Sector-based morphological analysis of four planetary nebulae (J 320, NGC 2610, A 6672 and PRTM 1) observed with UVIT/ASTROSAT

Four planetary nebulae (PNe), Jonckheere 320, NGC 2610, A 6672, and PRTM 1, have been imaged in the far-ultraviolet (FUV) filter F148W (λeff = 1481 Å) using the Ultraviolet Imaging Telescope (UVIT) onboard ASTROSAT. The FUV images reveal faint extended emission structures surrounding the bright central star. A novel sector-based structural analysis technique was employed to examine the detailed FUV morphologies of PNe. In this method, sector strengths were calculated for each nebula and used to classify their structures. Our results show that Jonckheere 320 exhibits a polypolar morphology, while NGC 2610 and PRTM 1 display spherical structures. In contrast, A 6672 shows irregular or elliptical features in the FUV.

The paper introduces a novel sector-based pair strength modeling technique designed to quantitatively characterize the morphological properties of the four PNe. To ensure reliable detection of faint nebular emission, a careful global background subtraction was applied by averaging the signal from 10 source-free regions distributed across each image, such that the residual emission contains only stellar and nebular contributions. First, the stellar signal was modeled using a Moffat point spread function (PSF), I(r) = I0 (1 + r/γ)(-α), where I0 is the background-subtracted central intensity and the shape parameters γ and α were empirically determined from bright, isolated field stars in the same UVIT images. This PSF represents the expected radial decline of a pure point source in the absence of extended emission. Any observed flux exceeding the PSF model at a given radius was interpreted as a genuine nebular excess emission.

All the PNe were divided into N = 12 equal angular sectors, each spanning 30° in azimuth and centered on the nebula's centroid. For each angular sector, radial profiles were computed in concentric annuli of five-pixel width. The observed flux, Fobs was compared directly with the PSF model to calculate the fractional excess flux, f(r, θ) = (Fobs(r, θ) - FPSF(r)) / FPSF(r), which isolates extended nebular emission beyond wings of the stellar PSF. These fractional excess profiles were radially integrated to obtain a single quantitative measure of extended emission per sector.

To identify large-scale axial symmetries in the nebular emission, the sectoral strengths were combined into opposite-sector pairs. For a total of N = 12 sectors, each sector was paired with its diametrically opposite counterpart. For example, sector 1 (0°–30°) was paired with sector 7 (180°–210°), sector 2 (30°–60°) with sector 8 (210°–240°), and so on. The pair strength of each opposite-sector pair was defined as the sum of the values of the integrated strengths of the two opposing sectors. These pair strengths were then normalized by the total excess signal to obtain fractional pair contributions to represent the relative importance of each symmetry axis in shaping the nebula. This normalization allows a direct comparison between different nebulae and ensures that the sum of all pair fractions equals unity.

The fractional strengths of opposite sector pairs, along with the resulting morphological classifications based on this method, are presented in Table 2. For the first time, the paper provides a quantitative basis for the morphological classification of PNe using the fractional pair strength values. Specifically, a nebula is classified as bipolar if a single opposite-sector pair contributes ≥ 50% of the total excess flux, indicating a dominant symmetry axis. It is classified as polypolar if two or more opposite-sector pairs each contribute between 20% and 30%, consistent with multiple lobe axes or episodic outflows. A spherical morphology is assigned when almost all pairs contribute nearly equally (∼ 15–20%), with the standard deviation of all pair fractions being less than 0.1, reflecting near-isotropic emission. Finally, sources are classified as irregular/elliptical when no clear dominance or uniformity is observed, indicating complex or asymmetric structures.

Based on the above classification scheme, the four PNe are discussed as follows:

• J 320 (PN G190.3−17.7): For this planetary nebula, the nebular emission is distributed across multiple axes rather than being dominated by a single bipolar pair. Two pairs of opposite-sectors contribute substantially to the total flux, with fractional contributions of ∼ 0.256 and 0.202, respectively. This morphology is a hallmark of polypolar PNe. In addition, the presence of multiple pairs of point-symmetric, high-velocity knots surrounding the main lobes is consistent with relics of episodic, precessing jets launched at different epochs and orientations, as proposed for multipolar nebula (Harman et al., 2004).

• NGC 2610 (PN G239.6+13.9): The opposite-sector pair fractions for NGC 2610 exhibit very small dispersion, indicating an almost perfectly isotropic distribution of excess flux and the absence of any preferred symmetry axis. This high degree of uniformity is characteristic of the rare round PNe and matches expectations for slowly evolving AGB-star ejecta that have not been strongly shaped by binary interactions or magnetic fields. High-resolution imaging and kinematic studies in the literature likewise describe NGC 2610 as a smooth, symmetric, high-excitation shell lacking low-ionization knots or filaments (Harrington, 2006), in excellent agreement with the results.

• A 6672 (PN G059.7−18.7): This is an evolved elliptical planetary nebula exhibiting a weakly defined symmetry axis, rather than a strongly bipolar morphology. The observed structure may be influenced by projection effects (Schwarz et al., 1992; Corradi and Schwarz, 1995). There is a clear but non-dominant preferred axis, with one pair of opposite sectors contributing a significantly larger fraction of the excess emission (∼0.32) compared to the other pairs (∼ 0.11 − 0.15). This indicates anisotropic expansion in the absence of a single dominant bipolar axis. Overall, the emission pattern points to mild elongation combined with irregular clumpiness, rather than a canonical bipolar morphology.

