- Email: [email protected]
The use of kernel density estimates to assist in the identification of volcanic tephra
70
Abstracts / Quaternary International 279-280 (2012) 9–120
Megachad, which is now completely desiccated leaving the lake bed sediments exposed to deflation. The deflation is attributed to a combination of strong near surface winds and a large reservoir of readily eroded, and easily entrained, low density sediment. Erosional remnants exposed as yardangs on the former lake floor indicate that locally up to 4 m of sediment has been removed by deflation. Dating of geomorphological features indicate that in some areas rates of deflation may be greater than or equal to 40 mmyr-1. Deflation appears to be greatest in an arc around the northeastern margins of the lake basin which is probably due to a combination of high wind velocities, an upwind source of abrasive sand, and thinner deposits around the lake margin. Satellite image analysis shows that outcropping diatomite covers 24,049km2 of the Bodélé Depression. Furthermore, the area from which diatomite has been removed by deflation is around 14,000 km2 . Combining this information with the ages and deflation rates mentioned above indicates that some 61,000 km3 of diatomite has been eroded from the Bodélé Depression during the past 1,000 years. The dust is blown across the Atlantic Ocean towards South America. The mineral dust contains micronutrients such as Fe and P that have the potential to act as a fertilizer, increasing primary productivity in the Amazon rain forest as well as the equatorial Atlantic Ocean. Chemical analysis of 28 dust samples collected from the source area, which indicates that up to 6.5 Tg of Fe and 0.12 Tg of P are exported from the Bodélé Depression every year. This suggests that the Bodélé may be a more significant micronutrient supplier than previously proposed.
chronologies for this period. Equally important, however, is the part they can play in helping us to understand aspects of the earth's system. Most of the existing radiocarbon archives for this period, prior to the continuous tree-ring records of Northern Europe and North America, are measured on samples that have reservoir offsets relative to the atmosphere. In the case of marine samples, such as coral and foraminifera, the difference between the atmosphere and the marine environment depends on ocean circulation and rates of exchange in surface waters. Speleothems, which provide another important archive, have ground-water reservoir effects which might (or might not) vary with local changes in climate. These archives have provided an important proxy for atmospheric variation in radiocarbon and give us the ability to calibrate radiocarbon over this age range, albeit with some inherent uncertainties. Lake Suigetsu with its new varve-based chronology, provides an independent and purely terrestrial radiocarbon record (SG06: see Staff et al this session). This enables us to test some of the assumptions underlying the use of non-terrestrial radiocarbon records for calibration and, perhaps equally importantly, enables us to use the combination of terrestrial, marine and other records to tell us about changes in carbon cycling. This presentation will look at the implications of the new radiocarbon dataset for both chronological and climatic research.
INVESTIGATING THE TIMESCALES OF GLACIAL LANDSCAPE EVOLUTION USING NUMERICAL MODELLING
Christopher Bronk Ramsey. University of Oxford, United Kingdom
Simon Brocklehurst. University of Manchester, United Kingdom
Identification of tephra through major and trace element data is normally done through the visual comparison of measurements on unknown tephra to those on identified eruptions. This visual inspection can be supplemented by statistical analysis using methods such as principle components analysis and discriminant analysis. Because the diagnostic features of different tephra are different for each case, this process tends to be timeconsuming and is, by its nature, not quantitative. This poster will explore the prospects for using multi-dimensional kernel density estimates to evaluate the likelihood of tephra correlations. The poster will describe an on-line tool which has been developed, that enables large numbers of possible source eruptions to be tested rapidly, with quantitative information on the possible identifications. Such statistical methods will not replace detailed consideration of the volcanic systems in question, but they may provide a useful tool in narrowing down possible correlations and help to give better criteria for positive matches.
