It's hard for me to say without reading the actual research article, but my guess is that temperature is referring to energy of each of the photons relative to one another. Meaning, if you were shining a laser beam down the z axis, typically the photons would also have some small velocity in the x and y directions (this would mean the light beam is diverging). So looking in a frame of reference moving along the z axis at the speed of the photons in the z direction, you would notice that most of the photons would be stationary in the z direction but are slowly moving in the x and y directions. The temperature is probably referring to the average kinetic energy of the photons in the x and y directions. So in order to get the photons all into a single state, you would want their x and y velocities to all be zero. That's my guess, I know sometimes temperature can refer to similar scenarios with a beam of atoms. Like with the LHC, they often refer to the atoms as having a very low temperature. However, the atoms are moving around in circles at very nearly the speed of light. With such a high kinetic energy, you would expect the atoms to have a high temperature. But really, the atoms are moving very slowly relative to one another, and thus temperature is referring to this kinetic energy.