A Filamentous Thread Network in Lunar Regolith, Visual Documentation and a Proposed Structural Mechanism for Multi Scale Surface Patterns on the Moon

Introduction A former observation I carried out (Observations of Multi Scale Surface Patterns on the Moon: Recurring Morphological Features in Apollo and LRO Imagery Across Millimetre and Kilometre Scales.Zenodo. DOI: 10.5281/zenodo.20485712) documented recurring surface patterns in Apollo photography and NASA Lunar Reconnaissance Orbiter imagery. Investigating the patterns I saw repeating at different scales, I identified spatial organisation distribution as non random and clustered from millimetres to kilometres. Clark Evans nearest neighbour R analysis and Donnelly corrected z-scores confirmed this organisation was present in both original and enhanced imagery, and across two independent imaging instruments, at p < 0.001 in all cases. The study showed that the spatial arrangement of surface features is non random, but did not address what structural mechanism might be responsible for producing this organisation. This study begins to investigate that question. During continued examination of Apollo close up surface photography at millimetre scale, I identified recurring thread like linear features running through the lunar regolith, interlocking and creating a layerd mesh organisation, within the patterns. Some of these features interact with objects within the scene, passing over rocks, casting shadows, and in several cases appearing to pass directly through individual grains and glass globules, emerging on the other side. The threads interacting with individual grains rather than lying on the surface, distinguishes them from lens artefacts or film dust, which cannot interact with scene geometry. These features are documented visually in the accompanying image files with original, sharpened, and annotated versions provided for each example. Spatial statistical analysis was applied to the distribution of linear edge segments detected across two independent Apollo close up images using Canny edge detection and Hough line transformation. Clark Evans nearest neighbour analysis with Donnelly edge correction returned R = 1.045 (z = 4.07, p = 0.000047) for Image 1 and R = 1.145 (z = 16.33, p < 0.000001) for Image 2. Both results are statistically significant at p < 0.001 If the distribution were truly random, the probability of a result this extreme or more extreme by chance is less than one in a thousand. The results indicate a dispersed or regular spatial pattern, linear features are more evenly spaced than a random distribution would predict. Ripley's L function remained positive at every spatial scale tested from 10 to 300 pixels in both images, confirming that the regularity is not confined to a single characteristic scale but persists across the full range of scales examined. These results are consistent across two independent images of different areas of the lunar surface. The arrangement of grains around the thread-like features bears visual resemblance to two distinct biological processes documented in terrestrial organisms, biomineralisation, in which filamentous organisms precipitate mineral coatings along their surfaces, and biological agglutination, in which organisms collect and bind environmental grains around a secreted scaffold. These morphological comparisons are noted as observations only, do not constitute evidence of biological origin, and are discussed further in the visual observations section of this record Source Imagery and Methodology Two publicly available Apollo close up lunar surface photographs were used as the primary source imagery for this study. Both images were taken at millimetre scale of the lunar regolith surface at the Sea of Tranquility during the Apollo 11 mission and are accessible via the NASA/Flickr archive under image identifiers 54292241841 and 54292490693. The full resolution original files were used throughout. Closeup1 54292241841_apollo_11_thread1and2_mm.jpg Closeup2_54292490693_apollo_11_thread3_mm.jpg A third image, 54290746152, is included as surface scale context photography from the same Apollo 11 mission, showing recurring pattern morphologies at centimetre scale. This image is used for cross scale comparison only and was not subjected to spatial statistical analysis. apollo_11_surface_patterns.jpg Image enhancement was applied using a blur then sharpen technique, the image is sharpened slightly, then lightly blurred to suppress noise, then sharpened to bring out fine detail. Both original and enhanced versions are provided for each visual example so that the reader can verify that observed features are present in the source data and not introduced by processing. Spatial statistical analysis was performed computationally using Canny edge detection to identify contrast boundaries, followed by Hough line transformation to extract linear segments. The midpoint of each detected segment was treated as a point in the spatial analysis. Clark Evans nearest neighbour R was calculated with Donnelly edge correction to account for boundary effects , the Donnelly correction adjusts the expected nearest neighbour distance to compensate