R. Mackay

Spatial variation in evapotranspiration and the influence of land use on catchment hydrology

Land use change has a direct effect on hydrology through its link with the evapotranspiration regime. In a large river basin there may be considerable variation in both climate and land use across the region. The potential impact of these variations on the evapotranspiration regime is assessed for the Tyne Basin in NE England, using an approach of physically based modelling. Using a simple hill-slope model, predictions of evapotranspiration are made for various scenarios of land use and climate, representative of those found within the catchment. A catchment model is used to demonstrate how the variations in evapotranspiration might affect the hydrology of a region. Simulations are performed applying various land uses to one upland and one lowland sub-catchment of the Tyne Basin. Results from the simulations show that the same land use change may have a significant effect on the hydrology of the lowland subcatchment, but an insignificant effect on the hydrology of the upland sub-catchment. This highlights the importance of considering the physical attributes of a region in predicting the consequences of land use change.

Modelling the hydrological impacts of open ditch drainage

Abstract Open ditch drainage has been observed to have a significant effect on catchment runoff regimes, yet the processes generating the changes are not well understood. An approach of physically based modelling is used, with a fine grid resolution, on a very simple hill-slope model to analyse how ditches affect runoff. Results from the simulations show that the most important process is the change in behaviour of the sub-surface, under near-saturated conditions, when there is a much higher and more variable sub-surface exchange flow into the stream network. The speed of the surface runoff response is also affected. From the results of the analysis a method using effective parameterisation has been developed to recreate the effects of the drainage in a coarser resolution catchment model. The method involves calculating effective saturated hydraulic conductivities for the sub-surface flows and effective Strickler coefficients for the surface flows. The application of the method is demonstrated for a sub-catchment of the River South Tyne at Alston. Results from these simulations show that predictions of flow for the catchment can be improved by using the effective parameterisation to create a pseudo drainage representation of the sub-grid-scale channels within the model. Simulations of more intensive drainage scenarios show that the effects of open ditch drainage on hydrology are significant and that the direction of change is not intuitive.

Development and application of a nitrogen modelling system for large catchments

The NELUP Nitrogen Modelling System (NMS) has been developed to model the transport of nitrates within large catchments. It is one component of a suite of hydrological models that have been integrated with ecological and agroeconomic models in a Decision Support System. The NMS has been designed to be applied in four stages: determination of nitrogen input categories; simulation of plant uptake and decay using the farm management model EPIC; simulation of initial nitrate profiles in the unsaturated zone with the one-dimensional transport model (MP); and finally a simulation of nitrate transport using the SHETRAN-UK catchment flow and transport modelling system. n nThe NMS has been applied to model the transport of nitrate within the 3000 km2 Tyne basin for the period 1985–1989. Predicted in-stream nitrate concentrations throughout the catchment display a low base concentration with peaks corresponding to overland flow events following fertiliser application. Model results compare favourably with the available monthly NRA nitrate observation data at a number of different locations within the catchment. A short field study has been undertaken within the Tyne which demonstrates the occurrence of the transport of nitrate within overland flow events, a hitherto undocumented process.

The hydrological component of the NELUP decision-support system: an appraisal

Abstract The NELUP decision-support system (DSS) has been developed to provide a quantitative description of the main economic and environmental impacts arising out of rural land-use change at the river-basin scale. The system integrates models of economics, ecology and hydrology with relational and spatial databases, thereby permitting interactive evaluation of different future scenarios through a graphical user-interface. The models are coupled by means of data transfers via the database and allow, for example, the consequences of economic policies on agricultural production to be estimated in terms of predicted impact on water resources and ecological diversity within a basin. The two hydrological modelling systems employed in the NELUP DSS, namely SHETRAN and NUARNO, are described. The utility of the hydrological component is illustrated, making use of an application relating to the problem of stream-fed irrigation in the Cam river basin.

