The authors regret to discover an error in our above mentioned article. We would like here to address such error related to the averaging calculation of Eq. (3) which affected Fig. 7 in the article. The error is that the radial heat transfer coefficients should be linearly averaged based on a horizontal line along the radius (at constant height and azimuth) instead of azimuthally averaged based on the cross-sectional area, therefore Eq. (3) becomes as:[formula]. According to that, the averaging scheme of Fig. 7 has been re-plotted as follows, which represented the corrected figure to be for the above article. Now, after this correction, the results clearly show that the radial averaged heat transfer coefficients are occurred near r/R ∼ 0.5. The authors would like to apologise for any inconvenience caused.