See the snow, tracks and sunlit rock in Harwood’s photograph. No open image licence has been established for reproduction here.
A study against the weather record
Walter Ufer
Winter in New Mexico
Snow underfoot, sunlight on the rocks. Which Taos winters offered the combination Nephos reads in this picture?
Sun and snow documentedPainted winter unknown
The Nephos method
How this study was read
The method in fullWhat is written
Nephos reads snow marked by footprints, two shawled figures and sunlit ochre rocks beneath a partly clouded sky: a sunny winter interval after snowfall. This is Nephos’s reading of the picture, not a museum statement. Harwood’s c.1930 remains approximate.
Records of those days
- Taos, GHCN-Daily USC0029866822 December–February seasons, 1914–15 through 1935–36. Positive snow depth survives in 16; all counts of cover are observed minima.
- Richard F. Oakley’s original Taos observer sheetsFifteen monthly sheets supply daily Clear, Partly cloudy and Cloudy categories for the five winters ending 1928–1932; one illegible cell remains unknown.
- NOAA 1991–2020 Taos normalsDecember–February snowfall totals 46.5 cm in the provisional normals. No snow-depth normal is supplied.
Not found or not yet searched
- No object-specific session, dated sketch or precise viewpoint has been established.
- The accessible Harwood caption has no accession number; the detailed catalogue returned an access challenge.
- Notre Dame’s diaries were not inspected. No newspaper weather item for the focus winters was verified; this does not mean none survives.
Each claim tested
- Did snow lie at Taos on clear winter days? Taos records at least 37 clear days with snow lying in 1931–32, and 27 in 1928–29. These are strong environmental candidates for Nephos’s reading of the picture, not dates for the painting.
- Did clear weather follow snowfall while snow still lay? In 1929–30, 5 of 6 eligible snowfall dates were followed within three days by Clear with positive depth. The final snowfall is excluded because its three-day window crosses into March.
- Is there a then-versus-now comparison? The 18 complete historical winters average 44.5 cm of snowfall; the provisional 1991–2020 normals total 46.5 cm for the same months. This is descriptive, not a snow-cover trend.
Verdict
Taos records at least 37 clear days with snow lying in 1931–32, and 27 in 1928–29. These are strong environmental candidates for Nephos’s reading of the picture, not dates for the painting.
- Partially agreesSunlight with snow lying: Nephos’s visual reading The combination occurs repeatedly at Taos. The station is regional evidence, not the unidentified painted path.
- UnknownThe painted winter or an alteration of observed weather No dated making record or preparatory comparison establishes either.
What this cannot tell you. The five focus winters were chosen around c.1930, not because of their weather; the broader period is biographical. Neither is a new date range for the painting. Positive depth counts are minima. Gaps and rounded thin cover prevent ranking true snow-cover duration; a clear-day category is stricter than visible sunlight through clouds. Snowfall follow-up windows may overlap. They count recorded snowfall dates, not independent storms or a probability that Ufer painted then. Taos historical instrument locations and observing hours changed or remain incompletely reconstructed; no stations are pooled. The modern normals are provisional, and the modern depth record is too incomplete for a thirty-year snow-cover comparison.
Step 1 · Start with the painting
Sunlight on snow that has stayed
Nephos’s reading of the picture, not a museum statement: Snow lies on the foreground path or possible creek bed, marked by tracks read as footprints. Two figures in dark shawls walk toward the viewer between ochre-gold rock walls; scrub and a juniper stand above. Blue sky is partly covered by a large grey cloud, while sunlight catches the rocks and foreground snow. We read this as a sunny winter interval after snowfall, with snow still lying on the ground.
The tracks make the snow a travelled surface; the warm, lit rock faces and blue shadows in the foreground give the snow a different light from the cliffs. The large cloud matters: the painting does not require an entirely clear day. Nor does it establish that the path is a frozen creek, how deep the snow is, or how recently it fell. Open Harwood’s image and compare this reading with the picture.
That reading makes a real test possible: did Taos record snow lying on days with clear weather, and did clear weather follow snowfall before the ground became bare? We join accepted daily snow-depth observations to the original observer’s daily sky descriptions. His sky column summarizes sunrise to sunset; depth was read at the stated afternoon observation time. The join pairs an observing date’s sky and depth, rather than measuring continuous daylight cover. “Clear” is a deliberately strict test; “Partly cloudy” is kept separate because it does not by itself prove sunlight at the moment Ufer depicts.