Modelling of contaminant migration in a chalk aquifer

Abstract A numerical model is formulated and used to investigate chloride migration in the chalk aquifer beneath effluent lagoons to the north of Royston, Hertfordshire, UK. The aquifer is characterised as a dual-porosity medium in which lateral advection and hydrodynamic dispersion are dominant in the fissure system whereas molecular diffusion is dominant in the chalk matrix. The Galerkin finite element method is used to solve the equations of flow and transport in the fissure system. A recurrence relationship is found which permits the coupling to the fissure model of a truncated series approximation to the analytical solution for one-dimensional diffusion in the surrounding porous matrix. The combined model has been calibrated using data collected between 1975 and 1979, and predictive simulations have been performed for the period from 1979 to 1986. The results are compared with data collected in 1986. The model is found to produce acceptable estimates of the spread of the chloride plume. The value of the model and the causes and potential implications of the observed model errors are discussed.

CONTAMINANT TRANSPORT IN HETEROGENEOUS POROUS MEDIA: A CASE STUDY. : 2. STOCHASTIC MODELLING

Abstract A stochastic modelling approach has been used to assess the uncertainty in predictions of contaminant migration in geological media surrounding a hypothetical buried waste repository in Bedfordshire, UK. In Part 1 of this two-part paper, the physical setting and the development of the conceptual and numerical models for groundwater flow and transport are described. In this part, the development of the stochastic modelling approach and the analysis of the set of Monte Carlo experiments used to characterise the uncertainty in the migration pathways are presented. Eleven Monte Carlo experiments have been performed. Nine have been undertaken to evaluate the effect of uncertainty in the parameters of the geostatistical model of hydraulic conductivity of the principal aquifer, and two have been performed to examine the influence of the location of the repository release zone. The study has demonstrated the use of the stochastic modelling approach in probabilistic risk assessments of the potential hazards arising from the burial of wastes.

CONTAMINANT TRANSPORT IN HETEROGENEOUS POROUS MEDIA: A CASE STUDY. : 1. SITE CHARACTERISATION AND DETERMINISTIC MODELLING

Abstract The aim of this study is to illustrate the contribution that geostatistical techniques can make to risk assessment for buried waste repositories. For a hypothetical repository site in Bedfordshire, UK, multiple realisations of the hydraulic properties of the principal aquifers are constructed. For each realisation a numerical model describing the groundwater flow regime at the site is used to predict a possible pattern of migration based on the available hydrogeological data. A statistical analysis of the results is carried out to provide quantitative measures of the uncertainty in the patterns of migration deriving from a lack of knowledge of the media properties. In the first part of this two-part paper, the formulation of the conceptual and numerical models describing the regional groundwater flow regime is presented. A detailed assessment of the site has been undertaken to determine the principal hydrogeological formations within which contaminant migration would occur and the principal controls governing regional groundwater flow and contaminant transport. From the resulting conceptual model a quasi-three-dimensional groundwater flow and transport model has been developed. Part 2 of the paper describes the simulation experiments carried out using this model and the analysis of the results.

Solution of multispecies transport in the unsaturated zone using a moving point method

Abstract Strongly advective contaminant transport can occur in highly permeable, unsaturated media such as gravel, fractured clay and soil-containing macropores. Standard numerical solution techniques are both highly inaccurate and unstable under such fast flow regimes. The moving point technique has been applied to contaminant transport in groundwater and shown to provide an accurate solution to the convection-dispersion equation for advection-dominated transport. In this paper, the technique has been adapted for the simulation of transport of multiple species through sorbing, dual porosity, unsaturated media. In order to model one-dimensional vertical transport, moving points are translated at the effective retarded velocities, simulating convection and retardation of each species. The contaminant concentrations in the immobile region and the changes in mobile concentrations due to diffusion, dispersion and chemical reactions are then calculated by finite difference methods on a fixed grid. The moving point model has been validated against available analytical solutions for solute transport. Comparisons have been made with standard finite difference solutions. The moving point method is shown to be considerably more accurate than the finite difference solutions for convection-dominated flow.