SELIS V2 / research / findings

Topic B2 — Site choice for the Type A proof of concept: findings

15 November 2026; all sources opened that day. Second pass on B-data-sets.md (follow-up 1) and answer to A3-uncertainty-in-planning.md follow-up 2. Read those two files and C-users-and-needs.md first; this file does not repeat them. There is no topic file; the question came from Ivaylo's request. Bracketed numbers refer to section 8. In section 8, "(full text)" marks a document read as text extracted from the file, "(page summary)" a page read through an automated summary, "(extract)" a source read as a search extract only, "(data)" a data file or interface queried directly. "Judgement" marks an assessment, not a sourced fact. "Not found" means searched and not found. "Own count" marks a figure computed in this chat from archive metadata, by the method in section 3.1; it is not a published figure. Compiled with AI (Claude); not yet re-checked by a person.

Length: over the two-page limit of research-brief.md, because the request had five sub-questions for eleven sites.

Abbreviations not in glossary.md: NAC = Narrow Angle Camera of the Lunar Reconnaissance Orbiter Camera (LROC); PDS = Planetary Data System; ODE = PDS Orbital Data Explorer; PGDA = NASA Planetary Geodesy Data Archive; LDEM = DEM made from LOLA data; SfS = shape-from-shading; DTM = digital terrain model; RoI = region of interest; PSR = permanently shadowed region.

Question: which South Pole site has the best measured data to verify a predicted lit or shadowed state at a dated time, and the stated confidence of that prediction?

1. Feeds

Open decision O9 (site). 1.2 Requirements; 1.4.4 Implementation aspects; 1.5 Technical Problem Areas and Risks; 1.6.1 programmatic compliance (link to missions).

2. Key findings

  1. Answer (judgement): Connecting Ridge, with Malapert Massif as fallback. Connecting Ridge has the highest image count at its reference point (403, own count), the only published clone-based illumination table in a main text [3], and five terrain products from four producers (table 3.2). The ranked shortlist is in section 4.
  2. NASA has already co-registered 322 dated NAC images to the Connecting Ridge terrain and published them. The Zenodo archive of the NASA SfS DEMs holds, for each of 13 regions, the input NAC images orthorectified to the 5 m grid, the adjusted camera for each image, and a status list of every overlapping image [5]. For Connecting Ridge the list has 3,061 images: 322 used for SfS, 1,241 aligned and not selected, 1,352 filtered out, 146 not aligned [5]. The licence is CC BY 4.0 [5]. B-data-sets.md did not record this content.
  3. Image counts at a point (own count). Images whose footprint contains the site reference point: Connecting Ridge 403; Shackleton rim 278; peak near Shackleton 264; de Gerlache rim 248; Malapert Massif 222; Mons Mouton 183 to 184; Haworth 159; Mons Mouton Plateau 140; Nobile Rim 2 131 [1]. Table 3.1 gives dates and Sun directions.
  4. Near the pole, almost all images date from 2009 to 2011. At Connecting Ridge 377 of 403 images were taken by the end of 2011, and 16 after 2019 (own count) [1]. NASA states that the LRO orbit inclination has decreased and "the immediate polar latitudes are less accessible" [8]. Sites at 84 to 86°S have images up to 2025 or 2026 (table 3.1). A test at Connecting Ridge therefore relies on images 15 to 17 years old; the terrain has not changed, but no new images can be requested (judgement).
  5. Sun direction is sampled all round at the near-pole sites. At Connecting Ridge all twelve 30° bins of sub-solar longitude hold at least 10 images; 239 images were taken with the sub-solar point south of the equator and 158 north of it (own count) [1]. Nobile Rim 2 has four empty bins.
  6. The six JAXA sites are the ESA Lunar Lander RoIs. They are CR1 (Connecting Ridge), GR1 and GR2 (de Gerlache Rim in the ESA table; GR2 lies in NASA's Connecting Ridge Extension), LP1 (Leibnitz beta Plateau, on Mons Mouton), MP1 and MP2 (Malapert Peak) [6][7][12]. Each DTM is 1 m/pixel, about 1 to 2 km on a side, and built from six NAC left-right pairs [6][7]. The authors call the uncertainty maps "a sensitivity diagnostic, not a posterior standard deviation" [7].
  7. NAC stereo DTMs exist at three candidates. The LROC PDS archive holds 11 NAC stereo DTMs poleward of 83.4°S [1]. Three were made by DLR at 2 m for the ESA Lunar Lander project: Connecting Ridge (CR1, eight image pairs), Shackleton rim (SR12, seven pairs), Malapert (MP1, eight pairs), all from images of September and October 2010 [9]. LROC added a 3.5 m DTM of Connecting Ridge in April 2024 (LOLA RMS 1.06 m) [10]. The others are Malapert 01 to 03 and Nobile 01 to 04 at 3 to 4 m [1].
  8. Published illumination percentiles from clones exist for four sites only. NASA's site LDEM page lists 27 sites, each with 100 clones and uncertainty maps [2]. The clone-based illumination study covers Sites 1, 4, 7 and 11: Connecting Ridge, Shackleton rim, peak near Shackleton, de Gerlache rim [2][3]. The main-text table covers the seven RoIs of Site 1 only [3]. Malapert Massif, Mons Mouton, Nobile and Haworth have clones and no published percentiles.
  9. Mission interest points to the Shackleton and Connecting Ridge area. NASA: Moon Base I "will land on the Shackleton Connecting Ridge" (release of 26 May 2026) [15]. Space Applications Services: first LUVMI-M mission "targeting the Shackleton Crater" (29 September 2025) [16]. Chang'e-7: Shackleton rim, launch delayed to 2027 (news, extract) [19]. A site for Argonaut Mission 1 or MAGPIE: not found [17][18]. Connecting Ridge is not in NASA's nine Artemis III regions of 28 October 2024 [14].
  10. The terrain products are not independent of each other or of the test images. NASA SfS, JAXA SfS and the stereo DTMs are all tied to LOLA heights [4][7][9]. The SfS products are computed from NAC images. An image used to build an SfS DEM is not an independent check of illumination predicted on that DEM. The LOLA clones use no images [2], so every NAC image is an independent check of them (judgement).

