Wavelength sampling - NIRSpec G395H

[ncrouzet]

Hi!

I have found a strange thing that might be a small bug in Pandexo.

My simulations are for one transit with NIRSpec G395H. The issue is: There are way too many points between wavelengths 5.097 and 5.1 microns.

I ran simulations for several planets and input spectra, they all have that issue. The input spectra I use do not have that. The issue appears in the spectrum resampled by Pandexo, then propagates to the final Pandexo output.

I attach one example, for HAT-P-11b, with:

• input model spectrum (met_office_um_transmission_spectrum_hatp11b_equilibrium.txt)
• spectrum resampled by Pandexo (met_office_um_transmission_spectrum_hatp11b_equilibrium_resampled.txt)
• Pandexo simulated spectrum (met_office_um_transmission_spectrum_hatp11b_equilibrium_pandexo.txt)
• plot of the wavelength in the resampled and simulated Pandexo spectra (plot-wavelengths.png). The jump at index 1270 is the detector gap, so it's fine, but the flat part between index 3200 and 3900 that is stuck around 5.1 is weird. Indeed in the txt files there are many many lines between 5.097 and 5.1 microns.

I have this issue for the current Pandexo version, and it was already there in the previous version I used, which dates back from February last year.

I found it for G395H but I have not checked for any other NIR modes.

It's not urgent, it affects the result only in a very small fraction of the spectrum.

Do you also see that issue? If not, it may come from my side...

Best,
Nicolas Crouzet

met_office_um_transmission_spectrum_hatp11b_equilibrium.txt
met_office_um_transmission_spectrum_hatp11b_equilibrium_resampled.txt
met_office_um_transmission_spectrum_hatp11b_equilibrium_pandexo.txt