Stellar mixing length theory with entropy rain

TL;DR

This paper extends stellar mixing length theory by incorporating a non-gradient flux term from entropy rain, explaining small-scale flows and downdrafts in stellar convection zones, and analyzing their impact on convective structures.

Contribution

It introduces a modified MLT including the Deardorff flux term, modeling nonlocal heat transport and small-scale convection structures in stars.

Findings

Entropy rain can cause upward enthalpy transport in stable layers.

Fewer, thicker downdrafts are most plausible at greater depths.

Standard large-scale flow structures may not be explained by the modified theory.

Abstract

The effects of a non-gradient flux term originating from the motion of convective elements with entropy perturbations of either sign are investigated and incorporated into a modified version of stellar mixing length theory (MLT). Such a term, first studied by Deardorff in the meteorological context, might represent the effects of cold intense downdrafts caused by the rapid cooling in the granulation layer at the top of the convection zone of late-type stars. Such intense downdrafts were first seen in the strongly stratified simulations of Stein & Nordlund in the late 1980s. These downdrafts transport heat nonlocally, a phenomenon referred to as entropy rain. Moreover, the Deardorff term can cause upward enthalpy transport even in a weakly Schwarzschild-stably stratified layer. In that case, no giant cell convection would be excited. This is interest in view of recent observations, which could be explained if the dominant flow structures were of small scale even at larger depths. To study this possibility, three distinct flow structures are examined: one in which convective structures have similar size and mutual separation at all depths, one in which the separation increases with depth, but their size is still unchanged, and one in which both size and separation increase with depth, which is the standard flow structure. It is concluded that the third possibility with fewer and thicker downdrafts in deeper layers remains most plausible, but it may be unable to explain the suspected absence of large-scale flows with speeds and scales expected from MLT.