Comparing First Ionisation Potential bias diagnostics in the solar atmosphere

This study evaluates three different Extreme ultraviolet Imaging Spectrometer (EIS) line pair diagnostics for measuring the First Ionisation Potential (FIP) bias in the solar atmosphere, demonstrating that while signal-to-noise cutoffs influence the distribution's high-value tail, the median FIP bias remains stable, thereby highlighting the need for a nuanced approach rather than a one-size-fits-all methodology.

The Gist

The Solar "Recipe" Check: Why One Measurement Isn't Enough

Imagine the Sun as a giant, bubbling pot of soup. In this soup, there are different ingredients (elements) like Hydrogen, Helium, Iron, and Silicon. Scientists have long known that the "soup" at the bottom of the pot (the photosphere, or the Sun's visible surface) has a different recipe than the "steam" rising from the top (the corona, or the Sun's outer atmosphere).

Specifically, the steam tends to be richer in certain heavy ingredients (like Silicon and Iron) and poorer in others (like Neon and Argon). This phenomenon is called the FIP Effect (First Ionization Potential effect). Think of it like a magical sieve that only lets the "heavy" ingredients float up into the steam while leaving the "light" ones behind.

The ratio of these heavy-to-light ingredients is called the FIP bias. It's a vital clue for scientists trying to understand where solar wind comes from and how the Sun's weather affects Earth.

The Problem: One Size Does Not Fit All

For a long time, scientists have tried to measure this "recipe" using a single, standard tool. They've assumed:

• Active Regions (sunny spots with strong magnetic fields) always have a high FIP bias (a very "heavy" recipe).
• Quiet Sun (calmer areas) always has a medium bias.
• Coronal Holes (dark, empty spots) always have a low bias (a "light" recipe).

It's like assuming every pizza in a restaurant has exactly the same amount of cheese, just because that's the standard recipe. But what if the pizza chef uses different ovens for different pizzas? What if the cheese melts differently depending on the heat?

The Experiment: Checking Three Different Thermometers

In this paper, the researchers (led by David Long and colleagues) decided to stop using just one tool. They used data from the Hinode spacecraft, which acts like a giant camera and spectrometer orbiting the Sun.

Instead of just one measurement, they looked at three different pairs of ingredients to calculate the FIP bias:

1. Si X / S X: Like checking the temperature of the soup at a "medium" heat.
2. Ca XIV / Ar XIV: Like checking the temperature at a "very hot" heat (only found in the hottest parts of the storm).
3. Fe XVI / S XIII: Like checking the temperature at a "warm" heat.

They took a full picture of the Sun and zoomed in on two specific neighborhoods: a Quiet Sun area and a chaotic Active Region.

The Surprising Results

Here is where the story gets interesting. If you just looked at the "average" number, you might think everything was fine. But when they looked at the distribution (the spread of all the individual measurements), they found something surprising:

• The "Heavy" Recipe isn't always Heavy: Even in the Active Regions (which are supposed to be super "heavy" with a bias of ~3), the measurements varied wildly. Some spots were very heavy, but many were only moderately heavy.
• The "Medium" Recipe isn't always Medium: In the Quiet Sun, the bias wasn't a single number like 1.5. It was a whole range of values, with a long "tail" of weirdly high numbers.
• The Tools Matter: The three different pairs of ingredients gave slightly different answers. The "hot" tool (Ca/Ar) saw the Active Region clearly but saw almost nothing in the Quiet Sun (because it was too cold for that tool to work). The "medium" tool (Si/S) saw everything but gave a different average number than the "hot" tool.

The Noise Factor: Static on the Radio

The researchers also tested what happens if you try to listen to the signal through a lot of static (noise). They set a rule: "Only count the data if the signal is loud enough."

• Strict Rule (High Signal-to-Noise): You get fewer data points, but they are very clean. You lose the "weird" high numbers at the edges.
• Loose Rule (Low Signal-to-Noise): You get lots of data points, including some that are just random noise. This creates a long "tail" of fake high values.

