Diffuse transfers of pesticides from agricultural land to ground and surface waters can lead to significant drinking water quality issues. This thesis describes the development and application of a parameter-efficient, numerical model to predict pesticide concentrations in raw water sources within an integrated hydrological framework. As such, it fills an unoccupied niche that exists in pesticide fate modelling for a computationally undemanding model that contains enough process complexity to be applicable in a wide range of catchments and hydrogeological settings in the UK and beyond. The model represents the key processes involved in pesticide fate (linear sorption and first-order degradation) and transport (surface runoff, lateral throughflow, drain flow, percolation to the unsaturated zone, calculated using a soil water balance) in the soil at a daily time step. Soil properties are derived from the national soil database for England and Wales and are used to define the boundary conditions at the interface between the subsoil and the unsaturated zone. This is the basis of the integrated hydrological framework which enables the application of the model to both surface water catchments and groundwater resources. The unsaturated zone model accounts for solute transport through two flow domains (accounting for fracture flow and intergranular matrix flow) in three hydrogeological settings (considering the presence and permeability of superficial deposits). The model was first applied to a small headwater sub-catchment in the upper Cherwell. Performance was good for drainflow predictions (Nash Sutcliffe Efficiency > 0.61) and performed better than the MACRO model and as well as the modified MACRO model. Surface water model performance was evaluated for eight pesticides in five different catchments. Performance was generally good for flow prediction (Nash Sutcliffe Efficiency > 0.59 and percentage bias below 10 %, in the validation period for all but two catchments). The 90th percentile measured concentration was captured by the model in 62 % of catchment-pesticide combinations. In theremaining cases predictions were within, at most, a factor of four of measured 90th percentile concentrations. The rank order of the frequency of pesticides detected over 0.1 μg L-1 was also predicted reasonably well (Spearman’s rank coefficient > 0.75; p < 0.05 in three catchments). Pesticide transport in the unsaturated zone model was explored at the point scale in three aquifers (chalk, limestone and sandstone). The results demonstrate that representing the unsaturated zone processes can have a major effect on the timing and magnitude of pesticide transfers to the water table. In comparison with the other catchment scale pesticide fate models that predict pesticide exposure at a daily time-step, the model developed stands out requiring only a small number of parameters for calibration and quick simulation times. The benefit of this is that the model can be used to predict pesticide exposure in multiple surface and groundwater resources relatively quickly which makes it a useful tool for water company risk assessment. The broad-scale approach to pesticide fate and transport modelling presented here can help to identify and prioritise pesticide monitoring strategies, to compare catchments in order to target catchment management and to highlight potential problems that could arise under different future scenarios.

Hydroelectric reservoirs account for 59% of the installed electricity generating capacity in Canada and 26% in Ontario. Reservoirs also provide irrigation capacity, drinking water, and recreational opportunities. Further, they continue to be built in northern Canada, neighbouring boreal countries, and around the world. Yet given their socio-economic importance, they are understudied with respect to greenhouse gas emissions, nutrient and mercury cycling, and aquatic metabolism. As one of many electricity generating options, hydroelectricity is viewed as well-tested because of its long history and diverse applications in mega-projects, run-of-the-river dams, and small, local applications. It is also considered renewable from a fuel stand-point because an adequate long-term supply of water is assumed. One of several significant criticisms of hydroelectric development is that reservoirs may be a significant source of greenhouse gases to the atmosphere relative to the amount of electricity produced due to flooding the landscape. As a result of the dearth of information on reservoir development and both greenhouse gases and aquatic metabolism, a pair of whole-ecosystem reservoir experiments were conducted staring in 1991. Three upland boreal forest reservoirs with differing amounts of pre-flood stored organic carbon were built in northwestern Ontario and flooded for five years. The rates of net greenhouse gas production in these reservoirs were determined by calculating mass budgets for carbon dioxide and methane. Additionally, rates of biological processes were determined by combining the mass budgets with measurements of the stable isotopes of carbon and oxygen. Assembling mass and isotope-mass budgets required three related projects on gas exchange, methane oxidation, and oxygen isotopes. To estimate the gas exchange coefficient for each of the upland reservoirs, a comparative-methods study was undertaken. Methane oxidation enrichment factors were determined in upland and wetland boreal reservoirs so that the importance of methane oxidation in these ecosystems could be assessed. In order to interpret the diel changes in both oxygen concentrations and their