SJET: An Interactive Solar Jet Extraction Tool

This paper introduces SJET, an interactive Python-based tool that combines multiple thresholding algorithms with morphological operations and a novel circular-region method for endpoint detection to standardize the extraction and geometric analysis of complex solar jet features across diverse observational conditions.

The Gist

Imagine the Sun as a giant, chaotic fireworks factory. Sometimes, instead of a random explosion, it shoots out long, focused streams of super-hot gas, like a laser-guided hose spraying water. Astronomers call these solar jets. They are crucial for understanding how the Sun heats its outer atmosphere and how it pushes solar wind out into space.

However, studying these jets is like trying to count and measure individual raindrops in a hurricane while wearing foggy glasses. The jets are messy, they twist and turn, and sometimes they look like they are coming from the side, and other times they look like they are coming straight at us, making them hard to spot.

Enter SJET (Solar Jet Extraction Tool). Think of SJET not as a robot that does all the work for you, but as a high-tech, interactive magnifying glass and ruler that helps scientists do their job faster and more consistently.

Here is how it works, broken down into simple concepts:

1. The Problem: The "Needle in a Haystack"

For years, scientists had to look at thousands of images of the Sun and manually draw lines around these jets to measure them. It was slow, boring, and everyone did it slightly differently. If Scientist A measured a jet one way and Scientist B measured it another way, their data didn't match, making it hard to compare results.

2. The Solution: A "Smart Sketchpad"

SJET is a computer program (built with Python) that acts like a smart sketchpad. Instead of forcing a computer to guess where a jet is (which often fails because the Sun is so noisy and complex), SJET puts the scientist in the driver's seat but gives them superpowers.

• The "Filters" (Thresholding): Imagine you have a photo of a jet that is a bit blurry or has a messy background. SJET offers five different "filters" (like photo editing apps) to make the jet stand out. You can choose a filter that brightens the jet, one that ignores the background noise, or one that finds the perfect contrast automatically.
• The "Eraser and Glue" (Morphology): Sometimes the jet looks broken or has little holes in it. SJET has tools to "glue" the broken pieces back together and "erase" tiny specks of noise that aren't part of the jet. It's like using a digital glue stick to fix a torn paper cutout.

3. The "Magic Ruler": Finding the Start and End

One of the hardest parts of a jet is knowing exactly where it starts and where it ends, especially if it's curvy.

• The Analogy: Imagine a jet is a snake. SJET finds the two furthest points of the snake (the head and the tail). Then, it draws two giant circles around those points.
• The Trick: The circle around the "head" (where the jet starts) will cover more of the snake's body than the circle around the "tail" because the jet is usually wider at the base. SJET uses this math trick to objectively decide, "Okay, this is the start, and that is the end," removing human guesswork.

4. The "Flexible Ruler": Measuring the Curve

Jets aren't always straight; they often curve like a banana.

• SJET draws a smooth, flexible line (called a Bézier curve) right down the middle of the jet, following its twists and turns perfectly.
• Once the line is drawn, it can instantly measure:
  • Length: How long is the jet?
  • Width: How thick is it?
  • Curvature: How much does it bend?
  • Speed: By looking at a series of images, it can calculate how fast the jet is flying.

5. Why This Matters: The "Standardized Recipe"

Before SJET, if you asked 10 scientists to measure the same jet, you might get 10 different answers because they all used different methods.

• SJET is the Standardized Recipe: It ensures that everyone uses the same "ingredients" and "steps." Even though a human still has to click a few buttons to say, "Yes, that's the jet I want to study," the way the tool measures it is consistent.
• This allows scientists to build huge databases of jet measurements. Instead of studying just one or two jets, they can now study thousands to find patterns, much like how a meteorologist needs thousands of weather reports to predict a storm.

The Bottom Line

SJET is a bridge between human intuition and computer speed. It acknowledges that the Sun is too complex for a computer to handle alone, but too vast for humans to measure manually. By giving scientists a toolkit to interactively clean up images and measure jets with a "magic ruler," it helps us understand the Sun's explosive behavior better, faster, and more accurately than ever before.

In short: It turns the chaotic, blurry mess of solar data into clean, measurable facts, helping us decode the Sun's secret language.

Technical Summary

1. Problem Statement

Solar jets are dynamic, collimated plasma flows crucial for coronal heating and solar wind acceleration. However, their complex and diverse morphologies pose significant challenges for automated analysis:

• Morphological Complexity: Jets exhibit varied structures (e.g., C-shaped, bidirectional, fragmented) that are difficult to capture with a single universal algorithm.
• On-Disk vs. Limb: While limb jets are relatively easy to identify, on-disk jets suffer from projection effects and background contamination, making automatic extraction highly error-prone.
• Methodological Inconsistency: Existing studies rely heavily on manual visual identification or disparate semi-automated methods (e.g., SAJIA, AJIA, citizen science), leading to a lack of standardized workflows. This hinders large-sample statistical studies and the comparison of results across different datasets.
• Limitations of Current Automation: Fully automated deep learning approaches (like U-NET) have improved precision but still struggle with the subtle intensity contrasts and rapid temporal evolution of jets against variable backgrounds.

