Just got the GPS data and cross-referenced it with the earthquake chronology. The data are very interesting.
9 March 2011, the eagle is at the top of a mountain ridge some 8 km SW from Ashbetsu town, Hokkaido. The fix is taken at noon, which is 2 am UTC. At 02:45:20 AM the global seismological network registers first earthquake of Hokkaido with magnitude 7.2 (source: http://earthquake.usgs.gov/earthquakes/recenteqsww/Quakes/quakes_big.php).
This transliterates into magnitude 3 tremor in some parts of Hokkaido (source Japan Meteorological Agency http://www.jma.go.jp/en/quake/20110309115106384-091145.html). The bird takes off in a direction opposite to that of the earthquake, and the next fix comes on 10 March from the top of a small mountain ridge some 10 km east of Rumoi, a town on the west coast of Hokkaido. Direct line travel was 55 km.
During this time there were a number of earthquakes off the coast of Honshu, but the first one which was accompanied by a tremor of magnitude 3 on Hokkaido was reported on 15:32, i.e. after the fix. The bird was continuing to fly north-west, and the Doppler transmission (no GPS fix yet) was received from a mountain ridge some 15-20 km north of town Haboro. Direct line travel is 63 km.
On the attached picture: yellow line - gps track. White dots in close to the coast show latest doppler coordinates taken at 3 am UTC, that is roughly 2 hours before the main hit of 8.9 magnitude stroke (2011/03/11 05:46:24) off coast of Honshu.
So, can a bird predict an earthquake? Probably not, but certainly our Steller's Sea Eagle flew away from a modest tremor, and by the time the major hit came on 11 March, it was away from the zones where tremors occured. It is likely that the bird will continue to move. The general picture is that by moving away from first tremors the bird was farther away from the epicenter by the time the main quake occured. Perhaps birds are more observant to small signs of on-coming earthquakes?