Tuesday, 5 December 2017

The Curious Incident of the ICBM that Launched by Night

image: KCNA

North Korea conducted a test launch of a new ICBM, the Hwasong-15 (KN-22)  on 28 November 2017. The launch was at 18:17 UT from a field just north of Pyongsong, not far from Pyongyang. It was a "lofted" test, reaching an incredible 4475 km apogee before coming down near Japan, 950 km east of the launch location and some 250 km out of the Japanese coast. It is a beast of a mobile launched ICBM:

images: KCNA

After the launch, North Korea's KCNA press/propaganda agency published several pictures, showing Kim Jung Un directing the readying of the TEL with missile, and the launch.

Several of these images, both from the missile erection sequence before launch and the launch itself, show stars. As part of the verification of a geolocation attempt, Jeffrey Lewis (@armscontrolwonk on twitter) of the Middlebury Institute of International Studies, a well known wonk of the North Korean (and other) missile program, asked me to look into these starry backgrounds. Could I say something about image orientations?

I could, and it became very interesting. I initially looked at and measured these two pre-launch images (Jeffrey provided me with high-res versions of these: the ones shown here are the low-res versions from the KCNA website):

images: KCNA

These two images appear to be real (although, given what I will point out below, all images remain suspect, because they clearly aren't all real. With "real", I mean "untampered with" here). I used them to determine azimuth directions and the Local Sidereal Time (and from that UTC time) these images apparently were taken, by creating an astrometric grid over the image. In the image below, each dotted star is a reference star measured. The two images below it show the reconstructed azimuth range for each picture.

Of course, I now have reason to doubt the validity of this whole exercise. Because (hold on):

The real fun started when, yesterday evening, I started to look at the pictures of the actual launch moment. The fact that some of these show stars in itself is already something, as these images necessitate short exposures (unlike the pre-launch images above, which are long duration exposures), so you do not expect stars. But the real fun came when I looked at these stars visible. There, things clearly were not right!

Take these two images, which I have put next to each other for comparison:

Image: KCNA

The shape of the exhaust cloud and exhaust flame (and the number decal, extremities and paint job on the missile) clearly indicate they were taken from the same viewpoint, probably within a fraction of a second of each other. But take a look at the stars in the background: these then should show the same sky area, right?

But they don't!

One shows Orion, which is south-southeastwest. The other shows Andromeda with the Andromeda galaxy (this is a bit more clear in a higher resolution version I have), which is northwest. So these two images from the same viewpoint, show dramatically opposite sky areas.

Below is another example, doing basically the opposite. The mirror character of these two images, from the exhaust shape plume, exhaust flame shape, and the lack of number decal on the missile in one of the images, indicates they were taken from opposite viewing points. So the sky should show opposite sky areas on each of these, right?

images: KCNA
(NOTE: earlier version of image replaced with version correcting error in labelling)

Again, they don't.

The top one shows Orion (but with Betelgeuze missing). The bottom one shows Canis major (but with Sirius missing). Orion and Canis Major are very close to each other, south-southwest at the time of launch. The images should show opposing sky areas, but don't.

So clearly, the starry sky background was added to the imagery and is not original.

So why should North Korea have done this?

The most likely reason is simply that they did it for aesthetics. An ICBM soaring into the stars makes for good propaganda images. They apparently just didn't care enough to do it correctly.

Aesthetics seem to be important in North Korean propaganda pictures. They frequently photoshop the ears of Kim Jung Un in pictures, for example.

Or maybe they wanted to play a prank on analysts as well: they know these images will be analyzed by the west. Fooling around with clues as to the orientation of images makes it harder to glean information from them on 3-dimensional missile shape, and launch site geolocation.

To be clear (because some hare-brained individuals on social media seemed to think that was implicated, even though it nowhere is): nobody disputes that the launch was real.

It was, and it shows that North Korea now has an ICBM that can reach the whole US mainland. It can even reach my country, the Netherlands (although we are not a very likely target, I must ad):

But at least some of the launch images have been clearly doctored, and show elements that were added later. This is, of course, something we have seen earlier with North Korean propaganda images (some examples are given in the CNN article linked below).

End note: I still think the pre-launch imagery showing the TEL with missile being erected in the field is unaltered. So maybe my azimuth and timing determinations from these is are valid. But I cannot proof it, and of course now all the images must be considered suspect.

They do tally with some other evidence though, including this still I extracted from the KCNA released video, which briefly offers a glimpse of the moon, low in the west-southwest. If the video isn't doctored as well (!), this should be around 16:00-16:15 UT, some two hours before launch:

Still taken from KCNA video: moon indicated
If the video was not altered, then together with earlier images showing the missile on the TEL leaving the plant, this orientation means that any of the launch pictures showing the number decal on the missile should actually be looking Northeast to East.

My findings on the doctoring of the launch imagery featured in this CNN article from 5 December 2017 by Joshua Berlinger:  North Korea missile: Inconsistencies spotted in Hwasong-15 images

"Yay! We fooled those imperialistic coward dogs!" (image: KCNA)

Sunday, 19 November 2017

Introducing TLE from Proxy

A simple way to estimate orbital elements for an upcoming launch, is to use elements from a previous, similar launch as a proxy.