3. Evidence

3.1 Dated NAC images per site (own count)

Method. Source: ODE REST interface, product type EDRNAC4 (PDS4 NAC experiment data records), queried on 15 November 2026 [1]. Seven footprint queries returned 38,708 distinct images; the query for footprints reaching 87.7°S or further south returned 24,604 records. Each record carries the row of the LROC index table; the column order was read from the LROC index label [11]. The footprint is the quadrilateral of the four corner coordinates, projected to south polar stereographic. "At point" counts images whose footprint contains the reference point. "In box" counts images whose footprint intersects the NASA site LDEM extent, read from the GeoTIFF headers [2]. Reference points: the mean of NASA's RoI centroids for Sites 1, 4, 7 and 11, which fall within 0.4 km of ESA's CR1, SR1, SV1 and GR1 [12]; the JAXA DTM centre for Malapert (MP1) and Leibnitz beta (LP1); the LDEM centre elsewhere; the LROC mosaic centre for Mons Mouton Plateau. Left and right NAC images are counted separately. "Bins" are 30° bins of sub-solar longitude, out of 12, with the smallest bin count. Summer means sub-solar latitude south of the equator. The last two columns are NASA's own figures from the status lists in [5], for the SfS region of the same name.

Site (NASA LDEM) Reference point At point Days First to last date By end 2011; after 2019 Bins; minimum Summer; winter In box NASA: all; co-registered
Connecting Ridge (Site01) 89.46°S, 222.5°E 403 158 2009-07-16 to 2024-02-28 377; 16 12; 10 239; 158 2,805 3,061; 322
Shackleton rim (Site04) 89.78°S, 203.5°E 278 135 2009-07-05 to 2024-02-28 252; 16 12; 12 120; 150 2,609 no SfS region
Peak near Shackleton (Site07) 88.81°S, 123.7°E 264 114 2009-07-04 to 2023-04-19 197; 3 12; 11 142; 109 2,044 1,700; 172
de Gerlache rim (Site11) 88.68°S, 291.6°E 248 110 2009-07-20 to 2024-02-28 116; 8 12; 6 104; 140 2,415 2,647; 194
de Gerlache Rim 2 (SL2) 88.23°S, 295.4°E 181 104 2009-07-19 to 2024-02-28 99; 5 12; 6 80; 101 2,693 3,123; 214
Connecting Ridge Extension (SL3) 89.01°S, 264.3°E 154 81 2009-07-22 to 2024-02-28 85; 11 12; 6 61; 92 2,366 2,855; 259
Malapert Massif (Site23) 85.97°S, 357.9°E 222 132 2009-10-03 to 2025-03-19 66; 4 12; 10 127; 89 2,900 2,761; 265
Mons Mouton: Leibnitz beta Plateau (Site20v2) 85.42°S, 31.9°E 183 130 2009-08-07 to 2025-05-05 27; 57 12; 5 101; 80 2,995 1,444; 120
Mons Mouton: Nobile Rim 1 (Site06) 85.44°S, 37.4°E 184 126 2009-09-30 to 2025-05-05 27; 39 12; 6 87; 94 2,218 1,768; 233
Haworth 86.76°S, 337.2°E 159 81 2009-07-29 to 2023-05-10 42; 7 12; 4 71; 88 3,260 4,609; 342
Mons Mouton Plateau 84.30°S, 31.1°E 140 105 2010-02-01 to 2026-02-27 16; 62 11; 0 66; 74 1,401 (16 km box) no SfS region
Nobile Rim 2 (DM2) 83.95°S, 58.8°E 131 85 2009-09-29 to 2026-01-29 15; 59 8; 0 72; 58 1,336 1,002; 160

Limits of the count. Corner coordinates are given to 0.01°, about 300 m at the pole. The paged download returned 65 duplicate records in 24,604, so a similar number may be missing. Between 0 and 9 images per site have a line exposure above 3 ms (long exposures aimed at shadow); the bins and seasons exclude them. The count does not test whether the reference point is lit, saturated or usable in each image. Median pixel scale is 1.2 to 1.6 m. The "in box" and NASA columns differ because the extents and the cut-off dates differ; NASA used the LROC release of March 2024 [13]. Cross-checks: the image NASA and JAXA used at Connecting Ridge (M139811097LE, 22 September 2010) [3][7] falls in the Connecting Ridge set; the two images of Zhong et al. cited in A3-uncertainty-in-planning.md fall in the peak near Shackleton set.

The 322 co-registered Connecting Ridge images in [5] span 5 July 2009 to 21 December 2011, on 117 days, with 8 to 52 images in each of the 12 bins (own count from the status list joined to [1]). The 265 Malapert images span 2 October 2009 to 1 August 2018; 255 were matched, on 113 days, with 2 to 74 per bin.

3.2 Terrain products and published illumination results per site

"Clones" means the PGDA 5 m LDEM with 100 clones and uncertainty maps [2]. "SfS 5 m" means a region in [4][5], which also carries an elevation uncertainty map per region. "JAXA 1 m" means a DTM with uncertainty map in [6]. Stereo DTMs are from [1][9][10]. MADNet sites are from [20].

