Search

the Siberian craton to the north and the TarimNorth China cratons to the south, is a complex collage of microcontinental blocks, island arcs, oceanic crustal remnants and continental marginal facies rocks. It is one of the largest and most complex accretionary orogenic belts and the most important site of Phanerozoic continental growth on the Earth [Jahn et al., 2000, 2004; Kovalenko et al., 2004] The widespread occurrence of large volumes of granitoids, mostly with juvenile sources, is a typical characteristic of the CAOB. These granitoids have been intensely studied (e.g. [Jahn et al., 2000, 2004; Kovalenko et al., 2004; Sorokin et al., 2004; Vladimirov et al., 2001; Han et al., 2010; Wang et al., 2006, 2015; Wu et al., 2011; Li et al., 2013; Yarmolyuk et al., 2002]). However, these studies mainly focused on some certain countries or regions.

Mongolia lies in the central part of the Central Asian Orogenic Belt [Mossakovsky et al., 1994; Zorin, 1999; Jahn, 2004; Khain et al., 2003; Badarch et al., 2002; Windley et al., 2007; Zhang et al, 2008], or Altaids [Şengör et al., 1993; Şengör, Natal’in, 1996; Wilhem et al., 2012], which is fringed by the Siberian craton in the north and by the Tarim and Sino-Korean Cratons in the south. According to the recent tectonic subdivision, the territory of Mongolia is subdivided into Northern and Southern domains which are separated by the so called Mid Mongolian Tectonic Line [Tomurtogoo, 2012]. The Herlen Massif is one of the important tectonic units of the South Mongolian domain in the Argun-Idermeg super terrane extending through the territories of Russia and China [Parfenov et al., 2009; Tomurtogoo, 2014b]. The Herlen massif, also known as Herlen superterrane [Tomurtogoo, 2012] or Idermeg terrane [Tomurtogoo, 2014a] is composed of Ereendavaa, Undur-Khaan, Idermeg and Gobian Altay-Baruun Urt terranes converged at the end of the Cambrianbeginning of the Ordovician [Badarch et al., 2002; Tomurtogoo, 2014b].

The paper reviews previously published and presents new data on the Darhad Basin, Mongolia, which is a key locality for reconstruction of Quaternary glaciation and environmental changes in northern Central Asia. The previously published data include those obtained in the 20th century by Soviet geological surveys and academic researches of the Darhad Basin, which are, as a rule, not accessible for an international reader, and those obtained by several international teams during the last decade. The new results include geomorphologic, sedimentological and geochronological data obtained prior to and within the International Darhad Drilling Project (DDP-2010). These data show that the Darhad sedimentary sequence has been formed since the Pliocene and represents a detailed archive of environmental changes due to a high content of lacustrine beds. Lakes formed several times in the Darhad Basin by basaltic, glacial and sedimentary dams, which blocked the water outlet of the basin. Of special interest in this paper are the late Pleistocene damming events, which are well-documented in the sediments and landforms, however, their chronology is still obscure. There have been two stages of deep lake. The first lake was dammed by a glacier during late MIS 5 ( Krivonogov et al., 2005 ) or, alternatively, during early to middle MIS 3 ( Gillespie et al., 2008a ). The second damming, glacial or sedimentary, formed another deep lake during MIS 4 or MIS 2. The level of the lake was greatly variable up to its complete disappearance. Our new data from the DDP10-3 and DN-1 boreholes and from the Hodon outcrop illustrate the Holocene history of the paleolake, which was dammed by the sediments and was kept at low levels. We propose a first sedimentation model for the Holocene part of the lacustrine sedimentary sequence. The lake was minimal or disappeared at ca 12–9.6 and after 4.5 ka cal. BP, relatively deep at 9.6–7.1 and 6.4–4.5 ka cal. BP and shallow at 7.1–6.4 ka cal. BP. Conclusively, we highlighted solved and unsolved geological problems of the Darhad Basin, the most important of which are our recognized stages of the paleolake development (solved) and the detailed chronology and environments of the lacustrine and glacial events (unsolved).

