Paleoceanography on sediment core MD01-2416

IMAGES core MD01-2416 (51°N, 168°E) provides the first centennial-scale multiproxy record of Holocene variation in North Pacific sea-surface temperature (SST), salinity, and biogenic productivity. Our results reveal a gradual decrease in subarctic SST by 3-5 °C from 11.1 to 4.2 ka and a stepwise long-term decrease in sea surface salinity (SSS) by 2-3 p.s.u. Early Holocene SSS were as high as in the modern subtropical Pacific. The steep halocline and stratification that is characteristic of the present-day subarctic North Pacific surface ocean is a fairly recent feature, developed as a product of mid-Holocene environmental change. High SSS matched a salient productivity maximum of biogenic opal during Bølling-to-Early Holocene times, reaching levels similar to those observed during preglacial times in the warm mid-Pliocene prior to 2.73 Ma. Similar productivity spikes marked every preceding glacial termination of the last 800 ka, indicating recurrent short-term events of mid-Pliocene-style intense upwelling of nutrient-rich Pacific Deepwater in the Pleistocene. Such events led to a repeated exposure of CO2-rich deepwater at the ocean surface facilitating a transient CO2 release to the atmosphere, but the timing and duration of these events repudiate a long-term influence of the subarctic North Pacific on global atmospheric CO2 concentration.

Benthic foraminiferal assemblages and epifaunal isotope record of the Agulhas Plateau

Benthic foraminiferal assemblages and the carbon isotope composition of the epifaunal benthic foraminifera Epistominella exigua and Fontbotia wuellerstorfi have been investigated along core MD02-2589 located at the southern Agulhas Plateau (41°26.03'S, 25°15.30'E, 2660 m water depth). This study aims to evaluate changes in the benthic paleoenvironment and its influence on benthic d13C with a notable focus on E. exigua, a species associated with phytodetritus deposits and poorly studied in isotope paleoceanographic reconstructions. The benthic foraminiferal assemblages (>63 µm) show large fluctuations in species composition suggesting significant changes in the pattern of ocean surface productivity conceivably related to migrations of the Subtropical Convergence (STC) and Subantarctic Front (SAF). Low to moderate seasonality and relatively higher food supply to the seafloor are indicated during glacial marine isotope stages (MIS) 6, 4, and 2 and during MIS 3, probably associated with the northward migration of the SAF and confluence with the more stationary STC above the southern flank of the Agulhas Plateau. The lowest organic carbon supply to the seafloor is indicated from late MIS 5b to MIS 4 as a consequence of increased influence of the Agulhas Front (AF) and/or weakening of the influence of the STC over the region. Episodic delivery of fresh organic matter, similar to modern conditions at the core location, is indicated during MIS 5c-MIS 5e and at Termination I. Comparison of this paleoenvironmental information with the paired d13C records of E. exigua and F. wuellerstorfi suggests that organic carbon offsets d13C of E. exigua from ambient bottom water d13CDIC, while its d13C amplitude, on glacial-interglacial timescales, does not seem affected by changes of organic carbon supply to the seafloor. This suggests that this species calcifies preferentially during the short time span of the year when productivity peaks and phytodetritus is delivered to the seafloor. Therefore E. exigua, while offset from d13CDIC, potentially more faithfully records the amplitude of ambient bottom water d13CDIC changes than F. wuellerstorfi, notably in settings such as the Southern Ocean that experienced substantial changes through time in the organic carbon supply to the seafloor.

Planktonic foraminiferal faunas and stable isotope ratios of sediments from the Benguela Upwelling System

Two sediment cores retrieved from the continental slope in the Benguela Upwelling System, GeoB 1706 (19°33.7'S 11°10.5'E) and GeoB 1711 (23°18.9'S, 12°22.6'E), reveal striking variations in planktonic foraminiferal abundances during the last 160,000 years. These fluctuations are investigated to assess changes in the intensity and position of the upwelling centres off Namibia. Four species make up over 95% of the variation within the core, and enable the record to be divided into episodes characterized by particular planktonic foraminiferal assemblages. The fossil assemblages have meaningful ecological significance when compared to those of the modern day and the relationship to their environment. The cold-water planktonic foraminifer, Neogloboquadrina pachyderma sinistral [N. pachyderma (s)], dominates the modern-day, coastal upwelling centres, and Neogloboquadrina pachyderma dextral and Globigerina bulloides characterize the fringes of the upwelling cells. Globorotalia inflata is representative of the offshore boundary between newly upwelled waters and the transitional, reduced nutrient levels of the subtropical waters. In the fossil record, episodes of high N. pachyderma (s) abundances are interpreted as evidence of increased upwelling intensity, and the associated increase in nutrients. The N. pachyderma (s) record suggests temporal shifts in the intensity of upwelling, and corresponding trophic domains, that do not follow the typical glacial-interglacial pattern. Periods of high N. pachyderma (s) abundance describe rapid, discrete events dominating isotope stages 3 and 2. The timing of these events correlates to the temporal shifts of the Angola-Benguela Front (Jansen et al., 1997) situated to the north of the Walvis Ridge. Absence of high abundances of N. pachyderma (s) from the continental slope of the southern Cape Basin indicates that Southern Ocean surface water advection has not exerted a major influence on the Benguela Current System. The coincidence of increased upwelling intensity with the movement of the Angola-Benguela Front can be interpreted mainly by changes in strength and zonality of the trade wind system.

