Coastline Time Machine: Academic Background

Sources (external links)
日本語 · English · 繁體中文
Back to the time machine

The following external resources are in Japanese.

Representative sea levels and references (revised 30 September 2026)

The sea levels assigned to the 13 eras of the time machine are educational estimates compiled by this site from the literature. The literature does not give a value for each era’s year directly; these are representative values this site read from the reported values and rates of rise, then rounded. The reference differs by era. Before about 10,000 years ago (eras 0–4) the values are global mean sea level (distant coral records and ice-volume equivalents). From about 8,000 to 2,000 years ago (eras 5–9) they are relative sea level on the central Kanto Plain (the Jomon transgression was a regional phenomenon caused partly by hydro-isostasy; global mean sea level at the time is estimated to have still been below today’s). After about 2,000 years ago (eras 9–12), smaller ups and downs have not been confirmed, so the Medieval Warm Period (+0.6 m, a regional comparison hypothesis) and the Little Ice Age (−0.3 m, the low end of the global range) are comparison settings, and the Yayoi & Kofun era assumes today’s level. The Asuka–Nara and Meiji–Taisho eras were removed because no reliable values exist. Sea level differs by several metres to more than ten metres between regions, so coastlines outside Japan are approximate. Years are rounded; “about n years ago” on screen counts from AD 2000 (BP in the literature counts from 1950).

#EraYear (approx.)ValueRangeReferenceMain sources
0Paleolithic (Last Glacial Maximum)20,000 years ago−120 m−134 to −116 mGlobal meanFairbanks 1989 (121±5 m lower, Barbados); Peltier & Fairbanks 2006 (about 120 m). The lower end is the ice-volume equivalent of Lambeck et al. 2014 (−134 m)
1End of the Ice Age (Younger Dryas)12,500 years ago−65 m−75 to −60 mGlobal meanDeschamps et al. 2012 (Meltwater Pulse 1A: 14–18 m rise between 14,650 and 14,310 years ago); read from Lambeck et al. 2014 and Smith et al. 2011
2Beginning of the Initial Jomon11,000 years ago−55 m−60 to −40 mGlobal meanSmith et al. 2011 (about 60 m rise between 11,650 and 7,000 cal BP); Lambeck et al. 2014 (about 12 m per thousand years on average)
3Initial Jomon10,000 years ago−40 m−45 to −30 mGlobal meanAs above
4Middle of the Initial Jomon9,000 years ago−25 m−30 to −20 mGlobal meanAs above
5End of the Initial Jomon8,000 years ago−4 m−5.5 to −2.5 mKanto RSLTanabe 2022 (−4±1.5 m at 8 ka); Tanabe 2020 (stepwise rise north of Tokyo Bay in the Early Holocene)
6Jomon transgression peak6,300 years ago+3 m+1.5 to +4.5 mKanto RSL (mean sea level)Tanabe 2022 (a plateau of +3±1.5 m mean sea level at 7–4 ka; the +4.5 m in the same study is mean high water spring, not mean sea level). +4 m lies within the range and appears in overviews such as a 2025 University of Tokyo guidance booklet, but there is no basis for making 6,300 years ago alone 1 m higher than 5,000 years ago. Hokuriku is +5 m (Fuji 2004) and Harima-nada about +1.5 m at most after tectonic correction (Sato 2008), so this is not a nationwide value
7Middle Jomon5,000 years ago+3 m+1.5 to +4.5 mKanto RSLTanabe 2022 (same plateau)
