Foreword 1. Introduction 2. Runoff, erosion and delivery to the coastal ocean 3. Temporal variations 4. Human impacts Appendices. Global River Database: Appendix A: North and Central America Appendix B: South America Appendix C: Europe Appendix D: Africa Appendix E: Eurasia Appendix F: Asia Appendix G: Oceania References Index.

/pdf/river-discharge-to-the-coastal-ocean-a-global-synthesis-571z1g19ts.pdf

River Discharge to the Coastal Ocean: A Global Synthesis

We present a new time-slice reconstruction of the Eurasian ice sheets (British–Irish, Svalbard–Barents–Kara Seas and Scandinavian) documenting the spatial evolution of these interconnected ice sheets every 1000 years from 25 to 10 ka, and at four selected time periods back to 40 ka. The time-slice maps of ice-sheet extent are based on a new Geographical Information System (GIS) database, where we have collected published numerical dates constraining the timing of ice-sheet advance and retreat, and additionally geomorphological and geological evidence contained within the existing literature. We integrate all uncertainty estimates into three ice-margin lines for each time-slice; a most-credible line, derived from our assessment of all available evidence, with bounding maximum and minimum limits allowed by existing data. This approach was motivated by the demands of glaciological, isostatic and climate modelling and to clearly display limitations in knowledge. The timing of advance and retreat were both remarkably spatially variable across the ice-sheet area. According to our compilation the westernmost limit along the British–Irish and Norwegian continental shelf was reached up to 7000 years earlier (at c. 27–26 ka) than the eastern limit on the Russian Plain (at c. 20–19 ka). The Eurasian ice sheet complex as a whole attained its maximum extent (5.5 Mkm2) and volume (~24 m Sea Level Equivalent) at c. 21 ka. Our continental-scale approach highlights instances of conflicting evidence and gaps in the ice-sheet chronology where uncertainties remain large and should be a focus for future research. Largest uncertainties coincide with locations presently below sea level and where contradicting evidence exists. This first version of the database and time-slices (DATED-1) has a census date of 1 January 2013 and both are available to download via the Bjerknes Climate Data Centre and PANGAEA (www.bcdc.no; http://doi.pangaea.de/10.1594/PANGAEA.848117).

The last Eurasian ice sheets - a chronological database and time-slice reconstruction, DATED-1

A review of tufa and travertine deposits of the world

Late Quaternary dynamics of southern Africa's winter rainfall zone

Global maps of lake-level fluctuations since 30,000 yr B.P.

Chronological data for glacier advances in the European Alps between the Last Interglacial (Eemian) and the Holocene are summarised (115 to 11 ka). During this time glaciers were most extensive, extending tens of kilometres out onto the forelands, between 30 and 18 ka, that is, synchronous with the global ice volume maximum of Marine Isotope Stage (MIS) 2. Evidence for ice expanding to just past the mountain front for an earlier major glacier advance comes from Swiss sites, where advances have been luminescence dated to MIS 5d (100 ka) and MIS 4 (70 ka). Up to now no such evidence has been found in the Eastern Alps. By 18 ka, more than 80% of the Late Wurmian ice volume had gone. Subsequently glaciers readvanced, reaching into the upper reaches of the main valleys during the Lateglacial Gschnitz stadial, which likely occurred around 17 ka, with final moraine stabilisation no later than 15.4 ka. The link of the Egesen stadial with the Younger Dryas climate deterioration is supported by exposure ages from four sites as well as minimum-limiting radiocarbon dates from bogs within former glacier tongue areas. Key questions on the spatial and temporal variability of ice extents throughout the last glacial cycle have yet to be answered. Copyright # 2008 John Wiley & Sons, Ltd.

https://uibk.ac.at/geographie/personal/kerschner/ivyjqs2008.pdf

Chronology of the last glacial cycle in the

Chronological data for glacier advances in the European Alps between the Last Interglacial (Eemian) and the Holocene are summarised (115 to 11 ka). During this time glaciers were most extensive, extending tens of kilometres out onto the forelands, between 30 and 18 ka, that is, synchronous with the global ice volume maximum of Marine Isotope Stage (MIS) 2. Evidence for ice expanding to just past the mountain front for an earlier major glacier advance comes from Swiss sites, where advances have been luminescence dated to MIS 5d (100 ka) and MIS 4 (70 ka). Up to now no such evidence has been found in the Eastern Alps. By 18 ka, more than 80% of the Late Wurmian ice volume had gone. Subsequently glaciers readvanced, reaching into the upper reaches of the main valleys during the Lateglacial Gschnitz stadial, which likely occurred around 17 ka, with final moraine stabilisation no later than 15.4 ka. The link of the Egesen stadial with the Younger Dryas climate deterioration is supported by exposure ages from four sites as well as minimum-limiting radiocarbon dates from bogs within former glacier tongue areas. Key questions on the spatial and temporal variability of ice extents throughout the last glacial cycle have yet to be answered. Copyright © 2008 John Wiley & Sons, Ltd.