for the fact that points near the image edge have fewer potential neighbours, which would otherwise bias the result. The z-score measures how many standard deviations the observed result is from what would be expected under complete spatial randomness, a large positive or negative z-score indicates the pattern is unlikely to have occurred by chance. Ripley's L function was computed across radii from 10 to 300 pixels. All analysis was performed on the full image content with no selective cropping of features. Visual Observations While examining Apollo 11 close up surface photography at millimetre scale, I identified recurring thread like features running through the lunar regolith. These features are visible across the full scene and appear to form an interlocking network, weaving between and around grain clusters and creating a layered mesh organisation. I propose it may be the structural mechanism underlying the surface patterns which I documented in my previous study (Evidence of Disordered Hyperuniform Organisation Across Millimetre and Kilometre Scales. Zenodo. DOI: 10.5281/zenodo.20485712) Not all thread like features visible in the imagery can be confirmed as genuine scene content, some may be artefacts of the film, lens, or digitisation process. However, a subset of features can be distinguished from artefacts by their demonstrable interaction with objects within the three dimensional scene. These include features that come out of crevices, pass over rocks and cast visible shadows, features that change direction around grain clusters, and features that appear to pass directly through individual grains and glass globules, emerging on the other side. A thread that passes through a grain cannot be a surface artefact. Three documented examples are provided as separate image files, each showing the original image section, the same section enhanced with sharpening, and an annotated version with the identified thread features marked. There are many thread features that are not marked as I only marked the finest dark threads that are easily viewable, however, I believe there are more which are lighter in appearance and more difficult to isolate with confidence at this resolution. thread1-annotated-apollo_11_closeup_mm.jpg thread2-annotated_apollo_11_closeup_mm.jpg thread3-annotated_apollo_11_closeup_mm.jpg In addition to the thread like linear features, a separate image document is included showing just a few examples of recurring surface patterns identified across Apollo 11 close up photography at millimetre scale. These patterns, including web like, curvilinear, and mesh organised grain arrangements, which are consistent with the morphological types documented in my previous study and are presented here as further visual evidence that the thread network described above produces a characteristic and a recurring surface organisation. apollo_11_closeup_mm_patterns_sharpened_sections.jpg An annotated image of a particularly clear example of an ovate mesh structure is also included, showing the internal mesh geometry and its cross-scale recurrence at centimetre and kilometre scales in independent Apollo surface and LRO photography. mesh_structure_ovate_apollo11_closeup_mm.jpg I note that at these larger scales the ovate form appears more geometrically defined, structurally consolidated, and visually stable than at millimetre scale. My observation has been that the larger scale examples may represent more mature expressions of the same structural process, where the mesh organisation has had more time or material to develop into a stronger and more clearly bounded form. Additional Observations During the course of my investigation, two recurring features were observed particularly in image Closeup2_54292490693_apollo_11_thread3_mm.jpg which I have not yet enhanced and annotated, pending further characterisation. Given the relevance of these observations to the findings presented, they are documented here. First, a moss like webbing material is observed across the regolith surface thicker in some areas than others. This material appears to form a connected network, it appears to be produced at crevice sites and sheltered interfaces, and has a texture and spatial organisation that differs from the surrounding regolith. Its distribution appears non random, resembling webs. Second, glass globules present in the same frames appear in several instances to show spatial association with the above described material, with the network apparently extending toward and in some cases into contact with globule surfaces. The glass globules seem to have threads going running through them. This spatial relationship could be consistent with the network utilising glass globules as localised resources. Given that recent work has established that lunar impact glass beads contain substantial solar wind derived water releasable at mild temperatures (He et al., 2023, Nature Geoscience), I will follow this documentation with examples of both features. Morphological Comparisons - a Note on Glass Filaments Before presenting morphological comparisons, I want to note an observation that I find difficult to reconcile with purely physical explanations. Impact