8Transition to Yayoi2,800 years ago−2 m−2.5 to −0.5 mKanto / Tone lowland RSL (debated)The −2 m on this map is an adopted value, not a directly reported one. Tanabe et al. 2016 (T.P. −2 m at 3.0 cal kyr BP); Tanabe 2022 (−1.5±1 m at 3 ka). The “Yayoi regression” is not confirmed in many regions and its cause is debated
9Yayoi & Kofun2,000 years ago0 m (assumed)−1 to +1 m (Kanto)Model assumptionTanabe 2022 (0±1 m at 2–0 ka). There is no basis for a finer value such as −0.5 m, so today’s level is assumed (0 m is not a measurement either)
10Medieval Warm Period (mid-Heian to Kamakura)AD 1100+0.6 mGlobal mean −0.3 to 0 m; regional records in the Tokyo lowland and Seto Inland Sea +0.6 to 0.7 m (another Seto Inland Sea source +0.4 to 1.2 m)Regional comparison hypothesisKatsushika City history (about 60 cm higher around 1100; the original study, reference year and method cannot be identified); Kimura (high-tide level of AD 901 +60–70 cm; not a direct measurement of mean sea level and not peer reviewed). See also Naruse et al. 1986 and Fuji 2004. Global mean: Kopp et al. 2016 (a fall of 8±8 cm in 1000–1400). This does not establish +0.6 m as Japan’s mean sea level in 1100. Nakatsuka 2022 criticises inferring “higher because it was warm”
11Little Ice Age (Edo period)AD 1650−0.3 mGlobal mean −0.3 to −0.1 mComparison setting (low end of the global range)Kopp et al. 2016; IPCC AR6 (Little Ice Age 1450–1850). One of the lowest periods of the past 3,000 years. −0.3 m, the low end of the global range, is used as a comparison setting. No value for Japan or Kanto in 1650 is confirmed (Hirai 1987 gives an upper bound of −0.15 m or lower about 300 years ago on the Okhotsk coast), so this is not an established value for Japan. Nakatsuka 2022 criticises inferring “lower because it was cold”
12PresentAD 20000 mReferenceSea level in 2000The app’s reference. Whether 0 m in the elevation data matches mean sea level in 2000 or T.P. (based on observations at Reiganjima in 1873–1879) is unconfirmed (GSI “mean sea level”). Japanese coastal observations show 1906–1925 only about 3.5 cm below 1996–2005 (JMA, mean of four long-record stations)
Verification status: The bibliographic details and values above were checked against abstracts and summaries; comparison with the original texts has not yet been done by this site (as of 29 September 2026). The results of an external check made on 30 September 2026 are given in the references below. Until the comparison is done, each value is treated as an estimate. DOIs are given only where certain.
Limits of the map: The map simply applies a water level to today’s terrain. It does not account for sediments deposited after the Ice Age (such as the alluvium of the Tokyo lowland) or for uplift and subsidence. For example, research suggests that a strait about 20 km wide remained between the Korean Peninsula and Tsushima even at the coldest time of the last glacial period (Park et al. 2000), but because of the coarse seabed data and the way it is drawn, the −120 m map shows the Korean Peninsula, Tsushima and Kyushu joined by land. The Tsugaru Strait (sill about −128 m in the data) opens at −120 m but may look closed at small scales such as on phones.

