Meta Research Bulletin (ISSN 1086-6590, USA)
March 15, 2000
Vol. 9, No. 1, pp.9-13.


Alexey V. Arkhipov
Institute of Radio Astronomy, 4 Krasnoznamennaya str., Kharkov, 61002, Ukraine


As a continuation of preliminary analysis of Clementine lunar images [2], an automated computer survey for ruin-like objects on the Moon has been executed. The finds are now classified and catalogued. It is shown that majority of these formations could be interpreted as collapsed subsurface cavities. Such local formations are puzzling from a  geological perspective, and seem promising candidates for archaeological objects. Besides, such subsurface cavities in polar regions could be interesting for other reasons, such as colonization of the Moon or as lava tubes.


          Our thesis is that the Moon could be used as indicator of extraterrestrial intelligence (ETI) visits to our unique "alive" planet [1]. ETI, as well as NASA, could understand the  strategic significance of Moon-ports for interplanetary communications. That is why it is reasonable to search for alien artifacts (e.g. ETI bases of 0-4 Gyr age) on our satellite.
          Various computer algorithms were proposed and tested for the archaeological reconnaissance of the Moon [2]. About 20,000 Clementine lunar orbital HIRES images have been processed, and a few ruin-like formations were found. Now the results of similar automated survey of all HIRES polar images (~80,000 files) are presented.


          As in the preliminary search, the orbital images of high-resolution camera (HIRES, 9-30 m/pixel) of the Clementine space probe [3] were analysed. Only the polar lunar regions of ±75° to ±90° latitudes were processed in this survey because of their oblique lighting. Basic tests used for image selection are described in [2]. These are the preliminary fractal, rectangular, geological tests and the SAAM filter. Moreover, two new tests were added.
          1. The false alarm probability was decreased by discarding of excessively shadowed images (shadow filter). If >5% of pixels are dimmer than 10% of the maximum brightness amplitude, that image was ignored. Files of <13 KB size were discarded too.
          2. For filtering of shadow interference after the preliminary fractal test, the following "FREX" procedure was used. The fractal a-parameter (a measure of artificiality) was computed as in [2], but for only 1 of every 5 points to speed up the analysis of the images. The average linear regression relating a of the random image set and zenith angle of the Sun (Zsun) was calculated by this simple algorithm. If a of the image was lower than the a-Zsun regression minus 1/2 of its standard deviation, the image was selected.
          In summary, the preliminary fractal test, shadow filter, FREX and rectangular tests selected ~5% of the images as interesting. The selected files were SAAM filtered and tested visually. About 97% of the selections were ignored after SAAM testing. The remaining 128 finds are catalogued (see Table I). Only 47 catalogued images were still selected after  the geological test. Their orientations are different (>10 deg.) from the significant directions of background lineaments (details in [2]). Finally, only 18 files of these 47 were selected as most interesting after the full fractal test for artificiality. Their a-factors are deviate from a-Zsun regression for 100 random images by more than 3 times of its standard deviation. Such images of top interest are marked by asterisks in Table I.
          However, it is not reasonable to ignore other catalogued finds. Human activity, for example, correlates with geological lineaments (e.g. valleys, rivers, deposits around faults). That is why a negative result of geological test is not evidence of natural object; but a positive result would be an additional argument for POSSIBLE artificiality. Moreover, eroded objects could be of low contrast on images taken from orbit. Their fractal properties might be  insufficiently different from background. Hence, the fractal test could undervalue the find. That is why all finds in Table I are of potential interest for archaeological reconnaissance of the Moon.
          The finds in the catalogue are described as systems of simple quasi-rectangular elements: d - depressions; f - furrows; h - quadrangle hills; p - rectangular pattern of craterlets; r - ridges. Thus, an abbreviation such as "dr" in the last column of  Table I means "a system with quasi-rectangular depression(s) and quasi-rectangular ridges". This method of description is convenient for morphological analysis.


          There are two main types of ruin-like formations.
          1. Quasi-rectangular patterns of depressions ("recdeps"). About 69% of ruin-like finds could be attributed to this type. Usually recdep is a cluster of rectangular depressions with rectangular ridges between them. This wafer pattern may be seen in the examples shown  in Fig. 1.

