In 2012, North Korea announced in advance launch of the third Unha carrying the Kwangmyangsong 3 satellite. There are some very distinct differences between the two launch profiles.

NOTE - Google Earth users can cut and paste the coordinates from following paragraphs in order to 'fly to' the approprate area.

Press reports at the time and various web sites report that the launch was from the Tonghae launch centre and give a location in the north of the country. Other sources place Tonghae much further south. A search in Google Earth produces a result for Tonghae and there seems to be a launch pad there. It can be found near coordinates 37°35'38"N, 129°5'5"E.

This pad is built quite near to a major highway that can be seen cutting across the lower left hand corner of the picture. the road was, presumably, built after the the launch pad and may indicate that the pad is no longer used.

It is not the Unha 2 launch site.

NOTE - thanks to Martyn Williams of North Korea Tech for pointing out that this particular Tonghae is, in fact, in South Korea, which would most likely also explain the highway! It is most definitelty not a North Korea launch site and the structure may not be a launch tower.

In the far north of the country is another location, sometimes called Musadan Ri. Its location fits the known trajectory of the second Unha rocket.

Google Earth shows two launch pads. The one from which Unha 2 probably lifted off can be found near 40°51'21"N, 129°39'57"E.

The second pad is about 1 km to the north west and can be found by following the roads in Google Earth. From both pads, the roads lead to what is probably an assembly and clearing area for missiles and rockets..

North Korea announced at the start that the Kwangmyangsong 3 satellite was to be launched from the Sohae (West Wind) site located in the north west of the country.

Again, Google Earth helps greatly in that it shows the launch pad, located near 39°39'36"N, 124°42'19"E. As with the Musadan Ri site, there are other buildings and work areas nearby that resemble parts of a rocket/missile assembly and launching complex.

At Sohae there is a greater amount of building so it may be a more-permanent launch site.

North Korea released information on where the first and second stages of each of the Unha 2 and Unha 3 rockets were expected to fall. Those for Unha 2 were in international waters but straddled Japan while Unha 3's were over international waters but no part of the implied trajectory took the rocket over any other countries. Japan's reaction to the earler launch is a partial explanation of why North Korea was building the Sohae site in order to avoid any possibility of rocket/missile components falling on another country. Certainly if the same route were used now, Japan indicates it would have no compunction in shooting down the rocket.

The main differences between the two trajectories are the dowrange distances to the impact zones and the sizes of the zones. Compared with Unha 3, the Unha 2 first stage zone is narrower but three times the length, indicating a greater uncertainty in the range of velocities at shutdown. The second stage zone for Unha 2 is both wider and longer, indicating again a greater expected potential error, not only along track but across track too.

There is a slightly different presentation of the Unha 3 ground track here.

It raises an interesting question of whether the smaller impact zones for Unha 3 genuinely reflect greater confidence in the accuracy of the launch vehicle or whether they have been squeezed into the areas of international waters available!

In comparison:

There are multiple ways of interpreting these variations depending on whether the ascent trajectories of stages 1 and 2 are similar between the two missions, or if there is a fundamental difference.

The shorter ranges from launch site to impact zones could mean that Unha 3 will be travelling more slowly than Unha 2 when the third stage is released, in which case stage three would seem to require more power to reach orbit than its predecessor.

A second possibility is that Unha 3 will follow a more 'lofted' trajectory, gaining height more swiftly than Unha 2, resulting in stage 3 separation occurring nearer the launch site but at a greater altitude.

Nowadays a 500 km circular orbit would usually be achieved by using a re-start of the third stage after coasting for about half a circuit of the Earth in an elliptical orbit. This is unlikely to be the aim with Unha. It would be a very ambitious manoeuvre and is unlikely to be possible with the existing third stage.

Starting in the early 1960s, the Soviet Union frequently launched meteorological satellites and electronic intelligence gathering vehicles to 650 km circular orbit using a steep ascent directly to orbital altitude. This may be what North Korea is planning to do with Unha 3 for the Kwangmyongsong 3 launch. It is one possible explanation of the relatively short distance from the launch site to the first and second stage impact zones.

There is more to come in the Unha/Kwangmyongsong saga but clarity will only really come with a successful orbital mission.