Stable isotope ratios, carbon content and primary production of three sediment cores from the subantarctic eastern Atlantic

We present three new benthic foraminiferal delta13C, delta18O, and total organic carbon time series from the eastern Atlantic sector of the Southern Ocean between 41°S and 47°S. The measured glacial delta13C values belong to the lowest hitherto reported. We demonstrate a coincidence between depleted late Holocene (LH) delta13C values and positions of sites relative to ocean surface productivity. A correction of +0.3 to +0.4 [per mil VPDB] for a productivity-induced depletion of Last Glacial Maximum (LGM) benthic delta13C values of these cores is suggested. The new data are compiled with published data from 13 sediment cores from the eastern Atlantic Ocean between 19°S and 47°S, and the regional deep and bottom water circulation is reconstructed for LH (4-0 ka) and LGM (22-16 ka) times. This extends earlier eastern Atlantic-wide synoptic reconstructions which suffered from the lack of data south of 20°S. A conceptual model of LGM deep-water circulation is discussed that, after correction of southernmost cores below the Antarctic Circumpolar Current (ACC) for a productivity-induced artifact, suggests a reduced formation of both North Atlantic Deep Water in the northern Atlantic and bottom water in the southwestern Weddell Sea. This reduction was compensated for by the formation of deep water in the zone of extended winter sea-ice coverage at the northern rim of the Weddell Sea, where air-sea gas exchange was reduced. This shift from LGM deep-water formation in the region south of the ACC to Holocene bottom water formation in the southwestern Weddell Sea, can explain lower preformed d13CDIC values of glacial circumantarctic deep water of approximately 0.3 per mil to 0.4 per mil. Our reconstruction brings Atlantic and Southern Ocean d13C and Cd/Ca data into better agreement, but is in conflict, however, with a scenario of an essentially unchanged thermohaline deep circulation on a global scale. Benthic delta18O-derived LGM bottom water temperatures, by 1.9°C and 0.3°C lower than during the LH at deepest southern and shallowest northern sites, respectively, agree with the here proposed reconstruction of deep-water circulation in the eastern South Atlantic Ocean.

Thorium-234, POC and PON fluxes during POLARSTERN cruise ANT-XXVIII/3, January-March 2012

Carbon fixation by phytoplankton plays a key role in the uptake of atmospheric CO2 in the Southern Ocean. Yet, it still remains unclear how efficiently the particulate organic carbon (POC) is exported and transferred from ocean surface waters to depth during phytoplankton blooms. In addition, little is known about the processes that control the flux attenuation within the upper twilight zone. Here, we present results of downward POC and particulate organic nitrogen fluxes during the decline of a vast diatom bloom in the Atlantic sector of the Southern Ocean in summer 2012. We used thorium-234 (234Th) as a particle tracer in combination with drifting sediment traps (ST). Their simultaneous use evidenced a sustained high export rate of 234Th at 100 m depth in the weeks prior to and during the sampling period. The entire study area, of approximately 8000 km**2, showed similar vertical export fluxes in spite of the heterogeneity in phytoplankton standing stocks and productivity, indicating a decoupling between production and export. The POC fluxes at 100 m were high, averaging 26 ± 15 mmol C/m**2/d, although the strength of the biological pump was generally low. Only <20% of the daily primary production reached 100 m, presumably due to an active recycling of carbon and nutrients. Pigment analyses indicated that direct sinking of diatoms likely caused the high POC transfer efficiencies (~60%) observed between 100 and 300 m, although faecal pellets and transport of POC linked to zooplankton vertical migration might have also contributed to downward fluxes.