https://www.uibk.ac.at/geographie/personal/kerschner/ivyjqs2008.pdf

Chronology of the last glacial cycle in the European Alps

The glacial chronology of Mexico is based on glacial landforms of Iztaccihuatl (5,282 m a.s.l., 19°10′N, 98°40′W), Nevado de Toluca (4,558 m a.s.l., 19°08′N, 99°45′W), and La Malinche (4,461 m, 19°14′N, 98°00′W). Morphostratigra-phy, tephrochronology and 81 cosmogenic 36Cl exposure ages from Iztaccihuatl reveal new insights into the glacial sequence in Mexico.

The most extensive recorded advance (Nexcoalango) occurred during Marine Isotope Stage 6 (MIS 6), probably at 195,000 BP and reached c. 3000 m. Iztaccihuatl lacks clear evidence of Wisconsinan glaciation prior to 20,000 cal yr BP. The local Late Pleistocene glacial maximum occurred after the climax of the Last Glacial Maximum (LGM). A first pulse (Hueyatlaco 1 advance) peaked at 20,000–17,500 cal yr BP and a second one (Hueyatlaco 2) at c. 17,000–14,000 cal yr BP. Valley glaciers reached to c. 3400–3500 m a.s.l.

Recessional moraines developed from 14,000 to 13,000 cal yr BP, followed by a rapid glacier retreat at 13,000–12,000 cal yr BP. Subsequently, glaciers peaked again at c. 12,000 cal yr BP reaching c. 3800 m (Milpulco 1), and built recessional moraines until c. 10,000 cal yr BP. Then between c. 8300–7000 cal yr BP glaciers formed small but distinctive moraines above 4000 m a.s.l. (Milpulco 2). No evidence of glacier expansion has been found between Milpulco 2 deposits and the massive moraines of <1000 cal yr BP (Ayoloco) that occur at 4300–4700 m a.s.l.

Possible correlations between the glacial record of Iztaccihuatl and other glacial sequences (La Malinche, Nevado de Toluca) are presented. The equilibrium line altitudes (ELA) of glaciers reveal ELA depressions for the five late Quaternary advances of 1030 m, 930 m, 730 m, 550 m and 250 m, respectively. The overall pattern of glaciation is similar to those of mid-latitude North America and tropical South America, thus supporting the general synchroneity of major climatic events.

A temperature decrease of 5–9°C estimated from Hueyatlaco ELA supports marked cooling over tropical land and oceans during the LGM. Although inconclusive, evidence does not indicate glacier expansion during, but rather immediately following the Younger Dryas Chronozone. However, a global early-to-mid-Holocene event (8200 cal yr event) is coeval with Milpulco 2 advance, and the Ayoloco advance is contemporaneous with the Little Ice Age.

Late Quaternary glaciation of México

The variation of δ18O that results from nearly all physical, biological and chemical processes on the Earth is approximately twice as large as the variation of δ17O. This so-called ‘mass-dependent’ fractionation is well documented in terrestrial minerals1,2. Evidence for ‘mass-independent’ fractionation (Δ17O = δ17O - 0.52δ18O), where deviation from this tight relationship occurs, has so far been found only in meteoritic material and a few terrestrial atmospheric substances3. In the rock record it is thought that oxygen isotopes have followed a mass-dependent relationship for at least the past 3.7 billion years (ref. 4), and no exception to this has been encountered for terrestrial solids5. Here, however, we report oxygen-isotope values of two massive sulphate mineral deposits, which formed in surface environments on the Earth but show large isotopic anomalies (Δ17O up to 4.6‰). These massive sulphate deposits are gypcretes from the central Namib Desert and the sulphate-bearing Miocene volcanic ash-beds in North America. The source of this isotope anomaly might be related to sulphur oxidation reactions in the atmosphere and therefore enable tracing of such oxidation. These findings also support the possibility of a chemical origin of variable isotope anomalies on other planets, such as Mars6.

