{
  "id": "last-glacial-maximum",
  "title": "Last Glacial Maximum",
  "lang": "en",
  "source_lang": "en",
  "summary": "The Last Glacial Maximum was the interval about 26,500–19,000 years ago when global ice sheets reached their greatest extent during the last glacial cycle.",
  "category": "geography",
  "subcategory": "",
  "type": "event",
  "aliases": [
    "LGM",
    "Last Glacial Maximum interval"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/last-glacial-maximum/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/last-glacial-maximum/",
      "json": "https://aiwiki.page/zh/last-glacial-maximum.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "The Last Glacial Maximum (LGM) was the interval during the most recent glacial cycle when continental ice sheets collectively reached their greatest extent and global sea level stood near its lowest position. It is commonly dated to approximately 26,500–19,000 years before present, although individual ice sheets and mountain glaciers attained their maximum positions at different times. Climate reconstructions and model experiments often use conditions around 21,000 years ago as a representative LGM time slice. The interval occurred late in the Pleistocene, rather than marking the beginning or the entirety of the last ice age.\n\nAt the LGM, enormous ice sheets covered much of northern North America and northern Europe, atmospheric greenhouse-gas concentrations were substantially lower than before industrialization, and global mean sea level was roughly 125–135 metres below its modern position. Exposed continental shelves altered the outlines and connections of landmasses, while colder oceans, expanded sea ice, and widespread changes in vegetation transformed environments far beyond the ice margins.\n\nThe LGM is a major reference state for understanding climate change. Its geological record reveals how ice sheets, oceans, the atmosphere, and living systems interact under sustained cooling. It also provides a demanding test for climate models: a successful explanation must account not only for global temperature but also for regional rainfall, ocean circulation, carbon storage, and the geography of ice."
    },
    {
      "heading": "Definition, chronology, and regional timing",
      "text": "The Last Glacial Maximum is a globally defined interval associated primarily with maximum land-ice volume, not a single moment when every location was simultaneously coldest or most extensively glaciated. The widely used range of approximately 26.5–19 thousand years ago emerged from syntheses of dated ice-sheet margins and other geological evidence. Within that interval, ice margins sometimes advanced, stabilized, or retreated while global ice volume remained close to its maximum. Consequently, an “LGM age” assigned to a landform must be interpreted in its regional setting.\n\nAges are commonly expressed as ka BP, meaning thousands of years before present, with “present” conventionally fixed at 1950 in radiocarbon chronology. Calendar ages and uncalibrated radiocarbon ages are not interchangeable. Because atmospheric radiocarbon abundance has varied, a sample dated to 18,000 radiocarbon years BP is appreciably older when converted to calendar years. Modern syntheses therefore specify whether ages are calibrated and include uncertainties arising from laboratory measurement, calibration, and geological interpretation.\n\nIn marine stratigraphy, the LGM falls within Marine Isotope Stage 2, a broader interval identified through oxygen-isotope variations in ocean sediments. Marine isotope stages are useful for correlating records across ocean basins, but they are not equivalent to mapped terrestrial glacial limits. The last glacial period itself began much earlier, following the last interglacial approximately 115,000 years ago, and included repeated changes in temperature and ice volume before the LGM.\n\nRegional differences are fundamental. Some mountain glaciers reached their greatest late-Pleistocene extent before the global maximum in ice volume, particularly where moisture availability changed strongly. Parts of the northern Eurasian ice-sheet system also advanced and retreated out of phase with one another. The timing of maximum glaciation in the Himalayas and adjacent Tibetan Plateau, for example, depended on elevation, exposure to monsoon moisture, and regional circulation; there was no single synchronous maximum across the entire region.\n\nThe end of the LGM is likewise not identical to the end of glacial conditions. Large-scale retreat began around 19,000 years ago, but extensive ice sheets persisted for many millennia. The transition into the Holocene, approximately 11,700 years ago, occurred substantially later. Distinguishing the maximum from the subsequent deglaciation prevents later events—such as major meltwater discharges or abrupt North Atlantic warming episodes—from being incorrectly treated as characteristics of the LGM itself."