Compilation of Mg/Ca ratios of Neogloboquadrina pachyderma (sinistral) and SST estimates from the northern North Atlantic

A map of estimated calcification temperatures of the planktic foraminifer Neogloboquadrina pachyderma sinistral (TNps) for the Nordic Seas and the northern North Atlantic for the Last Glacial Maximum was produced from oxygen isotopes with support of Mg/Ca ratios. To arrive at the reconstruction, several constraints concerning the plausible salinity and ?18O-fields were employed. The reconstruction indicates inflow of temperate waters in a wedge along the eastern border of the Nordic Seas and at least seasonally ice-free waters. The reconstruction from oxygen isotopes shows similarities with Mg/Ca based paleotemperatures in the southern and southeastern sector, while unrealistically high Mg/Ca values in the central Nordic Seas prevent the application of the method in this area. The oxygen isotope based reconstruction shows some agreement with temperature reconstructions based on the modern analogue technique, but with somewhat lower temperatures and a stronger internal gradient inside the Nordic Seas. All told, our results suggest a much more ice-free and dynamic high latitude ocean than the CLIMAP reconstruction.

Wave transformation at isolated vertical piles in shallow water /

"March 1978."

NEUROSAT: an overview

This report gives an overview of the work being carried out, as part of the NEUROSAT project, in the Neural Computing Research Group at Aston University. The aim is to give a general review of the work and methods, with reference to other documents which provide the detail. The document is ongoing and will be updated as parts of the project are completed. Thus some of the references are not yet present. In the broadest sense, the Aston part of NEUROSAT is about using neural networks (and other advanced statistical techniques) to extract wind vectors from satellite measurements of ocean surface radar backscatter. The work involves several phases, which are outlined below. A brief summary of the theory and application of satellite scatterometers forms the first section. The next section deals with the forward modelling of the scatterometer data, after which the inverse problem is addressed. Dealiasing (or disambiguation) is discussed, together with proposed solutions. Finally a holistic framework is presented in which the problem can be solved.

Regularisation of mixture density networks

Mixture Density Networks are a principled method to model conditional probability density functions which are non-Gaussian. This is achieved by modelling the conditional distribution for each pattern with a Gaussian Mixture Model for which the parameters are generated by a neural network. This thesis presents a novel method to introduce regularisation in this context for the special case where the mean and variance of the spherical Gaussian Kernels in the mixtures are fixed to predetermined values. Guidelines for how these parameters can be initialised are given, and it is shown how to apply the evidence framework to mixture density networks to achieve regularisation. This also provides an objective stopping criteria that can replace the `early stopping' methods that have previously been used. If the neural network used is an RBF network with fixed centres this opens up new opportunities for improved initialisation of the network weights, which are exploited to start training relatively close to the optimum. The new method is demonstrated on two data sets. The first is a simple synthetic data set while the second is a real life data set, namely satellite scatterometer data used to infer the wind speed and wind direction near the ocean surface. For both data sets the regularisation method performs well in comparison with earlier published results. Ideas on how the constraint on the kernels may be relaxed to allow fully adaptable kernels are presented.

First year qualifying report: neural networks for extracting wind vectors from satellite scatterometer data

The ERS-1 Satellite was launched in July 1991 by the European Space Agency into a polar orbit at about km800, carrying a C-band scatterometer. A scatterometer measures the amount of radar back scatter generated by small ripples on the ocean surface induced by instantaneous local winds. Operational methods that extract wind vectors from satellite scatterometer data are based on the local inversion of a forward model, mapping scatterometer observations to wind vectors, by the minimisation of a cost function in the scatterometer measurement space.par This report uses mixture density networks, a principled method for modelling conditional probability density functions, to model the joint probability distribution of the wind vectors given the satellite scatterometer measurements in a single cell (the `inverse' problem). The complexity of the mapping and the structure of the conditional probability density function are investigated by varying the number of units in the hidden layer of the multi-layer perceptron and the number of kernels in the Gaussian mixture model of the mixture density network respectively. The optimal model for networks trained per trace has twenty hidden units and four kernels. Further investigation shows that models trained with incidence angle as an input have results comparable to those models trained by trace. A hybrid mixture density network that incorporates geophysical knowledge of the problem confirms other results that the conditional probability distribution is dominantly bimodal.par The wind retrieval results improve on previous work at Aston, but do not match other neural network techniques that use spatial information in the inputs, which is to be expected given the ambiguity of the inverse problem. Current work uses the local inverse model for autonomous ambiguity removal in a principled Bayesian framework. Future directions in which these models may be improved are given.

