The geochemical compositions of <I>Porites lutea</I> coral skeletons were determined in specimens collected from a range of different reef environments in Phuket, South Thailand. Specimens were analysed for stable carbon and oxygen isotopes and trace and minor elements. Geochemical variations between colonies from the same site equalled those observed between different sites. This possibly reflects genetic or growth rate differences between corals. Large variations in the trace element concentration ranges between colonies led to widely varying estimates of distribution coefficients between seawater and coral. No seasonal variations in Mg, Rb, Sr, Ba and U were found in any of the corals analysed by ion microprobe or XRF. Corals are very heterogeneous with respect to these elements on a sub-millimetre scale and the entire concentration range of an element is frequently observed over a skeletal distance corresponding to less than one months growth. This heterogeneity far exceeded the Sr range expected from the Sr palaeothermometer. Heterogeneity may reflect variations in the composition and/or quantity of the organic matrix or regulation of coral tissue metal levels by the zooxanthellae. Anomalous high concentrations of magnesium and barium were observed on microborings and centres of calcification which permeate the coral skeletons. These may relate to organic materials, precipitates or sedimentary particles. Thorough cleaning of skeletal material prior to analysis is essential to remove this contamination of the coral aragonite. δ<SUP>18</SUP>O is inversely related to coral growth rate and may drive the annual oxygen isotopic trend. No evidence of the isotropic trends expected from seawater temperature data were found. δ<SUP>13</SUP>C and coral tissue chlorophyll a content are positively related, reflecting the impact of the algal symbiants on the dissolved inorganic carbon pool available for calcification.

A theoretical and experimental study has been carried out into the measurement of two-phase flows, with particular emphases placed on the analysis and assessment of factors which affect the data obtained from the experiments. Statistical theory has been developed to account for the recorded intensity of side-scattered light from a particle due to: (a) the integration effects of the pixels in the CCD sensor, (b) the quantization effects of the electronics associated with CCD cameras and (c) the position of the particle in a scanning-beam PIV illumination system. The theory was found to agree well with experiment and indicates that for typical PIV experimental parameters, the variation of light recorded due to a particle's random position in the laser sheet is an order of magnitude higher than the variation of light intensity due to the integration effects of the CCD sensor. A further experiment indicated that the variation of intensity due to the light scattering characteristics of quartz sediment particles is of the same order of magnitude as the variation due to the position of the particle in the laser sheet. An automatic system was developed to locate and analyse the particle images of both phases. The primary difficulty for the system was found to be in separating the phases. Experiments were performed to determine statistical confidence levels for this purpose. The analysis method was developed and applied to experiments involving the measurement of two-phase (sediment/water) flow over a rippled-sand bed. For this purpose, a flexible, convenient and robust PIV illumination system was built. The results obtained are discussed to indicate the validity of the method and show that it works well when applied to a practical two-phase flow experiment.

Loch Bradan, a drinking water reservoir in the Galloway Hills, south-west Scotland, frequently exhibits dissolved Mn concentrations of greater than 50 mg l^-1, the EC maximum admissible concentration for Mn. Situated in in a mineralised area overlying greywacke bedrock and granite, the catchment is peaty and hence highly organic in nature. To investigate the underlying cause and specific nature of what, given the well-oxygenated status of the loch, are surprisingly high dissolved Mn concentrations, peat cores and input stream water from the catchment along with sediment cores, pore water and loch water from Loch Bradan were collected. The concentrations of a number of parameters in the peat cores and thinly sliced sediment samples were determined, including pseudo-total Mn, easily reducible Mn, humic substances and humic-associated Mn. Concentrations of dissolved Mn at different depths in the Loch Bradan water column, pore water and input streams were measured in filtered and subsequently fractionated (< kDa and > 1 kDa) samples. The behaviour of Mn in the Loch Bradan system is controlled by a number of factors including classic redox cycling in the catchment peat and bottom sediments of Loch Bradan with enrichments of up to 2% and 7 %, respectively, occurring in the near-surface zones. The presence of insoluble inorganic species such as carbonates and associations with Fe oxides are also important influences upon the behaviour of Mn, especially at greater depth within both the sediment and the peat. The most important interaction with respect to the high dissolved Mn in the water column is, however, the association of Mn with humic substances. In the water column 50 - 60 % of the Mn was found to be associated with organics, with these species entering the water column both from the catchment (where ~ 45 - 55 % of the dissolved Mn in input streams was found to be associated with humic substances) and from the bottom sediment (where ~ 35 % of the dissolved Mn in the surface pore waters were found to be humic associated), with the humic substances in the sediment being principally terrigenous in nature.