Site Clones SfS 5 m JAXA 1 m NAC stereo DTM MADNet 1 m Published illumination results Mission interest
Connecting Ridge Site01 Yes CR1 DLR 2 m (ESALL_CR1); LROC 3.5 m (SHACKRDGE02) No Clone percentiles, main text [3]; ESA study with NAC snapshots [12]; 92.27% at 2 m, 95.65% at 10 m over 19 years [21] Moon Base I [15]; ESA Lunar Lander RoI [12]
Shackleton rim Site04; LM1 No No DLR 2 m (ESALL_SR12) No Clone percentiles, supplement [3]; ESA study [12] LUVMI-M "Shackleton Crater" [16]; Chang'e-7 (news) [19]
Peak near Shackleton Site07 Yes No None found Yes Clone percentiles, supplement [3]; two-image check by Zhong et al. (in A3) Chang'e-7 candidate area (in A3; news)
de Gerlache rim Site11 Yes GR1 None found No Clone percentiles, supplement [3]; ESA study [12] ESA Lunar Lander RoI
de Gerlache Rim 2 SL2 Yes No None found No None found NASA nine regions [14]
Malapert Massif Site23 Yes MP1, MP2 DLR 2 m (ESALL_MP1); LROC 4 m (MALAPERT01 to 03, overlap with Site23 not checked) Yes ESA study [12]; no clone percentiles NASA nine regions; ESA Lunar Lander RoI
Mons Mouton (Nobile Rim 1, Leibnitz beta) Site06, Site20v2 Yes, two regions LP1 ShadowCam 5 to 6 m DTMs (nine products, "nobile-mouton", "nobile-west-rim") [1] Yes ESA study (LP1) [12]; no clone percentiles NASA nine regions
Mons Mouton Plateau No No No LROC 3 to 4 m (NOBILE01 to 04, overlap not checked) No None found NASA nine regions
Nobile Rim 2 DM2 Yes No None found Yes None found NASA nine regions
Haworth Haworth Yes No None found Yes None found NASA nine regions

All sites: NASA average illumination and Earth-visibility maps over 18.6 years, at 60 m for 85 to 90°S and 120 m for 75 to 90°S [8]; Nobile Rim 2 and Mons Mouton Plateau lie outside the 60 m map. All sites south of 80°S: the LROC wide-angle illumination movie, 8,170 frames from 20 September 2009 to 11 December 2011 at 100 m/pixel [22]; it is a measured cross-check at a scale 20 times coarser than the 5 m analysis. NASA's 5 m LOLA mosaic covers 87 to 90°S [23], which includes the first six rows of table 3.1.

Further facts. NASA SfS: selection kept "at least three distinct illumination directions per pixel where possible" [4]; validation used about 380 NAC images, half of them not used for SfS [13]. JAXA 1 m: over 97% of height uncertainties below 0.20 m; median absolute deviation from LOLA about 0.9 to 1.7 m; shadowed pixels are NoData (10 to 31% of each DTM, read from the files) [6][7]. DLR stereo DTMs: tied to LOLA with 0.5 m RMS; orthophoto offsets of "up to 15m or greater" at CR1 [9]. LOLA at Site 1: about 2,500 tracks and one million returns; the other three sites have at most 25% fewer [3]. MADNet: seven sites, 20 to 54% coverage each, release planned through the ESA guest storage facility; a released product and its licence: not found [20]. ShadowCam: no DTM at Connecting Ridge, Shackleton or Malapert in the ODE list; image counts per site not obtained.

3.3 Licence or terms for commercial use

Product Terms Source
NAC images and LROC NAC DTMs taken from the PDS archive "Data products available through the LROC Data Node of the PDS or through NASA PDS interfaces are in the public domain." Commercial use of LROC images from other LROC web pages needs permission. [24]
PDS data in general, including the illumination maps "Data sets archived in the PDS are in the public domain." [25]
PGDA site LDEMs, clones, 5 m mosaic No licence text on the pages; a citation is requested. NASA's general guideline: content is generally not copyrighted in the United States; commercial use must not imply NASA endorsement. [2][23][26]
NASA SfS DEMs, orthorectified images, adjusted cameras CC BY 4.0 [5]
JAXA 1 m DTMs, orthomosaics, uncertainty maps CC BY 4.0 [6]
DLR-made stereo DTMs (ESALL_CR1, SR12, MP1) Archived in the LROC PDS volume; the readme carries no separate terms. Public domain under the LROC statement (judgement). [9][24]
MADNet 1 m DTMs Not found; the abstract states no data licence. [20]
ShadowCam Not re-checked; see B-data-sets.md. —

4. What this means for the Type A scope

All points are judgements.

