Is it possible to have two separate light sources generate coherent light at different locations? for example,   me and my brother are standing apart. we have watches and we have agreed to generate coherent laser at a particular time.
is it possible for us to generate coherent photon? 
 A: Yes.
These three parameters ensure that the two light beams generated at the two sources are coherent, hence they will constructively or destructively interfere into a stationary (spatially or temporally constant) interference pattern.
(1) you have the a constant phase difference
Temporal and spatial coherence lengths quantify regions or intervals over which the phase relation is fixed. In this regions you will observe interference.
Changing the phase of the two beams means shifting the two sinusoidal fields, and they would not line up exactly to destructively interfere. So a random phase noise on your laser beam phases would wash out the interference pattern. 
Experimentally this could be implemented by a PPL loop between the two sources.
(2) you are producing the same frequency
Well frequency $f = \mathrm{d}\, \phi / \mathrm{d}\,t$, so a fixed phase also makes sure the frequencies have a constant relation.  They would need to be the same, however, in order for the minima and maxima of the field to be spaced $(\propto \lambda \propto 1/f)$ equally, and to allow perfect cancellation (destructive interference).
Experimentally, this is guaranteed by the lasers emitting the same frequency. 
(3) You produce the same waveform, i.e. the shape of the signal
I.e. they are both square waves, or both cosine/sines.  If you use the same laser in both locations, this would be guaranteed. 
This is needed in order for the waves to be able to match up and to perfectly cancel each other out (destructive interference).
Extra details for fun:


*

*If you superimpose enough distinct laser frequencies together at different powers, you are effectively making a Fourier series and the shape of the waveform could become quite weird - square, pulses etc.   This would happen with or without coherence, since in the latter case you'd just be adding intensities, that also oscillate sinusoidally.

*While requiring the same waveform guarantees interference in the propagating direction ("forward"), it does not tell you much about its extent along the radial direction, i.e. how large the interference region will be in the perpendicular direction to that of propagation.   For this to be infinite, you need the wavefronts (the surface, perpendicular to the direction of propagation, where all points have the same phase) of the laser beams to be the same.  Experimentally this is guaranteed if your laser beams are in the same mode when they interfere. Which means they were either generated in the same mode, or all the mode distortions from air, optical components and non-linearities during their travel cancel out.
Single photons
All I said above is correct for continuous waves... Single photons are not really my area of expertise but I don't see why it should be different.
However you will see quantum effects intrinsic to the indistinguishability (or not) of photons, for example.
But again I cannot say much about this, maybe someone can expand on my answer in the comments and I can edit it later.