glass threads and agglutinate glass filaments are well documented in lunar regolith and remain legitimate candidate explanations for the thread like features observed here. However, glass threads form passively drawn out by impact ejection forces and cooled in place. They do not navigate around obstacles. The features I observe appear to curve around grain clusters, emerge from rock crevices, and continue through individual grains in a consistent direction. This directional behaviour, the thread maintaining continuity and orientation across and through scene geometry rather than simply lying where it landed, is what I find difficult to explain through passive physical processes alone. It is this specific property that leads me to consider the biological morphological comparisons that follow. Morphological Comparisons The visual morphology of the thread like features and the ovate mesh structures documented in this record bear resemblance to structures produced by two distinct biological processes documented in terrestrial organisms. These comparisons are offered as morphological observations only and do not yet form evidence of biological origin Biomineralisation Some filamentous organisms precipitate mineral coatings directly along their surfaces, using the filament as a scaffold around which mineral material accumulates. The result is a mineralised thread, chemically transformed from the original biological filament but retaining its linear form and spatial organisation. Euendolithic microorganisms are a particularly relevant example, being documented as capable of penetrating directly into mineral grains and precipitating mineralised coatings along their filament surfaces within the grain interior. This process is consistent with the observation of threads appearing to pass through individual grains rather than around them. I also note that lunar regolith grains, particularly agglutinate glass particles, contain interconnected internal vesicle networks formed by gas bubbles from impact melting events, as well as vesicular structures associated with volcanic glass. These pre existing internal void networks could in principle provide pathways through which a filamentous organism might extend, using the grain's own internal structure as a channel rather than boring through solid mineral from scratch. If so, the thread passing through a grain may reflect colonisation of an existing internal pathway rather than active penetration of solid material. Statistical Analysis Spatial statistical analysis was applied to two independent Apollo 11 close up images of the lunar surface (image identifiers 54292241841 and 54292490693). Linear edge segments were detected across both full images using Canny edge detection followed by Hough line transformation. The midpoint of each detected segment was treated as a point in the spatial analysis. No selective cropping or feature selection was applied, all detected linear edges across the complete image were included regardless of whether they correspond to thread like features specifically. Clark Evans nearest neighbour R was calculated with Donnelly edge correction for both images. Ripley's L function was computed across radii from 10 to 300 pixels. Results are summarised in the table below and in the accompanying figure files. Spatial analysis summary Metric Image 1 (54292241841) Image 2 (54292490693) Reference (random) Clark Evans R 1.0451 1.1454 1.000 Z(Donnelly corrected) 4.0699 16.329 0 p-value 4.70e-05 <0.000001 - Significant at p < 0.0001? Yes Yes - Pattern Dispersed / regular Dispersed /regular Ripley's L - all positive? Yes Yes No (L = 0) Lines detected 3,191 3,790 Results - Spatial organisation of linear features Both images returned Clark Evans R values above 1.0 — R = 1.045 for Image 1 and R = 1.145 for Image 2 nearest_neighbour_analysis .png This indicates that the detected linear features are more evenly spaced than a random distribution would predict. This is described as a dispersed or regular spatial pattern. Both results are significant at p < 0.001, meaning that if the distribution were truly random, the probability of observing a result this extreme by chance is less than one in a thousand. Ripley's L function remained positive at every spatial scale tested from 10 to 300 pixels in both images. This confirms that the regularity is not confined to a single characteristic scale but persists across the full range of scales examined, from fine grain level spacing up to broad patch level organisation. ripleys_L_analysis.png The multi scale persistence is consistent with a hierarchically organised structure rather than a single process physical explanation, which would typically produce a signal at one characteristic scale only. Both results are consistent across two independent images of different areas of the lunar surface, which strengthens the case that the spatial organisation detected reflects a genuine property of the regolith rather than a single image artefact. It should be noted that the spatial analysis measured the distribution of all detected linear edge segments across both images, not thread like features specifically. The algorithm cannot distinguish between