E-mail address: [email protected]
Glacial landscapes reflect a long history of Quaternary climate cycles, but geological evidence for the timescales over which distinctive glacial landforms developed remains largely elusive. Some stratigraphic evidence indicates that the onset of Northern Hemisphere glaciation immediately prior to the Quaternary was responsible for considerable sediment production. However, recent chronological studies have suggested that, in certain settings, the Middle Pleistocene Transition (MPT) from 40- to 100kyr glacial cycles caused a significant change in glacial erosion and sediment production. We use a numerical model of glacial longitudinal profile evolution to explore the timescales for modifying previously fluvial topography, and the conditions under which the MPT would have been more important for glacial landscape evolution than the initial glaciation. The model is driven by simplified temperature fluctuations; smaller amplitude, symmetrical 40-kyr cycles at the start of the Quaternary, and larger amplitude, asymmetric 100-kyr cycles post-MPT. Tectonic rock uplift is imposed, alongside the isostatic response to erosion. We demonstrate that the landscape response to the MPT is heavily dependent on the tectonic setting and the behaviour of the fluvial system downstream of the glacier. With no imposed tectonic rock uplift, the major change in the landscape is the initial cirque development; the MPT would have had little impact on this morphology. However, with rapid tectonic uplift, and inefficient fluvial transport downstream of the glacier, a broad, relatively high region develops downstream of the maximum pre-MPT advance. Hence post-MPT glaciers advance further, glacial erosion is more substantial, and sediment production increases dramatically. This complex interaction motivates further study of the interplay between fluvial and glacial erosion, and improved records of landscape modification and sediment production in alpine glacial settings during the Quaternary. ASSESSING THE IMPORTANCE OF TERRESTRIAL RADIOCARBON ARCHIVES FOR THE GLACIAL PERIOD: IMPLICATIONS OF THE SUIGETSU (SG06) RADIOCARBON RECORD Christopher Bronk Ramsey. University of Oxford, United Kingdom E-mail address: [email protected]
Suigetsu 2006 Project Members Long-term radiocarbon records contribute in two main ways to our understanding of the Quaternary. Most obviously they provide us with the necessary data for the calibration of radiocarbon dates to provide
THE USE OF KERNEL DENSITY ESTIMATES TO ASSIST IN THE IDENTIFICATION OF VOLCANIC TEPHRA
E-mail address: [email protected]
THE CHANGING GEO-ECOLOGY OF THE DIJLE FLOODPLAIN (BELGIUM) DURING THE HOLOCENE IN RELATION TO HUMAN IMPACT Nils Broothaerts. K.U. Leuven, Belgium E-mail address: [email protected]
Floodplain deposition rates have increased intensively under influence of anthropogenic land use throughout the late Holocene in many West- and Central European catchments. These variations in sedimentation rates have also changed the geomorphology and ecology of many floodplains as is shown here for the River Dijle in the Belgian Loess Belt. In the main valley of the 760 km2 catchment, a transition from organic accumulation and peat formation to deposition of detritic material can be observed. For one valley cross-section (1020 m wide), a detailed reconstruction of the fluvial architecture was made based on 57 hand augerings and 22 AMS radiocarbon dates. Peat growth starts at around 9-10 ka cal BP and reaches a thickness between 1 and 3 m. This unit was deposited in a marshy environment where no clear river channel is present. The peat growth ceased and was replaced by a unit dominated with silty to clayey overbank deposits (1-4 m thick), which contains at the lower part still a higher organic content. The transition between the peaty layer and the upper layer is rather abrupt but not erosive. The top of the peat layer is diachronic, ranging between 2500 and 950 cal BP. It is suggested that the diachronic transition between the two stratigraphic units is due to the gradual built-up of levees that laterally expended over the backswamp areas when the sediment input increased due to the increasing intensity of cultivation in the catchment. After around 1000 AD, the entire valley is dominated by
Abstracts / Quaternary International 279-280 (2012) 9–120
Megachad, which is now completely desiccated leaving the lake bed sediments exposed to deflation. The deflation is attributed to a combination of strong near surface winds and a large reservoir of readily eroded, and easily entrained, low density sediment. Erosional remnants exposed as yardangs on the former lake floor indicate that locally up to 4 m of sediment has been removed by deflation. Dating of geomorphological features indicate that in some areas rates of deflation may be greater than or equal to 40 mmyr-1. Deflation appears to be greatest in an arc around the northeastern margins of the lake basin which is probably due to a combination of high wind velocities, an upwind source of abrasive sand, and thinner deposits around the lake margin. Satellite image analysis shows that outcropping diatomite covers 24,049km2 of the Bodélé Depression. Furthermore, the area from which diatomite has been removed by deflation is around 14,000 km2 . Combining this information with the ages and deflation rates mentioned above indicates that some 61,000 km3 of diatomite has been eroded from the Bodélé Depression during the past 1,000 years. The dust is blown across the Atlantic Ocean towards South America. The mineral dust contains micronutrients such as Fe and P that have the potential to act as a fertilizer, increasing primary productivity in the Amazon rain forest as well as the equatorial Atlantic Ocean. Chemical analysis of 28 dust samples collected from the source area, which indicates that up to 6.5 Tg of Fe and 0.12 Tg of P are exported from the Bodélé Depression every year. This suggests that the Bodélé may be a more significant micronutrient supplier than previously proposed.
chronologies for this period. Equally important, however, is the part they can play in helping us to understand aspects of the earth's system. Most of the existing radiocarbon archives for this period, prior to the continuous tree-ring records of Northern Europe and North America, are measured on samples that have reservoir offsets relative to the atmosphere. In the case of marine samples, such as coral and foraminifera, the difference between the atmosphere and the marine environment depends on ocean circulation and rates of exchange in surface waters. Speleothems, which provide another important archive, have ground-water reservoir effects which might (or might not) vary with local changes in climate. These archives have provided an important proxy for atmospheric variation in radiocarbon and give us the ability to calibrate radiocarbon over this age range, albeit with some inherent uncertainties. Lake Suigetsu with its new varve-based chronology, provides an independent and purely terrestrial radiocarbon record (SG06: see Staff et al this session). This enables us to test some of the assumptions underlying the use of non-terrestrial radiocarbon records for calibration and, perhaps equally importantly, enables us to use the combination of terrestrial, marine and other records to tell us about changes in carbon cycling. This presentation will look at the implications of the new radiocarbon dataset for both chronological and climatic research.
