Model dependence is confined to the white-dwarf mass step: cooling tracks with thick hydrogen envelopes (MESA models); the paper argues a thin-envelope alternative would require anomalous fine-tuning and a ~20 Myr cooling age inconsistent with the system. Timing: 8,121 TOAs from GBT, Arecibo, and Effelsberg; assuming GR, the intersection of q and orbital decay gives the consistent M_PSR = 2.07 +0.20/-0.21 Msun.
Methodology
Two independent mass channels on PSR J0348+0432 (39-ms pulsar in a 2.46-hour orbit with a white dwarf): (1) phase-resolved VLT spectroscopy of the companion (K_WD = 351 +/- 4 km/s; Teff = 10120 +/- 47 +/- 90 K; log g = 6.035 +/- 0.032 +/- 0.060) combined with radio timing (K_PSR = 30.008235 km/s) gives mass ratio q = 11.70 +/- 0.13; white-dwarf model atmospheres plus cooling tracks (MESA, thick hydrogen envelope) give M_WD = 0.165-0.185 Msun (99.73% CL), hence M_PSR = q x M_WD = 2.01 +/- 0.04 Msun, independent of Shapiro delay. (2) Radio timing with GBT, Arecibo, and Effelsberg (8,121 TOAs) measures the orbital decay; assuming GR, the intersection of q and Pb-dot gives a consistent M_PSR = 2.07 +0.20/-0.21 Msun.
Link to original