| We modeled the seismic structure in the solid inner core to a depth of about 550 km below the inner core boundary using differential travel times, amplitude ratios and waveforms of pairs of broadband core phases collected in the 129°-141° and 149°-161° epicentral distance range. Uniformity of the F region that exists in the bottommost 200 km of the liquid outer core was also tested using the same data. Source effects and mantle attenuation were deconvolved from the seismograms using Effective Source Time Functions (ESTF) that were obtained from inverting P waves in the 30° to 90° distance range.;P velocity in the inner core between the longitudes 45°E and 180°E is greater than or equal to that of AK135-f reference model whereas that between 180°W and 45°E is less than the reference model. The region between 180°W and 0° show a greater reduction in velocity than that between 0° and 45°E indicating that the latter region is the transition between the fast and the slow regions that show a velocity contrast of about 1%. Observation of this heterogeneity to a depth of 550 km below the inner core and the existence of a transition rather than a sharp boundary favor either very slow inner core super-rotation or an oscillation.;There is a general positive relationship between attenuation and P velocity. If the grain size in the inner core is less than 100 m, attenuation is likely to be dominated by intrinsic mechanisms. Attenuation at the top of the inner core is generally high ( Q-1a ∼ 0.0033 -- 0.0040). While an increasing trend in attenuation with depth in the Eastern hemisphere is observed, attenuation remains nearly constant in the Western hemisphere between 25 km and 80 km below the inner core boundary. It is not clear at this time what attenuation mechanisms are responsible for the observed hemispherical difference. In deeper regions, Q-1a is relatively smaller than the top 100 km of the inner core and falls in the 0.0014 - 0.0040 range.;Evidence from numerical geodynamo simulation models suggests that the observed degree-one heterogeneity in the inner core might be coupled to the core-mantle boundary thermal heterogeneities and topography. Seismically fast region may be colder and growing faster than the slow region. It is possible that the heat flux at the core-mantle boundary has remained nearly unchanged for the past 500 Ma, roughly the time required to increase the inner core radius by 500 km. Restrictive conditions required by the inner core convective translation makes it less desirable to explain heterogeneity penetrating deep in to the inner core.;We conducted two tests to investigate the seismic structure in the F region. Test for the velocity profile in the F region confirmed that it is uniform globally and agrees well with that of AK135-f. A steeper gradient, such as that in PREM, is not needed to explain our data. We do not however, exclude the possibility of an intermediate gradient, which can be closer to AK135-f. Models of attenuation on the other hand, show that some localized heterogeneity can exist. This may be resulting from fine scale convection patterns that exist in the outer core. The F-region may be maintained by episodic melting of the inner core induced by a combination of outer core convection locked to thermal heterogeneities in the CMB and cyclones induced by CMB topography. |