For example, for a newly to be launched SpaceX DRAGON cargo spacecraft to the ISS launching from Cape Canaveral pad 39A, you can use a previous DRAGON launch from Cape Canaveral and then modify the elements to the new launch date and launch time. The method is described here by Ted Molczan in a Seesat-L mailinglist post from June 2002.

Basically, the method takes the elset from the previous launch and adjusts the epoch and RAAN values (all else being kept equal) based on the time difference between the original launch and the new launch.

To aid in this and make it as simple as a few buttonclicks, I have written TLE from Proxy. The program runs under the Windows .NET framework, and can be downloaded on my website.

Using the program is very simple, involving these five simple steps:

  1.  Obtain a TLE for a previous similar launch from Space-Track;
  2.  Paste line 1 and line 2 of this elset into the input box;
  3.  Fill in the date and time (in UT) of that launch;
  4.  Fill in the date and time (UT) of the new launch;
  5.  Press the button.

A new TLE will now be generated.

Note that in order for this to work, the launch must be from the same launch site, towards the same launch azimuth, and with a same launch-to-destination scenario.

Friday, 17 November 2017

[UPDATED] Tomorrow's SpaceX Zuma launch

click map to enlarge

If nothing interferes (the launch has been postponed twice already), SpaceX will launch the classified Zuma satellite from Cape Canaveral Pad 39A in the early hours (UT) of  November 18.

Zuma  was originally scheduled for November 16, but was delayed a day to November 17, and then yet another day to November 18.

The published Maritime Area Warnings give a window from 00:55 to 03:37 UT for the launch. From the Area Warnings, the de-orbit of the Falcon 9 Upper stage happens some 2 hours after launch over the southern Indian Ocean, during the 2nd orbital revolution.

The launch and Upper stage de-orbit hazard zones (I plotted them in red on the map above) strongly suggest a launch into a 50-degree inclined, ~400 km orbital altitude Low Earth Orbit.

The map above plots the trajectory for the first ~1.5 revolutions in such an orbit. As can be seen in the map, such an orbit lines up well with the direction of the launch hazard zones, and with the Falcon 9 upper stage de-orbit hazard zone in the Indian Ocean. The fact that the first stage will return to the Cape for a landing argues for a launch into Low Earth Orbit too.

If a ~50-degree inclined, ~400 km altitude orbit sounds familiar to you: that is because this orbit would be very similar to that of the enigmatic classified satellite USA 276 which was launched - also by SpaceX - in May 2017. This is the one that made all those peculiar close approaches to the ISS in June (see some previous posts from June and my Space Review article here). Perhaps, but this is pure speculation based on suspected potential orbital similarities only, Zuma is up for a similar mission.

It is very interesting that Zuma seems to have been contracted via a similar procedure as USA 276, and that like for USA 276, it has not been made public which Agency will operate the Zuma satellite. So there appear to be similarities from that aspect as well.

It will therefore be interesting to see how the orbit of Zuma, once launched, compares to that of USA 276 and the ISS. The orbital plane of the ISS will be overhead for Cape Canaveral near 2:38 UT on the 18th, so a launch exactly into the ISS orbital plane is possible - and will stay possible for several days to come in case the launch is postponed again (the moment of the ISS orbital plane passing over the Zuma launch site happens ~24 minutes earlier each day).

On the 18th, the orbital plane of USA 276 will be overhead for Cape Canaveral some 10 minutes before the launch window opens. With the newest delay, a launch exactly into the orbital plane of USA 276 is therefore no longer feasible.

But by launching directly at the opening of the launch window on the 18th, the orbits of Zuma and USA 276 would nevertheless still be quite close (launch at 1:00 UT would result in a difference in RAAN of 3 degrees), and differential rates of precession of the RAAN might still slowly drift the two orbits towards each other over the next weeks and months, depending on what the actual orbital altitude and inclination Zuma ends up in would be.

Therefore a launch exactly into the orbital plane of either USA 276 or the ISS, strictly speaking is not necessary to engineer close approaches (indeed, USA 276 itself was not launched exactly into the ISS orbital plane in May).

So it might be worth monitoring Zuma and its behaviour in relation to both USA 276 and the ISS in the weeks after launch. Still, it is also very well possible that Zuma has nothing to do with both spacecraft whatsoever.

UPDATE 1  17 Nov 2017, 13:00 UT:

The  maps below show a comparison of the hazard zones (from Maritime Area Warnings) for the launch of USA 276 in May 2017, and for Zuma.

click maps to enlarge

The USA 276 de-orbit area is shifted more West-wards, because the Falcon 9 upper stage de-orbit from that launch was de-orbitted one orbital revolution later than apparently planned for Zuma. The small difference in size might point to slightly different orbital altitudes for the upper stage (e.g.due to  a somewhat different collision avoidance manoeuvre after payload separation)

UPDATE 2  17 Nov 2017, 13:00 UT:

SpaceX has released a statement that, while not taking a launch tonight off the table, might indicate a further prolonged delay.


