Valuable information about North Korea's nuclear test comes from seismological recordings of the shock waves produced in the event.One of the interesting things is that because the North Korean nuclear tests are basically done in the same place, seismograms of the events at long distances have quite similar characteristics. Their amplitudes, however, do vary with the increasing yield of the events.
Here are a few links I have come across:
- Physics Today commentary
- IRIS Special event page on the event
- IRIS Recent Earthquake Teachable Moments on the event
- US Geological Survey report on the event
- NORSAR records of the last three DPRK nuclear tests (2006, 2009, and 2013)
- Comparison for the nearest real-time broadband seismometer (in China) for 2009 and 2013 nuclear tests (by Andy Frassetto, IRIS)
How do we know it was an explosion and not an earthquake? First of all, that part of North Korea is not known for historic earthquakes. Secondly and more to the point, with seismometers arrayed around the source area regionally and globally, the explosion will cause all of the first motions to be AWAY from the source. If it is an earthquake, two quadrants will be pulled IN towards the source and two quadrants will be pulled out ("double couple").
--disclaimer: this is not my area of expertise, but I do find it fascinating. To some degree, monitoring of nuclear tests using seismology has driven the development of seismology and seismic networks and that lets us learn a lot more about earthquakes globally.