Note the mushroom-like veil of cloud that forms and propagates very rapidly away from the explosion. This veil of clouds is moving much faster than the smoke generated by the explosion and quickly moves out of view. If you watch and listen to the video carefully, you'll notice that the cloud passes out of the view of the camera before the sound of the explosion is recorded. Thus, the cloud appears to be moving at speeds at or above the speed of sound.
Phenomena of this type are outside my normal areas of study but I suspect that the cloud is being produced by the shock or blast wave generated by the explosion. Such a wave is characterized by a sharp increase in pressure at its leading edge followed by a decrease in pressure to sub-ambient values. These pressure changes occur in less than a second as the wave passes, with the cloud forming in the area of lower pressure due to expansional cooling. Shock waves move faster than the speed of sound.
If you look carefully at the footage at the moment of explosion, you can also see some very low stratocumulus clouds. Thus, I suspect the relative humidity was fairly high, which helped enable cloud formation. In a drier airmass, the drop in pressure probably would not have produced the cloud.
It's also interesting that you can see smoke rising before the expansion cloud appears. I don't know if this might indicate that it was not the first but a second explosion that produced a shock wave sufficient to generate a cloud.
Please comment if you have some background in explosives or fluid mechanics and can add to this discussion. As a meteorologist, I typically deal with phenomenon acting on time scales of minutes to days rather than microseconds.