These are the Area Warnings published for the launch. They are graphically depicted in the map in the top of this post and the two maps above.

NAVAREA IV 1067/17

160055Z TO 160337Z NOV, ALTERNATE 
170055Z TO 170337Z NOV IN AREAS BOUND BY: 
A. 28-38N 080-43W, 29-12N 080-06W, 
30-04N 079-00W, 29-56N 078-52W, 
28-41N 080-10W, 28-26N 080-21W, 
28-22N 080-38W. 
B. 30-04N 079-00W, 30-52N 
078-17W, 31-32N 077-25W, 
31-54N 076-49W, 31-49N 076-45W, 
31-36N 076-57W, 30-44N 077-53W, 
29-56N 078-52W. 
2. CANCEL THIS MSG 170437Z NOV 17.// 

Authority: EASTERN RANGE 072156Z NOV 17. 

Date: 110428Z NOV 17 
Cancel: 17043700 Nov 17 

HYDROPAC 3895/17 

DNC 03, DNC 04. 
160300Z TO 160637Z NOV, ALTERNATE 
170300Z TO 170637Z NOV IN AREA BOUND BY 
30-27S 064-51E, 30-44S 067-03E, 
38-10S 082-43E, 47-22S 108-39E, 
50-30S 124-39E, 51-55S 126-03E, 
53-32S 125-05E, 54-24S 116-01E, 
53-34S 101-27E, 47-46S 082-05E, 
39-58S 069-31E, 31-56S 063-23E. 
2. CANCEL THIS MSG 170737Z NOV 17.// 

Authority: EASTERN RANGE 072155Z NOV 17. 
Date: 110407Z NOV 17 
Cancel: 17073700 Nov 17

Friday, 3 November 2017

Introducing IOD Entry: software to aid observers in creating IOD formatted observational data [UPDATED]

International amateur satellite observers (well: apart from the British, who use their own format) generally use the IOD format to communicate positional measurements on satellites. The IOD format however can be cumbersome and error-prone to manually write.

In the old days, there was a neat little DOS program called ObsEntry to help you turn your data into IOD format. Unfortunately, this no longer works on newer Windows machines.

Time for something new to replace it: so I present to you IOD Entry 1.0!

IOD Entry 1.0 is software that runs under the Windows .NET framework, which is a standard component of Windows 7 and later (otherwise, the .NET framework can be downloaded here). I wrote it in Visual Basic using Microsoft Visual Studio 2017, as part of self-teaching me to code .NET windows applications in Visual Basic.

The program and how to work with it is described in detail in this Satobs.org mailinglist post. The program can be downloaded as a .zip file through my astronomy software page at http://software.langbroek.org. It is (of course) freeware.

UPDATE: IOD Entry version 1.1 has now been released. It allows to choose the format of both the Right Ascencion and Declination entries. For the RA, the choice is between HH MM SS.s, or decimal degrees. For the declination, the choice is between degrees, arcminutes and arcseconds, or decimal degrees.

Version 1.1 can be downloaded at the same link above.

Wednesday, 4 October 2017

The Space Age turns 60

click to enlarge
Today at 19:28:24 UT, it is exactly 60 years ago that Sputnik 1, the first artificial satellite of the Earth, was launched from Tyuratam (Baikonur) in the Soviet Union. Thus, the Space Age was born.

Sputnik 1 would stay in orbit for  four months and then re-entered into the atmosphere. The image above shows the very first two orbits (and is based on a TLE from the website of Jonathan McDowell. Strictly speaking, this TLE is for the upper stage of the rocket, but during the first few orbits the orbits of both objects were very similar).

Monday, 2 October 2017

The North Korean Hwasong-12 test of 14 September: georeferencing Kim Jong Un's map

(This post has been a while in the making. Due to various reasons, including the 21 Sept fireball, I did not come to write it up on this blog until now. The same analysis however was earlier presented as a Twitter-thread, on 23 September)

"I love the smell of UDMH in the morning..." (photo: KCNA)

Over the past months, I have spent a number of posts on North Korea's recent increasingly bold missile tests. The latest of these tests happened on September 14 (local September 15 in North Korea). At 21:57 UT that day, North Korea launched a Hwasong-12 IRBM from Pyongyang Sunan airfield towards the east-northeast. It crossed over Japan (causing air-raid alarms to go off) with, according to western military sources, a range of ~3700 km and apogee at 770 km, before impacting in the Pacific Ocean. The test was the second one to launch the Hwasong in a 'regular' rather than 'lofted' trajectory, over Japan.