  1. Ranked shortlist by the rule of 15 November 2026 (count of dated NAC images, count of independent terrain products):
    1. Connecting Ridge: 403 images at the point; clones, SfS 5 m, JAXA 1 m, two stereo DTMs; clone percentiles in a main-text table; Moon Base I.
    2. Malapert Massif: 222 images spread from 2009 to 2017; clones, SfS 5 m, two JAXA 1 m DTMs, one stereo DTM at the point; in NASA's nine regions. No published clone percentiles.
    3. de Gerlache rim: 248 images; clones, SfS 5 m, JAXA 1 m; clone percentiles in the supplement. No stereo DTM.
    4. Peak near Shackleton: 264 images; clones, SfS 5 m; clone percentiles in the supplement; one published two-image check.
    5. Shackleton rim: 278 images; clones and one stereo DTM; no SfS region and no JAXA DTM.
    6. Mons Mouton (Leibnitz beta, Nobile Rim 1): 183 to 184 images, up to 2025; clones, SfS 5 m, JAXA 1 m at LP1.
    7. Haworth, Nobile Rim 2, Mons Mouton Plateau: fewer images, two terrain products or fewer.
  2. Recommended site: Connecting Ridge. Fallback: Malapert Massif. If the fallback must allow a comparison with NASA's published percentiles, use de Gerlache rim instead; it shares the method and the data family with Connecting Ridge.
  3. The test design should separate image sets. Test the LOLA clones against all usable images. Test SfS-based terrain only against images outside its input list; the status lists name 1,241 such aligned images at Connecting Ridge [5].
  4. Registration effort falls. The 322 orthorectified images and adjusted cameras remove most of the image registration work for a first test at Connecting Ridge. Registration of further images, and the 5 m grid of the published copies against a 1 m analysis, remain open work.
  5. Positioning. Connecting Ridge is the site of the ESA Lunar Lander study of 2012 and of NASA's Moon Base I, and it is not an Artemis III candidate region. State all three facts in 1.6.1. Do not claim a link to Argonaut Mission 1: no site was found.

5. Ready-to-paste sentences

Re-check each against its source before use. Define acronyms again in the section where a sentence is used. Sentences 1 and 2 rest on the own count of section 3.1 and need a person to repeat the query.

  1. The Planetary Data System (PDS) holds 403 Narrow Angle Camera (NAC) images that cover the Connecting Ridge reference point, taken between July 2009 and February 2024 [1].
  2. Of these 403 images, 377 date from 2009 to 2011 [1].
  3. The National Aeronautics and Space Administration (NASA) published 322 NAC images of Connecting Ridge, orthorectified to a 5 m grid, under the Creative Commons Attribution 4.0 licence [5].
  4. Four image-based terrain products from four producers cover Connecting Ridge: NASA at 5 m, the Japan Aerospace Exploration Agency (JAXA) at 1 m, the German Aerospace Center (DLR) at 2 m and the camera team at 3.5 m [5][6][9][10].
  5. NASA stated in May 2026 that its Moon Base I mission will land on the Shackleton Connecting Ridge [15].

6. Open risks and unknowns

Contradictions with the findings files and CLAUDE.md:

