Friday, 18 July 2014

SBIRS, SIGINT and the MH17 tragedy (updated)



Yesterday 17 July near 13:15 UT, 298 people including at least 173 189 192 of my countrymen perished when Malaysian Airlines flight MH17 on its way from Amsterdam to Kuala Lumpur crashed over the eastern Ukraine, reportedly after being hit by a missile.

This is a terrible tragedy. Among the victims are complete families, including children. It is the start of the holidays in the Netherlands, and the flight carried many Dutch families on their way to their holiday destinations in southeast Asia. My thoughts are with these highly stricken families.

For me personally, it is an unnerving fact that I was about to fly the same route from Amsterdam to southeast Asia only a few days later.

In the wake of the incident, accusations fly between the Ukrainians, pro-Russian separatists and Russians, all accusing each other of being responsible for this tragedy. At the moment it is difficult to say which bits of information floating around are true and which are false. I strongly suspect that the current suspicion against Russian-backed separatists will hold though. Some less ambiguous evidence (e.g. the location of the crash, which is close to the locations where separatists earlier downed two other (military) aircraft) certainly seem to suggest this. But we will see: at the moment, nothing is certain.

Of interest to this blog, is that US Intelligence officials have confirmed that the aircraft was hit by a surface-to-air missile, according to several US media. Senior US officials appear to have told CNN that they detected a radar signal from a surface-to-air missile system being turned on right before the crash, and that they also detected a 'heat signature' at the time the aircraft was lost.

If the CNN report is correct, it is highly likely that the 'heat signature' detection was a space-born detection by the SBIRS system of infra-red early warning satellites. I have written about this satellite system before, in the context of that other recent tragedy with a Malaysian Airlines flight, the disappeared flight MH370.

click image to enlarge

Three of the four SBIRS satellites, SBIRS GEO 1 (2011-019A) and SBIRS GEO 2 (2013-011A) in geostationary orbit and USA 184 (2006-027A) in HEO, had coverage of the area where MH17 went down at the time this happened (17 July 14:15 GMT, see image above).

SBIRS and SIGINT platform USA 184, imaged on 20 March 2014

SBIRS GEO 2 imaged on 20 June 2014

It is possible that the quoted detection of a missile radar tracking system activation around the time of the disaster was done by satellites too. Several SIGINT and ELINT satellites cover this area, including various MENTOR (ORION) satellites and one MERCURY satellite in GEO, and USA 184, which is both a TRUMPET-FO SIGINT satellite and a SBIRS platform, in HEO. That these SIGINT satellites amongst others serve to detect and monitor signals from military radar and missile systems, is known. Given the interest of the USA and NATO in closely watching military developments in the Ukraine conflict, it is almost certain that some of these are targetting the area.

The question is, whether these satellites can help pinpoint the location from where the missile was launched, and hence provide an indication of who did it (Ukrainian forces, separatist militia, or the Russians).

I suspect they can. If the SIGINT detections were indeed done by satellites, it is known that the US recently made large progress in geolocating the origin of detected signals. In a speech from September 2010 available on the NRO website, NRO director Bruce Carlson specifically remarked on the NRO's increasing capability to geolocate using SIGINT:

"I will tell you that just in the last 24 months, we’ve improved the accuracy of geo-location by nearly an order of magnitude, and we’re going to continue to do that and bring it down. We’re getting to the point where here very, very shortly, within the very near term, we will be able to target using signals intelligence". 

If they indeed have a SIGINT detection of the missile's radar system (and the CNN quote seems to say that), the character of the signature might yield information on what missile system was used (i.e. if it was indeed an SA-17/BUK).

Likewise, and although as far as I know no exact public information is available on the accuracy of this kind of detections (update: but see the update at the end of this post!) , I suspect that the  'heat signature' detections of the missile launch,  if indeed SBIRS infra-red detections, are also accurate enough to geolocate the launch site (and whether that is in Ukranian held, or separatist held territory).

A SBIRS platform has two sensors: one in staring mode, and one in scanning mode. The staring scanning mode sensor watches for heat signatures over a wide semi-global area. The scanning staring sensor targets specific regions, and when the staring scanning sensor detects a signature, the scanning staring sensor (at least according to some sources) can be employed to further pinpoint and track this event (more sources amongst others here, here and here). The goal of SBIRS reportedly is to be able to track launches, pinpoint launch sites and accurately predict potential target locations from the tracking data. That needs quite accurate tracking.

