- Model a galaxy's 3D mass distribution
- raw surface brightness --> mass pixels --> mass voxels
- W/m^2 --> kg/m^2 --> kg/m^3
- Try various gravity formulas
- Compare a_pred to a_obs
index.html- See 3D mass model of a galaxy (loads random galaxy on refresh)
- Load a specific galaxy by /index.html?name=NGC5005
dashboard1.html- Check quality of 3D mass model by comparing sim a_classic to SPARC-authors-claimed a_classic (from their reported v contribution, eq 1 https://arxiv.org/pdf/1803.00022)
dashboard2.htmldashboard3.html- Compare many a_gravformula to a_obs
See code for details. Informally
- Disk
- (r, surface brightness) --> mass pixels --> interpolation --> z height fx estimate --> mass voxels
- Bulge
- (r, surface brightness) --> sphere model --> mass voxels
- Gas
- (r, reported v contribution) --> (r, accel) --> nonnegativeleastsquares fit for radialsurfacedensity --> mass pixels --> z height fx estimate --> mass voxels
(Substantially aided by a GPT (Gemini 4))
SPARC: Mass Models for 175 Disk Galaxies with Spitzer Photometry and Accurate Rotation Curves https://astroweb.case.edu/SPARC (Table2.mrt)
- Classic: a = GM/r^2
- MOND (Milgrom): a = GM/r^2 / (1 - e^(-sqrt((GM/r^2) / (c^2/87Gly))))
- Conformal (Mannheim): a = GM/r^2 + k1 M + k2 + k3 r_galaxycenter, with k1 ≈ 1.2e-52 m/(s^2 kg), k2 ≈ 1.4e-11 m/s^2, k3 ≈ -8.6e-33 1/s^2
- Polynomial: a = GM/r^2 + k1 GM/r + k2 GM + k3 GMr + k4 GMr^2, with k1 ≈ 5.4e-21 1/m, k2 ≈ -5.8e-42 1/m^2, k3 ≈ 3.1e-63 1/m^3, k4 ≈ -5.0e-85 1/m^4
- Accel-dep term: a = GM/r^2 (1 + (c^2/87Gly) / a), and solve w quadratic formula
- Doesn't fit observed accelerations
- Doesn't produce a_total in galactic mass scene when near planet. "low-accel requirement" is a hack.
- k2 direction is hardcoded. r_galaxycenter (note, it's not r) is a hardcode.
- GMr and GMr^2 terms mean huge accel and tidal accel from a galaxy 2Mly away.
- Doesn't produce accel boost to galaxycenter when near planet