Data assimilation (DA) is an emerging research area in palaeoclimatology. Here, ensemble-based DA schemes are implemented and evaluated for the reconstruction of the climate of some of the key periods from the past millennium. The study is among the first to employ a General Circulation Model for palaeoclimate DA. An off-line and an on-line DA method are first compared, assimilating continental proxy-based temperature reconstructions and using the 17th century as testing period. Both schemes provide simulations that follow the assimilated targets on large scales better than without DA. The on-line scheme has the advantage of temporal consistency of the analysis, and is subsequently used to reconstruct the climate for 1750-1850 AD. The assimilation performs well on large-scale temperatures, but there is no agreement between the DA analysis and reconstructions for regional temperature patterns. Evidence is presented to suggest that this lack of information propagation to smaller spatial scales is likely due to the fact that the Northern Hemisphere continental mean temperatures are not the best predictors for large-scale circulation anomalies, or that the assimilated reconstructions include noise. The lack of regional skill is again found when instrumental data for 1850-1949 AD are assimilated. Based on these results, it is argued that a potential way of improving the performance of DA is the assimilation of temperature reconstructions with higher spatial resolution.

Groundwater quality sampling guidance typically requires representative samples to be obtained. Such guidance is not always clear what this means and which sampling methods are most appropriate. The situation is complicated by increasing well screen/open interval length. Uncertainty, resulting particularly from observations of vertical flow in wells has led to calls for the use of long-screen (> 3 m) wells to be abandoned for groundwater quality monitoring. Here, four complementary field and modelling studies at various scales are used to examine appropriate groundwater quality sampling in such wells. Numerical modelling demonstrates that literature reported vertical flows in wells < 10 m in length are sufficient to bias pumped groundwater quality sampling. Bias starts for vertical well flow rates less than 50 % of the pumping rate. Vertical flow measurements explain differences and similarities in historical passive sampling between four boreholes and allow vertical aquifer concentration distributions to be quantified. However, such quantification requires per-borehole flow measurement. New technology (Active Distributed Temperature Sensing) provides a versatile alternative to existing borehole flow characterisation methods under ambient and pumping conditions. Data from contrasting field environments demonstrate that even without comprehensive flow investigation long-screen wells can still provide useful information about groundwater concentrations and trends.

The research described in this thesis primarily investigates the scope of IP (induced polarisation) as a method of determining the permeabilities of Bunter sandstones in situ. Although a limited amount of preliminary fieldwork has been completed, the project has been largely confined to the laboratory, directed toward identification of the petrophysical factors involved, rather than determination of aquifer geometry. In contrast: to all of the previous studies, the conclusions are supported by a large number of measurements. IP has been measured in the time domain exclusively. Compared to the frequency domain method this offers the advantages of a wider scope of information from which a suitable measure of chargeability may be selected, together with the facility to monitor the potential decay for spurious results. The variation of IP with permeability was investigated for a large number of sandstones with suitable pore water control, and because of the diversity of the results a representative selection of the samples was then subjected to further measurements of IP over a range of pore water salinities. All these data have been described by empirical expressions which relate IP, permeability and matrix conductivity, and some geological implications of these results have been considered. Preliminary fieldwork has been done in the Fylde to find out whether an empirical relationship determined in the laboratory can be used to evaluate field permeabilities directly. The problems involved have been considered and further work in this direction is suggested.

The Antarctic Peninsula (AP) has been one of the most rapidly warming regions on this planet. This warming has been accompanied by major glaciological changes such as tidewater glacier retreat, ice-shelf retreat and collapse alongside acceleration of outlet glaciers in response to ice-shelf removal. As faster owing glaciers deliver more ice from the ice sheet's interior to the margins, the AP has been identified as an important contributor to global sea-level rise (SLR). However, comprehensible SLR projections of the AP induced by ice dynamics over the next three centuries are still lacking. In this thesis, numerical ice-sheet models are utilised to present scenario-based ice dynamic SLR projections for the AP.

Urban waste water production increases day by day and its safe treatment and disposal need efficient procedures. In many areas, such effluents are discharged to open water bodies such as lakes, rivers and sea coastal areas. Since there are no perennial streams in arid and semi-arid regions the disposal of treated outfalls is often to dry wadis. However developing an understanding of complex urban systems, where processes act at different space and time scales, is not easy. Dynamical systems approaches have been used for many years in complex feedback systems. So in this work the use of dynamical system modelling is investigated to see if this approach can help develop at least in semi-quantitative way an understanding good enough to aid managers of urban water systems where wadis are involved. The approach taken was to develop a flow and then a solute transport model for the urban system of Riyadh City - Wadi Hanifah. The softwares used was ‘Stella’, and a representation of two aquifers, two soil systems, the sewerage system, the water supply system, the non-urban catchments and the urban drainage system was set up and run using daily meteorological data for about 20 years. The model was compared with limited field data on water levels, flows, flooding, and water quality and modified until results were consistent with field data. Model was then investigated by looking at effects of changing a wide range of hydrogeological and other parameter values, including pipe leakage rates, rainfall, and water supply rate.