isotopic ratios, a dynamic model was developed. This model, PoRGy, was successfully applied to the upland boreal reservoirs as well as prairie rivers and ponds. Further, PoRGy was used to understand the interplay between the key parameters that control oxygen concentrations, to compare aquatic ecosystems, to make quantitative estimates of ecosystem metabolism, and to assess the vulnerability of aquatic ecosystems under various environmental stressors. Carbon isotope-mass budgets were used to conclude that community respiration rates declined quickly in the upland reservoirs and had declined by half over five years. This suggested that the most labile organic carbon is quickly consumed but decomposition continued for the five-year life of the project. Net primary production rates were similar for three years, with a small peak in the second or third year, before declining by half by the fifth year. Together, these results indicated that aquatic metabolism slowed over five years while the reservoirs remained a source of greenhouse gases to the atmosphere each year. Net methane production was greatest in the third year of flooding then decreasing by about half by the fifth year. Methane ebullition also peaked in the third year and declined by two-thirds by the fifth year. Together, these results indicated that methanogenesis was greatest in the third year of flooding. The flux of methane to the atmosphere grew in importance relative to that of carbon dioxide over the five years of the experiment. Community respiration and primary production could not be estimated directly from the oxygen isotope-mass budgets since the oxygen respiration enrichment factor remains poorly constrained. Instead, three estimates were made, each based on a different assumption. In general, these estimates suggested that rates of community respiration and primary production decreased slightly for three years and most rapidly in the final two years. The oxygen isotope-mass budgets provided a new method for assessing and constraining community metabolism and greenhouse gas fluxes to the atmosphere. One of the major hypotheses of the whole-ecosystem reservoir experiments was that pre-flood organic carbon stores less tree boles were positively related to greenhouse gas fluxes. Within the three upland boreal forest reservoirs, this hypothesis did not hold true. Over five years, community respiration in the three reservoirs was within 5% of each other. When methane is included, to assess total greenhouse gas fluxes to the atmosphere, the reservoirs were within 1% of each other. Organic carbon stores were therefore poor short-term predictors of carbon lability and greenhouse gas fluxes. This research presented two methods for determining biological rates at the whole-ecosystem scale: one using carbon isotopes and one using oxygen isotopes. Temporal evolution of greenhouse gas cycling within the upland reservoirs was different than in the wetland reservoir and should inform how reservoir development is done vis-à-vis the amount of flooded land of each type versus electricity production. Medium-term estimates of greenhouse gas fluxes suggest that upland reservoirs do not have adequate pre-flood organic carbon stores to sustain elevated levels of decomposition the way wetlands do. The strong evidence of continued production of dissolved organic carbon in the upland reservoirs should concern operators of municipal drinking water reservoirs since elevated dissolved organic carbon can make disinfection difficult.

There is a great need to reduce nutrient leaching from arable land into lakes and oceans. By using several different types of models it has previously been possible to describe nutrient losses in a catchment area with a minimum unit of sub-catchment level. At present, it is instead desirable to model a smaller catchment with an opportunity to re-connect the results to the corresponding fields in the catchment. Such models already exist but they are not fully able to properly describe Swedish conditions and land characteristics in our region. With the approach of creating such a model, SLU has developed a project with this work as its first stage. The model is expected to be created under the working name SWE-model which stands for Soil Water Environment and is in this first stage supposed to apply the SOIL model in catchment scale. During the procedure to describe the first step in the process of developing such a model adapted to Swedish conditions and which works in the catchment scale with an area of about 10-30 km2, focus has been set on calculating the transport of water flow from different hydrological response units. Regardless of the processes occurring in the soil after the water has been added, it is assumed that all the water which flows from each simulated unit is drained. In the first step the hydrologic response units were identified based on land use and soil type in the study area. With the help of a script with functions that retrieve and transform data, certain units were chosen for simulation. The script was also created in this project. Finally, the model results were aggregated and summarized for each unique unit, for each sub-catchment, and also for the whole catchment. From the results it is possible to see similarities in the flow dynamics between modeled and measured data. The efficiency coefficient has been calculated to correspond to the mean of the measured values for the whole simulation period. With an automated calibration process the model should be able to perform better. The volume error gives an indication of overestimation from the model.