INVESTIGATING THE TIMESCALES OF GLACIAL LANDSCAPE EVOLUTION USING NUMERICAL MODELLING
Christopher Bronk Ramsey. University of Oxford, United Kingdom
Simon Brocklehurst. University of Manchester, United Kingdom
Identification of tephra through major and trace element data is normally done through the visual comparison of measurements on unknown tephra to those on identified eruptions. This visual inspection can be supplemented by statistical analysis using methods such as principle components analysis and discriminant analysis. Because the diagnostic features of different tephra are different for each case, this process tends to be timeconsuming and is, by its nature, not quantitative. This poster will explore the prospects for using multi-dimensional kernel density estimates to evaluate the likelihood of tephra correlations. The poster will describe an on-line tool which has been developed, that enables large numbers of possible source eruptions to be tested rapidly, with quantitative information on the possible identifications. Such statistical methods will not replace detailed consideration of the volcanic systems in question, but they may provide a useful tool in narrowing down possible correlations and help to give better criteria for positive matches.
E-mail address: [email protected]
Glacial landscapes reflect a long history of Quaternary climate cycles, but geological evidence for the timescales over which distinctive glacial landforms developed remains largely elusive. Some stratigraphic evidence indicates that the onset of Northern Hemisphere glaciation immediately prior to the Quaternary was responsible for considerable sediment production. However, recent chronological studies have suggested that, in certain settings, the Middle Pleistocene Transition (MPT) from 40- to 100kyr glacial cycles caused a significant change in glacial erosion and sediment production. We use a numerical model of glacial longitudinal profile evolution to explore the timescales for modifying previously fluvial topography, and the conditions under which the MPT would have been more important for glacial landscape evolution than the initial glaciation. The model is driven by simplified temperature fluctuations; smaller amplitude, symmetrical 40-kyr cycles at the start of the Quaternary, and larger amplitude, asymmetric 100-kyr cycles post-MPT. Tectonic rock uplift is imposed, alongside the isostatic response to erosion. We demonstrate that the landscape response to the MPT is heavily dependent on the tectonic setting and the behaviour of the fluvial system downstream of the glacier. With no imposed tectonic rock uplift, the major change in the landscape is the initial cirque development; the MPT would have had little impact on this morphology. However, with rapid tectonic uplift, and inefficient fluvial transport downstream of the glacier, a broad, relatively high region develops downstream of the maximum pre-MPT advance. Hence post-MPT glaciers advance further, glacial erosion is more substantial, and sediment production increases dramatically. This complex interaction motivates further study of the interplay between fluvial and glacial erosion, and improved records of landscape modification and sediment production in alpine glacial settings during the Quaternary. ASSESSING THE IMPORTANCE OF TERRESTRIAL RADIOCARBON ARCHIVES FOR THE GLACIAL PERIOD: IMPLICATIONS OF THE SUIGETSU (SG06) RADIOCARBON RECORD Christopher Bronk Ramsey. University of Oxford, United Kingdom E-mail address: [email protected]
Suigetsu 2006 Project Members Long-term radiocarbon records contribute in two main ways to our understanding of the Quaternary. Most obviously they provide us with the necessary data for the calibration of radiocarbon dates to provide
THE USE OF KERNEL DENSITY ESTIMATES TO ASSIST IN THE IDENTIFICATION OF VOLCANIC TEPHRA
E-mail address: [email protected]
THE CHANGING GEO-ECOLOGY OF THE DIJLE FLOODPLAIN (BELGIUM) DURING THE HOLOCENE IN RELATION TO HUMAN IMPACT Nils Broothaerts. K.U. Leuven, Belgium E-mail address: [email protected]
Floodplain deposition rates have increased intensively under influence of anthropogenic land use throughout the late Holocene in many West- and Central European catchments. These variations in sedimentation rates have also changed the geomorphology and ecology of many floodplains as is shown here for the River Dijle in the Belgian Loess Belt. In the main valley of the 760 km2 catchment, a transition from organic accumulation and peat formation to deposition of detritic material can be observed. For one valley cross-section (1020 m wide), a detailed reconstruction of the fluvial architecture was made based on 57 hand augerings and 22 AMS radiocarbon dates. Peat growth starts at around 9-10 ka cal BP and reaches a thickness between 1 and 3 m. This unit was deposited in a marshy environment where no clear river channel is present. The peat growth ceased and was replaced by a unit dominated with silty to clayey overbank deposits (1-4 m thick), which contains at the lower part still a higher organic content. The transition between the peaty layer and the upper layer is rather abrupt but not erosive. The top of the peat layer is diachronic, ranging between 2500 and 950 cal BP. It is suggested that the diachronic transition between the two stratigraphic units is due to the gradual built-up of levees that laterally expended over the backswamp areas when the sediment input increased due to the increasing intensity of cultivation in the catchment. After around 1000 AD, the entire valley is dominated by