I was abroad at the time of the test,  meaning it was several days before I could look at this test and start an analysis. Some results were earlier presented on 23 September as a Twitter-thread.

photo: KCNA

As part of the propaganda photographs published by KCNA following the test, there was (just like after the test of 29 August) a picture of Kim Jong Un sitting behind a desk, with a map in front of him. As was the case for August 29, the map shows a missile trajectory, presumably the trajectory aimed for in this test:

photo: KCNA

I obtained a high resolution version of this photograph for analysis with the much appreciated help of Ankit Panda. First, I roughly squared this image using the skew-tools in Photoshop:

Next, I georeferenced this image against an actual map using QGIS. The resulting map is the one below (the map projection I have used is a Lambert azimuthal equal area projection centered on Sunan. The grid is WGS84). :

click map to enlarge

The result of this georefencing is that the drawn trajectory starting at Sunan ends at approximately 39.60 N, 168.05 E. The range from Sunan to this impact location is 3615 km.

Note that I have added, in yellow, country outlines to the map to show the validity of the georeferencing. I also added a few annotations. I have added an STK-modelled ballistic trajectory to the map as a dashed red line: it is highly similar to the original trajectory line drawn on the map, as you can see, with apogee position corresponding to what appears to be a text in red on the map: what was depicted on Kim Jong Un's map hence appears to be a real ballistic trajectory.

The resulting range of 3615 km is close enough to the range reported by Western military sources (about 3700 km) to conclude that the 14 September test went much as intended.

This is a difference with the 29 August test. In that case, the map showed a ~3300 km range while western military sources said the actual range flown was ~2700 km: it also flew more north than the trajectory that was depicted on the 29 August map. See my earlier blog post here. The match for the 14 September test might indicate that the 29 August test perhaps did not go as intended.

The map below shows both trajectories: that of 29 August as depicted on Kim Jong Un's map (i.e. not the actually flown trajectory!), and that of 14 September as depicted on Kim Jong Un's map:

click map to enlarge (map by author)

It is interesting to look at the range distances for both tests. The range distances of 3316 km and 3615 km ring a bell. The distance from Pyongyang Sunan (the launch location) to Henderson AFB on Guam is ~3415 km. The August 29 map trajectory falls almost exactly 100 km short of that range, while the September 14 trajectory almost exactly 200 km overshoots this range.

I am going back-and-forth on whether this is significant or not. I tend to think that all these kind of map images released are intentional propaganda meant to convey a message: the map is meant to be analysed and "read" by the West.

As an illustration of how it potentially could be read: would we change the launch directions so that they would be towards Guam, then this (the two black crosses) would be the result in terms of the pattern of impact locations relative to Guam:

Remember that North Korea has "threathened" to conduct a quadruple demonstration with four missiles impacting in international waters in an 'envelope' around Guam, if the Trump administration does not stop B1 bomber demonstration flights from Guam. Do we see some kind of test mock enactment for this here? Is the message: "We are already practising for this, imperialist barking dottard dog!".

It would become very interesting if North Korea were to launche two other missiles the coming month, and the pattern where they (were implied to) land relative to the earlier two tests might be very interesting.

photo: KCNA
(I thank Ankit Panda for his help with obtaining a high resolution version of the photograph with Kim Jong Un and the map)

Tuesday, 26 September 2017

OT: The brilliant fireball over the Netherlands of 21 September 2017, 19:00 UT, a piece of comet Encke

The fireball as photographed from Ermelo, the Netherlands. Image (c) Koen Miskotte

In the evening of 21 September 2017 at 21:00:10 CEST (19:00:10 UT), a brilliant fireball, as bright as the first quarter moon, appeared over the Netherlands. It was widely seen and reported and garnered quite some social media and press attention (e.g. here). The next day I was live in a Dutch TV program to talk about it.

The fireball was captured by six all-sky camera stations of the Dutch-Belgian all-sky meteor camera network operated by amateurs of the Dutch Meteor Society and KNVWS Meteor Section: stations Ermelo, Oostkapelle, Borne, Utrecht, Twisk and Wilderen, operated by respectively Koen Miskotte, Klaas Jobse, Peter van Leuteren, Felix Bettonvil, Marco Verstraaten and Jean-Marie Biets.

The image in the top of this post shows the photograph taken by the all-sky camera in Ermelo (courtesy Koen Miskotte), where the fireball appeared almost right overhead. The image below was taken by the all-sky camera in Utrecht (courtesy Felix Bettonvil), showing it slightly lower in the sky (click the images to enlarge).

The fireball as photographed from Utrecht, the Netherlands. Image (c) Felix Bettonvil

In the photographs above, the "dashed" appearance of the fireball trail is caused by an LCD shutter between the lens and the camera CCD, which briefly interupts the image at a set interval. For Ermelo this was 14 interuptions per second, for Utrecht 10 interuptions per second.

Knowing the shutter frequency you get the duration of the fireball by counting the number of shutter breaks in the trail: in the case of this fireball, it lasted over 5.3 seconds. Together with triangulation information on the path of the trail in the atmosphere, it gives you the speed of the fireball in km/s, which is necessary to calculate the orbit in the solar system. It also provides you with information about the deceleration of the meteoroid in the atmosphere. In this case, it entered the atmosphere with a speed of 31 km/s and by the time it had completely burned up at 53 km altitude, the speed had decelerated to 23 km/s.