Rb-Sr studies of the major alkaline igneous complexes within the Precambrian Gardar Province of south Greenland indicate that the magmatic activity is divisible into three phases, referred to as Early (1300 m.y.), Mid (1250 m.y.) and Late (1165 m.y.) Gardar, respectively. Initial 87Sr/86Sr ratios for the majority of the Gardar magmas were low (0.702-0.704) suggesting derivation from a primitive (mantle) source; some complexes have moderately elevated ratios (0.704-0.707) attributed to varying degrees of contamination by older granitic basement. Two complexes (Ilimaussaq and Ivigtut), both highly differentiated and having high volatile contents, had high (0.707) initial ratios that cannot be explained by such bulk assimilation mechanisms. For these, it is suggested that preferential leaching of B7Sr from the surrounding crust before and during intrusion can account for the elevated values. Rb-Sr investigations of the Red Wine (Labrador) and Norra Farr (Sweden) agpaitic complexes show that the former (1350 m.y.) may well have been emplaced during the Early Gardar Period but the latter is considerably older (1580 m.y.). Both have been affected by the Grenville metamorphism. Rare earth element data for the Tugtutoq younger giant dyke complex of the Gardar Province support previous suggestions that anorthosite inclusions found in the gabbro dykes represent early fractionates disrupted and carried upward by the gabbro magmas. Rare earth and other evidence suggests that the anorthosites may be related to the development of peralkaline magmas at Ilimaussaq.

The Firth of Forth is a wide, shallow, marine-dominated coastal embayment on the east coast of Scotland. The concentration of suspended particulate matter ranges from <1 mg 1-1 at the seaward end to >30 mg 1-1 in the west, with intermediate values along the south shore and in bottom waters. Studies of the major-element chemistry (Al, Si, Ca, Fe, Ti, IC, Mg, P and S) of local suspended matter are reported. An attempt is made to account for the observed areal, vertical and temporal changes in the chemistry of this material in terms of its concentration, gross composition, mineralogy, particle size characteristics and the partitioning of major elements among terrigenous, biogenic and organic components. The extent to which these factors are influenced by hydrodynamic processes and resuspension of bottom sediments is examined. Suspended matter in the Forth is dominated by terrigenous debris, particularly to the west. Locally, variations in inter-element ratios of this material as a function of water depth indicate increasing particle size and loading of terrigenous matter in bottom waters. Particulate Al resides almost exclusively in detrital aluminosilicates. All the elements examined are present to some degree in the terrigenous fraction, their abundance relative to Al being a function of particle size-specific mineralogy. Biogenic associations of Si, Ca (skeletal debris) and P (organics) are quantified. Pollutant phases of particulate P and Fe occur in direct association (organics and possibly ferriphosphates) and in independent forms (sewage organics and ferruginous waste). Seaward and northward flux of particulate matter occurs in surface waters, ensuring some dispersal of pollutant material and mixing with terrigenous debris. As this settles, however, it is subject to net landward transport in the bottom waters along with material winnowed from local sediments. Some degree of hydrodynamic sorting of particles is expressed in their chemistry. Relative to the amount of material in suspension, only negligible escape of particulate matter from the Forth appears to occur.