Harwood dates the painting c.1930. The five winters ending 1928–1932 are a comparison around that approximate date, chosen before reading their skies. They are not a new dating interval. The broader December–February record runs from 1914–15 to 1935–36, between Ufer’s first Taos visit and his death. The station represents Taos weather; the painted path remains unlocated.
Step 2 · Test snow lying against the sky record
Two winters offer especially strong candidates
1931–32 records at least 67 days with snow lying, including 37 classified Clear. In 1928–29 the corresponding minima are 39 and 27. These are the strongest documented counts of the sun-and-snow combination among the five winters. They make those winters useful environmental candidates, never a match to the canvas.
Every focus winter contains clear days with snow lying. Even 1929–30, with only 7 measurable snowfall dates, has at least 9 clear days with snow still on the ground. The image’s combination was a recurring Taos possibility, rather than a weather contradiction. The counts do not rank the true frequency of these conditions: winters with less complete depth reporting can be undercounted much more severely.
Five winters: coverage, snow lying and sky categories
| Winter | Depth entries / days | Snow lying | Of these, Clear | Of these, Partly cloudy | New snowfall dates |
|---|---|---|---|---|---|
| 1927–28 | 15/91 | 12+ | 7+ | 1+ | 11 |
| 1928–29 | 59/90 | 39+ | 27+ | 4+ | 15 |
| 1929–30 | 33/90 | 29+ | 9+ | 8+ | 7 |
| 1930–31 | 25/90 | 17+ | 12+ | 2+ | 8 |
| 1931–32 | 69/91 | 67+ | 37+ | 12+ | 14 |
After snowfall: how often did the sky clear?
In 1929–30, 5 of 6 eligible snowfall dates were followed within three calendar days by a Clear day with positive snow depth. The last snowfall, on 27 February, is excluded because its full follow-up window extends into March. This is a test of the sequence suggested by the picture; three days is our stated comparison window, not an age assigned to its snow.
The table also counts clear weather without requiring a surviving depth entry. A later clear day can serve more than one snowfall date: these are overlapping sequences, not independent storms. Unknown outcomes retain missing depth; they are not failures. Observer days and observation hours do not necessarily align with midnight-to-midnight weather.
The next-day and three-day tests in every focus winter
| Winter | Eligible dates | Next day Clear | Clear within 3 days | With snow | Unknown with snow | Boundary exclusions |
|---|---|---|---|---|---|---|
| 1927–28 | 11 | 6 | 10 | 4+ | 6 | 0 |
| 1928–29 | 13 | 8 | 11 | 7+ | 2 | 2 |
| 1929–30 | 6 | 2 | 5 | 5+ | 0 | 1 |
| 1930–31 | 7 | 3 | 4 | 1+ | 3 | 1 |
| 1931–32 | 14 | 8 | 12 | 8+ | 3 | 0 |
Step 3 · Listen to the contemporary observer
“Thawing”, with snow still on the ground
On 25 January 1930, Taos observer Richard F. Oakley wrote “Thawing”. His sheet still records 2 inches of snow on the ground. He classified that day as Cloudy, so it is not one of the chart’s clear-day cells. The note nevertheless makes a useful distinction: winter snow could persist while conditions softened. Five days earlier he had recorded continuing snowfall.
“Thawing”
Richard F. Oakley, Taos cooperative observer sheet, 1930-01 · 25 January, miscellaneous phenomena; snow-depth column reads 2 inches
“Still snowing at time of observation”
Richard F. Oakley, Taos cooperative observer sheet, 1930-01 · Remarks, entry for the 20th; date prefix transcribed separately
These are local, contemporary observations, read on Oakley’s signed January sheet. They are the human account behind the station data, not an independent second instrument and not testimony about Ufer’s painting. The Weather Bureau’s statewide review separately describes a snowy January and a warm, dry February; that wider description cannot stand in for Taos.
A Taos newspaper witness remains a specific gap. UNM’s digitized Revista de Taos / Taos Valley News collection ends in 1922, before the focus winters. No relevant newspaper weather item was verified here. This is a limit of the sources inspected, not evidence that no such issue or report survives. Notre Dame lists Ufer diaries and correspondence; their contents were not inspected.
