A Statistical Comparison between Photospheric Vector Magnetograms Obtained by SDO/HMI and Hinode/SP

TL;DR

This study compares vector magnetic field measurements from SDO/HMI and Hinode/SP to understand differences arising from instrument modes, spectral sampling, and inversion methods, aiming to improve combined data analysis for solar magnetic studies.

Contribution

It provides a detailed comparison and transformation of Hinode/SP data to match SDO/HMI observations, highlighting the impact of filling factors and sampling on magnetic field measurements.

Findings

Filling factor assumptions significantly affect magnetic property estimates.

Spectral and angular sampling differences influence measurement consistency.

Standardizing disambiguation methods enhances data comparability.

Abstract

Since May 1, 2010, we have been able to study (almost) continuously the vector magnetic field in the Sun, thanks to two space-based observatories: the Solar Dynamics Observatory (SDO) and Hinode. Both are equipped with instruments able to measure the Stokes parameters of Zeeman-induced polarization of photospheric line radiation. But the observation modes, the spectral lines, the spatial, spectral and temporal sampling, and even the inversion codes used to recover magnetic and thermodynamic information from the Stokes profiles are different. We compare the vector magnetic fields derived from observations with the HMI instrument on board SDO, with those observed by the SP instrument on Hinode. We have obtained relationships between components of magnetic vectors in the umbra, penumbra and plage observed in 14 maps of NOAA AR 11084. Importantly, we have transformed SP data into observables comparable to those of HMI, to explore possible influences of the different modes of operation of the two instruments, and the inversion schemes used to infer the magnetic fields. The assumed filling factor (fraction of each pixel containing a Zeeman signature) produces the most significant differences in derived magnetic properties, especially in the plage. The spectral and angular samplings have the next largest effects. We suggest to treat the disambiguation in the same way in the data provided by HMI and SP. That would make the relationship between the vector magnetic field recovered from these data stronger, what would favor the simultaneous or complementary use of both instruments.