    },
    {
      "heading": "Geological evidence and reconstruction methods",
      "text": "Reconstructing the LGM requires combining evidence that records different aspects of the environment. An ice margin, an ocean temperature estimate, and an atmospheric gas measurement do not describe the same variable, and each has its own chronological limitations. A climate proxy is useful because a measurable property—such as a fossil assemblage or chemical ratio—has a physical or ecological relationship to past conditions. Robust reconstructions compare several independent proxies rather than relying on one indicator.\n\nOn land, moraines, glacially transported boulders, streamlined bedrock, and sediment deposited beneath moving ice establish former glacier extent and flow direction. Terminal moraines can identify a prolonged or repeated ice-margin position, although the outermost preserved ridge need not mark the absolute maximum: earlier deposits may have been eroded. Offshore seismic profiles and seabed landforms extend these reconstructions across continental shelves, where grounded ice once occupied areas now submerged.\n\nRadiocarbon dating provides ages for organic materials associated with glacial deposits, including wood, peat, and shells. Usually it brackets rather than directly dates ice occupation. Organic material below a till can establish that an advance occurred later, while vegetation growing on a deglaciated surface can constrain when retreat had already occurred. Marine samples require corrections for reservoir effects because ocean carbon may be older than atmospheric carbon, with offsets that vary geographically and through time.\n\nCosmogenic-nuclide dating measures isotopes such as beryllium-10 produced when cosmic rays strike exposed rock. It is especially important for dating boulders on moraines and bedrock uncovered by retreating ice. Interpretation depends on exposure history: inherited nuclides can make a surface appear too old, whereas erosion, snow cover, or later burial can lower the measured concentration. Luminescence methods, which estimate when sediment grains were last exposed to light, provide complementary ages for windblown and water-laid deposits.\n\nOcean sediments preserve microfossils, chemical temperature indicators, and oxygen-isotope measurements. The oxygen-isotope composition of benthic foraminiferal shells reflects both deep-water temperature and the isotopic composition of seawater, the latter influenced by global ice volume. Separating these influences is essential. Polar ice cores provide records of snowfall chemistry, local temperature indicators, dust, and trapped atmospheric gases; the gases are younger than the surrounding ice because air remains mobile within porous snow before bubbles become sealed.\n\nReconstructions also depend on spatial sampling. Early marine compilations, notably the CLIMAP project’s 1981 reconstruction, established a global picture but left substantial uncertainty in tropical temperatures and seasonal sea ice. Later work incorporated more proxies, improved calibrations, and statistical methods that combine observations with physically consistent model fields."
    },
    {
      "heading": "Orbital forcing and climate feedbacks",
      "text": "The LGM developed through the interaction of slow changes in Earth’s orbit with feedbacks involving ice, greenhouse gases, vegetation, dust, and ocean circulation. Milankovitch cycles alter the seasonal and latitudinal distribution of incoming sunlight. They do not simply reduce solar energy everywhere at once. For the growth of northern continental ice sheets, summer conditions at high northern latitudes are especially important because they determine whether winter snow survives the melt season.\n\nOrbital eccentricity changes the shape of Earth’s orbit, obliquity changes the tilt of its rotational axis, and precession changes the seasonal position of Earth along its orbit. Their characteristic timescales are approximately 100,000 years, 41,000 years, and 19,000–23,000 years, respectively. These variations influence the balance between accumulation and melting over many millennia. The large ice sheets present at the LGM therefore reflected a long history of climate forcing and ice dynamics, not just the instantaneous orbital configuration at 21,000 years ago.\n\nIce amplified the initial cooling. Snow and ice reflect a greater fraction of incoming sunlight than forests, bare ground, or open water, increasing planetary albedo. As ice sheets thickened, their high surfaces became colder, helping preserve snow. Their elevation and shape also reorganized atmospheric circulation. These effects were partly counteracted by processes such as faster ice flow, reduced snowfall over very cold interiors, and melting or calving at their margins.