I'm betting that a pile of explosives has a different seismic signal from a nuke. Here's why.

A nuke contains a small amount of conventional explosives that concentrate tremendous, focused compression waves on the fissile mass in order to get it super-critically dense. The chemical explosion is initiated with high-precision timed detonators and is followed almost instantaneously by an extremely violent, initially relativistic expansion of the nuclear blast.

Chemical explosives have a burn rate measured in feet per second. IIRC high explosives burn at something over 6, 000 m/sec. So let's guess NK piled 1,000 - 4,000 tons of high explosives into a cavern. TNT is 1.6 g/cc =6,000 m^3 (good God, is that right?). So to cross from the center of the pile (assume sphere to keep me from going nucking futs) to the outside is 77 meters so it should take what, a millisecond, more or less? I'm ball-parking here so check my math.

Contrast that with the nuke expansion of 300,000,000 m/sec. In a millisecond the shockwave moves 300,000 meters.

So a seismic analysis should show an energy progression of approximately a millisecond for the chemical explosion and an instantaneous rise for the nukular explosion.

Ha, call me Edward Teller!

B