  1. B-data-sets.md finding 2 calls NAC stereo "the exception at the pole" and its table lists two stereo DTMs (DLR "10 pairs", USGS 4 m). The PDS holds 11 NAC stereo DTMs poleward of 83.4°S, including a 2024 LROC product at Connecting Ridge [1][10]. The readme of the DLR Connecting Ridge DTM lists eight image pairs, not ten [9]. The statement that stereo cannot be the main terrain method still holds.
  2. B-data-sets.md table row "LUPEX DTMs" says "six images per site". The paper uses six left-right pairs, 12 images [7]. The paper names the sites by ESA Lunar Lander RoI and mentions LUPEX only as an example of mission interest [7]; "LUPEX DTMs" is the archive's label, not re-opened here. The five site names missing in B-data-sets.md section 6 are in finding 6.
  3. B-data-sets.md finding 6 names LROC mosaics as the measured data for checks. The SfS archive is a better source at the candidate sites: per-image orthorectified copies with dates, not mosaics [5].
  4. B-data-sets.md finding 8 says LRO still acquires data. That is true and of little use near the pole: 16 of 403 Connecting Ridge images are later than 2019 (own count).
  5. B-data-sets.md site view counts five "independent" terrain products at Connecting Ridge. They share LOLA control, and two share NAC images (finding 10). The count is now six with the LROC 3.5 m DTM; the USGS DEM named there was not re-opened.
  6. C-users-and-needs.md cites the EPSC 2026 abstract for the LUVMI-M target "Shackleton crater". The abstract says only "the lunar South Pole region" [27]; the crater is named in the company news item [16]. The two sources also differ on the date: Q4 2028 and Q3 2028.
  7. A3-uncertainty-in-planning.md: no conflict. Its follow-up 2 is answered by table 3.1 at the level of image counts; the number of independent checks per cell is still open.
  8. CLAUDE.md: no conflict.

What could not be counted or opened:

Decisions for Ivaylo: (a) confirm Connecting Ridge and the fallback; (b) accept that the test images date from 2009 to 2011; (c) whether a written NASA statement on PGDA terms is needed (follow-up 2 of B-data-sets.md).

7. Suggested follow-up topics

  1. Usable checks per cell at Connecting Ridge: for the seven NASA RoIs, how many of the 322 co-registered images show each 5 m cell lit, shadowed or saturated. It decides whether the sample supports a test of the stated confidence (1.4.4, 1.5).
  2. Size of the image registration error against the width of the predicted lit-to-shadow transition at 1 to 5 m height. It decides whether a mismatch can be attributed to terrain uncertainty (1.5).
  3. none.

8. Sources

Accessed 15 November 2026.