Figure 1
"Recdep" examples in evolutionary order: (a) LHD0316A.083; (b) LHD0470B.112; ( c ) LHD5443Q.291; (d) LHD5472Q.287; (e) LHD5661R.068.

Presumably, an isolated, single rectangular depression could be considered as an extreme form of recdep. Moreover, there are transitional forms from rectangular pattern of craterlets to recdep (e.g. Fig. 1b). So, recdeps in the Table I have descriptions with d, dr or p elements. The typical size of recdeps is ~1-3 km. The size of these rectangular depressions is 0.1-2 km. Quasi-rectangular patterns of depressions correlate with plain terrains (e.g., inter-crater space, or the bottom of the large-scale craters).
          2. Quasi-rectangular lattices of lineaments ("reclats"). These comprise 30% of the  ruin-like formations here. A reclat is a complex of interlacing, broken ridges or furrows, which form the quasi-rectangular pattern (Fig. 2).

Figure 2
"Reclats" examples in evolutionary order: (a) LHD0558B.072; (b) LHD5559Q.279; ( c ) LHD6749R.318; (d) LHD6158R.320.

This morphological type is present in  Table I as complexes of r and/or f elements without d. These lineaments have a typical width of ~50 m and cover territory of ~1 km. Reclats correlate with slopes and hill tops, where the regolith layer must be thinnest. Apparently, what we see is subsurface structure rather than some organization of regolith.
          Besides recdeps and reclats, quadrangle hills are worthy of separate  description. The hills are located in formations of both morphological types. The dimensions of such hills are 0.3-1 km. Usually the quadrangle hill has a craterlet on its top. Sometimes the top depression is so large  that the hill appears hollow (Fig. 3).

Figure 3
The hollow hill is bounded by a rectangular depression: a candidate for embankment (LHD5345Q.059).

The rectangular depressions around the hill on Fig. 3 are a rarity for the Moon, but they are common for man-made mounds on the Earth.


          The origin of ruin-like formations could be reconstructed from images of various stages of their evolution.
          Thus, the reconstruction of recdep evolution is shown in Fig. 1. The simplest, probably the first stage formation, is a regular pattern of craterlets (Fig. 1a). Apparently, this is  the collapse or regolith drainage into subsurface caverns. Expanding craterlets became angular. The rectangular lattice of ridges appears between them (Fig. 1b). The rectangular lineaments around such formation (Fig. 1c) show the regular and local nature of subsurface caverns. Such a cavern system is seen after its total collapse (Fig. 1d). The bottom collapses (Fig. 1e) and slope terraces [1] in rectangular depressions argue for several levels of caves.
          The reclet evolution could be interpreted in terms of erosion too (Fig. 2). Apparently, the first (simplest) stage of reclet is the quasi-rectangular system of narrow furrows-cracks (Fig. 2a). The cracks expand (Fig. 2b) and transform into quasi-rectangular pattern of ridges (Fig. 2c). The Fig. 2d shows the quadrangle mesa-like hill surrounded by the ridge system (using the high-pass filter of Adobe Photoshop). Obviously, such ridges are a relatively stable aspect  of  the hill they reside on.
          These rectangular systems of depressions and ridges are resemble terrestrial ruins. Recdeps and reclats are too localized and regularized for tectonic features or  jointing pattern of impacts. Subsurface, rectangular, multilevel caves are not known in lunar geology. However, they are usual considered in modern plans for lunar bases. The rectangular systems of ridges could be interpreted in terms of archaeology too. Of course, suggesting this possibility is not a form of evidence, but rather an argument for archaeological reconnaissance in situ.


          The systematic survey for lunar ruin-like objects is realised. The results follow.
          1. New ruin-like formations are found.
          2. A catalogue of promising objects for archaeological reconnaissance of the Moon is compiled. Even if catalogued finds are natural, they are interesting examples of unusual lunar geology.
          3. Catalogued rectangular systems of depressions and ridges (recdeps and reclats) are  landscape forms not described in other catalogues.
          4. It is argued that recdep could be interpreted as a collapse of a subsurface system of caves. Such rectangular, multilevel caverns could be interesting for archaeology, geology and sites for lunar base.
          Therefore, the archaeological reconnaissance of the Moon appears to be a viable, active,  interdiscipline field of science.