Hydroelectric reservoirs account for 59% of the installed electricity generating capacity in Canada and 26% in Ontario. Reservoirs also provide irrigation capacity, drinking water, and recreational opportunities. Further, they continue to be built in northern Canada, neighbouring boreal countries, and around the world. Yet given their socio-economic importance, they are understudied with respect to greenhouse gas emissions, nutrient and mercury cycling, and aquatic metabolism. As one of many electricity generating options, hydroelectricity is viewed as well-tested because of its long history and diverse applications in mega-projects, run-of-the-river dams, and small, local applications. It is also considered renewable from a fuel stand-point because an adequate long-term supply of water is assumed. One of several significant criticisms of hydroelectric development is that reservoirs may be a significant source of greenhouse gases to the atmosphere relative to the amount of electricity produced due to flooding the landscape. As a result of the dearth of information on reservoir development and both greenhouse gases and aquatic metabolism, a pair of whole-ecosystem reservoir experiments were conducted staring in 1991. Three upland boreal forest reservoirs with differing amounts of pre-flood stored organic carbon were built in northwestern Ontario and flooded for five years. The rates of net greenhouse gas production in these reservoirs were determined by calculating mass budgets for carbon dioxide and methane. Additionally, rates of biological processes were determined by combining the mass budgets with measurements of the stable isotopes of carbon and oxygen. Assembling mass and isotope-mass budgets required three related projects on gas exchange, methane oxidation, and oxygen isotopes. To estimate the gas exchange coefficient for each of the upland reservoirs, a comparative-methods study was undertaken. Methane oxidation enrichment factors were determined in upland and wetland boreal reservoirs so that the importance of methane oxidation in these ecosystems could be assessed. In order to interpret the diel changes in both oxygen concentrations and their isotopic ratios, a dynamic model was developed. This model, PoRGy, was successfully applied to the upland boreal reservoirs as well as prairie rivers and ponds. Further, PoRGy was used to understand the interplay between the key parameters that control oxygen concentrations, to compare aquatic ecosystems, to make quantitative estimates of ecosystem metabolism, and to assess the vulnerability of aquatic ecosystems under various environmental stressors. Carbon isotope-mass budgets were used to conclude that community respiration rates declined quickly in the upland reservoirs and had declined by half over five years. This suggested that the most labile organic carbon is quickly consumed but decomposition continued for the five-year life of the project. Net primary production rates were similar for three years, with a small peak in the second or third year, before declining by half by the fifth year. Together, these results indicated that aquatic metabolism slowed over five years while the reservoirs remained a source of greenhouse gases to the atmosphere each year. Net methane production was greatest in the third year of flooding then decreasing by about half by the fifth year. Methane ebullition also peaked in the third year and declined by two-thirds by the fifth year. Together, these results indicated that methanogenesis was greatest in the third year of flooding. The flux of methane to the atmosphere grew in importance relative to that of carbon dioxide over the five years of the experiment. Community respiration and primary production could not be estimated directly from the oxygen isotope-mass budgets since the oxygen respiration enrichment factor remains poorly constrained. Instead, three estimates were made, each based on a different assumption. In general, these estimates suggested that rates of community respiration and primary production decreased slightly for three years and most rapidly in the final two years. The oxygen isotope-mass budgets provided a new method for assessing and constraining community metabolism and greenhouse gas fluxes to the atmosphere. One of the major hypotheses of the whole-ecosystem reservoir experiments was that pre-flood organic carbon stores less tree boles were positively related to greenhouse gas fluxes. Within the three upland boreal forest reservoirs, this hypothesis did not hold true. Over five years, community respiration in the three reservoirs was within 5% of each other. When methane is included, to assess total greenhouse gas fluxes to the atmosphere, the reservoirs were within 1% of each other. Organic carbon stores were therefore poor short-term predictors of carbon lability and greenhouse gas fluxes. This research presented two methods for determining biological rates at the whole-ecosystem scale: one using carbon isotopes and one using oxygen isotopes. Temporal evolution of greenhouse gas cycling within the upland reservoirs was different than in the wetland reservoir and should inform how reservoir development is done vis-à-vis the amount of flooded land of each type versus electricity production. Medium-term estimates of greenhouse gas fluxes suggest that upland reservoirs do not have adequate pre-flood organic carbon stores to sustain elevated levels of decomposition the way wetlands do. The strong evidence of continued production of dissolved organic carbon in the upland reservoirs should concern operators of municipal drinking water reservoirs since elevated dissolved organic carbon can make disinfection difficult.

