JWST Data Driven 3D fly-by wire Exoplanet Chart Plotter

#astronomy #showdev #learning #buildinpublic

I won a 27ft sailboat years ago on a for surplus auction peanuts. I went all in on sailing for a little minute: Dyneema standing rigging, a DIY LiPo battery bank. Man, I miss that boat. I called it Social Distance (I got it around the middle of Covid, blah blah) sold it. The point: I was listening to exoplanet podcasts and that, that boat and the $4300 MFD package I should not have put in a $800 sail boat from 1972 “ Ericsson slope....the hulls were hand laid back then”, anyways voila, Cosmoplot popped out of my butt today. It is still rough, the graphics, the controls, the design, but the bones are solid. If you like this kind of stuff and want to help, that would be awesome.

This is my 3D space chart plotter. You launch from the Sun and fly out to real objects, and the planets and stars you see are rendered downstream from actual physics, computed from real NASA survey data, the Deep sky nebula wouldn't be able to do them justice so they are Nasa archive images not painted by hand.

The inputs are real NASA surveys: the NASA Exoplanet Archive for system and stellar parameters, and JWST spectra served through MAST for atmospheres. Pick a planet and you get a derived physics profile: likely interior composition, habitable-zone placement, atmospheric escape regime, a transmission-spectrum inference from real JWST data, and how detectable its heat would be to JWST.

Live: cosmoplot.io. Source: github.com/H-XX-D/Cosmoplot.

Every physical parameter comes from a real NASA survey: the NASA Exoplanet Archive for system and stellar parameters, and JWST spectra via MAST for atmospheres. Everything the app computes is derived from those values with a published relation, and it is labeled as derived, not observed.

Every value carries its provenance. Each number belongs to a tier, and the tier is shown next to it observed, derived, infered, or proxy, .A rendered planet can be mistaken for a photo, and a population estimate shouldn't be passed off measurement.

The Astrophysics

Interior composition is read from the mass-radius point against the reference curves of Zeng, Sasselov & Stewart (2016), R/Re = C (M/Me)^(1/3.7), with C set by composition (iron 0.86, Earth-like 1.00, rock 1.07, water-rich above).

Habitable zone follows the stellar-flux limits of Kopparapu et al. (2013), with the boundary distance d = sqrt((L/Lsun) / S_eff).
Earth Similarity Index uses Schulze-Makuch et al. (2011) over radius, density, escape velocity, and equilibrium temperature.

Atmospheric escape is screened with the Jeans parameter evaluated at the exobase, not the surface, because for a puffy hot atmosphere the large scale height lifts the exobase well above the surface and changes the answer.
Transmission spectra are inverted with scale-height physics, H = kT/(mu g), on real reduced JWST spectra to estimate the mean molecular weight and separate a light hydrogen envelope from a heavy or cloudy one.

White dwarfs use the degenerate mass-radius relation of Nauenberg (1972) shown against the real Tremblay et al. (2019) Gaia sample.

Rather than reporting single derived values, the app draws each input from its survey error bars with a per-planet seeded sampler, recomputes the full derived chain thousands of times, and reports the 16th, 50th, and 84th percentiles. You get an interval, not false precision.

Testing then deleting my own theory

I started with a phenomenological “correction” to trying to use binding energy to radiation pressure mass radius for atmosphere volume height etc Then I tested it against real catalogs. It failed negative R-squared on one dataset, a coupling parameter that only worked if I let it change sign per situation, cross-checks that did not hold. So I threw it out published relations, and I went back through the committed system write-ups and stripped the framework in favor of actual escape physics.

That was the moment the project became more than puff piece. I was 20 systems and why waste good data I put the physics and the data on a pedestal, because feelings, beliefs, and confidently wrong AI have replaced the scientific method for the majority. Building the provenance system forced me to be honest about which was which.

Source and Appreciation

NASA Exoplanet Archive TAP service, with JWST transmission spectra from MAST for supported targets.
Thank you to the dedicated hardworking scientists and engineers who keep asking the best questions and producing the most clever solutions.

To the Reader

Thanks for checking it out. I got carried away because I thought it was cool, and I shared it for the same reason. I hope you enjoy it, find it useful, and learn, as much as I did building it for you. If you want to help, whether that is a fix, a dataset, or just a bug report, you become officially a non-zero participant in a brighter tomorrow.

P.S.
If you like the web app and want to throw me a bone,