Improved multi-beam neural network scatterometer forward models

Current methods for retrieving near surface winds from scatterometer observations over the ocean surface require a foward sensor model which maps the wind vector to the measured backscatter. This paper develops a hybrid neural network forward model, which retains the physical understanding embodied in ¸mod, but incorporates greater flexibility, allowing a better fit to the observations. By introducing a separate model for the mid-beam and using a common model for the fore- and aft-beams, we show a significant improvement in local wind vector retrieval. The hybrid model also fits the scatterometer observations more closely. The model is trained in a Bayesian framework, accounting for the noise on the wind vector inputs. We show that adding more high wind speed observations in the training set improves wind vector retrieval at high wind speeds without compromising performance at medium or low wind speeds.

Online approximations for wind-field models

We study online approximations to Gaussian process models for spatially distributed systems. We apply our method to the prediction of wind fields over the ocean surface from scatterometer data. Our approach combines a sequential update of a Gaussian approximation to the posterior with a sparse representation that allows to treat problems with a large number of observations.

Regularisation of mixture density networks

Mixture Density Networks are a principled method to model conditional probability density functions which are non-Gaussian. This is achieved by modelling the conditional distribution for each pattern with a Gaussian Mixture Model for which the parameters are generated by a neural network. This thesis presents a novel method to introduce regularisation in this context for the special case where the mean and variance of the spherical Gaussian Kernels in the mixtures are fixed to predetermined values. Guidelines for how these parameters can be initialised are given, and it is shown how to apply the evidence framework to mixture density networks to achieve regularisation. This also provides an objective stopping criteria that can replace the `early stopping' methods that have previously been used. If the neural network used is an RBF network with fixed centres this opens up new opportunities for improved initialisation of the network weights, which are exploited to start training relatively close to the optimum. The new method is demonstrated on two data sets. The first is a simple synthetic data set while the second is a real life data set, namely satellite scatterometer data used to infer the wind speed and wind direction near the ocean surface. For both data sets the regularisation method performs well in comparison with earlier published results. Ideas on how the constraint on the kernels may be relaxed to allow fully adaptable kernels are presented.

Multiphase modelling of the characteristics of close coupled gas atomization

Not withstanding the high demand of metal powder for automotive and High Tech applications, there are still many unclear aspects of the production process. Only recentlyhas supercomputer performance made possible numerical investigation of such phenomena. This thesis focuses on the modelling aspects of primary and secondary atomization. Initially two-dimensional analysis is carried out to investigate the influence of flow parameters (reservoir pressure and gas temperature principally) and nozzle geometry on final powder yielding. Among the different types, close coupled atomizers have the best performance in terms of cost and narrow size distribution. An isentropic contoured nozzle is introduced to minimize the gas flow losses through shock cells: the results demonstrate that it outperformed the standard converging-diverging slit nozzle. Furthermore the utilization of hot gas gave a promising outcome: the powder size distribution is narrowed and the gas consumption reduced. In the second part of the thesis, the interaction of liquid metal and high speed gas near the feeding tube exit was studied. Both axisymmetric andnon-axisymmetric geometries were simulated using a 3D approach. The filming mechanism was detected only for very small metal flow rates (typically obtained in laboratory scale atomizers). When the melt flow increased, the liquid core overtook the adverse gas flow and entered in the high speed wake directly: in this case the disruption isdriven by sinusoidal surface waves. The process is characterized by fluctuating values of liquid volumes entering the domain that are monitored only as a time average rate: it is far from industrial robustness and capability concept. The non-axisymmetric geometry promoted the splitting of the initial stream into four cores, smaller in diameter and easier to atomize. Finally a new atomization design based on the lesson learned from previous cases simulation is presented.

Outlier detection in scatterometer data:Neural network approaches

Satellite-borne scatterometers are used to measure backscattered micro-wave radiation from the ocean surface. This data may be used to infer surface wind vectors where no direct measurements exist. Inherent in this data are outliers owing to aberrations on the water surface and measurement errors within the equipment. We present two techniques for identifying outliers using neural networks; the outliers may then be removed to improve models derived from the data. Firstly the generative topographic mapping (GTM) is used to create a probability density model; data with low probability under the model may be classed as outliers. In the second part of the paper, a sensor model with input-dependent noise is used and outliers are identified based on their probability under this model. GTM was successfully modified to incorporate prior knowledge of the shape of the observation manifold; however, GTM could not learn the double skinned nature of the observation manifold. To learn this double skinned manifold necessitated the use of a sensor model which imposes strong constraints on the mapping. The results using GTM with a fixed noise level suggested the noise level may vary as a function of wind speed. This was confirmed by experiments using a sensor model with input-dependent noise, where the variation in noise is most sensitive to the wind speed input. Both models successfully identified gross outliers with the largest differences between models occurring at low wind speeds. © 2003 Elsevier Science Ltd. All rights reserved.