In recent years, the increasing awareness of scarcity of water resources, indications of likely climate variability, and the increasing pressure to use available fresh water resources more efficiently have together reinforced the need to look at infrastructure solutions with due regard to environmental considerations and social impacts, present and future. There is a vital need to apply an integrated approach to catchment management to implement sustainable solutions to resolve issues such as water supply and sewerage, drainage and river flooding. Many potentials solutions are available to control water demand and manage flood problems. Greywater recycling and rainwater harvesting are novel technologies. However, their catchment scale impacts on hydraulic and hydrological flows are poorly understood. The research aim is to identify the hydrologic and hydraulic impacts of scaling up such technologies at catchment scale. For this particular study, a computer simulation model will be used to evaluate how increasing urbanisation, climate change and the implementation of greywater recycling and rainwater harvesting may alter the water balance within a representative catchment. To achieve these aims data from the Carrickmines catchment in Ireland have been collected; a simulation model has been adapted to carry out the study, the model has been calibrated and validated, results have been analysed, and finally, a sensitivity analysis has been carried out. The results show that rainwater harvesting systems are comparatively more effective than greywater recycling techniques in reducing flood frequency and intensity. Under five year return period rainfall events, the implementation of rainwater harvesting at any scale and number of units is a useful technique to control river flow and floods. However, the study also shows that under extreme conditions the efficiency of rainwater harvesting systems decreases. The study concludes that implementing the two technologies within a single catchment is not a solution to several forms of hydrological problem. The study shows that implementing rainwater harvesting or re-use technologies are a very useful way to protect local freshwater reserves and therefore conserve our environment.

Les concentrations élevées en pesticides dans les eaux de surface drainant des bassins versants agricoles sont devenues une préoccupation majeure en Europe depuis une cinquantaine d'années. Les pesticides sont transférés dans l'environnement par différentes voies (le ruissellement de surface et de sub-surface, le flux de nappe), soit en solution soit adsorbés aux particules de sol en suspension dans l'eau. Les eaux de ruissellement et de percolation entraînent avec elles des charges de contaminants dont les concentrations en solution peuvent s'avérer toxiques pour la faune et la flore aquatique et rendre l'eau impropre à la consommation humaine si le réseau de drainage est une source de captage pour l'alimentation en eau potable. Les crues constituent donc des événements hydrologiques de première importance dans la contamination des eaux continentales par les pesticides. Les objectifs de cette thèse ont été de (1) caractériser, à l'aide d'un modèle agro-hydrologique, la dynamique des transferts de pesticides à l'échelle du bassin versant dans une région agricole, notamment en période de crue ; (2) identifier les facteurs de contrôle du transfert de pesticides et (3) améliorer, le cas échéant, les équations formalisées dans le modèle. Deux approches ont été menées de front afin de répondre aux questions posées : l'analyse de données mesurées et modélisées sur le bassin versant agricole de la Save (sud-ouest de la France). Une étude de faisabilité réalisée en préliminaire a montré que le modèle Soil and Water Assessment Tool (SWAT - Arnold et al., 1998) était adapté à la modélisation du transfert de pesticides, dans les phases dissoute et particulaire, à l'échelle du bassin versant. L'hydrologie et les concentrations à l'exutoire des phases dissoute et particulaire (respectivement les nitrates et les matières en suspension) ont été calibrées. Les voies privilégiées de transfert des pesticides en fonction des conditions hydrologiques ont été identifiées. La modélisation a ensuite été mise en œuvre avec des itinéraires techniques plus détaillés en entrée du modèle et des mesures sub-journalières de pesticides en crue. Les différentes voies de transfert des pesticides dans les deux phases, ainsi que leurs facteurs de contrôle environnementaux, ont été étudiés. Deux facteurs de contrôle, respectivement dépendant des pratiques agricoles (la date d'application des pesticides, qui est un facteur anthropique) et intrinsèque aux molécules de pesticides (le coefficient Kd de partition entre phases dissoute et particulaire, qui est un facteur physico-chimique) ont été abordés plus en détail. Le rôle de la typologie du bassin versant sur les transferts est discuté. Des cartes de risque de contamination des eaux de surface par les pesticides sont présentées pour le bassin de la Save. Dans la perspective d'améliorer le formalisme des modèles de transfert des pesticides, une équation qui relie le coefficient Kd au coefficient de distribution octanol/eau Kow et à la concentration en matières en suspension a été proposée. / Rising pesticide levels in streams draining intensively managed agricultural land has become a widespread problem throughout Europe in recent decades. Pesticides are transferred into the environment through various pathways (surface and sub-surface runoff, groundwater return flow), either in solution or sorbed onto particles. Runoff and percolating water carry contaminants loads which concentrations in solution may be harmful to terrestrial and aquatic ecosystems rendering water unfit to human consumption if the draining network is a source for drinking water. Floods are hydrological events of major importance in continental waters contamination by pesticides. The objectives of this PhD thesis were (1) to characterise pesticides transfer dynamics at catchment scale in an agricultural area during floods; (2) to identify the factors controlling pesticides transfer and (3) to improve modelling by changing formalism with more suitable equations. Two approaches were set up: analysing both measured and simulated data sets, stemming from the River Save catchment (south-western France). A preliminary feasibility study showed that the Soil and Water Assessment Tool (SWAT - Arnold et al., 1998) was adapted for pesticides transfer modelling in both dissolved and sorbed phases, at catchment scale. Water discharge, dissolved and sorbed phases (respectively nitrate and suspended sediments) were calibrated. Pesticides transfer preferred pathways depending on hydrological conditions were identified. Modelling was then carried on more detailed management practices as input and on sub-daily pesticides concentration measurements during flood events. The various transfer pathways in both phases together with the environmental controlling factors were assessed. At last, two controlling factors, respectively depending on management practices (application date, an anthropogenic factor) and on an intrinsic pesticide molecule property (the partition coefficient Kd which is a physico-chemical factor) were studied. The role of catchment typology was discussed. Surface water contamination risk maps were drawn on Save catchment. In order to improve the formalism of pesticide transfer models, an equation was proposed that relates Kd to the octanol/water partition coefficient Kow and to suspended matter concentration.