The fireball fragmented into pieces quite early during its atmospheric entry. Some of these fragmentation events can be seen as brief brightenings (flares) in the images.

Triangulation of the six all-sky images yields the following atmospheric trajectory:

Atmospheric trajectory of the fireball, calculated by the author. Camera stations in yellow.

The  fireball moved almost due east-west. It started over Deventer, crossed over southern Amsterdam and Schiphol airport, and ended over sea. The end altitude at 53 km and end speed of 23 km/s indicate that nothing was left of the original meteoroid by the time the fireball extinguished: no meteorites reached earth surface, it completely ablated away.

The apparent radiant of the fireball was located low in the sky, at 16 degrees elevation and almost due east. The grazing entry into the atmosphere resulted in a long trajectory length of over 150 km.

The geocentric radiant of the fireball is located on the Pegasus-Pisces border, just north of the ecliptic. The radiant and speed, and the resulting orbit in the solar system, show that this was an early member of the northern branch of the Taurid stream complex, a meteor stream complex associated with comet P/Encke. It is active from September to December with a  peak in activity in November. The stream is broken up in several substreams, and the early Northern Taurids from September are sometimes called Northern delta Piscids, one of these substreams in the Taurid complex.

The radiant position and heliocentric orbit for this fireball are shown below.

apparent (observed) and geocentric radiant of the fireball

calculated heliocentric orbit of the meteoroid

Acknowledgement: I thank the photographers (Koen Miskotte, Klaas Jobse, Peter van Leuteren, Felix Bettonvil, Marco Verstraaten and Jean-Marie Biets) for providing their imagery for this analysis.

Monday, 4 September 2017

OT: imaging the close flyby of Amor asteroid (3122) Florence

On 1 September 2017 near 12:06 UT the Amor class Near Earth Asteroid (3122) Florence made a relatively close approach to Earth. The nominal pass distance, 0.047 AU or about 18 Lunar distances, was a safe distance. It does not often come this close: the asteroid had last been about similarly close in 1930, and will not be really close again until 2057.

(3122) Florence is one of the more rare larger NEA's. It is about 5 km in diameter and currently has a theoretical MOID (Minimum Orbit Intersection Distance) of 0.044 AU, only slightly less than the pass this year.

While this is a safe distance for now, orbital disturbances in the (distant) future could bring it closer, and the size of the object is such that it would be of future impact concern. It is therefore on the PHA list, where PHA stand for "Potentially Hazardous Asteroid" (for some reason, I always make "Potentially Hostile Asteroid" of this in my mind, probably because of the similar "Potentially Hostile Aircraft").

(3122) Florence became quite bright during this close pass, and was visible in binoculars at a maximum brightness of about magnitude +8.7 just before the moment of closest approach.

About 1.5 days after the moment of closest approach, near local midnight of September 2-3, I imaged the fast moving asteroid from Leiden with my Celestron C6 telescope. It was about mag +9.1 at that time. The movie above, and the stacked image below, was made from 87 images (each an exposure of 10 seconds with a 5 second interval) taken over a 17-minute period between 22:31:32 - 22:48:47 UT (Sep 2).

The asteroid was at a distance of 0.048 AU at that time and moving fast through Delphinus, at an angular speed of about 22".5 per minute.

The equipment used was a Celestron C6 (15 cm) Schmidt Cassegrain telescope with F6.3 focal reducer, and a Canon EOS 60D DSLR at ISO 3200 in the prime focus of the telescope.

(the satellite that can also be seen zipping past in the movie, is a Russian GLONASS, Kosmos 2425)

stack of 87 images taken over a 17-minute time interval
(click image to enlarge)

Friday, 1 September 2017

The other subliminal message in Kim Jong Un's missile test map

image: KCNA
The image above (from North Korea's KCNA) shows the North Korean leader Kim Jong Un and a map. It was likely taken on the day of the 29 August missile test that shot a Hwasong-12 MRBM over Japan, into the Pacific Ocean. I analyzed this map and what it shows in detail in my previous post, showing that the trajectory which seems to be sketched on the map deviates from the real missile flight path.

But there is more to this map, it turns out after further analysis. There are not one, but two subliminal messages contained in this map.

The first subliminal message, already discussed at the end of my previous post, is that it depicts a 3300 km trajectory (the real flight trajectory appears to have been shorter, 2700 km). This is the approximate distance to Guam.

But there is a second subliminal message. The map also contains a veiled threath to Hawaii.

click to enlarge

As can be seen in the annotated georeferenced version of the map image above (see also my previous post), the trajectory skeched on the map (white) is very close in direction to what a trajectory towards Hawaii would be.

In the map above, there are two lines representing such a trajectory targetting Hawaii:

- The green dashed line shows what a real, 7400 km range trajectory towards Pearl Harbour would be (this takes earth curvature and earth rotation into account);

- The yellow dashed line shows what the trajectory is if launched into the same azimuth as for  targetting Pearl Harbour, but with a shorter 3300 km range, as on the map in front of Kim Jong Un.

