Just wanted to share this new paper on the protoplanets around PDS 70 based on new Keck data that I've been working on with a great team of early career scientists: https://arxiv.org/abs/2004.09597 . Here's a tweet summary for those interested.

Using the new infrared pyramid wavefront sensor at Keck, we were able to get the cleanest images of PDS 70 b and c in L'-band. To characterize them, we had to remove contamination from the protoplanetary disk, thanks to modeling by @renrenbin and @Nicole_Wallack.

We don't have much orbital motion, so I had to use dynamically motivated constraints of non-crossing orbits and being < 20° misaligned from the disk to get realistic orbits. I used the custom likelihood function in orbitize! which allows users to add orbital constraints like this

We fit some atmospheric models to the planets' SEDs, but found that a blackbody is preferred with the current data. We also found that the total inferred luminosity is pretty consistent regardless of which atmosphere model is used. We'll come back to if a blackbody is realistic.

I teamed up with fellow @HSFdn #51Pegb fellows Sivan Ginzburg and @PlanetaryGao to interpret these results. Sivan used the my luminosities and his evolutionary models for protoplanets to infer that these planets are 1-4 Jupiter masses (some of the lowest mass imaged planets!)

@PlanetaryGao used the mass accretion rates inferred by Sivan to show that dust accretes so fast that it should shroud the atmosphere of this planet. This means that a blackbody is not a terribly unphysical model for what we're seeing. It makes sense, for now...