  1. PDS Geosciences Node, Orbital Data Explorer REST interface (data): product types https://oderest.rsl.wustl.edu/live2/?query=iipy&odemetadb=moon&output=JSON ; footprint queries of the form https://oderest.rsl.wustl.edu/live2/?query=product&target=moon&ihid=LRO&iid=LROC&pt=EDRNAC4&results=mp&output=JSON&minlat=-90&maxlat=-87.7&westlon=0&eastlon=360&loc=f ; also product types SDNDTM, BDRROI and ShadowCam DTM and CMOS with maxlat=-83.4
  2. NASA PGDA, "High-Resolution LOLA Topography for Lunar South Pole Sites" (full text of the page; GeoTIFF headers and RoI shapefiles read): https://pgda.gsfc.nasa.gov/products/78 ; readme: https://pgda.gsfc.nasa.gov/data/LOLA_5mpp/README
  3. Barker et al. 2021, Planetary and Space Science 203, 105119 (full text, NTRS manuscript): https://ntrs.nasa.gov/api/citations/20205009660/downloads/Mike%20Barker_%20Paper_%20Improved%20Lola.pdf
  4. NASA PGDA, "Enhanced Topography Models for Selected Lunar South Pole Regions with Shape-from-Shading" (full text of the page): https://pgda.gsfc.nasa.gov/products/104
  5. Bertone et al. 2025, Zenodo archive, version 1.0.0, 16 December 2025 (data: record metadata, README.md, file lists of all 13 archives, the Connecting Ridge readme, and all 13 image status lists; rasters not downloaded): https://zenodo.org/records/17954508
  6. Hemmi and Inoue 2025, Zenodo archive, 25 September 2025 (data: DataS1.zip downloaded and read): https://zenodo.org/records/17153447
  7. Hemmi et al. 2025, Planetary Science Journal (page summary, whole page in two parts; Table 1 coordinates garbled in the text read): https://iopscience.iop.org/article/10.3847/PSJ/ae10a4
  8. NASA PGDA, "Lunar Polar Illumination" (full text of the page): https://pgda.gsfc.nasa.gov/products/69
  9. LROC PDS archive, readme files of the DLR stereo DTMs (full text): https://pds.lroc.im-ldi.com/data/LRO-L-LROC-5-RDR-V1.0/LROLRC_2001/DATA/SDP/NAC_DTM/ESALL_CR1/NAC_DTM_ESALL_CR1_README.TXT ; https://pds.lroc.im-ldi.com/data/LRO-L-LROC-5-RDR-V1.0/LROLRC_2001/DATA/SDP/NAC_DTM/ESALL_SR12/NAC_DTM_ESALL_SR12_README.TXT ; https://pds.lroc.im-ldi.com/data/LRO-L-LROC-5-RDR-V1.0/LROLRC_2001/DATA/SDP/NAC_DTM/ESALL_MP1/NAC_DTM_ESALL_MP1_README.TXT
  10. LROC, "Shackleton - de Gerlache Connecting Ridge DTM", SHACKRDGE02 (page summary): https://data.lroc.im-ldi.com/lroc/view_rdr/NAC_DTM_SHACKRDGE02 ; ESALL pages (page summaries): https://data.lroc.im-ldi.com/lroc/view_rdr/NAC_DTM_ESALL_CR1 ; https://data.lroc.im-ldi.com/lroc/view_rdr/NAC_DTM_ESALL_SR12
