysical and instrumental systematics.
- PandExo provides community access to JWST and HST observation simulations through an open-source Python package and online tool.It uses portions of STScI’s Pandeia exposure-time calculator and supports observatory-supported time-series spectroscopy modes.
2. PANDEIA: SIMULATING NOISE SOURCES
Pandeia models detector scenes and instrument noise for transit spectroscopy, while PandExo adds exoplanet-specific inputs and supports alternative signal and noise formulations. The resulting simulations account for detector effects, backgrounds, saturation, and observing efficiency, with JWST benchmarks generally agreeing with instrument-team calculations within 10%.
- Pandeia simulations: Pandeia uses three-dimensional, pixel-based astronomical scenes to model detector flux for supported observing modes.Scenes specify spectral energy distributions across spatial and spectral dimensions.
- Noise formulations: MULTIACCUM fits each point up the ramp, whereas PandExo’s LMF signal subtracts the first readout from the final readout.MULTIACCUM accounts for correlations involving groups and averaged frames when calculating pixel noise.
- Detector noise: Correlated detector noise requires covariance-aware error propagation because read noise in one pixel depends on read noise in other pixels.Ignoring these correlations would underestimate the noise, especially in the fast-read direction.
- PandExo inputs: PandExo combines Pandeia with stellar and planetary spectra, transit timing, saturation limits, and a user-defined noise floor.Pandeia updates are automatically incorporated when users update the underlying Python package.
- Benchmarking: Within 10% agreement, PandExo LMF results match the instrument teams’ simulations for the benchmarked JWST modes.MIRI values may be slightly conservative at small group counts because its readout is more efficient than the near-infrared formulation assumes.
5. PANDEXO: SIMULATING HST OBSERVATIONS
PandExo extends HST/WFC3 simulations beyond uncertainty estimation by supporting setup optimization and scheduling, using scaled measurements from prior observations. Its WASP-43b example produces uncertainties close to published data while calculating feasible observation phases.
- PandExo simulates realistic HST/WFC3 transmission and emission uncertainties, optimizes instrument setups, and generates scheduling requirements.These outputs help estimate the number of transit or eclipse visits needed for meaningful constraints.
- The HST/WFC3 implementation scales measured flux, variance, and exposure times from previously observed systems before estimating channel and transit/eclipse-depth uncertainties.
- PandExo requires stellar H-band magnitude, transit/eclipse duration, visit count, channel count, disperser, scan direction, subarray, and schedulability as inputs.Users may additionally optimize HST orbits per visit and WFC3 parameters such as NSAMP and SAMP-SEQ.
- 37.6 ppm simulated uncertainty compares with 36.5±3.5 ppm mean published uncertainty for WASP-43b observed with WFC3/G141.
- WASP-43b simulations place feasible observation starts between phase values 0.3071 and 0.3241.The earliest and latest start times correspond to the computed minimum and maximum phases.
- Future HST work includes STIS G430 and G750 support, wavelength-dependent uncertainties, and methods beyond scaling previous observations.
6. CONCLUSION
PandExo is an open-source package for modeling instrumental noise across JWST and HST transit time-series spectroscopy modes. Its LMF calculations agree well with instrument-team simulations, while photometry modes remain unsupported and the tool is available with online documentation.
- 37.6 ppm simulated uncertainty and 36.5±3.5 ppm mean published uncertainty are shown for the WASP-43b HST/WFC3 example.
- PandExo models instrumental noise for JWST NIRISS, NIRCam, NIRSpec, and MIRI LRS, plus HST/WFC3 transit time-series spectroscopy.
- PandExo supports both Last-Minus-First subtraction and MULTIACCUM ramp fitting with correlated-noise accounting.
- Instrument-team calculations are in good agreement with PandExo noise calculations using the LMF method.
- PandExo currently excludes photometry modes.
- PandExo is available for download with associated documentation, tutorials, and an online interface.