References

  1. Fairbanks, R. G. (1989) A 17,000-year glacio-eustatic sea level record: influence of glacial melting rates on the Younger Dryas event and deep-ocean circulation. Nature 342: 637–642. doi:10.1038/342637a0
  2. Peltier, W. R. & Fairbanks, R. G. (2006) Global glacial ice volume and Last Glacial Maximum duration from an extended Barbados sea level record. Quaternary Science Reviews 25: 3322–3337.
  3. Lambeck, K., Rouby, H., Purcell, A., Sun, Y. & Sambridge, M. (2014) Sea level and global ice volumes from the Last Glacial Maximum to the Holocene. PNAS 111(43): 15296–15303. doi:10.1073/pnas.1411762111 (PMC)
  4. Deschamps, P., Durand, N., Bard, E. et al. (2012) Ice-sheet collapse and sea-level rise at the Bølling warming 14,600 years ago. Nature 483: 559–564. doi:10.1038/nature10902
  5. Smith, D. E., Harrison, S., Firth, C. R. & Jordan, J. T. (2011) The early Holocene sea level rise. Quaternary Science Reviews 30: 1846–1860. doi:10.1016/j.quascirev.2011.04.019
  6. Tanabe, S. (2020) Stepwise accelerations in the rate of sea-level rise in the area north of Tokyo Bay during the Early Holocene. Quaternary Science Reviews 248: 106575. doi:10.1016/j.quascirev.2020.106575
  7. Tanabe, S. (2022) A 3-kyr plateau of middle Holocene higher sea levels on the central Kanto Plain, Japan. Quaternary Science Reviews 295: 107769. doi:10.1016/j.quascirev.2022.107769
  8. 田辺 晋・堀 和明・百原 新・中島 礼 (2016) 利根川低地における「弥生の小海退」の検証. 地質学雑誌 122(4): 135–153. doi:10.5575/geosoc.2016.0007 — Testing the “Yayoi regression” in the Tone lowland; −2.2 m at 3,000 cal BP.
  9. 松井裕之・多田隆治・大場忠道 (1998) 最終氷期の海水準変動に対する日本海の応答. 第四紀研究 37(3): 221–233. — Response of the Sea of Japan to last-glacial sea-level change.
  10. Park, S.-C., Yoo, D.-G. et al. (2000) Last glacial sea-level changes and paleogeography of the Korea (Tsushima) Strait. Geo-Marine Letters 20: 64–71.
  11. Kopp, R. E., Kemp, A. C., Bittermann, K., Horton, B. P., Donnelly, J. P., Gehrels, W. R. et al. (2016) Temperature-driven global sea-level variability in the Common Era. PNAS 113(11): E1434–E1441. doi:10.1073/pnas.1517056113 (PMC); correction doi:10.1073/pnas.1613396113. Pre-industrial global amplitude about ±8 cm; the main palaeo figure uses a 1900 baseline; global rise 1900–2000 about 13.8 cm.
  12. IPCC (2013) Climate Change 2013: The Physical Science Basis. Working Group I, Chapter 5 (variation of global mean sea level over the past 2,000 years within ±0.25 m).
  13. IPCC (2021) Climate Change 2021: The Physical Science Basis. Working Group I, Summary for Policymakers A.1.7 and A.2.4.
  14. 田邉 晋「沖積低地における地形と遺跡の編纂に基づいた過去4千年間の洪水の履歴と規模，要因の解明」Asahi Group Foundation research report, FY2023 (source) — the past 4,000 years in Kanto.
  15. 平井幸弘 (1987) サロマ湖の湖岸・湖底地形と完新世後半のオホーツク海の海水準変動. 東北地理 39: 1–15 (J-STAGE) — a low sea level of −0.15 m or lower about 300 years ago (an upper bound; uncalibrated radiocarbon dates; Okhotsk coast).
  16. Japan Meteorological Agency, long-term trend of sea level along the Japanese coast (JMA) — no monotonic rise like the global series since 1906; in the four-station mean of the public CSV, 1906–1925 is about 3.5 cm below 1996–2005 (calculated in the external check).
  17. Geospatial Information Authority of Japan, “Mean sea level” (GSI) — the height datum T.P. is based on observations at Reiganjima in 1873–1879; mean sea level in 2000 does not necessarily equal T.P. 0 m.
  18. University of Tokyo, Department of Civil Engineering (2025) Komaba guidance booklet, p. 60 (source) — an educational overview describing the Jomon transgression as “+4 m or more above present mean sea level”.
  19. 「荒川・妻沼低地と中川・渡良瀬低地における沖積層の形成過程とその対比」Abstracts of the Association of Japanese Geographers, spring 2012 (J-STAGE) — on the Arakawa side, marine beds reach about 60 km from the river mouth; the muds near Kumagaya (about 80 km) are fluvial.
  20. 「海水準変動とは」earthresources.sakura.ne.jp (source) — a collection of older studies from Hokuriku, Sendai, Kanto and elsewhere (updated 2016), used as an entry point to the original sources.

Part of the bibliographic details and the scope of verification comes from an external check (30 September 2026) of the settings for each era against the original sources. The original texts have not been checked directly by this site, so the values for each source follow that check.

Additional survey of the Medieval Warm Period (“Heian transgression”) (30 September 2026)

We searched again for studies claiming that sea level was higher than today in the Medieval Warm Period (the original texts have not been checked; based on search-result abstracts).