What Ufer chose
Nephos reads a deliberate contrast between the dark figures, pale tracked snow and golden rock. Ufer’s own 1928 account describes arranging a model in relation to light. That supports discussing composition, without claiming that he added snow or changed the weather. No dated sketch, earlier version or account of this canvas establishes such an alteration.
“placing him in the sun or indoors”
Macbeth Gallery, Exhibition of Recent Paintings by Walter Ufer (1928) · 1928 catalogue foreword, PDF p. 2
Step 4 · Then and now
Winter snowfall remains a substantial part of Taos weather
The 18 complete historical winters average 44.5 cm of new snow in December–February. NOAA’s 1991–2020 Taos monthly normals total 46.5 cm (18.3 inches) for those same months. That is a comparison of similar amounts of new snowfall. It does not establish a trend in snow lying on the ground.
| Measure | 1914–15 to 1935–36 | 1991–2020 normals |
|---|---|---|
| Mean new snowfall | 44.5 cm | 46.5 cm |
| Mean measurable snowfall days | 9.9 | 8.1 |
| Mean days with snow lying | Not recoverable from complete winters | No Taos depth normal supplied |
The historical means use complete winters only. NOAA labels the modern snowfall normals provisional: their December, January and February totals use 13, 12 and 15 observed years respectively; snowfall-day frequencies use 19, 18 and 16 years. Missing months could not be filled from neighbouring stations. The two samples and historical station exposures are not a homogenized climate comparison.
For actual modern snow cover, 2006–07 records 40 days with measurable snow lying. It is the only one of the 29 whole winters inside 1991–2020 with complete accepted daily depth in this file. That is a dated example of snow persisting in the modern period, not a thirty-year average or evidence of increase or decline.
The full historical span: all winters, including gaps
| Winter | New snowfall, cm | Depth entries / days | Days with snow lying, minimum |
|---|---|---|---|
| 1914–15 | Incomplete | 0/90 | 0+ |
| 1915–16 | 67.0 | 1/91 | 1+ |
| 1916–17 | 57.5 | 0/90 | 0+ |
| 1917–18 | 50.1 | 1/90 | 0+ |
| 1918–19 | Incomplete | 0/90 | 0+ |
| 1919–20 | Incomplete | 2/91 | 1+ |
| 1920–21 | 49.1 | 0/90 | 0+ |
| 1921–22 | 34.9 | 1/90 | 1+ |
| 1922–23 | 24.1 | 1/90 | 0+ |
| 1923–24 | 44.0 | 52/91 | 51+ |
| 1924–25 | Incomplete | 79/90 | 78+ |
| 1925–26 | 42.3 | 31/90 | 20+ |
| 1926–27 | 56.9 | 38/90 | 33+ |
| 1927–28 | 25.7 | 15/91 | 12+ |
| 1928–29 | 71.3 | 59/90 | 39+ |
| 1929–30 | 30.2 | 33/90 | 29+ |
| 1930–31 | 33.0 | 25/90 | 17+ |
| 1931–32 | 72.9 | 69/91 | 67+ |
| 1932–33 | 43.1 | 52/90 | 43+ |
| 1933–34 | 34.3 | 31/90 | 23+ |
| 1934–35 | 28.9 | 15/90 | 7+ |
| 1935–36 | 35.4 | 16/91 | 9+ |
Modern snow-depth coverage, 1991–92 through 2019–20
| Winter | Accepted depth entries / days | Days with snow lying |
|---|---|---|
| 1991–92 | 26/91 | Incomplete |
| 1992–93 | 1/90 | Incomplete |
| 1993–94 | 9/90 | Incomplete |
| 1994–95 | 3/90 | Incomplete |
| 1995–96 | 33/91 | Incomplete |
| 1996–97 | 31/90 | Incomplete |
| 1997–98 | 2/90 | Incomplete |
| 1998–99 | 87/90 | Incomplete |
| 1999–00 | 44/91 | Incomplete |
| 2000–01 | 21/90 | Incomplete |
| 2001–02 | 14/90 | Incomplete |
| 2002–03 | 8/90 | Incomplete |
| 2003–04 | 66/91 | Incomplete |
| 2004–05 | 47/90 | Incomplete |
| 2005–06 | 65/90 | Incomplete |
| 2006–07 | 90/90 | 40 |
| 2007–08 | 41/91 | Incomplete |
| 2008–09 | 58/90 | Incomplete |
| 2009–10 | 1/90 | Incomplete |
| 2010–11 | 0/90 | Incomplete |
| 2011–12 | 0/91 | Incomplete |
| 2012–13 | 0/90 | Incomplete |
| 2013–14 | 0/90 | Incomplete |
| 2014–15 | 0/90 | Incomplete |
| 2015–16 | 0/91 | Incomplete |
| 2016–17 | 31/90 | Incomplete |
| 2017–18 | 0/90 | Incomplete |
| 2018–19 | 0/90 | Incomplete |
| 2019–20 | 0/91 | Incomplete |
What the test can and cannot establish
The picture’s snow-and-sun combination has repeated, local documentary support. Weather alone cannot select its winter. Oakley’s daily descriptions cover the town station, not an identified canyon or the whole day at the painted path. The GHCN depth series rounds observations and encodes some thin cover as trace; the chart therefore counts positive numeric observations, never interpolated cover. Missing values, quality flags and “missing presumed zero” entries are excluded. Taos Canyon and Santa Fe have not been pooled with Taos.
