On the estimate of magnetic non-potentiality of sunspots derived using Hinode SOT/SP observations: Effect of polarimetric noise (1007.2505v1)

Abstract: The accuracy of Milne-Eddington (ME) inversions, used to retrieve the magnetic field vector, depends upon the signal-to-noise ratio (SNR) of the spectro-polarimetric observations. The SNR in real observations varies from pixel to pixel, therefore the accuracy of the field vector also varies over the map. The aim of this work is to study the effect of polarimetric noise on the inference of magnetic field vector and the magnetic non-potentiality of a real sunspot. To this end, we use Hinode SOT/SP vector magnetogram of a real sunspot NOAA 10933 as an input to generate synthetic Stokes profiles under ME model assumptions. We then add normally-distributed polarimetric noise of the level 0.5\% of continuum intensity to these synthetic profiles and invert them again using ME code. This process is repeated 100 times with different realizations of noise. It is found that within most of the sunspot area (> 90% area) the spread in the (i) field strength is less than 8 Gauss, (ii) field inclination is less than 1 degree, and (iii) field azimuth is less than 5 degrees. Further, we determine the uncertainty in the magnetic non-potentiality of a sunspot as determined by the force-free parameter alpha_g and Spatially Averaged Signed Shear Angle (SASSA). It is found that for the sunspot studied here these parameters are alpha_g = -3.5 +/- 0.37 (x 10^{-9} m^{-1}) and SASSA = -1.68 +/- 0.014 degrees. This suggests that the SASSA is a less dispersion non-potentiality parameter as compared to alpha_g. Further, we examine the effect of increasing noise levels viz. 0.01, 0.1, 0.5 and 1% of continuum intensity and find that SASSA is less vulnerable to noise as compared to alpha_g parameter.