# Benchmarking Substellar Evolutionary Models Using New Age Estimates for HD 4747 B and HD 19467 B

## TL;DR

This study benchmarks substellar evolutionary models using new age estimates for two brown dwarf companions, revealing significant luminosity and mass prediction discrepancies and highlighting the impact of cloud physics.

## Contribution

It provides new age estimates for HD 4747 A and HD 19467 A and evaluates the accuracy of SSEMs against these benchmarks, identifying model limitations.

## Key findings

- SSEMs under-predict luminosities by ~0.75 dex for HD 4747 B and ~0.5 dex for HD 19467 B.
- SSEMs over-predict masses by ~12% and ~30% for the two brown dwarfs.
- Including cloud effects reduces luminosity under-prediction to ~0.6 dex and mass over-prediction to ~8%.

## Abstract

Constraining substellar evolutionary models (SSEMs) is particularly difficult due to a degeneracy between the mass, age, and luminosity of a brown dwarf. In cases where a brown dwarf is found as a directly imaged companion to a star, as in HD 4747 and HD 19467, the mass, age, and luminosity of the brown dwarf are determined independently, making them ideal objects to use to benchmark SSEMs. Using the Center for High Angular Resolution Astronomy Array, we measured the angular diameters and calculated the radii of the host stars HD 4747 A and HD 19467 A. After fitting their parameters to the Dartmouth Stellar Evolution Database, MESA Isochrones and Stellar Tracks, and Yonsei-Yale isochronal models, we adopt age estimates of $10.74^{+6.75}_{-6.87}$ Gyr for HD 4747 A and $10.06^{+1.16}_{-0.82}$ Gyr for HD 19467 A. Assuming the brown dwarf companions HD 4747 B and HD 19467 B have the same ages as their host stars, we show that many of the SSEMs under-predict bolometric luminosities by $\sim$ 0.75 dex for HD 4747 B and $\sim 0.5$ dex for HD 19467 B. The discrepancies in luminosity correspond to over-predictions of the masses by $\sim$ 12\% for HD 4747 B and $\sim$ 30\% for HD 19467 B. We also show that SSEMs that take into account the effect of clouds reduce the under-prediction of luminosity to $\sim 0.6$ dex and the over-prediction of mass to $\sim 8\%$ for HD 4747 B, an L/T transition object that is cool enough to begin forming clouds. One possible explanation for the remaining discrepancies is missing physics in the models, such as the inclusion of metallicity effects.