The solar surface is covered with tiny structures of magnetic field and their interactions cause various solar activities. However, the formation and maintenance of the magnetic field there is still in hidden. One of the key process there is a transport of the magnetic field there.
In Hinode-8, we reported the analysis of displacement of magnetic patches by use of the longest magnetogram data taken by Hinode. The result shows that the sub-diffusion scaling is seen in the scale larger than the network. It is critical for the transport in the global scale, e.g. solar dynamo model, and the estimated value of the diffusion co-efficient is smaller than that required from the theoretical model. We discussed that this inconsistency between small and large scales may root in the removal of magnetic patch interactions, especially merging and splitting, from the consideration.
In this presentation, we evaluate the transport by magnetic patch interactions with the actual magnetograms observed by Hinode. In the model, the transport of the magnetic field is mainly decided by the frequencies of merging and splitting at the network field in the model. We found three characters in the analysis of the magnetograms. One is that the frequencies of merging and splitting have strong dependence on the magnetic flux. Another finding is that the smaller flux is less re-merged to the network field. The last one is that the smaller flux has larger velocity than the larger one. From these results, we concluded that the splitting and merging of smaller magnetic flux is a dominant process in the transport over the network field.