\n\nAtmospheric carbon dioxide concentrations were approximately 185–195 parts per million during much of the LGM, compared with about 280 parts per million before industrialization. Methane concentrations were generally around 350–400 parts per billion, also markedly below preindustrial values. Lower greenhouse-gas concentrations reduced the atmosphere’s ability to retain outgoing infrared radiation and helped transmit cooling beyond the northern ice sheets.\n\nA simplified expression for carbon dioxide’s radiative forcing is\n\n\\[\\Delta F \\approx 5.35\\ln\\left(\\frac{C}{C_0}\\right),\\]\n\nwhere \\(C\\) and \\(C_0\\) are the compared concentrations and \\(\\Delta F\\) is measured in watts per square metre. Comparing 190 with 280 parts per million gives a forcing of roughly \\(-2.1\\) watts per square metre from carbon dioxide alone. This is not the total LGM forcing, because ice-sheet albedo, other gases, dust, and geographical changes also mattered.\n\nThe resulting temperature response depended on feedbacks involving water vapour, clouds, snow, and sea ice. These processes make the LGM useful for evaluating climate sensitivity, but an estimate derived from it must distinguish the effects prescribed as external boundary conditions from those treated as feedbacks. Ice sheets, for example, may be fixed in a model experiment even though they evolved interactively in the real climate system."
    },
    {
      "heading": "Continental ice sheets and mountain glaciers",
      "text": "The largest Northern Hemisphere ice mass was the Laurentide Ice Sheet, which covered most of Canada and extended into the northern United States. Its southern margins reached regions now occupied by the Great Lakes and parts of the upper Midwest and northeastern United States. It was several kilometres thick in its major interior domes. Its weight depressed the underlying crust, while its surface elevation formed a continental-scale obstacle to atmospheric flow.\n\nThe Cordilleran Ice Sheet occupied much of western Canada and adjoining mountainous terrain. During maximum conditions it coalesced with the Laurentide system in parts of western Canada, limiting ice-free passage between the northern interior and areas farther south. This configuration is important both for reconstructing ice drainage and for evaluating possible routes of human movement. Large portions of Alaska and northeastern Siberia remained outside the continental ice sheets, despite severe cold, because low snowfall restricted ice accumulation.\n\nIn northern Europe, the Fennoscandian Ice Sheet covered Scandinavia and the Baltic region and extended into surrounding lowlands. Its margins and connections changed through time. The British–Irish Ice Sheet covered much of Britain and Ireland and extended offshore, while the Barents–Kara ice-sheet complex occupied parts of the Eurasian Arctic continental shelves. These were dynamic systems rather than fixed ice caps, with fast-flowing corridors transporting ice toward marine margins.\n\nThe Antarctic ice sheet expanded particularly across continental shelves, but the amount of additional ice varied by sector. Reconstructions of Antarctica remain less certain than those of some northern land margins because much of the relevant evidence lies beneath modern ice or deep water. Greenland also supported a larger ice sheet, with outlet glaciers extending farther across coastal shelves than today. Sea ice surrounding these ice masses must be distinguished from grounded land ice: floating sea ice contributed little directly to lower sea level.\n\nMountain glaciers expanded in the Alps, Andes, Southern Alps of New Zealand, and many other ranges. Their extent depended on both cooling and snowfall. A glacier could remain relatively restricted in a cold but dry region, while a wetter mountain range supported a large ice field. The glacier mass balance—the difference between accumulation and loss—therefore provides a more complete explanation than temperature alone.\n\nIce streams, basal sliding, and ice calving controlled how these ice masses discharged. Meltwater pressure beneath glaciers could reduce friction, while soft sediments could deform under moving ice. Marine-based margins were especially sensitive to water depth and bed geometry. These processes produced lobes, rapid flow corridors, and uneven retreat, leaving the landforms used to reconstruct the former ice sheets."