          The author is grateful to Dr. Y.G.Shkuratov for access to the Clementine CDs. I also thank Dr. L.N.Litvinenko and Dr. T. Van Flandern for support.


          1. A.V. Arkhipov, "Earth-Moon System as a Collector of Alien Artefacts", J. Brit. Interplanet. Soc., 51, 181-184 (1998).
          2. A.V. Arkhipov, Preliminary Search For Ruin-Like Formations on the Moon, Meta Research Bulletin, 8, No. 4, 49-54.
          3. DoD/NASA, "Mission to the Moon", Deep Space Program Science Experiment, Clementine EDR Image Archive. Vol. 1-88. Planetary Data System & Naval Research Laboratory,  Pasadena, 1995 (CDs).

          Table. Catalogue of Ruin-Like Finds

          Note: The central coordinates of images are in degrees. The last column contains the quasi-rectangular elements (see text for definitions).


| Longitude | Latitude |     File     | Elements |
|    deg.   |   deg.   |              |          |
   11.05        89.16    LHD5814R.295      d
   13.63        85.57    LHD5741R.295      d
   16.08       -76.10    LHD0480B.030      f
 * 20.03       -81.24    LHD0395A.160      p
   20.69       -79.70    LHD0159B.293      dr
   22.50        80.63    LHD5686R.160      r
   25.38        75.50    LHB5443Q.291      prf
   28.25       -76.50    LHD0132B.290      dr
 * 28.35        79.10    LHD5502Q.290      f
   31.16        80.78    LHD5833R.157      f
 * 31.21        78.82    LHD5256Q.293      d
   32.97        79.60    LHD5538Q.289      f
   33.55        77.27    LHD5715Q.156      dr
   33.57        77.05    LHD5713Q.156      dr
   35.45        81.20    LHD5555R.289      rfd
   37.00        77.58    LHD5472Q.287      pr
   37.18        79.86    LHD5525Q.287      df
   41.93       -82.88    LHD0280A.151      fd
   43.09        86.94    LHD5724R.286      dr
   44.05       -75.87    LHD0445B.151      r
   51.34       -83.68    LHD0233A.147      f
 * 53.95       -83.54    LHD0287A.146      rd
   56.88        87.01    LHD5705R.282      dr
   60.29        79.20    LHD5559Q.279      d
   60.30        85.14    LHD5636R.280      p
   108.97      -76.82    LHD0412B.127      rhf
   109.85      -82.38    LHD0344A.126      d
   113.40       82.50    LHD5350R.260      fdr
   123.50       86.07    LHD5652R.126      df
   124.55      -82.47    LHD0282A.121      d
   128.05       80.00    LHD5375R.254      ?
   128.25      -78.26    LHD0162B.253      f
   128.41      -76.13    LHD0191B.253      r
   128.83       82.91    LHD5459R.254      dr
   130.26      -82.91    LHD0073A.252      d
   130.33      -82.75    LHD0274A.119      rp
   130.52       79.32    LHD4691Q.253      pf
   130.71       80.68    LHD4722R.253      dr
   131.20      -78.77    LHD0111B.252      dr
   135.66       80.05    LHD4807R.251      ?
   137.97      -84.74    LHD0276A.116      dr
   139.41      -86.30    LHD0184A.115      f
   145.91       77.84    LHD5288Q.247      f
   148.00      -81.36    LHD0248A.113      f
   148.41      -79.04    LHD0305B.113      d
   149.69      -84.26    LHD0231A.112      f
   150.71      -81.43    LHD0315A.112      rd
   151.29      -77.99    LHD0415B.112      d
   151.44      -76.24    LHD0470B.112      pr
   154.36       83.95    LHD6979R.244      p
   155.35       83.91    LHD5605R.112      dp
   156.86       83.25    LHD5564R.243      f
   159.68      -78.18    LHD0343B.109      pr
   164.46       76.18    LHD4993Q.240      rf
   164.51       81.34    LHD5173R.240      fd
