Limb Shift of the Fe I 6569 Å line on the Sun

TL;DR

This study addresses how convective granulation shapes the center-to-limb Doppler shifts in the Fe I 6569 Å line, using CHASE observations and Bifrost-based radiative-MHD simulations. It compares two limb-shift extraction methods—spectral-averaging and velocity-averaging—and finds both reproduce the observed trend at CHASE resolution, while revealing a convective blueshift offset of about $0.36 km s^{-1}$ that calibrates the wavelength reference. At high spatial resolution, the methods diverge, with the Dopplergram component dominating toward the limb and the contrast component waning, leading to a two-component interpretation (contrast and Dopplergram). The results quantify how spatial and spectral resolution shape limb-shift measurements and clarify the physical origin of the effect via density inhomogeneities and corrugation of the line-formation surface, providing guidance for applying limb-shift corrections in solar-disk spectroscopy. The findings have practical implications for precision Doppler measurements and the interpretation of solar photospheric dynamics with current and future high-resolution instruments.

Abstract

The convective motions of solar granules generate a center-to-limb variation of Doppler velocity in the photospheric lines, known as the limb shift effect. This study presents a comprehensive analysis of this effect for the Fe I 6569 Å line using both observational data from the CHASE mission and numerical simulations from the Bifrost code. We employ two independent methods to derive the limb shift curve: a spectral-averaging method (Method 1) and a velocity-averaging method (Method 2). By comparing synthetic and observed data, we determine the convective blueshift, which is not accounted for in the CHASE observations. The simulations reproduce the observed trends for both methods at the instrument's spatial resolution of 1.2 arcsec. However, at resolutions below 1 arcsec, Method 2 produces limb-shift curves that depart significantly from both Method 1 results and traditional limb-shift profiles, whereas Method 1 remains in agreement with classical behavior. Further analysis finds that the results from Method 1 comprise two distinct components: a contrast contribution caused by the correlation between velocity and line depth, and a Dopplergram contribution caused by density inhomogeneities and corrugation effects.

Figures (12)

• Figure 1: Left: solar image in the Fe I line center. The concentric circles from center to limb represent the $\mu$ varying from 0.9 to 0, which divide the disk into 10 parts. Right: Dopplergram of the same line derived by bisectors; the color denotes the blueshift and redshift. Three white rectangles in both panels are the sunspot regions deduced when averaging.
• Figure 2: Limb shift curves and their error bar as functions of the cosine of the heliocentric angle calculated with two methods. Positive value denotes blueshift. Velocity at the extreme limb is not shown.
• Figure 3: Left: Synthetic Fe intensity images in the far wing (+0.78 Å). Middle: Corresponding images in the line center. Right: Derived Dopplergrams and their colorbars. The images are shown with native simulation spatial resolution (top) and with the CHASE spatial resolution (bottom).
• Figure 4: Limb shift curves and their error bars from simulation and observation derived from Method 1 and Method 2, which are shown both with (left) and without calibration (right).
• Figure 5: Left: Velocity at the disk center derived from Method 1 (blue) and Method 2 (orange); vertical dashed line denotes the CHASE's spatial resolution ($1.2^{\prime\prime}$). Right: Limb shift curves of Method 1 (black dashed line) and Method 2 (colored solid line).