I believe that the Particle SOP is using a basic Perlin noise internally.
With a bit of tweaking, the closest I could get to replicating basic turbulence in the Particle POP was by using Curl 3D noise combined with some velocity damping on the particles.