The solar photosphere is the visible surface of the Sun, where numerous numbers of bright granules, called granulation, surrounded by narrow dark intergranular lanes, are observed anywhere. The stratification, e.g. the structure along the height direction, is critical for convective instability. It is, however, difficult to obtain it from observations and hence the detailed convective structure in the height direction is still unclear. We tackle this problem by using bisector analysis, derivation of the Doppler velocity at each intensity level in a photospheric absorption line. At the LWS meeting (Hinode8) last year, we reported the velocity field derived with this analysis. The issue in the analysis was that the 5-minute oscillations were included in the velocity structure; The 5-minute oscillations may give velocity signals in the same order of magnitude with that of convection. This time, we removed the 5-minute oscillations signals from the data by using the k-ω diagram and investigated the height structure of the pure convective motions.
The spectral data for the bisector analysis was acquired with the spectro-polarimeter (SP) of the Hinode / Solar Optical Telescope (SOT). The data set is a 2-second cadence time series of the sit-and-stare Stokes IQUV measurements for 45 minutes, with the simultaneous series of BFI blue continuum images. The SP slit was placed at the fixed position in the quiet region at the disk center. We examined the Stokes I profiles in the pixels where no Stokes QUV signals are observed.
To investigate the mechanisms of convection, we focused on the dependence of the vertical velocity amplitude on height. The average amplitude decreases from 0.65 km/s to 0.40 km/s with height in granules where the upward mass motions may be observed in general. On the other hand, it increases from 0.30 km/s to 0.50 km/s with depth in intergranular lanes, where the downward motions are dominantly observed. Theoretically, if the photosphere is in convective stable condition, the amplitude of the velocity decreases with depth. Hence our results cannot be explained with the convective stable condition and we need the extra force to break the convective stable condition in the intergranular region. We suggest that radiative cooling is very effective in the downflow region and the energy loss through it may result in accelerating downward motions.