Earlier quoted context omitted.
In this case, the atmosphere is essentially a thin (compared to the radius of the Earth) two dimensional sheet wrapping the Earth, so it's probably valid to think of the sound as propagating on a surface. There wouldn't be much loss of energy to space. Would the atmospheric interface with the ground reflect or absorb? The divergence is probably even less that 1/r because the Earth is roughly sperical, and the sound w…
Would gravity affect soundwaves at all? If sound is traveling parallel to the Earth's surface, will it follow along the curvature of the Earth (implying gravity has an effect on it) or was it able to propagate around the Earth four times simply because the intensity was so great that it was able to still be heard despite the loss of energy in the "up-down" directions relative to the origin of the sound? If I had to c…
The radius of the Earth is roughly 6000 km, while the thickness of the atmosphere is about 120km. Thus, the ratio of the two heights is about 1/60. Because the ratio of atmosphere thickness to terrestrial thickness is so small, for the purposes of modeling wave propagation through the atmosphere we can approximate the atmosphere as a thin spherical shell around the Earth.
Furthermore, the Earth's surface forms a hard boundary below the shell, further confining the sound to the atmosphere, while the force of gravity also prevents the sound wave from escaping into space via the atmosphere.
As other posters have stated, this forms a 2-dimensional propagation layer which results in a linear falloff rather than an inverse square falloff.