Ufer was not necessarily in Taos in every winter: the Woolsey exhibition history describes his winter returns east. The museum accession and basis of c.1930 remain unconfirmed in the accessible caption. An object file, dated study or letter could narrow the making history. Until then, the candidate winters explain what the painting makes plausible; they do not become a painted day.
Sources
Every source on this page
Each file was downloaded and given a fingerprint (its sha256). The page is rebuilt from these files and will not build if a fingerprint changes, so every number above comes from them rather than being typed in.
- Harwood Museum of Art, Early 20th Century & Taos Society of Artists
sha256 1b65d1e4b1fe35ff498e601326c3105df46854d4b5819f32be15c1a547b8b99a - Harwood image metadata
sha256 bfdc228482b546ef77064ae9d2902330d3bb25ea91347890fcb1f2009bc9ffd7 - Smithsonian American Art Museum, Walter Ufer biography
sha256 582942cc4f39e74b5e25673496fd4a51b49c6a2f5264fa8bf5842d7f6a09e82e - Macbeth Gallery, Exhibition of Recent Paintings by Walter Ufer (1928)
sha256 2fa3e96016fd8f06e4f7ff4c8db7c115f666174557c5151520cf5cf937e0f083 - Text extraction of Macbeth Gallery catalogue
sha256 f0d6722d4c282c4b5fb531b657ecb4625ad63a003ed3fd89f66ea741827b6cca - Met Libraries catalogue record for the 1928 exhibition
sha256 0b9fba9551882a0019883b5fa498a9f556ab3c54179ed876d0580f456112d2c3 - James E. May, Hoosiers in Taos: The Woolsey Brothers
sha256 0e97d5b052ec8c97ea87e90d87b1e836906ede1c45f9edc5ee3b986c9305ee50 - James E. May, Hoosiers in Taos, notes
sha256 07788f57e1bceb619b99eae87cac8385ea043394d9f299842cacf3a6cd40fd48 - University of Notre Dame Archives, Walter Ufer Papers, UFR
sha256 38883fda8ea078fb0b281d29e3c7fc2844ce1b8abb87ef85cbc8d4c7c8406d63 - GHCN-Daily, Taos, USC00298668
sha256 d0681a31d6a3081e2a8fcdf7d7ce763dab641511a03cd44d746cca85457769f3 - GHCN-Daily, Taos Canyon, USC00298673
sha256 0cb047163d994a411de7cfde5d16db82f40f883c52de586147523ac0dc1fbdf8 - GHCN-Daily, Santa Fe, USC00298072
sha256 29640cceeb8bd12a2e46dc7e7334c66456f12f08b79651b38961871f9e77c49a - GHCN-Daily station metadata, three-station extract
sha256 2279d50984ddba91bf150d45ad900014adf1fba4532b1fd724aca2a5404e68b2 - GHCN-Daily element inventory, three-station extract
sha256 9b9ebc96ce9c3f4ef1af8cb7f0f70eca06b271ec8175d9c671e9d29b97e6f23c - GHCN-Daily format, units and flags
sha256 0e4e6f0e6947cf1add884c328a01760f31c56ef5fb86b85ed57dd5dc4c4aeaa6 - US Weather Bureau, Climatological Data, New Mexico, Annual 1930
sha256 1870492656b42cf26c5d910dea599a600ca7e513179ec354ec7126ed5ce723f4 - Nephos transcription of two 1930 Weather Bureau phrases and one annual total
sha256 45ce3ff3b751aeef9d964da3f5336e0d7d4b43b24dd496dc7128b34702803bc4 - UNM, Revista de Taos / Taos Valley News digital collection