• PRTM 1 (PN G243−37.1): For PRTM 1, the sector-based analysis shows nearly uniform pair fractions (∼0.15−0.19) with no dominant axis. This reinforces the overall spherical symmetry of the system. Morphologically, the surrounding planetary nebula is almost perfectly spherical and has several shells. According to (Boffin et al., 2012), it lies in a high-ionization environment characterized by thin inner arcs along the major axis, while lacking low-ionization structures such as jets, knots, or fliers. The absence of these features indicates minimal asymmetrical mass ejection. Taken together, these results indicate that PRTM 1 is best described as a nearly spherical planetary nebula with minor asymmetrical outflows.

Overall, the paper studies the UV morphology and sizes of four PNe observed with UVIT. A novel sector-based pair-strength modeling technique was employed to characterize their structures. The results show that Jonckheere 320 exhibits a polypolar morphology, NGC 2610 and PRTM 1 are spherical, and A 6672 displays an elliptical morphology. In the future, the authors aim to interpret the physical drivers behind these diverse nebular morphologies through lobe-specific mass-loss profile modeling, supported by detailed kinematic analyses and multi-wavelength observations.

Improvements to AI Systems:

1. Enhanced Morphological Classification for Astronomical Imaging: Implement the sector-based pair-strength modeling technique as a trainable feature-extraction module in computer vision systems. This allows AI to quantitatively classify nebular structures (bipolar, polypolar, spherical, irregular) with explicit thresholds (e.g., ≥50% for bipolar, 20–30% for polypolar, std <0.1 for spherical) rather than relying on subjective visual inspection.

2. Robust Point-Spread Function (PSF) Subtraction for Faint Source Detection: Integrate the Moffat PSF fitting (I(r) = I0 (1 + r/γ)(-α)) with empirical parameter estimation from isolated field stars into AI pipelines for image deconvolution. This improves the detection of faint extended emission around bright central sources, reducing false positives in low-signal-to-noise regimes.

3. Adaptive Background Estimation via Multi-Region Sampling: Adopt the global background subtraction method (averaging 10 source-free regions) as a preprocessing layer in AI vision systems. This reduces systematic bias from uneven illumination or detector artifacts, enabling more accurate flux measurements in crowded or heterogeneous fields.

4. Sectoral Symmetry Analysis for General Object Recognition: Generalize the opposite-sector pair strength normalization (sum of paired sectors, normalized to unity) into a symmetry-detection algorithm. This can be applied to any radially symmetric or asymmetric object (e.g., biological cells, industrial defects) to quantify axial dominance or isotropy, improving classification accuracy in non-astronomical domains.

5. Automated Morphology-Driven Physical Inference: Train a neural network to map fractional pair-strength distributions to physical interpretations (e.g., episodic precessing jets for polypolar, slow AGB ejecta for spherical). This enables AI to automatically generate hypotheses about underlying formation mechanisms from imaging data alone, without requiring manual kinematic follow-up.

6. Multi-Wavelength Fusion for Structure Validation: Use the pair-strength metrics as a consistency check across different wavelength bands (FUV, optical, IR). An improved AI system can cross-correlate morphological classifications from multiple filters to flag projection effects or contamination, as done for A 6672, improving robustness of final classifications.

7. Uncertainty Quantification in Morphological Classes: Incorporate the standard deviation of pair fractions (e.g., <0.1 for spherical) as a confidence metric. AI systems can output not just a class label but a confidence score based on how well the observed pair fractions match the defined thresholds, aiding in prioritizing follow-up observations.

What the Improved AI System Can Do:

• Automatically analyze high-resolution images of nebulae, galaxies, or other extended sources to produce quantitative morphological classifications with confidence intervals, replacing manual expert assessment.

• Detect and isolate faint extended structures around bright point sources in any imaging domain (e.g., medical fluorescence microscopy, satellite Earth observation) using PSF-aware subtraction.

• Provide physical interpretations (e.g., jet precession, binary interaction, isotropic ejection) directly from morphological data, accelerating astrophysical discovery.

• Validate classifications across multiple wavelengths or instruments, reducing false positives due to projection or noise.

• Serve as a plug-in module for existing astronomical survey pipelines (e.g., LSST, JWST) to automatically catalog morphologies of thousands of nebulae in real time.

Four planetary nebulae (PNe), Jonckheere 320, NGC 2610, A 6672, and PRTM 1, have been imaged in the far-ultraviolet (FUV) filter F148W (lambda eff = 1481,) using the Ultraviolet Imaging Telescope (UVIT) onboard ASTROSAT. The FUV images reveal faint extended emission structures surrounding the bright central star. A novel sector-based structural analysis technique was employed to examine the detailed FUV morphologies of PNe. In this method, sector strengths were calculated for each nebula and used to classify their structures. Our results show that Jonckheere 320 exhibits a polypolar morphology, while NGC 2610 and PRTM 1 display spherical structures. In contrast, A 6672 shows irregular or elliptical features in the FUV.