The latter yellow line obviously is very close to what is sketched on Kim Jong Un's map. In fact, I think that within the error margins of my georeferencing effort, they might well be the same.

The white line, my best effort to represent the trajectory sketched on the map, would be a missile launched into azimuth ~77.25 degrees. The dashed yellow and green lines, are both for missiles fired into azimuth ~76.9 degrees. 76.9 and 77.25 degrees are very close, a difference of only ~0.35 degrees.

To clarify further what the map depicts: the reason that, even though the launch azimuth is the same, the green and yellow lines do not overlap is because of earth rotation and earth curvature. Firing a missile into the same azimuth (~76.9 degrees) from Pyongyang Sunan, yields a slightly more southern trajectory if the range of the missile is ~3300 km (as on Kim Jong Un's map) instead of ~7400 km (the distance to Pearl Harbour). This is because a flight to Pearl Harbour takes longer in time, and the earth surface is meanwhile rotating under the flight path.

So the trajectory sketched on the map contains two subliminal messages: the length (~3300 km) is roughly the distance to Guam. The launch direction is the launch direction towards Pearl Harbour. Two veiled threaths in one. No wonder Kim Jong Un smiles so exuberantly: "Riddle this, you Imperialist agressor Puppets!".

I firmly believe that the North Korean propaganda machine deliberately includes these images into what it makes public. When they show a map or monitor with a trajectory, however oblique and vague, it is not unintentional but (I firmly believe) intentional.

(this post is a follow-on to my earlier analysis here)

Thursday, 31 August 2017

North Korea's 29 August Hwasong-12 test with Japan fly-over: did it go as planned?

image: KCNA
It are hectic times for those (like this author) interested in the North Korean rocketry program. The past months have seen a steady series of increasingly bold missile tests (along with a threath of test-firing missiles towards Guam), that have been the subject of three earlier posts on this blog the past months.

And now a new one has occurred, and it is perhaps the boldest of these test flights so far. Last Thursday, 29 August 2017, at 20:57 UT, North Korea launched a Hwasong-12 MRBM on a trajectory over Japan: a politically bold move that caused air-raid sirens to go off in Northern Japan.

images: KCNA

Western military sources say that the missile was fired from Pyongyang Sunan airfield. It reportedly travelled 2700 km, crossing over Hokkaido, coming down in the Pacific Ocean about 1180 km from Cape Erimo in southeast Hokkaido, Japan. Apogee of the trajectory is quoted as 550 km.

This allows us to produce the trajectory reconstruction below (with some leeway as the quoted ranges are probably balpark figures). It shows that the missile crossed Japan over the southern tip of Hokkaido, a trajectory that seems designed to minimize Japanese landmass overflown, mitigating risk levels. The two red circles on the second map are depicting a 2700 km range around Pyongyang Sunan and 1180 km range around Cape Erimo. Where they are closest, is the resulting impact point (near 41.92 N, 157.56 E). Using STK I get a flight-time of ~14 minutes.

click to enlarge

click to enlarge
That odd-looking second of the two maps above is based on an interesting photograph published by the North Korean State News Agency KCNA. It shows an exuberant Kim Jong Un sitting on a desk with a map in front of him:
image: KCNA

We have recently seen more of this kind of shots from KCNA published after previous tests, and they often yield interesting information. In this case, the interesting detail is that a trajectory appears to be depicted on the map, as a thin line (the arc above it ending at the same point likely is a 3D representation):

If the sketched line indeed is the intended trajectory, then the interesting point of it is that it does not match the observed trajectory according to Western military sources.

I used my GIS skills to georeference the map on the image (with QGIS). Like in the case of the 'Guam map', the low resolution of the image and very oblique angle of te map on it make this a challenge, and the georeferencing obviously is not perfect - but good enough for some conclusions. This is the result (look for the vague line that possibly represents a trajectory on the image):

click to enlarge
Here is the same map but with annotations added, and with the trajectory sketched on the map (white line) and the actual observed trajectory (black dashed line):

click to enlarge

It is immediately obvious that the two trajectories do not match well. The trajectory (if it is one) sketched on the Kim Jong Un map is located more to the south, and indicates a larger range, about 3300 km (interestingly, it does conform to a real ballistic trajectory with earth rotation and curvature taken into account, unlike the straight line drawn on the infamous "Guam map").

(the second map shows exactly the same map area, but with a DEM and country shapefile as background. It shows you how the topography matches that on the georectified map from the photograph)

The difference between the two trajectories amounts to a range difference of 600 km and a 45 km difference in apogee. It represents a difference in launch azimuth of 6 degrees and of 3.7 degrees in launch elevation. It points to an underperformance of around 0.4 km/s in the missile's burnout speed.

click to enlarge
It is interesting to note that while North Korea for some previous tests (and their proposed Guam shot) published very accurate figures for range, flight time and apogee, they did not do that this time. It is also interesting to note that some Western military sources report that the missile "broke into three pieces".

So there are multiple indications that the test did perhaps not go entirely as planned, with an underperformance of the rocket engine and an error in launch direction, and perhaps structural failure. This would imply that the targetting of the Hwasong-12 is not quite fail-safe yet, which has implications for Kim Jong Un's proposed Guam enveloping missile demonstration.