sha256 31c626e8430daf503dd06adf7790db6017b6d517390a4992a854da7528f9fe4c - Harwood Museum, photograph of Winter in New Mexico
sha256 c68ac140906577d2fd9d1977d0156c46834b99654721ebb8400ba97dcc4a2a87 - Richard F. Oakley, Taos cooperative observer sheet, 1927-12
sha256 447feadc6b6497d1e95397815eaa1a59924429795afbd5c57676393ac73f6c82 - Richard F. Oakley, Taos cooperative observer sheet, 1928-01
sha256 d506a57843fc08c156f84e24307fe44d4f8133cf28d864edfd3e46f24c90b4fa - Richard F. Oakley, Taos cooperative observer sheet, 1928-02
sha256 e2348164c4d6ca2a2cb209c94aa6cf6d7203a3e6a187b04c119127d283df519b - Richard F. Oakley, Taos cooperative observer sheet, 1928-12
sha256 ac5c5da6cca2c952e4053b3f0e8d64a34c2d14beb56b32e14361a589f701ebf0 - Richard F. Oakley, Taos cooperative observer sheet, 1929-01
sha256 244bdb6bec15bee64f35337a11c673d504974558d8785c631988167a3cdb202f - Richard F. Oakley, Taos cooperative observer sheet, 1929-02
sha256 d58103eec0589e974b436ce624a6108250437dcb3a87c0d66df274349e47cd40 - Richard F. Oakley, Taos cooperative observer sheet, 1929-12
sha256 07ee014ad0cb08c9d486ece0c4867afcf41d95d218e119ac29d83e99b6a3903c - Richard F. Oakley, Taos cooperative observer sheet, 1930-01
sha256 bbef2d04e2e9bb72851e2673b13d36c5ebf48c29faf7473293f07cb5d84eca67 - Richard F. Oakley, Taos cooperative observer sheet, 1930-02
sha256 a3e3dcddebcc2b9c7a0abe58b8121393ad10eec13e48bf54deab60a620c0b840 - Richard F. Oakley, Taos cooperative observer sheet, 1930-12
sha256 d30fe90f4d260db8202f54652e8c488ebce904f8ac3c89c21894282e472a9d58 - Richard F. Oakley, Taos cooperative observer sheet, 1931-01
sha256 bcd4c73feb6539aef5a629bc8eb022d6918b01e8591af2d0c75851e7b3cb01f1 - Richard F. Oakley, Taos cooperative observer sheet, 1931-02
sha256 a2a82697ab2b2db57d458f4ce13ba95f2bdc0171b010e1f774bc73fa39a71c76 - Richard F. Oakley, Taos cooperative observer sheet, 1931-12
sha256 f294200af178ccacc046cddf471e1ea37cfe0d6579bd13a943949dfa9ff0d570 - Richard F. Oakley, Taos cooperative observer sheet, 1932-01
sha256 acaaf706e2adda9cd0d447dc32c32e31f0f6a8f438ba1e433db8150aa695af03 - Richard F. Oakley, Taos cooperative observer sheet, 1932-02
sha256 483a7255f89c4f5edcf2bd1f8f65b3748c377bbaf5af7c1fb10be98f23a6e6e4 - Nephos transcription of fifteen Taos observer sheets
sha256 2373820312556e7c011c3433a8b4fc1e63bba94c157553fcb173ac1b8a7ff163 - NOAA NCEI, Taos 1991–2020 monthly normals
sha256 8351683bb37db498eada7f61c5d19210d0e241098acd57e9e51833548f36514c - NOAA NCEI, normals variables and completeness flags
sha256 feb4e1dd5914c8d35eccd017e114c6e5d352b1e217602cb622a67ad6670c3c21 - NOAA NCEI, Taos observer-sheet index, December 1927–February 1932
sha256 fe064754222fccd3400426ef443d199ca4e9a01bc56589ba32c3f56ff72015d9