   166.93       89.03    LHD5643R.114      dr
   167.15       80.91    LHD5286R.239      f
   169.86       81.35    LHD5175R.238      d
   169.87       79.18    LHD5107Q.238      dr
   171.02      -81.44    LHD0095A.238      p
 * 179.43       89.72    LHD5696R.248      fp
   190.15      -77.39    LHD0469B.098      rf
   191.53       83.32    LHD5417R.230      pr
 * 191.54       83.21    LHD5416R.230      r
   192.67      -80.56    LHD0308A.097      r
 * 192.83      -81.40    LHD0096A.230      dr
 * 192.90      -76.89    LHD0392B.097      f
   197.24       89.46    LHD5611R.108      drf
   200.20       78.82    LHD5279Q.227      dr
   224.67      -76.57    LHD0421B.085      dr
   224.72      -86.21    LHD0175A.083      r
   229.10      -80.45    LHD0316A.083      p
   230.32      -83.27    LHD0516A.082      pd
 * 232.01      -76.20    LHD0210B.215      f
   232.08       86.83    LHD5588R.217      fr
   242.82       87.26    LHD5629R.214      df
   243.37       82.05    LHD5628R.080      dr
   244.03      -81.12    LHD0146A.210      d
   244.99       85.05    LHD7605R.344      r
 * 246.08       81.88    LHD7638R.343      fh
   246.21      -82.25    LHD0142A.209      dr
 * 250.58      -85.48    LHD0193A.073      r
   251.14      -82.54    LHD0140A.207      r
   251.65       79.76    LHD5397Q.209      f
   254.56       79.99    LHD5250Q.208      f
   254.65      -80.58    LHD0148A.206      r
   258.78      -77.45    LHD0558B.072      f
 * 261.17       86.87    LHD5466R.208      dr
 * 266.18      -83.86    LHD0278A.068      r
   266.42       86.58    LHD5492R.206      dr
   268.33       87.79    LHD5595R.207      fp
 * 269.63       85.11    LHD5650R.072      d
   269.77       87.47    LHD5521R.206      dr
 * 272.70       82.72    LHD5562R.202      r
   273.41       79.55    LHD5545Q.069      d
   273.56       79.74    LHD5547Q.069      d
   281.47      -82.36    LHD0273A.063      fd
   284.08       87.80    LHD5717R.202      dr
   289.90      -80.94    LHD0149A.193      d
   290.49       87.58    LHD5661R.068      d
   291.22      -75.94    LHD0211B.193      d
   292.29       77.16    LHD5116Q.194      d
   292.30       77.07    LHD5110Q.194      d
   293.74      -80.73    LHD0315A.059      p
   296.28      -79.60    LHD0173B.191      dr
   297.82       84.15    LHD5528R.193      dr
 * 300.02       79.68    LHD5345Q.059      hd
   300.98       80.42    LHD5441R.191      d
   301.21       80.96    LHD5456R.191      dr
 * 301.28       85.55    LHD6749R.318      r
   301.55      -86.03    LHD0082A.320      h
   301.58      -88.19    LHD0119A.052      r
 * 306.10      -77.54    LHD0387B.055      dr
   311.45       86.05    LHD6158R.320      rh
   312.61       77.97    LHD5576Q.054      dr
   312.73       78.18    LHD5578Q.054      dr
   312.75       78.38    LHD5579Q.054      dr
   314.96       77.38    LHD5307Q.053      dr
   315.05       77.60    LHD5313Q.053      d
   315.37       77.84    LHD5314Q.053      d
   318.16       79.39    LHD5862Q.316      fdr
   320.67       79.28    LHD5916Q.315      dr
   323.28       86.62    LHD5574R.052      f
   329.05      -78.41    LHD0362B.047      fd
   338.05       86.90    LHD5972R.308      d
   341.12       81.88    LHA3621R.307      dr
   349.97       87.33    LHD5752R.303      pr
   351.42       85.96    LHD5165R.171      r

(This web page produced for Alexey Arkhipov by Francis Ridge of  The Lunascan Project)
 Home Page