In this project, field measurements and controlled experiments were used to investigate the magnitude of, and controls on, some terrestrial sources of CH₃Br to help in assessing global fluxes. The influence of factors such as temperature, soil properties, sunshine level, plant species and water table, and correlation to methane (CH­₄) fluxes, were investigated. Field measurements were carried out using eight static flux chapters installed in pairs in the lower and upper areas of a salt marsh in East Lothian (Scotland) and four at a woodland site in Edinburgh. Mean annual measured net CH₃Br emissions from the salt marsh were 350 ng m^-2 h^-1, with a large variability in fluxes from zero to ~4000 ng m^-2 h^-1, which increased seasonally. Summer emissions were on average about three times higher than winter emissions. Variation in CH₃Br fluxes was larger between individual chambers indifferent pairs at the same salt marsh area (upper or lower) than between the two areas. There was also a strong association between sunny conditions and the timings of the frequent peaks in CH₃Br emissions from the high emitting chambers above the general annual trend. There was a distinct diurnal variation in CH₃Br fluxes from the salt marsh chambers throughout the year, also strongly correlated to the solar flux, showing highest emissions at mid-day and lowest at night. Salt marsh CH₃Br fluxes were highly spatially variable, with a small proportion of “hot spots” accounting for the bulk of the spatially-integrated net emissions. There was no obvious explanation for the spatial heterogeneity of CH₃Br emissions. Net CH₃Br fluxes at the woodland site were an order of magnitude lower than at the salt marsh site, fluctuating between net uptake and net emission and with a mean in annual measured emissions (minimum and maximum in parentheses) of 27 (-73-279) ng m² h^-1. Net fluxes from the woodland site showed no seasonal trend and only fairly small diurnal variation. There were no or only modest correlations of net CH₃Br fluxes with air and soil temperature, CH₄ flux and water table depth. Scaling-up the salt marsh CH₃Br fluxes from this work globally gave estimated annual emissions of 1 (0.5-3) Gg y^-1, which was only ~10% of the global salt marsh emission estimate regularly quoted in the methyl bromide literature. In contrast, the emissions from the temperate woodland gave a global average of 2.7 (1.3-5) Gy y^-1 which contradicted the common assumption that woodland soils are only sinks for CH₃Br. Global annual net CH₃Br fluxes from leaf litter were calculated to be 0.9 (-1.3-4) Gg y^-1 which was within the range quoted in the literature, and from conifer needles to 5.3 (1.5-11.6) Gy y^-1, slightly higher than previously reported.

No description available.

The Gardar Alkaline Igneous Province is a Proterozoic, rift-related, province in south-west Greenland. It consists of lavas, central complexes and dyke swarms, dating from 1350 and 1150 million years ago. The majority of the intrusives were emplaced into a Proterozoic mobile belt which formed during the Ketilidian orogeny at -1850- 1740 Ma, but the north-western part of the province transgressed onto the edge of the South Greenland Archaen craton, and it is this area which has been under study. Fieldwork has been carried out as part of the present project. The Gardar intrusives within the craton include three central complexes, Grennedal-Ika, Ivigtut and Kungnat, as well as many dykes. Gronnedal-Ika is the oldest of the complexes, and consists of nepheline syenites cut by a carbonatite plug. The oldest dykes in the area are lamprophyres; these were succeeded by various groups of basic and alkaline salic dykes, including a major set of olivine dolerites known as the "Brown Dykes". The Ivigtut stock is an alkali granite which, prior to mining, contained the world's largest deposit of cryolite, and the granite shows as distinctive pattern of alteration related to the fluids from which the ore deposit formed. The Kungnat Complex consists of saturated and over-saturated syenites with an alkali grabbro ring-dyke. The petrography of the different rock-types is described together with the mineralogy of the Ivigtut granite. It is concluded that the lamprophyre magmas of the Gardar formed as volatile-rich small-degree partial melts within the asthenosphere, which were intruded with minimal lithospheric contamination. Larger volumes of melt were derived from lithospheric mantle which had been selectively enriched in incompatible elements, possibly through the action of fluids derived from a subducting slab at the close the of the Ketilidian orogeny. These melts were intruded as primitive dykes, or evolved through fractionation and variable crustal contamination to produce the spectrum of Gardar magmas.