  1. Kemp, A. C. et al. (2011) Climate related sea-level variations over the past two millennia. PNAS 108(27). doi:10.1073/pnas.1015619108 — +0.6 mm/yr for about 400 years from AD 950 in North Carolina (about 20 cm in total); the same group later concluded this was a regional effect of ocean circulation (Kemp et al. 2017; Walker et al. 2021).
  2. Grinsted, A., Moore, J. C. & Jevrejeva, S. (2010) Reconstructing sea level from paleo and projected temperatures 200 to 2100 AD. Climate Dynamics 34: 461–472. doi:10.1007/s00382-008-0507-2 — a semi-empirical model; a maximum around 1150 (+12 to 21 cm) and a minimum around 1730 (−19 to −26 cm); never higher than today.
  3. Walker, J. S. et al. (2022) Timing of emergence of modern rates of sea-level rise by 1863. Nature Communications 13. doi:10.1038/s41467-022-28564-6 — century-scale variations are dominated by regional processes.
  4. 藤 則雄 (2004) 縄文時代における自然環境（3）自然環境要因の相関性. 金沢星稜大学論集 37(3) — Fig. 6 “Holocene history of Hokuriku” shows a “small Heian transgression” (about 1,000 years ago, no amount given), +5 m and +2 °C in the Jomon transgression, and a regression of −2 m and −1 to −1.5 °C in the Yayoi–Kofun period, based on pollen analysis and buried forests in Hokuriku.
  5. 藤 則雄 (2002) 北陸海退 the Hokuriku Regression (CiNii 110000040706) — proposal of a sea-level fall from the Late Jomon to the beginning of the Kofun period.
  6. 佐藤裕司 (2008) 瀬戸内海東部、播磨灘沿岸域における完新世海水準変動の復元. 第四紀研究 47(4): 247– — a curve corrected for local tectonics; about +1 m from 6,000 to 3,000 years ago and nearly 0 m since about 2,000 years ago.
  7. 成瀬敏郎・小野間正己・村上良典 (1986) 瀬戸内海，播磨灘沿岸における完新世後期の海水準変化に関する資料. Bulletin of the Faculty of Letters, Hiroshima University — mean high tide in the Heian period +0.4 to 1.2 m above present (radiocarbon dates of wave-cut benches and terraces).
  8. 太田陽子 et al. (1982) Holocene sea-level curves in Japan (reproduced in Ichiki 2012, Fig. 3-8); 遠藤邦彦 et al. (1989a) sea-level curve of the Oku-Tokyo Bay area (same, Fig. 3-9) — the height of the Jomon transgression varies by region from +2 to +5 m.

Basis for +0.6 m in the Medieval Warm Period (30 September 2026)

  1. Katsushika City history, Chapter 1 “Climate and nature of Katsushika”, explanation of sea-level change (Katsushika City) — states that sea level was about 60 cm higher than today around 1100; the original study, reference year and method cannot be identified from the text and figure.
  2. Kimura, T., “Sea level of the Seto Inland Sea in the Heian period from coastal terraces around Sugawara no Michizane’s landing site” (a two-page presentation, PDF, not peer reviewed) — estimates the high-tide level of the time as +60–70 cm above present from a hypothesised landing site of AD 901 and coastal terraces (not a direct measurement of mean sea level; short-term ground movement extrapolated to the long term; compared with tide levels in 2006–2010).
  3. Caution: Nakatsuka, T. (2022) 樹木年輪セルロースの酸素同位体比からみた古代日本の気候変動. Bulletin of the National Museum of Japanese History 232 (PDF) — criticises the dating and basis of the Fairbridge (1961) type sea-level curve used to support the “Heian transgression”. In Kopp et al. (2016) the global mean for the same period stays within ±8 cm; the regional +60 cm is relative sea level including tectonic and tidal differences and must not be read as a global rise caused by warming.

Summary: there is no basis in global mean sea level for drawing the Medieval Warm Period higher than today (Kopp et al. 2016 even shows a fall of 8±8 cm in 1000–1400). In Japan, records from the Seto Inland Sea (Naruse et al. 1986) and Hokuriku (Fuji 2004, “small Heian transgression”) describe about +0.5 m or a qualitative small transgression, but many include tectonic effects or are not peer reviewed, and no peer-reviewed study establishes a nationwide rise. This time machine uses +0.6 m for the Medieval Warm Period (a regional comparison hypothesis based on the Katsushika City history and Kimura’s Seto Inland Sea survey, not a global rise), −0.3 m for the Little Ice Age (a comparison setting at the low end of the global range), and assumes today’s level for the Yayoi & Kofun era.