2. Methodology: The SJET Tool

The authors present SJET (Solar Jet Extraction Tool), an interactive Python-based software designed to bridge the gap between fully manual and fully automated analysis.

Core Architecture

• Framework: Built on open-source libraries including SunPy (data handling), scikit-image and OpenCV (image processing), SciPy/NumPy (geometry), and Tkinter/Matplotlib (GUI).
• Workflow: Follows an "Input-Processing-Interaction-Output" loop. Users import FITS files, define a Region of Interest (ROI), and iteratively adjust parameters with real-time visual feedback.

Key Algorithms

1. Multi-Algorithm Thresholding: SJET integrates five distinct thresholding methods to handle varying data characteristics:
   • Manual: Slider-based percentage adjustment.
   • Otsu: Automatic thresholding based on bimodal histogram variance.
   • Adaptive: Handles non-uniform backgrounds by considering local pixel neighborhoods.
   • Percentile: Sets thresholds based on intensity distribution percentiles (useful for long-tail distributions).
   • Log-Enhanced: Applies logarithmic transformation to enhance weak signals in high dynamic range data.
2. Morphological Optimization:
   • Opening/Closing: Removes noise and fills internal holes/connects fragments.
   • Region Merging: Uses distance-based and intensity-based filtering. Includes a "Force Merge" option (using convex hulls) for complex structures, with a warning about potential geometric bias in C-shaped jets.
3. Geometric Parameter Extraction (Novelty):
   • Start/End Point Identification: Instead of arbitrary selection, SJET uses a circular region method. It identifies the two most distant pixels in the binary mask, creates two overlapping circles, and counts pixels within them. The region with more pixels (wider base) is identified as the start point, objectively determining propagation direction.
   • Axis Modeling: The jet axis is modeled using a quadratic Bézier curve. The control point is calculated based on the geometric center of the mask but can be manually overridden by the user to correct for curved or force-merged structures.
   • Parameter Calculation: Calculates length, width (via perpendicular cross-sections along the Bézier curve), curvature ($\kappa$), and deflection angles.
   • Gaussian FWHM: Includes a built-in tool to fit Gaussian profiles to intensity slices perpendicular to the axis, providing a traditional width metric comparable to legacy studies.

3. Key Contributions

• Standardized Interactive Workflow: SJET provides a unified platform that standardizes the extraction process, ensuring reproducibility through comprehensive metadata preservation (FITS, PNG, and ASCII logs of all parameters).
• Objective Directionality: The circular region algorithm offers a mathematically objective method to determine jet propagation direction, reducing user bias in start/end point selection.
• Hybrid Human-Machine Approach: By combining automated segmentation with real-time user interaction, SJET leverages human expertise to handle complex morphologies that fully automated algorithms fail to process correctly.
• Dual-Width Analysis: The tool provides both mask-based boundary width (total morphological extent) and Gaussian FWHM (intensity core width), offering complementary physical insights.

4. Results and Validation

The tool was validated using two distinct datasets:

1. Solar Orbiter/EUI (HRIEUV 174 Å): High-resolution (195 km/pixel) data of an active region jet.
   • Outcome: Successfully separated the jet from overlapping coronal loops using log-thresholding. Extracted a velocity of 487 km/s and a width of 48 pixels.
2. SDO/AIA (304 Å): Lower resolution (435 km/pixel) data featuring complex, bidirectional "two-sided-loop" jets.
   • Outcome: Demonstrated the necessity of morphological closing operations to connect fragmented structures. Successfully quantified bidirectional propagation (forward: 141 km/s; backward: 10 km/s).

• Comparison with Traditional Methods:
  • SJET's Gaussian FWHM measurements (e.g., $8.00 \pm 2.17$ pixels) were consistent with traditional slice-fitting methods.
  • The tool highlighted that mask-based width (12.70 pixels) captures the full spatial extent, while FWHM captures the bright core, showing a ratio of ~63% consistent with physical expectations.

5. Significance and Future Impact

• Statistical Foundation: SJET addresses the critical need for standardized workflows in solar jet research, enabling reliable large-sample statistical studies that were previously hindered by methodological inconsistencies.
• Integration with CoSEE-Cat: The tool is poised for integration into the Comprehensive Solar Energetic Electron event Catalogue (CoSEE-Cat), facilitating the correlation of jet parameters with solar energetic particle acceleration.
• Community Resource: The authors plan to release the full source code, documentation, and test datasets on GitHub, fostering community adoption and further development.
• Design Philosophy: SJET serves as a reference model for handling other complex solar phenomena, demonstrating that integrating multiple algorithms into an interactive platform is a robust strategy for balancing automation efficiency with scientific rigor.

In conclusion, SJET does not claim to be a fully autonomous detector but rather a standardized, interactive extraction tool that empowers researchers to consistently and reproducibly analyze the complex morphologies of solar jets across different instruments and observational conditions.