The Small-scale Structure of Photospheric Convection Retrieved by a Deconvolution Technique Applied to Hinode/SP Data

• Published in: The Astrophysical journal Vol. 849; no. 1; pp. 7 - 17
• Main Authors: Oba, T., Riethmüller, T. L., Solanki, S. K., Iida, Y., Quintero Noda, C., Shimizu, T.
• Format: Journal Article
• Language: English
• Published: Philadelphia The American Astronomical Society 01.11.2017; IOP Publishing
• Subjects: ACCELERATION; AMPLITUDES; Astrophysics; ASTROPHYSICS, COSMOLOGY AND ASTRONOMY; Computer simulation; COMPUTERIZED SIMULATION; CONVECTION; Deceleration; Deconvolution; DIFFRACTION; DISTRIBUTION; Field of view; Granular materials; GRANULATION; HYDRODYNAMIC MODEL; hydrodynamics; Image degradation; IMAGE PROCESSING; Light diffraction; MASS BALANCE; Numerical simulations; Optical distortion; PHOTOSPHERE; PRESSURE GRADIENTS; SUN; Sun: atmosphere; Sun: granulation; Sun: photosphere; techniques: image processing; VISIBLE RADIATION
• ISSN: 0004-637X; 1538-4357
• DOI: 10.3847/1538-4357/aa8e44

Solar granules are bright patterns surrounded by dark channels, called intergranular lanes, in the solar photosphere and are a manifestation of overshooting convection. Observational studies generally find stronger upflows in granules and weaker downflows in intergranular lanes. This trend is, however, inconsistent with the results of numerical simulations in which downflows are stronger than upflows through the joint action of gravitational acceleration/deceleration and pressure gradients. One cause of this discrepancy is the image degradation caused by optical distortion and light diffraction and scattering that takes place in an imaging instrument. We apply a deconvolution technique to Hinode/SP data in an attempt to recover the original solar scene. Our results show a significant enhancement in both the convective upflows and downflows but particularly for the latter. After deconvolution, the up- and downflows reach maximum amplitudes of −3.0 km s−1 and +3.0 km s−1 at an average geometrical height of roughly 50 km, respectively. We found that the velocity distributions after deconvolution match those derived from numerical simulations. After deconvolution, the net LOS velocity averaged over the whole field of view lies close to zero as expected in a rough sense from mass balance.