The Big Discovery: Even though the shape of the data changed with the noise, the median (the middle number) stayed almost exactly the same. This means that while the "weird" outliers exist, the core truth of the recipe remains stable. However, relying on a single average number hides the fact that the data is messy and varied.

The Takeaway: Stop Using a "One-Size-Fits-All" Label

The main message of this paper is that we need to stop treating the Sun's atmosphere like a simple, uniform block.

• Don't just say: "Active regions have a FIP bias of 3."
• Do say: "Active regions have a FIP bias that mostly sits between 2.5 and 4.2, but it varies depending on how hot the plasma is and which chemical ingredients you are measuring."

The Analogy:
Imagine you are trying to describe the temperature of a room.

• The Old Way: You stick one thermometer in the corner and say, "It's 72°F."
• The New Way: You realize the room has a fireplace, an open window, and a drafty door. You use three different thermometers. You find that near the fire it's 90°F, near the window it's 60°F, and in the middle it's 75°F. You realize that saying "It's 72°F" is misleading. You need to describe the range and the distribution of temperatures to truly understand the room.

Why Does This Matter?

The Sun shoots out a constant stream of particles called the solar wind. This wind can mess up satellites, GPS, and power grids on Earth. To predict space weather, scientists need to know exactly where this wind is coming from and what it's made of.

If we use a "one-size-fits-all" recipe, we might misidentify the source of a dangerous solar storm. By understanding that the FIP bias is a complex, varied distribution rather than a single number, scientists can better track the plasma from the Sun's surface all the way to Earth, helping us predict and prepare for space weather.

In short: The Sun is more complex than we thought. To understand its "soup," we need to taste it with multiple spoons and accept that the flavor varies from spoon to spoon.

Technical Summary

Here is a detailed technical summary of the research article "Comparing First Ionisation Potential bias diagnostics in the solar atmosphere."

1. Problem Statement

The solar atmosphere exhibits a distinct difference in elemental abundance between the photosphere and the corona, known as the First Ionisation Potential (FIP) effect. Low-FIP elements (ionization potential < 10 eV) are enriched in the corona relative to high-FIP elements. This enrichment is quantified as the FIP bias (the ratio of low- to high-FIP abundances).

While the FIP bias is a critical diagnostic for tracing plasma origin (e.g., distinguishing solar wind sources) and understanding fractionation mechanisms (likely driven by ponderomotive forces), current practices rely on simplistic, single-value assumptions for different solar regions:

• Coronal Holes: FIP bias $\sim$ 1 (photospheric composition).
• Quiet Sun (QS): FIP bias $\sim$ 1.5–2.
• Active Regions (AR): FIP bias $\sim$ 3.

Furthermore, different instruments and studies use different spectral line pairs to estimate FIP bias, each with varying temperature sensitivities. There is a lack of comprehensive comparison regarding how these different diagnostics behave across distinct solar regions and how data filtering (specifically Signal-to-Noise Ratio, SNR) impacts the resulting distributions.

2. Methodology

The authors utilized full-disk mosaic observations from the Extreme ultraviolet Imaging Spectrometer (EIS) onboard the Hinode spacecraft, taken between October 18–20, 2015.

Data Processing:

• Regions of Interest (ROI): A Quiet Sun region and an Active Region were identified and isolated from the full-disk mosaic.
• Diagnostics Compared: Three distinct FIP bias diagnostics were analyzed:
  1. Si X / S X: Sensitive to plasma at $T \sim 1\text{--}2$ MK (coronal temperatures). Derived using a Differential Emission Measure (DEM) approach via Markov Chain Monte Carlo (MCMC) fitting.
  2. Fe XVI / S XIII: Sensitive to plasma at $T \sim 2\text{--}3$ MK. Also derived using the MCMC DEM approach.
  3. Ca XIV / Ar XIV: Sensitive to hotter plasma at $T \sim 3.5$ MK (AR cores). Derived using a direct intensity ratio method, as the contribution functions of these lines are similar.
• Statistical Analysis: Instead of averaging pixel values, the authors employed Kernel Density Estimation (KDE) to visualize the full distribution of FIP bias values within each ROI.
• SNR Filtering: The study systematically tested the impact of SNR cutoffs (ranging from 5 down to 0.0001) on the data. Pixels were filtered based on the ratio of spectral intensity to the standard deviation of residuals ($SNR = I_{fit} / \sigma_{res}$).