Anomalous 17O compositions in massive sulphate deposits on the Earth

The palaeogeographical and palaeoclimatic development of the Kalahari region provides strong evidence that South African climatic changes in precipitation and temperature, as well as in surface winds are closely related to variations in intensity of the general circulation rather than in shifts of the climatic belts. According to the great diversity of palaeoclimatic evidence, the palaeoclimatic reconstruction is incomplete. Further finds of datable material will supplement this record and possibly add to the complexity of the evidence. The Middle Wisconsin interstadial period is characterized by pluvial conditions in the Kalahari (>30,000-ca.19,000 yr BP). Last glacial maximum climatic conditions prevail between ca. 19,000 and 13,000 yr BP. A large increase in meridional and, over southern Africa, in W-E temperature gradients cause a change in the frequency of occurrence and in'velocity of winds. The last glacial maximum was more arid in South West Africa than today, but more humid in the southern Kalahari than today. It is not quite clear whether the phase of greater glacial humidity in the Kalahari is restricted to only a short time span, probably between ca. 17,000 and 15,000 yrBP. The period between ca.13,000 and 10,000 yr BP presents the transition from glacial climatic conditions to the post-glacial climatic optimum. During the Holocene temperature maximum around 9,500 yr BP geomorphic and palaeopedologic features indicate wetter conditions in the Kalahari. The results show that different palaeoclimatic indicators (geomorphic, sedimentological, palaeobotanical, palaeontological, archaeological, etc.) yield different time-stratigraphic periods for the Late Quaternary. That may be the reason, apart from the fact that many deposits have alternative explanations, that a great variety of palaeoclimatic and palaeogeographic reconstructions exist for southern Africa. I N T R O D U C T I O N Late Quaternary geomorphological and climatic reconstructions for the Kalahari region have, until now, depended largely upon the implications of evidence derived from areas outside the Kalahari, especially from the Cape Province, the Orange Free State, and Transvaal, and from East Africa (Figure 1). This evidence has been reviewed by Van Zinderen Bakker (1976) and Butzer et al. (1978). Butzer et al. (1978) come to the conclusion, that the palaeoclimatic record of much of the Kalahari region is long, complex informative, and distinctive; this broad climatic province responded differently to global circulation anomalies than did other African areas. Inter-regional comparisons show both similarities and deviations that should serve as a basis for a creative palaeoclimatic model. In glacial times the Benguela Current, which largely controls the arid climate of southwestern Africa, had a greater vigour and an even more aridifying effect upon the coastal regions (Namib desert) than it had in interglacial times (Gardner & Hays 1976, Selby, Hendy & Seely 1979). In the southern Kalahari moist intervals (glacial times) were primarily, although not exclusively, a response to variations in the penetration and persistance of summer, rather than winter rains (Butzer et al. 1978, Lancaster 1979). Furthermore, the Late Quaternary lake-level fluctuations of the Makgadikgadi basin indicate a huge palaeo-lake from about 30,000 to 19,000 yr BP and about 12,000 yr BP and a low lake-level (the lake dried out) between 19,000 and 12,000 yr BP (Heine 1978a, b). In this paper I present some conclusions from geomorphogical studies in the Kalahari region. I provide evidence that, although there are many contradictions concerning the interpretation of the geomorphological observations and facts, it is now possible to reconstruct the main phases of the Late Quaternary evolution of the Kalahari region. T H E K A L A H A R I R E G I O N The Kalahari extends from the Okavango river in the north to close to the Orange river in the Republic of South Africa in the south. The Kalahari is situated on a plateau generally more than 1,000 m above sea level in Botswana, eastern South West Africa/Namibia, the northern Cape Province, and northwestern Transvaal. It is flat or gently undulating, with sand dunes more frequently occurring in the southwest. The true Kalahari is a huge sand-filled basin (De Vos 1975). The zone is semiarid. Rainfall is erratic, confined mainly to the period November to April and decreases from about 500 mm annually in the north and east to about 200 mm in the southwest. From 22° latitude southward the zone is covered with dry grassland and savanna, receiving only little and unreliable rainfall (De Vos 1975, Diem 1977). The summer climate is dry and continental. Scarcity of surface water and poor soils are the main limiting factors to vegetative growth (Cole & Brown 1976, Werger 1978). Ground waters in the northern Kalahari are directly, and in some cases rapidly, recharged by rain (Verhagen, Mazor & Sellschop 1974). Surface water is available in only a few areas for a short time after the rains, when it collects in pans or shallow depressions. These pans play