  11. LROC PDS archive, EDR index label (full text): https://pds.lroc.im-ldi.com/data/LRO-L-LROC-2-EDR-V1.0/LROLRC_0001/INDEX/INDEX.LBL
  12. De Rosa et al. 2012, ESA Lunar Lander site characterisation (full text, arXiv PDF; Table 2 and section 2.4): https://arxiv.org/abs/1208.5587
  13. Bertone et al. 2026, Planetary Science Journal 7(5), 118 (page summary, first 100,000 characters): https://iopscience.iop.org/article/10.3847/PSJ/ae5b70
  14. NASA, Artemis III landing regions, 28 October 2024 (page summary): https://www.nasa.gov/news-release/nasa-provides-update-on-artemis-iii-moon-landing-regions/
  15. NASA, Moon Base update, 26 May 2026 (page summary): https://www.nasa.gov/news-release/nasa-provides-update-on-moon-base-rovers-landers-missions/
  16. Space Applications Services, news, 29 September 2025 (page summary): https://www.spaceapplications.com/news/space-applications-services-and-aerospace-applications-north-america-announce-first-lunar-surface-mission
  17. ESA, "Argonaut: a first European lunar lander", 30 January 2025 (page summary): https://www.esa.int/Science_Exploration/Human_and_Robotic_Exploration/Argonaut_a_first_European_lunar_lander
  18. ESA, MAGPIE, 26 October 2026 (page summary): https://www.esa.int/Science_Exploration/Human_and_Robotic_Exploration/MAGPIE_ESA_s_first_lunar_rover
  19. Scientific American, news, updated 24 August 2026 (extract): https://www.scientificamerican.com/article/chinas-change-7-moonshot-will-seek-water-ice-at-the-lunar-south-pole/
  20. EPSC 2024 abstract 1170, MADNet at seven Artemis sites (page summary): https://meetingorganizer.copernicus.org/EPSC2024/EPSC2024-1170.html
  21. Gläser et al. 2014, Icarus 243, DLR repository record (page summary of the abstract): https://elib.dlr.de/94240/
  22. LROC, WAC South Pole movie (page summary, first 100,000 characters): https://data.lroc.im-ldi.com/lroc/view_rdr/WAC_POLE_MOVIE_SOUTH
  23. NASA PGDA, "South Pole LOLA DEM Mosaic" (full text of the page): https://pgda.gsfc.nasa.gov/products/81
  24. LROC, terms of use (page summary): https://lroc.im-ldi.com/about/terms
  25. PDS Geosciences Node, FAQ (page summary): https://pds-geosciences.wustl.edu/faq.html
  26. NASA, images and media usage guidelines (page summary): https://www.nasa.gov/nasa-brand-center/images-and-media/
  27. Winters et al., EPSC 2026 abstract 432, LUVMI-M (page summary): https://meetingorganizer.copernicus.org/EPSC2026/EPSC2026-432.html