3. Key Results

A. Diagnostic Variability and Temperature Sensitivity

• Si X/S X: Detected measurable FIP bias in both Quiet Sun and Active Regions. The Active Region distribution showed a median of 3.29, while the Quiet Sun median was 2.64.
• Ca XIV/Ar XIV: Primarily detected in Active Regions (hot plasma). The median FIP bias for the AR was 2.21, significantly lower than the expected value of $\sim$3, while the Quiet Sun showed values $\le$ 1 (with many pixels effectively zero signal).
• Fe XVI/S XIII: Detected in both regions but with a narrower distribution. The median for the AR was 2.12, comparable to the Quiet Sun median of 1.81.
• Conclusion: No single diagnostic yields a uniform "textbook" value (e.g., 3 for ARs). The measured bias depends heavily on the temperature sensitivity of the specific line pair used.

B. Distribution vs. Single Values

• The distributions of FIP bias values are broad and asymmetric, often extending to values as high as 10, rather than clustering tightly around a mean.
• The Quiet Sun distributions consistently show a "tail" of values greater than 2, challenging the assumption that QS is strictly low-bias.
• The median value remains robust across different SNR cutoffs, whereas the mean is skewed by the high-value tails introduced by noisy pixels.

C. Impact of Signal-to-Noise (SNR) Filtering

• High SNR Cutoffs (e.g., 0.1): Remove a significant number of pixels, particularly in Quiet Sun and coronal holes for high-temperature diagnostics (Ca XIV/Ar XIV, Fe XVI/S XIII). This eliminates the "tail" of spurious high FIP bias values caused by noise.
• Low SNR Cutoffs (e.g., 0.0001): Retain noisy pixels, resulting in longer tails in the KDE distributions with artificially high FIP bias values.
• Key Finding: While SNR filtering drastically changes the number of usable pixels and the shape of the distribution tails, the median FIP bias value remains largely unchanged. This suggests that the median is a robust metric even in noisy data, provided the distribution is not averaged.

4. Key Contributions

1. Multi-Diagnostic Comparison: The study provides a direct, side-by-side comparison of three commonly used FIP bias diagnostics (Si X/S X, Fe XVI/S XIII, Ca XIV/Ar XIV) on the same dataset, highlighting their distinct temperature sensitivities and resulting value discrepancies.
2. Distribution-Based Analysis: It advocates for moving away from single-value reporting (e.g., "AR = 3") toward reporting distributions (quartiles and medians). This approach captures the intrinsic variability and asymmetry of plasma composition within solar regions.
3. SNR Sensitivity Analysis: The paper quantifies how SNR filtering affects FIP bias maps. It demonstrates that while low SNR introduces spurious high-bias tails, the median statistic is resilient, offering a practical guideline for data processing.
4. Methodological Rigor: It highlights the importance of checking spectral line fits (as shown in Figure 4) and understanding the limitations of intensity-ratio vs. DEM-based derivation methods.

5. Significance

• Solar Wind Connection: Accurate FIP bias diagnostics are essential for the Solar Orbiter mission, which aims to link in-situ solar wind measurements with remote-sensing observations of the Sun. Misinterpreting FIP bias values due to diagnostic choice or data filtering could lead to incorrect source identification for solar wind streams.
• Plasma Physics: The results suggest that plasma fractionation is more complex than a simple "on/off" switch based on magnetic field strength. The broad distributions imply a mix of plasma sources or dynamic processes within a single pixel or region.
• Future Instrumentation: The findings underscore the need for next-generation instruments (like Solar-C EUVST) to provide higher spectral and spatial resolution to better resolve these distributions and reduce the reliance on noisy, low-SNR data.
• Scientific Practice: The authors argue against a "one-size-fits-all" approach to FIP bias. Researchers must select diagnostics appropriate for the specific temperature regime of the region being studied and report statistical distributions rather than single averages to accurately characterize solar plasma evolution.