This was the first test of the Hwasong-12 on a "normal" trajectory and it shows why it is important for North Korea to test its missiles on such a normal operational, rather than a 'lofted' trajectory: it is a different regime of stress on the missile and has the engines perform under more realistic conditions than on a 'lofted' trajectory. Errors in targetting and missile performance become more apparent.

For North Korea to do such tests on a "normal" trajectory for its MRBM's and ICBM's, they have to fly it over Japan. We are going to see more of this kind of tests the coming months is my prediction.

Update 31 Aug 2017:

(1) The 38 North blog has an interesting post suggesting the possible failure of a post-boost vehicle.
(2) South Korean and US military analysts appear to come to the same analytical conclusions as I did, according to this Korean news bulletin
(3) As Ankit Panda rightfully remarked, the situation drawn on Kim Jong Un's map need not comply to reality. It could be merely meant as a propaganda message (3300 km is roughly the distance from the launch site to Guam).

Update 1 Sep 2017:

A further follow-on post is here, discussing how the map contains two veiled threaths, to Guam and to Pearl Harbour.

Sunday, 13 August 2017

[UPDATED] My take on North Korea's hypothetical threath to Guam

click to enlarge

The sabre rattling between North Korea and the United States is continuing, with increasingly colourful and volatile language and threaths from both sides.

In response to Trumps statements on August 8 that North Korea better stop its threaths or else "they will be met with fire and fury like the world has never seen", the North Korean General Kim Rak Gyom, commander of the Strategic Force of the Korean People's Army, released a  statement through KCNA on August 10 that included a surprisingly detailed threath about firing four Hwasong-12 missiles towards the US island (and important Pacific Naval and Air Base) of Guam.

Two relevant parts of the statement read (emphasis added):

As already clarified, the Strategic Force of the KPA is seriously examining the plan for an enveloping strike at Guam through simultaneous fire of four Hwasong-12 intermediate-range strategic ballistic rockets in order to interdict the enemy forces on major military bases on Guam and to signal a crucial warning to the U.S.
The Hwasong-12 rockets to be launched by the KPA will cross the sky above Shimane, Hiroshima and Kochi Prefectures of Japan. They will fly 3 356.7 km for 1 065 seconds and hit the waters 30 to 40 km away from Guam.

The details listed, allow to reconstruct the intended launch site for this scenario. It points to launch of the missiles in the vicinity of Sinpo, an important North Korean Naval Base that is home to its experimental SLBM force and according to satellite imagery analysts has recently seen "high levels of activity" (although that is not necessarily related to the issue under discussion).

The distance from the Sinpo submarine SSBA dockyard to Anderson AFB on Guam is ~3357.5 km, in agreement with the distance mentioned in the KCNA statement. Taking into account Earth rotation, the flight trajectories of the missiles when launched from Sinpo would indeed cross over the three Japanese prefectures mentioned, as can be well seen in the diagram in the top of this post. Together with the listed flight time, it would point to a flight apogee at ~865 km altitude and a launch elevation of ~31.4 degrees, i.e. a "normal" flight trajectory for an IRBM (as opposed to the "lofted", reduced range trajectories of previous tests).

click to enlarge

The map in the top of this post displays the flight trajectories based on this scenario, with the four missiles straddling Guam at ~40 km distance of its shores, uprange, downrange and lateral to both sides of the main flight direction needed to reach Guam.

While the language of the KCNA release is threathening, it is important to realize it is not threathening an actual "attack" on Guam. They talk about a strike that will "envelope" the island with the missiles, each of them impacting at a distance of 30-40 km of the Guam shore.

With the listed 30-40 km offshore impact distances, the missiles would land outside of Guam's territorial waters: but in its Exclusive Economical Zone (EEZ). While sensitive, this is something else than actually hitting Guam or the territorial waters directly surrounding it. The latter would be a real attack on US territory.

Targetting the EEZ has earlier precedents, as recent North Korean missile tests have landed in the Japanese EEZ. What makes it different this time is that it is the EEZ of the US itself, not of an ally, and involves a fly-over of Japan.

Rather than an attack, the scenario painted by the KCNA announcement would be best described as a missile test and demonstration. A test, because it would be the first flight of the Hwasong-12 on a "normal" battle trajectory rather than a "lofted" trajectory. A test, as it would for the first time have the Reentry Vehicle (RV) face the atmospheric reentry conditions it would face if used in an actual conflict. A demonstration, as it would demonstrate that the Hwasong-12 can indeed reach targets at the distance of Guam. A demonstration, as straddling Guam with missile impacts obviously sends a very clear message to the US and its allies.

It would also potentially test the Hwasong-12's performance against US missile defense measures. And it would be a (risky) political test to see how the US and its allies would react to such a demonstration of military power.

How likely is it that it will happen? On the face of it, not that likely. It most likely is hollow rethoric. Firing missiles towards a US military base and US territory is very risky as it could provoke a disastrous counter-attack. A full scale US counter strike would mean the end of the North Korean regime.