The aim of the study was to carry out a much needed field investigation on the physical and chemical processes affecting contaminant movement in the groundwater of coastal aquifers, prior to their discharge to coastal waters. The study site located in Ayrshire, UK, was selected as it contains a low-pH acidic and metal-contaminated plume migrating to the Garnock estuary. The tidal estuary is characterised by a very low relief sandy beach. A key research motivation was to determine the mechanisms of contaminant attenuation on the transport pathway, which significantly reduced the impact of groundwater contamination on the estuary. It was found that near to the shore, the groundwater flow was very responsible to changes in the tidal stage. The amplitude of the tidally induced water table fluctuations was rapidly dampened as it propagated inland due to the mild sloping beach. A zone of brackish groundwater of marine origin was found at depth and identified as part of the saltwater intrusion wedge. Little spatial and temporal changes observed during the monitoring period suggest that the saline water was at a quasi-steady state. Mixing between upper acidic and deep alkaline saline groundwaters was unlikely in the monitoring transect. The migration of the acidic plume towards the estuary was limited to the upper section of the aquifer, and was considerably affected by buffering processes involving mainly carbonates, AI and Fe oxy-hydroxides. In turn, the fate of metal contaminants was mainly controlled by pH, however the effect of sediment heterogeneity on the distribution of dissolved metals was shown to be significant. In addition, the results suggested that tide-induced fluctuations in the groundwater flow would have an impact on the distribution of metal contaminants near the shore.

Estimates of nitrous oxide (N₂O) emissions from grassland at field and national scale are presented. Concentrations of atmospheric N₂O are current rising due to increasing anthropogenic emissions. This is of concern as N₂O participates in reactions leading to the destruction of stratospheric ozone and adding to radiative forcing. Quantifying these emissions accurately is a challenge complicated by environmentally driven spatial and temporal variability of its main source, microbial production in soils. In the UK grasslands are the largest source of N₂O. In this project for the first time N₂O was measured from typically intensively managed grassland at Easter Bush, S. Scotland from June 2002 to June 2003 almost continuously. Measurements were made using the micrometeorological eddy covariance technique which couples an ultrasonic anemometer with a Tunable Diode Laser fast response gas. Analysis System (TDLAS) to measure the N₂O concentration (precision ±0.9%, frequency 10 Hz). The spatial coverage attainable using long-term eddy covariance N₂O flux measurements (10³-10⁴ m²) represents a valuable alternative to the commonly used static chamber method (< 1 m²), overcoming the problem of spatial and temporal variability of N₂O emissions associated with the latter technique. Four fertiliser applications made over this period were investigated in detail together with environmental conditions. Background activity was characterised by small fluxes (±50 ng N₂O-N m^-2s^-1). However, N₂O flux varied widely, with a series of daily emission peaks (≤ 3,795 ng N₂O-N m^-2 s^-1) following the fertilisation in June 2002 and less pronounced emissions after the other fertilizations in August 2002, March and June 2003. A multi-linear regression relating combinations of soil water content, soil temperature and the soil available nitrogen with N₂O fluxes was used to study the seasonal variation of the emissions. Good results (R² = 0.83-0.9) were obtained with all the parameters but the best predictors resulted soil water content and soil temperature at 7 cm depth. The extraordinarily large pulse observed on 8 June 2002 immediately after the fertilizer application (120 kg N ha^-1) represents 1.6% of the N applied. After this event clear diurnal emission cycles, peaking at noon, were recorded for about 10 days, allowing an analysis of the environmental factors controlling the loss. In this case turbulent parameters such as sensible heat flux and friction velocity were closely related to the N₂O flux suggesting that they could be fundamental drivers in the emissions. The estimate of the total cumulative flux was 5.3 kg N ha^-1 y^-1, representing an emission factor of 1.68%. The data also reveal a small, but significant, uptake during the winter months, highlighting the existence of N₂O sink activity on grassland.