(note added: a 1-hour timezone conversion error in the original version of this post has been corrected)

Update 19/07/2014: Daniel Fischer managed to dig up this unclassified presentation from 2006, which shows that SBIRS indeed can detect SAM. Pages 2 and 3 mention the capability to pinpoint the launch location. 
Rainer Kresken has raised the legitimate question of the cloud cover present at the time of the shootdown. Water vapour obscures Infra Red, which means the cloud cover might have blocked detection of the initial launch phase of the SAM. The SIGINT detection of the missile system radar does not suffer from this problem.

Wednesday, 9 July 2014

A bounty of GEO satellites on June 21


The night of June 21-22 was clear, and as I had trouble sleeping, I decided to take the short bicycle trip to my secondary site, Cospar 4355. This site is located in the polder only just outside of town, but the sky is better there than at my regular site 4353, which is in the town center (the secondary site is about 2 km south of my regular site). As a result, I can use twice as long exposures, which means I can image fainter GEO satellites than from my regular site. The site, being in a polder, also has less horizon obstruction. Below is a panorama of the site, split up in two parts, each slightly larger than 180 degrees. Azimuth directions are indicated.


Panoramic view at Cospar 4355

I took some 54 picture (20 second exposures with a Canon EOS 60D + SamYang 1.4/85mm at 800 ISO) over the course of an hour. My main focus was on approximately 20-30 degree (1-2 camera fields) wide equatorial areas near azimuth 120-130 deg, 160 deg and 200 deg.

I captured a nice batch of objects: 17 classified objects, two Unknowns (initially four but two got ID-ed as classifieds) and A LOT of unclassified objects. The image in the top of this post shows an only 2.7 degree wide stretch of one image, and look how many objects are already in it.
One of the objects in the image, the defunct Russian military comsat Raduga 1-M1/Kosmos 2434 (2007-058A) was flaring repeatedly in subsequent images (compare also the two images in the top of this post).

The images below show two other swaths of sky only a few degrees wide. Various commercial GEO sats are visible, as well as two old Ariane r/b, of which several were captured this night:

click image to enlarge

It also shows  the British military communications satellite Skynet 5B (2007-0056B).

One of the classified objects captured this night was AEHF 2 (USA 235, 2012-019A), part the new military communications satellite constellation that is gradually replacing the Milsat system. Another object imaged was the SBIRS GEO 2 (2013-011A) satellite, part of the new infra-red Early Warning constellation that is replacing the DSP constellation.


click images to enlarge


The lower of the two images above (it is slightly blurry because it is the edge of the image) also shows one of the initial UNID's of that night, "UNID 2", one that Cees and Ted later identified as the classified Italian military communications satellite Sicral 1 (2001-005A), which has recently been moved to 22 E.

Cees also managed to identify another UNID I imaged that night, "UNID 3":

click image to enlarge

It is the object we amateur trackers designate as Unknown 130929 (2013-772A), an object in a Molniya orbit which was last seen 132 days before my observations (i.e. we temporarily "lost" it). It was over West Africa at an altitude of 1270 km at the time of observation, moving away from perigee:



Two other UNID's of this night remain to be identified. One of these ("UNID 1") appears to be in GTO: the other one ("UNID 4") appears to be in LEO and was very faint.

The image below shows two classified objects (plus several commercial geosats), both US Military communications satellites: USA 236 (2012-033A) and WGS 3 (2009-068A). WGS 3 is the third satellite in the Wideband Global Satcom constellation. USA 236 is a geostationary SDS data relay satellite. It is believed that they notably relay imagery of IMINT satellites in LEO, for example optical imageryby  KH-11 Keyhole/CRYSTAL and radar imagery by Lacrosse and FIA.

click image to enlarge

Mentor 4 and Thuraya 2 change of configuration

A change is occurring in the configuration of Mentor 4 (USA 202, 2009-001A), a huge Mentor /ORION SIGINT satellite, and the commercial communications satellite Thuraya 2. For over 3 years, Mentor 4 was stationed (as seen from my observing location) slightly south of Thuraya 2. On my June 21 imagery, it has moved to slightly North of Thuraya 2. Compare the top image from last June 21 with some images shot in previous years:

21 June 2014:

8 December 2010:

18 November 2012:

29 December 2013:
click images to enlarge


(The first image also shows the still unidentified UNID 1, likely in GTO, and  a classified r/b from another Mentor/ORION launch, Mentor 3 r/b (2003-041B)).