Hitomi (ASTRO-H) scientific X-ray satellite suffered an orbit-altering event

On 26 March 2016 near 1:42 UT, the Japanese Hitomi/ASTRO-H satellite, a scientific Röntgen space telescope launched February 17, suffered a clearly energetic event that changed its orbit and resulted in a communications loss. JSpOC reported the release of at least 5 fragments from the spacecraft. The existence of these fragments was first detected by US tracking stations near 8:20 UT. Observations by amateur satellite observers reported on Seesat-L (here, here, here) suggest that the spacecraft is tumbling with a periodicity of about 23.5 seconds (flashes each ~11.75 seconds) according to a preliminary analysis by Ted Molczan here.

click to enlarge map

Analysis of the pre- and post-event orbital elements suggests that the event occurred near 1:40-1:50 UT on March 26 (see note at end of post (*). JSpOC estimates the event occured at 1:42 +/- 11 m UT. The satellite was just past perigee and had just passed the ascending node of its orbit at that moment, moving over Meso-America (see map above). The nominal time for the event, 1:42 UT, is about 12 minutes past nodal crossing.

The Japanese Space Agency JAXA reports that since noting loss of communications at 7:40 UT on March 26, they have managed to receive signals from the spacecraft twice, near 13h and 15:30 UT on March 28.

The event was energetic enough to alter the orbit of the satellite, slightly increasing its eccentricity and dropping slightly in altitude: it's semi-major axis changed by 2 km, with perigee dropping by 4.5 km. The inclination was changed too, by about 0.0045 degrees.

(click diagrams to enlarge)

data source: Space-Track

Based on the change in semi-major axis and inclination, and if my calculations are correct, the orbit change of ~2 km in semi-major axis and ~0.0045 degree in inclination is the result of a velocity impulse with delta V of about 1.2 m/s. The most likely source is a sudden venting or a small explosion of some sort. An explosion of some sort also explains the discharge of five fragments reported.

The time of the disruptive event is almost exactly 6 hours before the satellite was to turn operational.

How serious the event is, is unclear at the moment. The fact that the spacecraft still appears to be sending signals is a positive aspect. If JAXA can re-establish contact and stop the tumbling, perhaps the mission can be saved. Or perhaps not, depending on the damage to the spacecraft.


* Note: my initial analysis of pre- and post event orbital elements on March 28 yielded times near 1:40 UT (March 26), close to JSpOC's nominal 1:42 UT time for the event. Annoyingly, I had some odd brain malfunction next and tweeted "4:40 UT" instead. 

It was probably induced by a mix-up in my mind of the established 1:40 UT event time with the epoch time of the first abberative TLE, which was16086.196059 = 26 Mar 4:42 UT. I had jotted both times down and then mixed them up apparently...such things happen when you spent too much time coding and processing data in an excel spreadsheet.

The way in which I analyzed the probable time of the event, is by taking pairs of pre- and post-event elsets for the satellite and treating them as if they concerned two separate objects. I then used Rob Matson's COLA software to establish close encounters between the "two" objects: near the time of the event, the orbital positions from both elsets should nearly coincide. Depending on which elsets you pair, this yields times scattering between 1:30 and 2:06 UT . My initial test of four elset pairs had three of them yielding times near ~1:40 UT.