Frequently asked questions

What is the Jomon transgression?

The Jomon transgression is the rise in sea level after the Ice Age that brought the sea deep into today’s plains. Reconstructions for the central Kanto Plain put sea level about 3 m (±1.5 m) above today 7,000–4,000 years ago (the time machine uses +3 m; the value of +4.5 m sometimes quoted is the spring high-tide level, not mean sea level). The sea reached as far as Omiya in Saitama, and Tokyo Bay extended far inland. This height is a relative value for Kanto; global mean sea level is estimated to have still been below today’s. Many shell mounds were formed near the coastline of the time.

How low was sea level in the Ice Age?

At the coldest time of the last glacial period, about 20,000 years ago, sea level was about 120 m lower than today. Hokkaido was joined to the continent via Sakhalin, and Honshu, Shikoku and Kyushu, with the Seto Inland Sea as dry land, formed one large island (Paleo-Honshu). Narrow sea channels are thought to have remained in the Tsugaru Strait and between the Korean Peninsula and Tsushima. People are thought to have arrived in the Japanese archipelago from about 38,000 years ago by routes through Tsushima, Okinawa and Hokkaido, crossing the sea by boat on the Tsushima and Okinawa routes. Mammoths crossed into Hokkaido from Sakhalin. The youngest fossils of Naumann’s elephant date to about 24,000 years ago, and some studies suggest it died out earlier. Because the map applies a water level to today’s terrain, the Korean Peninsula, Tsushima and Kyushu look joined by land at −120 m (a limit of the seabed data and the drawing method).

What can I learn with this time machine?

You can freely change sea level between −150 m and +50 m and see where the coastline would be at that level. It helps explain why Jomon shell mounds lie inland, which routes people took to reach Japan in the Ice Age, and how coastal areas could change with future sea-level rise.

How are the sea levels for each era chosen?

They are educational estimates compiled by this site from the literature. Before about 10,000 years ago they are global mean sea level (Fairbanks 1989; Lambeck et al. 2014; Deschamps et al. 2012; Smith et al. 2011); from about 8,000 to 2,000 years ago they are relative sea level on the central Kanto Plain (Tanabe 2020/2022; Tanabe et al. 2016); after that they follow the global mean (Kopp et al. 2016; IPCC AR6). Sea level differs by several metres to more than ten metres between regions, so coastlines outside Japan are approximate. The values, ranges and references for each era are in “Representative sea levels and references” on this page.

Did sea level rise in the Medieval Warm Period?

A reconstruction of global mean sea level (Kopp et al. 2016) estimates variations of about 10 cm up or down from the start of the Common Era to the Industrial Revolution, with a possible slight fall in 1000–1400. The time machine uses +0.6 m for the Medieval Warm Period. This is a regional comparison hypothesis based on records from the Tokyo lowland (Katsushika City history, “Heian transgression and sea-level change”) and the Seto Inland Sea (Kimura’s coastal-terrace survey, +60–70 cm in high-tide level); the original study and method have not been confirmed, and it does not indicate a rise in global mean sea level. The Little Ice Age in the Edo period is a comparison setting of −0.3 m, the low end of the global range. The Yayoi & Kofun era assumes today’s level because the details are unconfirmed. For the sea-level rise called the “Heian transgression” in Japan, no peer-reviewed study supporting a rise of a metre or more has been confirmed, and views differ.

Where do the background map and elevation data come from?

Land, lakes, rivers and place names come from public-domain Natural Earth, simplified and served by this site; when you zoom in, coastlines switch to high-resolution GSHHG data (LGPL). Elevation is read on your device from the Terrain Tiles (Mapzen) on AWS Open Data, and the GSI hillshade and pale maps are overlaid in real time over Japan. Elevation tiles are neither stored nor redistributed.