A high-resolution pollen- and diatom-based bioclimatic reconstruction from Ugii Nuur lake core with a chronological support of 14 AMS dates revealed that a prolonged dry climate prevailed between 5830–3080 14C BP in central Mongolia, as indicated by a dramatic increase in Chenopodiaceae pollen percentages at the expense of Pinus, Poaceae, Cyperaceae and other mesophytic forbs pollen percentages. Higher values of pollen-based temperature index and lower values of pollen-based moisture index also support the notion that the mid-Holocene was persistently warm and dry relative to the preceding and following periods. This mid-Holocene drought is further confirmed by diatom and sedimentary variations, both the planktonic/benthic diatom ratio and the deposition rate being the lowest. Reviews of published regional palaeoclimate data suggest that a prolonged mid-Holocene drought might have prevailed extensively in the central-east Asian arid and hyper-arid areas. The prolonged and extensive drought might have been resulted from a well-documented large-scale temperature rise. The temperature rise-dictated enhancement of evaporation might have exceeded the precipitation increase (if any), resulting in the aridity increase.

A new experiment of the Hövsgöl Drilling Project was carried out in Lake Hövsgöl in 2004, as part of the international program of scientific drilling in the Baikal Rift. The reported data include a preliminary description of the recovered lake sediments (HDP-04 core), with their compositions and physical properties, and a tentative age-depth model based on the paleomagnetic polarity scale. This is the first evidence that the lithology of sediments deposited for the past ~1 Myr records periodic alternation of carbonate-free diatomaceous mud and carbonate-bearing silty clay. The diatom intervals in the record are interpreted as corresponding to interglacials by analogy with the periodicity known since the Last Glacial. The core bears signature of at least nine lithological change events. The sediment lithology records extremely low stand of Paleo-Hövsgöl (shallow-water facies produced by erosion of older sediments at the point where the today’s lake reaches a depth of 240 m). Correlation of the HDP-04 core data with reflection profiling evidence confirms the presence of quite a large gap in the Pleistocene sedimentary record from the Hövsgöl rift basin. The discovery of alternating carbonate-rich/carbonate-free cycles and evidence for sudden lake level changes impart special importance to the Hövsgöl archive: It can provide a deeper insight into the regional water budget and humidity history than it has been so far possible for the Middle and Late Pleistocene.

This study focuses on an eolian section (Shaamar section) in the Northern Mongolian Plateau and compares the eolian sequences in the Northern Mongolian Plateau with those along the southern boundary of the Gobi Deserts to better understand the dynamics of the Gobi deserts during the past ∼ 35,000 14C yr BP. Only a weak Entisol-like paleosol was formed around 29,000 14C yr BP in the exposed MIS 3 portion at the Shaamar section, whereas an eolian–colluvium–paleosol sequence at nearby Bureghkanga section exhibits three paleosols formed around 29,000 14C yr BP, 31,000 14C yr BP, 34,000 14C yr BP. Loess–paleosol sequences in the western Siberian Lowland exhibit four paleosols formed from > 40,000 to 25,000 14C yr BP (i.e., MIS 3). During MIS 2 two Mollisol-like paleosols were formed from ∼ 25,000 to ∼ 21,000 14C yr BP and from ∼ 16,000 to ∼ 13,000 14C yr BP and one Entisol-like paleosol was formed around 9500 14C yr BP at the Shaamar section. The MIS 2 was characterized primarily by silt deposition, except for the interval between ∼ 21,000 and ∼ 16,000 14C yr BP that was dominated by sand deposition. The Holocene began with the Mollisol-like paleosol formation (from ∼ 8600 to ∼ 7000 14C yr BP). The mid-Holocene (∼ 7000 to ∼ 3000 14C yr BP) was marked by a relatively poor vegetation cover and the late Holocene (since ∼ 3000 14C yr BP) by the densest vegetation cover of the entire Holocene. In summary, the maximal extent of hyperarid and arid areas (Gobi deserts and Gobi-like) occurred twice: (1) from ∼ 21,000 to ∼ 16,000 14C yr BP and from ∼ 13,000 to ∼ 8600 14C yr BP when the dominant eolian deposition conditions extended from 56° N to 33° N or even larger. The extent of hyperarid and arid areas retreated to the area between 38° N and 48° N or even much smaller several times during MIS 3 and during the early Holocene. Considering the uncertainties of dates, it seems that the Holocene bioclimatic conditions might have changed more or less synchronously between the north and the south and that the bioclimatic conditions varied more frequently in the north than in the south during MIS 3. Two Mollisol-like paleosols (IIc: ∼ 13,000 to ∼ 16,000 14C yr BP; and IIIa: ∼ 21,000 to ∼ 25,000 14C yr BP) were well developed during MIS 2 in the north, but no corresponding major paleosols were discovered in the south, suggesting that the climate did not change synchronously during MIS 2. The climates during the stadial–interstadial MIS 3–2 transition (∼ 25,000 to ∼ 21,000 14C yr BP) and during the last glacial maximum-deglacial transition (∼ 16,000 to ∼ 13,000 14C yr BP) were much more humid in the north than in the south.