Traditional and well-established applications of airborne remote sensing to archaeology involve standard aerial photographic recording, either oblique or vertical, of archaeological phenomena visible as shadow, soil or cropmarks, all of which require particular conditions before they become visible. Cropmarks have made the most important contribution to archaeological site discovery and are the most complex of the phenomena, whose occurrence relies upon the differential availability of moisture to a crop growing over buried archaeological remains, reflecting different soil depths and precipitation, to create differential growth patterns in times of moisture stress. Consistently successful aerial photographic reconnaissance for cropmarks relies on dry weather and well-drained soils in arable agriculture. There is, thus, in Scotland a bias in the discovery of archaeological sites in favour of drier eastern districts, supporting arable agriculture, compared to the west with its wetter climate focussed on pastureland. Because cropmarks are linked to moisture stress in growing plants they are potentially detectable in bands outside the visible part of the electro-magnetic spectrum. Although historically cropmark detection has used film, whose sensitivity closely approximates the human eye, hyperspectral scanning allows consideration of a wider range of different wavelengths, beyond the visible spectrum, many of which are more sensitive to changes in vegetation status. The main objectives of this research are to test to what extent hyperspectral and multispectral imagery can reveal otherwise invisible archaeological sites surviving as cropmarks; to assess the relative usefulness of the different sensors employed; and to investigate the potential of hyperspectral and multispectral imagery to augment cropmark detection rates in areas less conducive to their production. It uses a range of imagery (CASI 2, ATM and digital vertical photographic data) acquired by NERC ARSF from two case study sites in Lowland Scotland, one in the east and one in the west, selected to facilitate comparison between areas of good and poor cropmark production. Following processing, comparison and detailed analysis of this data, this thesis has demonstrated the high efficacy of this imagery in the identification of archaeological cropmarks; has established the most appropriate range of bandwidths and processing methods applicable to that imagery; has demonstrated its potential for the discovery of previously unrecognised archaeological sites in areas of lowland pasture; and has reinforced the value of systematic block coverage as compared to traditional ‘observer-directed’ archaeological aerial reconnaissance.

This thesis investigates torrent erosion in Lake District mountain catchments, Northern England. A nested research approach was used. Detailed investigations were undertaken at two case study sites (Iron Crag, Raise Beck) together with a survey of torrents across the Skiddaw and Helvellyn massifs. At Iron Crag an annual sediment budget was constructed by monitoring hillslope, channel and fan processes. Particle size characteristics of sediments, and the history of fan development were investigated. Results show channel and bank sediments are the main source of material supplied to the fan. Large rainfall events cause significant change in the channel, banks and fan. The impact of different meteorological conditions on sediment characteristics is complex, however a seasonal cycle of sediment production (winter) and exhaustion (autumn) exists. Historically, initial fan aggradation predates 36 BC, but a rapid phase of deposition began between 1200-1400 AD. Investigations at Raise Beck focussed on a flood that occurred in January 1995 and caused channel avulsion and shallow landsliding. This was reconstructed using a range of geomorphological and sedimentological evidence. Palaeohydrological methods give a discharge between 27- 74 in s-1. whereas as rainfall-runoff values range between 4-6 m3 s-1. The magnitude of the 1995 flood was smaller than two 19th Century events, but would still exceed the capacity of contemporary engineered channels. The regional survey considered the characteristics and importance of torrents,mountain streams, and debris flows; and provided a context for work at the case study sites. The case study sites are distinct members of the regional populations. Raise Beck being the largest (133 ha) and highest (858 m O. D. ); Iron Crag amongst the smallest (2.4 ha) and lowest (600 m O. D. ). Overall, torrents and hillslope debris flows are minor components of the landscape (aerially 2.1 % Helvellyn massif, 0.4 % Skiddaw massif). Sites are preferentially located in regard to altitude and slope. Debris flows are related to geological type. Large torrent floods are relatively rare and can be broadly related to regional flood episodes. Contemporary debris flow activity is of low magnitude and frequency.