threads, grain boundaries, surface cracks, and other linear features. The statistical results therefore reflect the spatial organisation of all linear edges in the image rather than the thread network in isolation. This is discussed further in the methodology and limitations sections of the documentation. Methodological Limitations The following limitations are acknowledged and should be considered when evaluating the observations and results presented in this record. Image resolution and compression. The source imagery is photographic film that has been digitised and compressed as JPEG. Compression artefacts and resolution constraints affect the visibility of fine features and may influence statistical results. I excluded a mesh analysis/ crossing angle analysis from this documentation because it produced a dominant peak at near zero degrees in both images, which is consistent with JPEG ringing artefacts alongside high contrast edges. The Clark Evans and Ripley's L analyses are considered relatively robust to compression artefacts as both methods are based on spatial distances between detected points rather than edge geometry. The Hough line transformation is more susceptible as it detects linear patterns and JPEG block boundaries may introduce a bias toward certain orientations, though the predominantly long segment lengths detected suggest genuine extended features rather than short block boundary artefacts. Thread and crack distinction. Fine surface cracks in the regolith can resemble thread like features. This distinction is further complicated by the observation that the threads appear to form the borders of the mesh layers themselves, meaning a feature that presents as a crack may in some cases be a thread creating a border, or the two may be aspects of the same underlying structure. Clean visual separation is not always possible at this resolution. Camera and lens distortion. Lens distortion, film flatness variation, and digitisation artefacts may introduce artificial curvature into detected features. The degree of this effect has not been characterised for these specific images. A tortuosity analysis returned a median value of approximately 2.0, consistent with the visually observed winding and non straight character of the thread like features. This result I noted as supportive but treated with caution as lens distortion may artificially inflate curvature measurements and this effect has not been corrected for. Enhancement methodology. The blur and sharpen technique enhances fine detail but may also introduce or emphasise artefacts. Original unenhanced versions are provided for all visual examples so the reader can verify features independently. Statistical methodology. The spatial analysis measured the distribution of all detected linear edge segments across the full image, not filamentous thread like features specifically. The algorithm cannot distinguish between threads, grain boundaries, surface cracks, and other linear features. The results therefore reflect the spatial organisation of all linear edges in the image rather than the thread network in isolation. Visual interpretation. All visual observations are my own interpretation of photographic imagery at the limits of available resolution. Independent verification by other observers is encouraged. Conclusion This study documents the visual identification of recurring thread like linear features in Apollo close up lunar surface photography at millimetre scale. These features appear to form an interlocking network running through the regolith creating a mesh, and I propose that this network is the structural mechanism underlying the multi scale surface patterns I documented previously. Some features can be distinguished from optical artefacts by their demonstrable interaction with scene geometry including cases where threads appear to pass directly through individual grains and glass globules. Spatial statistical analysis of two independent images returned statistically significant results consistent with a dispersed and regularly organised linear network, persistent across all spatial scales tested from 10 to 300 pixels. These results are consistent across both images. The visual morphology of the observed features bears resemblance to structures produced by both biomineralisation and biological agglutination processes in terrestrial organisms. These comparisons are offered as observations only and do not constitute evidence of biological origin. I would like to continue investigating these observations through further visual examination and computational analysis of Apollo close up imagery to see what further evidence emerges. References He, H., Ji, J., Zhang, Y., Hu, S., Lin, Y., Hui, H., Hao, J., Li, R., Yang, W., Tian, H., Zhang, C., Zhong, F., Liu, Y., Cao, X., Zeng, X., & Li, X. (2023). A solar wind-derived water reservoir on the Moon hosted by impact glass beads. Nature Geoscience, 16, 294–300. https://doi.org/10.1038/s41561-023-01159-6 Sørensen, Arezoo Shabani (2026). Observations of Multi-Scale Surface Patterns on the Moon: Evidence of Disordered Hyperuniform Organisation Across Millimetre and Kilometre Scales. Zenodo. DOI: 10.5281/zenodo.20485712