This study compares two pairs of adjacent lacustrine and eolian sections at sites in the southern and northern Mongolian Plateaus in order to test spatial climate variability during the Holocene. Based on the lithology, proxy data, and 14C dated and the interpolated ages, the following observations can be made. In the northern Mongolian Plateau, a best developed Holocene paleosol dated at 8672 14C yr BP at the Shaamar section and the carbonate-rich laminated layer in the Gun Nuur lake core mark the interval of warmer and dryer climate during the early Holocene. Younger paleosols at the Shaamar section and corresponding organic-rich layers in the Gun Nuur core were formed under distinctly cooler and more humid conditions. The Baahar Nuur lake core in the southern Mongolian Plateau and the Dingxi-type section in the northern part of the Western Chinese Loess Plateau appear to indicate that a prolonged interval of maximum humidity prevailed in this region during the early and mid-Holocene (9000–4000 14C yr BP). By contrast, in the northern Mongolian Plateau the most humid conditions seem to have occurred from 4500 to 2500 (possibly to 1650) 14C yr BP. This discrepancy implies that the concept of the Holocene climatic optimum has limitations and may have to be reconsidered if it is intended to have a large-scale connotation.

Two gravity cores (1.1 and 2.2 m long) of deep-water bottom sediments from Lake Khubsugul (Mongolia) were studied. The Holocene, biogenic silica and organic matter-rich part of the first core was subjected to AMS radiocarbon dating which placed the date of dramatic increase of pelagic diatoms (40 cm below sediment surface) at a calendar age of 11.5 cal ky BP. ICP-MS analysis of weak nitric acid extracts revealed that the upper Pleistocene, compared to the Holocene samples, were enriched in Ca, Cinorg, Sr, Mg and depleted of U, W, Sb, V and some other elements. Transition to the Holocene resulted in an increase of total diatoms from 0 to 108 g-1, of BiSi from 1% to 20%, of organic matter from 6%. The Bølling–Allerød–Younger Dryas–Holocene abrupt climate oscillations manifested themselves in oscillations of geochemical proxies. A remarkable oscillation also occurred at 22 cm (ca. 5.5 ky BP). The Pleistocene section of the second, longer core was enriched in carbonate CO2 (up to 10%) and water-extractable SO42- (up to 300 times greater than that in Holocene pore waters). All this evidence is in an accord with the earlier finding of drowned paleo-deltas at ca. 170 m below the modern lake surface of the lake [Dokl. Akad. Nauk 382 (2002) 261] and suggests that, due to low (ca. 110 mm) regional precipitation at the end of the Pleistocene, Lake Khubsugul was only 100 m deep, and that its volume was ca. 10 times less than today.