Here are the arguments for running blasé on the HPF spectra:
We are over-zealously masking the telluric lines, censoring the redshifted emission line wing. By using blasé in its telluric-correcting mode we could "claw back" some of the missing spectrum and have a clearer view of the gas dynamics surrounding the escaping exosphere.
Some ostensibly telluric lines (near 10825 Å) appear as sharp lines near the planet velocity at quadrature. If these are tellurics, they evaded our masking procedure and need to be rectified and attributed as such. If they are somehow actually related to the exosphere that's a pretty awesome signal we would like to understand (my guess is this scenario is low probability)
The 10830 Å emission appears relatively broad, meaning it has a comparable vsini~35 km/s as the star. That would make sense since the gas cloud appears so extended as to cover the entire stellar disk at almost all phases. But there still appears to be some velocity substructure attributable to the bulk motion of the gas. Disentangling the gas substructure from the vsini requires a deconvolution process. blasé automatically acts as a deconvolution tool, correctly taking into account both instrumental and rotational modulation. So running blasé on the helium region would yield a clearer view of the velocity structure of the gas and its dynamics.
We are currently searching pre-selected, well-known regions for spectral variability (He, Ca IRT). But the excess is so large that it's conceivable some other, weaker opacity sources may show some evidence for in-transit variability. We could do this in a data-driven way, but two frictions arise: A) continuum flattening the spectrum becomes difficult because its easier to flatten around pre-selected indices rather than flattening the entire spectrum, and B) separating telluric and stellar signals stars to be a chore. So blasé could alleviate these two friciton points, making it easier to search the whole-spectrum.
One of the outputs of blasé is the RV and vsini. So we could hypothetically measure the RV variations attributable to the planet to get towards a mass measurement.
As a stretch goal, it may be able to identify the RM-signal of the planet in-transit, which would be awesome.
Here are the arguments against running blasé on the HAT-P-67 HPF spectra:
Many of the goals can be achieved adequately through data-driven techniques: GP's for continuum flattening, for example.
It will take time to get up-and-running, calibrate, and debug. This risk is especially acute for the more experimental applications such as the R-M idea.
Here are the arguments for running blasé on the HPF spectra:
We are over-zealously masking the telluric lines, censoring the redshifted emission line wing. By using blasé in its telluric-correcting mode we could "claw back" some of the missing spectrum and have a clearer view of the gas dynamics surrounding the escaping exosphere.
Some ostensibly telluric lines (near 10825 Å) appear as sharp lines near the planet velocity at quadrature. If these are tellurics, they evaded our masking procedure and need to be rectified and attributed as such. If they are somehow actually related to the exosphere that's a pretty awesome signal we would like to understand (my guess is this scenario is low probability)
The 10830 Å emission appears relatively broad, meaning it has a comparable vsini~35 km/s as the star. That would make sense since the gas cloud appears so extended as to cover the entire stellar disk at almost all phases. But there still appears to be some velocity substructure attributable to the bulk motion of the gas. Disentangling the gas substructure from the vsini requires a deconvolution process. blasé automatically acts as a deconvolution tool, correctly taking into account both instrumental and rotational modulation. So running blasé on the helium region would yield a clearer view of the velocity structure of the gas and its dynamics.
We are currently searching pre-selected, well-known regions for spectral variability (He, Ca IRT). But the excess is so large that it's conceivable some other, weaker opacity sources may show some evidence for in-transit variability. We could do this in a data-driven way, but two frictions arise: A) continuum flattening the spectrum becomes difficult because its easier to flatten around pre-selected indices rather than flattening the entire spectrum, and B) separating telluric and stellar signals stars to be a chore. So blasé could alleviate these two friciton points, making it easier to search the whole-spectrum.
One of the outputs of blasé is the RV and vsini. So we could hypothetically measure the RV variations attributable to the planet to get towards a mass measurement.
As a stretch goal, it may be able to identify the RM-signal of the planet in-transit, which would be awesome.
Here are the arguments against running blasé on the HAT-P-67 HPF spectra:
Many of the goals can be achieved adequately through data-driven techniques: GP's for continuum flattening, for example.
It will take time to get up-and-running, calibrate, and debug. This risk is especially acute for the more experimental applications such as the R-M idea.
It sounds like scope creep / mission creep.
On balance I think it's a good idea, but I acknowledge the risks and tradeoffs. Here's a compromise surgical solution: let's restrict our focus to---