However, the oddly detailed August 10 KCNA announcement (which as some analysts on Twitter mentioned almost reads like a NOTAM) could also be seen as a way to mitigate this risk. It clearly signals the intention to not hit Guam itself, just fire near it in a demonstration of power. It is in essence a (tough worded) notice of an upcoming missile demonstration, not an actual war threath.

We should also realize that a US counter-attack would be disastrous for both North Korea and the US and its allies. If the US counter-attacks, North Korea will respond by shelling Seoul into oblivion and, if it can, firing missiles towards targets in South Korea, and possibly towards Japan and US bases in the Pacific. The carnage will be incredible. Hence, a military strike on North Korea as a gut-reaction to North Korean missiles landing near Guam is not as likely as it might at first glance seem. A sane US response would be to restrain from such a counter-attack unless there are strong reasons to do so, like missiles actually hitting targets on Guam. North Korea might have made a similar assessment. But it is a risky one (also taking potential missile malfunctions into account).

North Korea however has taken bold calculated risks before. They recently fired missiles into the EEZ of Japan, a key US ally. They took a risk in sinking the Cheonan in 2010, and shelling Yeonpyeong island that same year.

It is important to note the recent impact of North Korean missiles into the EEZ of Japan, given that the August 10 KCNA announcement points to targetting the EEZ of Guam. This firing into the EEZ of Japan did not lead to military retaliation by the US and allies (and in a sense perhaps might be seen as "testing the waters"). This might embolden North Korea to fire into the EEZ of Guam, expecting (rightfully or not) a similar non-response.

What could the US realistically do in response? As mentioned, a counter-attack in response to the missile firings would lead to dramatic results for the US and its allies. Seoul and Tokyo and who knows Honolulu burning seems a steep price to pay for "acting tough" when the missiles provoking the response harmlessly landed in the Pacific Ocean outside of Guam territorial waters. I fail to see how the US can justify such a result to its citizens and allies when the target of the missiles is in the Guam EEZ and not Guam itself.

Alternatively, the US or allies could perhaps attempt to intercept the missiles. That sounds like a good middle ground but isn't, as it has its own risks. Intercepting missiles on a trajectory that would not reach Guam territorial waters would provide political fuel to North Korea, who can claim their missile test was attacked. Moreover, there is a serious risk that an intercept attempt partly or wholy fails. That would seriously damage the credibility of the US and its capacity to defend itself and its allies, and with that seriously undermine the faith of US allies in the region (apart from the general loss of face). An attempted missile intercept would therefore be unwise in my opinion, as it would be a win-win situation for North Korea no matter the outcome.

So to me, it seems that North Korea is playing a high level (if dangerous) game of chess here, and has managed to checkmate the US. There seems very little the US realistically can do about it, apart from denouncing it. And the tougher the talk from Washington, the more face they will lose if it happens. "Hold my beer", you can hear Kim say in response, knowing he got them cornered.

In case North Korea does go forward with the announced missile demonstration, I expect Trump to do what he has done before: back-peddle on his earlier tough talk and grand-standing. And obscure this by making some unrelated outrageous policy decision or statement in order to deflect attention away from this political defeat, just as he has done with earlier political defeats.

Maybe the North Korean announcement is just a hollow threath. Maybe it is not. More likely than not it will not happen. But I would not exclude it entirely from happening, if North Korea made a similar assessment as outlined above. We will see.

UPDATE  15 aug 13:25:

The North Korean KCNA published a  bulletin on August 15, also carries by Rodong Sinmun. Apart from some interesting statements that might suggest North Korea has put a test fire towards Guam on hold for a while, the Rodong Sinmun newspaper contains some interesting photo's (pdf here via KNCA Watch):

images: KCNA/Rodong Sinmun

Two items on these photographs are of interest: the image on the wall at right, which depicts Anderson AFB on Guam. The other one is the map in front of Kim Jong Un (I put it upside-down here to have north up in the map):

image: KCNA/Rodong Sinmun
This map clearly depicts a missile trajectory towards Guam, that is very close to the trajectory I presented earlier in this blog post. The point of origin of the trajectory indeed seems to be Sinpo on the southeast coast.

UPDATE 2 & 3, 16 Aug 2017 00:30 & 2:45 UT:

Some claims appeared on Twitter that according to South Korean analysis, the map points to launch from Musudan-ri rather than Sinpo:

I do not agree. I used my GIS skills to georeference and reproject the map in front of Kim Jong Un. Below is the preliminary final result:

click to enlarge

What can be seen is that the drawn trajectory points clearly and unequivocally to Sinpo, not Musudan-ri.

What also can be seen is that the line drawn on Kim Jong Un's map is a simple straight drawn line rather than a real missile trajectory. The blue line I have drawn in the map is a true trajectory, including earth curvature and earth rotation and taking the map projection into account. The dashed red line I have drawn is merely a straight line (ignoring map projection).

The country outlines have been added in by me to show how the georeferenced image matches a true map.