Existing terrestrial carbon accounting models have mainly investigated atmosphere-vegetationsoil stocks and fluxes but have largely ignored the hydrological flux of organic carbon. It is generally assumed that biomass and soil carbon are the only relevant pools in a landscape ecosystem. However, recent findings have suggested that significant amounts of organic carbon can dissolve (dissolved organic carbon or DOC) or particulate (particulate organic carbon or POC) in water and enter the hydrological flux at the catchment scale. A significant quantity of total organic carbon (TOC) sequestered through photosynthesis may be exported from the landscape through the hydrological flux and stored in downstream stocks.¶ This thesis presents a catchment-scale case study investigation into the export of organic carbon through a river system in comparison with carbon that is produced by vegetation through photosynthesis. The Cotter River Catchment was selected as the case study. It is a forested catchment that experienced a major wildfire event in January 2003. The approach is based on an integration of a number of models. The main input data were time series of in-stream carbon measurements and remotely sensed vegetation greenness. The application of models to investigate diffuse chemical substances has dramatically increased in the past few years because of the significant role of hydrology in controlling ecosystem exchange. The research firstly discusses the use of a hydrological simulation model (IHACRES) to analyse organic carbon samples from stream and tributaries in the Cotter River Catchment case study. The IHACRES rainfall-runoff model and a regionalization method are used to estimate stream-flow for the 75 sub-catchments. The simulated streamflow data were used to calculate organic carbon loads from concentrations sampled at five locations in the catchment.¶ The gross primary productivity (GPP) of the vegetation cover in the catchment was estimated using a radiation use efficiency (RUE) model driven by MODIS TERRA data on vegetation greenness and modeled surface irradiance (RS). The relationship between total organic carbon discharged in-stream and total carbon uptake by plants was assessed using a cross-correlation analysis.¶ The IHACRES rainfall-runoff model was successfully calibrated at three gauged sites and performed well. The results of the calibration procedure were used in the regionalization method that enabled streamflow to be estimated at ungauged locations including the seven sampling sites and the 75 sub-catchment areas. The IHACRES modelling approach was found appropriate for investigating a wide range of issues related to the hydrological export of organic carbon at the catchment scale. A weekly sampling program was implemented to provide estimates of TOC, DOC and POC concentrations in the Cotter River Catchment between July 2003 and June 2004. The organic carbon load was estimated using an averaging method.¶ The rate of photosynthesis by vegetation (GPP) was successfully estimated using the radiation use efficiency model to discern general patterns of vegetation productivity at sub-catchment scales. This analysis required detailed spatial resolution of the GPP across the entire catchment area (comprising 75 sub-catchment areas) in addition to the sampling locations. Important factors that varied at the catchment scale during the sampling period July 2003 – June 2004, particularly the wildfire impacts, were also considered in this assessment. ¶ The results of the hydrologic modelling approach and terrestrial GPP outcome were compared using cross correlation and regression analysis. This comparison revealed the likely proportion of catchment GPP that contributes to in-stream hydrological flux of organic carbon. TOC Load was 0.45% of GPP and 22.5 - 25% of litter layer. As a result of this investigation and giving due consideration to the uncertainties in the approach, it can be concluded that the hydrological flux of organic carbon in a forested catchment is a function of gross primary productivity.