an important role in the ecology of the area (De Vos 1975). With the exception of the Botletle river, which takes off from the Thamalakane river at the margin of the Okavango swamps and flows east and then south into the Makgadikgadi basin, there are no rivers between the swamp zone in the north (Okavango swamps of Ngamiland, Chobe and Linyanti swamps and Zambezi swamps in Barotseland) and the Orange river in the south* In Botswana the Kalahari forms a broad watershed between the Okwa and Mmone dry river systems (BakalahariSchwelle), which are directed towards the Makgadikgadi depression (Okwa), and the Nossob and Molopo, both of them dry rivers, which form the southern border of Botswana. Here the principal geomorphic features are the thousand or so small pans which break the monotony of this otherwise almost featureless sand plain (Lancaster 1978). Detailed descriptions of the physiography of the Kalahari are available elsewhere (Passarge 1904, Wellington 1955, Grove 1969, Grey & Cooke 1977, Wright 1978). T H E M I D D L E W I S C O N S I N ( M I D D L E W E I C H S E L I A N ) I N T E R S T A D I A L P E R I O D The reconstruction of the palaeogeography and the palaeoclimatology of the Kalahari is based on radiocarbon dating (Van Zinderen Bakker 1976, Grey & Cooke 1977, Heine 1978a, Lancaster 1979); therefore the critical question is to what degree radiocarbon determinations are of real chronometric value. I will not discuss this problem here (Rafter 1975). From the Kalahari region only a few stratigraphically consistent 1 4 C dates are known. Cooke (1975), Grey & Cooke (1977) and Cooke & Verhagen (1977) report on 1 4 C dated sinter from Drotsky's cave in western Ngamiland (Botswana) and calcretes from the adjacent area; the dates range from 22,700 ± 500 yr BP to > 45,000 yr BP for the old calcrete CI and CII and from 29,300 ± 1,000 yr BP to > 45,000 yr BP for cave sinter SII; the ages are likely to be infinite and should be regarded with appropriate caution. According to Cooke (1975) and Grey & Cooke (1977) a major wet phase occurred in western Ngamiland before the SHI sinter deposition which is dated 16,000-13,000 yr BP. During this major wet phase a high water-table flooded the old cave passages, resulting in extensive enlargement and the concomitant removal of much of the SII sinter. This major wet phase prior to about 16,000 ± 200 yr BP (corrected age, Cooke & Verhagen 1977) or 17,300 ± 200 yr BP (conventional age, Cooke 1975) can probably be correlated with a major highstand of Lake PalaeoMakgadikgadi, which is dated between 19,170 ± 660 and > 30,250 ± 520 yr BP (Heine 1978a). The distribution of lacustrine deposits and molluscs in the Ngami/Makgadikgadi area indicate quite clearly that the 30,000-19,000 yr BP high lake-level belongs to the most extensive lake which existed during the Late Quaternary (ca.40,000-0 yr BP) and which might correspond to the 945 m level mentioned by Grey & Cooke (1977). The evidence for greater humidity during the period 30,000-19,000 yr BP is not confined to the Makgadikgadi basin. In the southwest Kalahari between the Auob and Nossob dry valleys (25°56'S, 20°25'30"E), a fossil soil buried under red sand dunes and a calcrete layer, was found (Figure 3). The l 4 C age of this soil of 2 8 , 0 0 0 + 3 $ $ y r B P ( H v 9 5 0 2 ) i s methodically reliable (personal communication Prof. M.A.Geyh, Hannover). The fossil A-horizon that is rich in organic matter indicates wet conditions during the period of soil formation. Furthermore, stromatolite-like concretions of the surface of Klein Awas Pan (26°34'S, 20°28'30"E) yield a radiocarbon age of 23,410 + §§§ yr BP (Hv 9885), thus documenting more humid conditions in the southwest Kalahari. In the northwest Kalahari, the age of a mollusc-rich tufa is 22,250 ± 330 yr BP (Hv 9883); the tufa is situated at the southern shore of the Etosha Pan, northwest of Homob. Although there is no evidence for a fossil lake shore, this dated tufa layer represents a higher water-level of the Etosha Pan. At the Gaap Escarpment (southern Kalahari margin, South Africa) a chronostratigraphy for the last 30,000 yr BP is provided by 1 4 C dating (Butzer et al. 1978); the last Pleistocene cold-moist interval began after 35,000 yr BP and ended 14,000 yr BP. About 150 km southwest of the Gaap Plateau lacustrine chalk of a large pluvial lake was dated at 24,890 ± 695 yr BP (Hv 9504); this means an additional indication for wetter conditions compared with today. Figure 2. Cross-section of Terrace Bay beach and nearshore region. According to l 4 C dating of molluscs, the youngest fluvial sediments of the Molopo river near Koopan Suid (27°14'S, 20°22'30"E) were deposited about 19,085 + 9g5 yr BP (Hv 9495). After the deposition of the mollusc bearing fluvial sands, white sand dunes were formed in the dry river valley (Figure 3), thus indi

Klaus Heine

Papers

Chronology of the last glacial cycle in the

Chronology of the last glacial cycle in the European Alps

Late Quaternary glaciation of México

Anomalous 17O compositions in massive sulphate deposits on the Earth

The Main Stages of the Late Quaternary Evolution of the Kalahari Region, Southern Africa