    },
    {
      "heading": "Lower sea level and transformed geography",
      "text": "The storage of water in expanded continental ice sheets lowered global mean sea level by roughly 125–135 metres relative to modern conditions. This estimate concerns the globally averaged redistribution of water, not a uniform fall of the same magnitude along every coast. Local relative sea level also depended on crustal deformation, changes in Earth’s gravity field, and the position of a site relative to major ice loads.\n\nGlacial isostatic adjustment produced especially large geographical differences. Beneath an ice sheet, the crust subsided under the load, while a peripheral forebulge could rise outside the ice margin. As ice later melted, depressed regions rebounded and some formerly uplifted regions subsided. These responses continue today. Reconstructing LGM coastlines therefore requires a model of both ice history and the mechanical response of Earth’s interior, rather than simply subtracting 130 metres from a modern elevation map.\n\nThe lowered ocean exposed broad shelves around many continents. Beringia linked northeastern Asia and northwestern North America across the present Bering Strait and included extensive adjoining lowlands. In Southeast Asia, Sundaland joined the Malay Peninsula with Sumatra, Java, Borneo, and large intervening shelf areas. Farther east, Sahul united Australia, New Guinea, and Tasmania into a larger landmass.\n\nNot all straits disappeared. Deep channels between the Sunda and Sahul shelves remained marine barriers, so movement between them still required water crossings. Similarly, lower sea level did not eliminate every separation between islands elsewhere. The effects depended on local seabed depth and topography. In northwestern Europe, large areas of the North Sea became exposed land or were occupied by ice, while river systems extended across terrain now submerged.\n\nExposed shelves changed more than the availability of land routes. Rivers travelled farther before reaching the ocean, incised valleys across shelves, and delivered sediment to different depositional environments. Estuaries and shallow coastal habitats shifted seaward. The narrower ocean passages and altered shelf areas also affected tides, regional circulation, and exchanges between marginal seas and the open ocean.\n\nThese changes complicate archaeological reconstruction because former coastal settlements and landscapes may now lie underwater. They also complicate sea-level estimation: coral growth depths, sediment compaction, tectonic movement, and glacial deformation must all be considered when interpreting an ancient shoreline. Reconstructions therefore combine far-field coral records, submerged landforms, sedimentary evidence, and physical models rather than treating a single coastal site as a direct global sea-level gauge."
    },
    {
      "heading": "Temperature, atmospheric circulation, and the water cycle",
      "text": "Global mean surface temperature during the LGM was several degrees Celsius below the preindustrial level. Estimates differ with the proxies used, the treatment of data gaps, and the method for combining land and ocean information. A 2020 reconstruction by Jessica Tierney and colleagues estimated approximately 6.1°C of global cooling, with a reported interval of about 5.7–6.5°C. Other reconstructions have produced smaller values, illustrating that the exact magnitude remains a quantitative research question rather than a directly measured global average.\n\nCooling was highly uneven. Ice-covered continental interiors experienced extreme local changes because of their elevation, reflectivity, and distance from open water. High northern latitudes generally cooled more than the tropics, and land temperatures changed more strongly than ocean temperatures. Tropical conditions nevertheless differed substantially from modern ones: lower sea-surface temperatures, colder mountain environments, and downslope shifts in vegetation all demonstrate that the tropics were not climatically unchanged.\n\nThe ice sheets redirected atmospheric circulation. Their great height altered planetary waves, storm tracks, and the jet stream, helping create regional patterns that cannot be explained by a simple pole-to-equator temperature shift. In western North America, changed circulation and lower evaporation contributed to large lakes in some enclosed basins. Lake Bonneville and Lake Lahontan are prominent examples, although their highest shorelines were not necessarily synchronous with the global ice maximum.\n\nThe global water cycle was generally weaker because a colder atmosphere held less water vapour and evaporation was reduced. This did not mean that every region received less precipitation. Regional rainfall depended on storm tracks, moisture transport, seasonal circulation, and proximity to changed coastlines. A lake could expand even with limited rainfall if colder conditions substantially reduced evaporation from its surface and catchment.\n\nMany tropical monsoon systems were weaker or geographically displaced. Changes in land–sea temperature contrasts, greenhouse gases, ice sheets, and exposed shelves all influenced their behaviour. Over parts of Asia, increased dust transport and changes in vegetation indicate cold, dry conditions, but the response varied between summer-rainfall and winter-rainfall regions. The LGM should therefore not be represented by a single global map of uniformly intensified aridity.\n\nDust deposition increased markedly in many ice-core and marine records. Reduced vegetation cover, exposed sediment, glacial grinding, and altered winds supplied and transported mineral particles. Deposits on the Loess Plateau and elsewhere preserve aspects of this circulation. Dust also affected radiation and supplied nutrients to ocean waters, linking atmospheric conditions to the carbon cycle."
