,
25 tweets,
3 min read
Read on Twitter
A short thread on why Monday's gravitational wave detection is so devastating to modified-gravity models of Dark Matter like MOND. 1/
It's really easy to come up with a modified gravity model which explains many things about the dynamics of galaxies and clusters. 3/
For a simple example, imagine that the gravitational potential energy as a function of radius r varies as log(r) instead of 1/r. 4/
This very simply explains flat galactic rotation curves like this plot of orbital velocity as a function of radius in galaxy M33. 5/ 
It also (sort of amazingly) explains the Tully-Fisher relation between the luminosity of a galaxy and its rotation velocity. 6/
It does a very good job with dynamics of galaxies in clusters as well. 7/
But there's a problem. If you change the force law between massive objects only, then (massless) photons will not be affected. 9/
This is completely ruled out by gravitational lensing measurements like this famous picture of the "Bullet Cluster". 10/
The pink regions show the distribution of atoms. The blue regions show the distribution of mass inferred from gravitational lensing. 11/
They don't match. 12/
All the lensing measurements we have give the same answer: the photons are being anomalously bent by the same force felt by the matter. 13/
So our simple log(r) theory fails. 14/
But people have come up with more complicated "modified gravity" theories to explain why the photons are bent in just the right way ... 15/
... to match the motion of massive objects, and reproduce both dynamics and lensing. 16/
This also means that propagation of photons through regions with mass in them will be *slowed down*. This is called the Shapiro Delay. 17/
Key point: in models of this type, gravitons don't feel the "extra" force, and aren't slowed as much as the photons. 18/
So we should see the gravitons from the neutron star merger before we see the photons. 19/
The photons did in fact arrive a little later: the gamma-ray burst was seen two seconds after the gravitational wave signal. 20/
This puts an upper bound on the difference in propagation speed of gravitons and photons of one part in 10^15 (ten to the fifteenth power)!
If MOND-like theories are to explain both dynamics and gravitational lensing, the photons would be slowed much more. 21/
Boran, et al. calculated the expected photon delay in these theories: it's 445 days instead of two seconds. 22/
Not even close. This is a convincing blow against any model for which photons and gravitons travel on different geodesics. 23/
Science. It's what's for breakfast. 24/24
