# How does a planet lose mass to its host star?

How do bodies like WASP-12b lose mass to its central body? The process is never really explained in popular media, instead showing visualisations of a thin veil of gas being "sucked" in a rather straight line (and then into a ring around the star) rising straight from the surface of the planet. I highly doubt the mechanics which these illustration insinuate: Gravitation isn't working like a vacuum cleaner on a dusty rug. Either the rug goes with it, or the dust stays on the rug. The only mechanism that I can think of are particles from the upper atmosphere which happen to pick up enough thermodynamic energy so that they are faster than the escape velocity of the planet. But is this a process which effects a big enough mass transfer to be of significance? Do I overlook something? How is such a mass defect going on really?

• There is no one answer. You need to narrow your question down to particular scenarios. e.g. red giants lose mass through winds, some of which is captured; close binaries lose mass through Roche-lobe overflow; exoplanets can lose their atmospheres through heating... Commented Sep 25, 2017 at 16:33
• It sounds like OP is asking specifically about Roche-lobe overflow. Commented Sep 25, 2017 at 17:43
• @RobJeffries narrowed the question down to the planet/star scenario. Commented Oct 4, 2017 at 13:18
• @Phiteros narrowed the question down to the planet/star scenario. Commented Oct 4, 2017 at 13:18

$\frac{r_1}{A}=.46224\sqrt[3]{\frac{M_1}{M_1+M_2}}$ for $\frac{M_1}{M_2}<.8$ (which it is)
where $A$ is the orbital separation and $r_1$ is the radius of the Roche-lobe around $M_1$.
Solving for $r_1$ we get $97788\pm 2318$ miles. The planet's radius is $77759 \pm 3909$ miles. Taking the best possible scenario, there would be $\approx$ 13,000 miles between the planet's surface and its Roche-lobe if the planet was a perfect sphere. It is unknown how much the star affects the planet, but it would not be a stretch to think that the major axis of the planet would be 13,000 miles longer than the minor axis, allowing for the planet's gas to be shoved out of the Roche-lobe and "sucked" into the star.