    },
    {
      "heading": "Ocean circulation, sea ice, and carbon storage",
      "text": "The glacial ocean was colder and, on average, saltier than today because freshwater was stored in continental ice. Cooling increased seawater density, while salt redistribution through freezing and melting influenced where dense waters formed. Sea ice expanded seasonally and regionally, particularly in the Southern Ocean and North Atlantic sector. Its exact extent remains uncertain because many marine indicators preferentially record a particular season rather than the annual mean.\n\nThe Atlantic meridional overturning circulation did not simply cease during the LGM. Evidence indicates that northern-sourced waters continued to form, but their depth, properties, geographical reach, and rate of renewal differed from modern conditions. Many reconstructions show a shallower northern-sourced water mass overlying a larger volume of southern-sourced deep water. Estimates of transport strength remain sensitive to proxy interpretation and the interval examined.\n\nSeveral geochemical tracers constrain this circulation. Carbon isotopes in benthic foraminifera help distinguish water masses and nutrient histories. Radiocarbon measurements indicate how long waters have been isolated from atmospheric exchange, although changing surface reservoir ages complicate interpretation. Ratios of sedimentary protactinium to thorium provide another circulation indicator, but particle scavenging and sediment transport also affect them. Agreement among different tracers is therefore more informative than any one measurement.\n\nThe low atmospheric carbon dioxide concentration required a redistribution of carbon among the atmosphere, ocean, land, and sediments. Colder water could dissolve more carbon dioxide, but temperature-dependent solubility alone cannot explain the entire glacial reduction. Changes in ocean stratification, deep-water ventilation, biological export, and seawater chemistry also contributed. On land, reduced forest extent and lower carbon dioxide altered vegetation and carbon storage, partly opposing the transfer of carbon out of the atmosphere.\n\nThe Southern Ocean was especially important because it is a major region where deep ocean waters interact with the atmosphere. More persistent sea ice and stronger stratification could reduce the escape of carbon-rich deep water to the air. Meanwhile, iron supplied by mineral dust could stimulate phytoplankton growth in iron-limited waters, strengthening parts of the biological pump that transfers carbon from surface waters to depth.\n\nNo single mechanism fully explains the glacial carbon cycle. A successful account must reproduce atmospheric gas concentrations, ocean nutrient patterns, carbon-isotope distributions, and evidence for deep-water isolation. It must also explain how these conditions changed during deglaciation. LGM oceanography is therefore reconstructed as an interacting physical and biogeochemical system, rather than as a circulation pattern inferred solely from temperature."
