Ionosphere disturbance which was seen prior to an earthquake which occurred on 11 March 2011 off the coast of Tohoku was studied. Ionosphere data which were used are: O + density acquired with US satellite DMSP, and NmF2, h'F, and fmin obtained by ionosondes. Global Ionosphere Model (GIM) is used as side evidence to compliment findings from DMSP data and ground- based data. Although magnetic disturbance is strong during the earthquake preparation period, global survey of ground based NmF2 shows that special disturbance is limited to earthquake area. Satellite data (DMSP and GIM) analysis shows four important findings before the occurrence of the earthquake; (1) Over geomagnetic equator enhancement of O + is found, (2) midlatitude trough (MLT) is formed before the earthquake and it moves toward lower latitude as EQ day approaches, (3) O + enhancement at the equator side edge of the MLT, and (4) no clear difference of O + behavior between east and west of the epicenter is identified. These four features are connected each other. Night time NmF2 at high latitude ionosonde stations such as Khabarovsk shows 2 days oscillation from 5 March and disappears on 12 March. As the latitude of the station is lower, 2 days oscillation becomes unclear, and the duration of the appearance is shorter. In order to explain both ground – based and satellite data consistently, enhanced east (west) ward dynamo electric field during daytime (nighttime) are proposed. Our speculation is that internal gravity waves of extremely small amplitude due to the ground motion interact with planetary scales waves below 10 km and is amplified. The amplified IGW propagates to the dynamo region, modifies the wind system, or conductivity, which modifies the electric field. Several observational facts which are favor to the generation of IGW are described. Finally, we stress the need of satellite constellation in order to obtain the global morphology of ionosphere disturbance and to identify the mechanism, which at least provides us the material to judge applicability to future earthquake prediction.