Of numerous aeolian studies around the world, few have been dedicated to dust trapping in the Saharan regions and none is known in Libya. This research aims to explore the extent of dust activities in the western part of Libya, the main factors that influence dust entrainment and deposition, the likely regional emission sources transporting aerosol to this part of the country and ultimately to establish base line information in space and time based on a study area larger than the size of the UK. Dust trapping was carried out at thirty sites on a monthly basis starting from the beginning June 2000 until the end of May 2001. A physical characterization of 274 aeolian sediments were based on grain size distribution analysis and the mineral composition was based on XRD testing of 24 dust samples taken from extreme locations near the Mediterranean Sea and about 800 km inland. An assessment of dust emissions within the area of study, the vertical deposition rates in relation to The Total Ozone Mapping Spectrometer (TOMS) data have been investigated. Moreover, an attempt identify potential Saharan dust emission sources that have impacted the study area has also been made utilizing TOMS data. The area of study has been divided into three regions northern, central and southern according to the dominant factors controlling mobilization of dust particles on the bed surface. This study concludes that deposition rates in the northern coastal region are largely dominated by human activities. Rates in the central region were strongly affected by topographic irregularities whereas in the south deposition rates were less affected by topography and anthropogenic activity. However, deposition rates and particle size distributions are strongly controlled by wind regimes and correlate with the average atmospheric temperatures. Local sediments seem to be strongly affected by saltating particles in most of the study area and no evidence of long range aerosol emission from western Libya was found. Nevertheless, fine dust (<10 gm) is wide spread in the local atmosphere, however it is more pronounced during late spring and summer. TOMS data and the prevailing wind directions reveal that the depressions of the Libyan Desert and the Bodele Depression in Chad were the main sources of aerosol transported over the study area during the highest months of emission, transport and deposition, July and April. It is hoped that this baseline information can pave the way for future studies on dust impacts on soil fertility, human health, desertification, climate change and the validation of present day computer models.

An investigation into the trajectory of river water quality as a strategic issue for the water industry was conducted within two water organisations. This research traced the profile of this issue on the agenda of the water authorities over twenty years and within the industry over a century. The aim was to gain insights into processes of strategic agenda building and organisational development and change, linking process to performance in the achievement of river quality. A contextualist theory of method was adopted in a comparative case study approach which sought to assess the performance of the two organisations in attaining improvements in river water quality during the time frame. The contextualist methodology necessitated examining the agenda building process from multiple levels and over time. To this end extensive archive research and 40 interviews were conducted.The influence of the wider social environment and the sector in the long term were found to be important in the formation of sector and organisational ideology which conditioned organisational developments. A structurationist approach demonstrated the key social structures and their properties implicated in the formation of organisational ideology and its change, as the water authorities were privatised. A process model of strategic agenda building was developed and extended, based on an earlier model by Dutton (1988). This emphasised the influence of multiple contexts, the role of organisational ideology, issue related activities and the actions of sponsors as important additions to the original model. Further, the use of structuration theory, underpinned by a Realist perspective, outlined a conception of agency based on the causal powers granted by the necessary relations of the organisational structure or ideology, and that agency was granted by organisational members' access to alternative structural systems outside the organisational context. This research concluded that the links between structure, process and performance are implicated in incremental and transformational change, and that the properties of structure were instrumental in the propensity for adaption and change. Finally, organisational processes should accurately reflect the rules of the system for change to work.

This thesis examines the dynamic of infectious diseases after natural disasters. Methods from epidemiology and geography intersect in the nexus of this research to form new insights into the risk of infectious disease in the aftermath of natural disaster and catastrophe. In the past decades, natural disasters have increased in frequency and magnitude, and with climate change progressing as it is, this trend is expected to continue. It is thus important to gain a fuller understanding of the dynamic between natural disaster and disease, and challenge the persisting problems in disaster and disease response efforts. Two approaches were taken to determine the risk of disease after disaster. Firstly, by pooling data from previously published literature, a form of meta-analysis was conducted to gain insight into risk patterns as well as to define relevant confounding factors that held significance in determining vulnerabilities of affected populations. For this analysis, a new tool was applied to identify relevant research, and this tool is expected to be useful in future study of the subject. Secondly, a set of empirical studies were conducted to determine the association between types of natural disasters, geographic region, and four distinct disease profiles. Cholera, malaria, tuberculosis, and the co-infection with HIV and tuberculosis served as examples for the types of diseases commonly observed after disasters (diarrhoeal diseases, vector-borne diseases, and acute respiratory infections). Logistic regression models were used to find the odds ratios for above average diseases at different tiers of disaster magnitude. It was shown in this research that the relative risk of infectious disease after natural disasters was 3.45, indicating a higher probability of disease after disasters. Specific results show that disasters affecting higher numbers of the population typically lead to increases in new infections. Most interestingly, tuberculosis relapses showed significant increases after natural disasters, especially meteorological and hydrological disasters.