    },
    {
      "heading": "Vegetation, soils, and animal communities",
      "text": "LGM ecosystems differed from modern ones through combinations of cold, dryness, seasonality, low carbon dioxide, and disturbance. Pollen preserved in lake and marine sediments provides much of the evidence for vegetation distributions. Plant macrofossils, charcoal, beetle remains, ancient DNA, and fossil mammals add information at different spatial scales. These records often reveal communities without close modern equivalents, limiting reconstructions based only on present-day vegetation zones.\n\nAcross extensive parts of northern Eurasia and Beringia, open herb-rich vegetation supported the mammoth steppe. This was not simply modern Arctic tundra displaced southward. It included grasses, sedges, and diverse flowering herbs under strongly seasonal conditions, with regional variations in moisture and soil fertility. Woolly mammoths, horses, reindeer, bison, and other large herbivores occupied different parts of this broad ecological mosaic.\n\nForests generally contracted or changed composition, but they did not disappear everywhere outside the tropics. Sheltered terrain and favourable regional climates supported refugial populations of trees. Southern European peninsulas were important refugial regions, while evidence also indicates survival in some more northerly locations. The distribution and size of such populations affected the routes and rates of forest expansion after the LGM.\n\nTropical responses were diverse. Cooler conditions displaced montane vegetation downslope, and reduced moisture favoured more open vegetation in some regions. However, the proposition that the entire Amazon forest fragmented into a few isolated refuges is not supported as a universal description. Different areas retained forest, experienced compositional change, or developed more open environments. Local pollen records, sediment history, and climate modelling are needed to distinguish these outcomes.\n\nLow atmospheric carbon dioxide imposed an additional physiological constraint on plants. Most trees and many temperate herbs use the C3 photosynthetic pathway, whose productivity is more strongly limited by low carbon dioxide than that of C4 plants. Yet low temperatures often favoured C3 vegetation relative to warm-adapted C4 grasses. Vegetation patterns therefore reflected interacting physiological and climatic controls rather than carbon dioxide concentration alone.\n\nFrozen ground extended beyond the continental ice sheets. Permafrost, seasonal frost disturbance, and sparse vegetation influenced soil development, slope movement, and river behaviour. Windblown loess accumulated where sediment supply, wind transport, and surface conditions permitted. These environments later affected the formation of fertile soils and preserved detailed records of changing dust and climate. Animal populations responded to the resulting habitat mosaics, but late-Pleistocene extinctions occurred at different times in different regions and cannot be attributed to the LGM as a single simultaneous event."
    },
    {
      "heading": "Human populations and archaeological landscapes",
      "text": "The LGM occurred within the Upper Palaeolithic in much of Eurasia, but archaeological period names are regional and should not be applied uniformly around the world. Human communities occupied environments ranging from tropical forests and savannas to cold continental interiors. Their archaeological record reflects changes in settlement, mobility, subsistence, and technology, although preservation and research intensity strongly influence what is visible.\n\nIn Europe, expanded ice and severe periglacial conditions restricted occupation in parts of the north. Population distributions shifted, with important concentrations in southern and southwestern regions. The Solutrean archaeological tradition, dated broadly to around 24,000–19,000 years ago, is associated principally with southwestern Europe. Its finely worked stone points represent a regional technological development, not a universal LGM culture or an automatic measure of adaptation to cold.\n\nShelter, clothing, fire, food storage, and flexible mobility were important components of survival in cold environments. Animal bones and stone tools indicate the exploitation of large herbivores, but plant foods and smaller animals could also contribute where conditions permitted. The relative importance of each resource varied considerably. Archaeological evidence rarely supports a single subsistence model for all glacial populations, and sites used briefly for hunting should not be treated as representative of an entire annual economy.\n\nThe geography of human migration was affected by both exposed land and ice barriers. Beringia connected Asia with North America, but the ice-sheet configuration restricted routes farther south. The timing of movement into the Americas, the availability of coastal routes, and the later opening of an interior passage are distinct questions. The existence of a land bridge alone does not establish when people crossed it, and an ice-free route was not necessarily immediately productive enough to support sustained travel.\n\nLower sea level also changed settlement opportunities in Southeast Asia and Sahul. Australia and New Guinea had been reached well before the LGM; their initial occupation therefore cannot be explained as a consequence of maximum shelf exposure. During the LGM, communities encountered altered coastlines, expanded dry interiors, and changing distributions of freshwater and food. Deep-water gaps persisted between some major landmasses despite the lower ocean.\n\nAncient genetic evidence helps reconstruct population relationships and changes in ancestry, but genetic lineages are not interchangeable with archaeological cultures or languages. Inferred population contractions and expansions depend on demographic models and sampling. Coastal inundation after the LGM further biases the record by concealing former shorelines. Understanding human responses consequently requires combining archaeology, genetics, palaeoenvironmental evidence, and reconstructed geography without assuming that climate alone determined cultural change."
    },
    {
      "heading": "Deglaciation and the LGM as a climate-model benchmark",
      "text": "The transition out of maximum glacial conditions began around 19,000 years ago, but it unfolded unevenly. Increasing northern summer insolation favoured ice loss, while changes in ocean circulation, atmospheric carbon dioxide, and albedo amplified and redistributed warming. Initial retreat could expose darker land and water, increasing absorbed sunlight. As ice surfaces lowered, they also became exposed to warmer air, reinforcing melting in some regions.\n\nDeglaciation included abrupt events superimposed on the longer-term warming. Heinrich Stadial 1, broadly about 18,000–14,700 years ago, involved major North Atlantic cooling and extensive disruption of ocean circulation. The Bølling–Allerød warming began around 14,700 years ago, followed later by the Younger Dryas, approximately 12,900–11,700 years ago. These events occurred after the principal LGM interval and differed greatly in their regional expression.\n\nGlobal sea level rose as ice sheets lost mass, but the rate was not constant. Meltwater Pulse 1A, around 14,600 years ago, involved a rapid rise of roughly 10–20 metres over several centuries, with estimates depending on chronology and reconstruction method. The respective contributions of different ice sheets remain debated. Large meltwater inputs changed coastal environments and potentially influenced ocean density and circulation, while retreating ice margins created lakes capable of sudden drainage.\n\nFor modelling, the LGM offers a climate state substantially different from the modern one but documented by a wide range of evidence. The Paleoclimate Modelling Intercomparison Project coordinates experiments using common boundary conditions. These commonly include reduced greenhouse gases, reconstructed ice-sheet elevation and extent, lower sea level, and the orbital configuration around 21,000 years ago. Experiments with general circulation models and more comprehensive Earth-system models can then be compared with proxy reconstructions.\n\nThese comparisons test more than whether a model produces sufficient average cooling. They examine tropical sea-surface temperatures, polar amplification, monsoon changes, dust, vegetation, and deep-ocean properties. A model may reproduce the global mean while misplacing regional rainfall or ocean water masses. Conversely, a mismatch at one site may reflect uncertain proxy seasonality, dating, or local processes rather than a failure of the large-scale simulation.\n\nImportant uncertainties remain in ice-sheet geometry, Antarctic ice volume, regional precipitation, and the mechanisms of carbon storage. Standard equilibrium experiments also simplify a climate that retained variability and inherited effects from earlier millennia. Transient simulations address these limitations by allowing forcing and ice geometry to change through time, while data assimilation combines model dynamics with geographically incomplete observations. These approaches are used to distinguish competing reconstructions and to quantify how strongly LGM evidence constrains climate feedbacks."
    }
  ],
  "links": [
    "climate-change",
    "himalayas",
    "tibetan-plateau",
    "holocene",
    "climate-proxy",
    "moraine",
    "radiocarbon-dating",
    "milankovitch-cycles",
    "carbon-dioxide",
    "greenhouse-gas",
    "climate-sensitivity",
    "laurentide-ice-sheet",
    "canada",
    "united-states",
    "cordilleran-ice-sheet",
    "europe",
    "antarctica",
    "greenland",
    "glacier-mass-balance",
    "ice-calving",
    "glacial-isostatic-adjustment",
    "beringia",
    "sahul",
    "jet-stream",
    "precipitation",
    "loess-plateau",
    "dna",
    "permafrost",
    "human-migration",
    "archaeology",
    "general-circulation-model"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}