Plant ecology of lowland Alnus Glutinosa woodlands:The management implications of species composition, requirements and distribution

Wet woodlands have been recognised as a priority habitat and have featured in the UK BAP since 1994. Although this has been acknowledged in a number of UK policies and guidelines, there is little information relating to their detailed ecology and management. This research, focusing on lowland Alnus glutinosa woodlands, aimed to address this data paucity through the analysis of species requirements and to develop a methodology to guide appropriate management for this habitat for the benefit of wildlife. To achieve these aims data were collected from 64 lowland Alnus glutinosa woodlands and a review of the literature was undertaken to identify species associated with the target habitat. The groundflora species found to be associated with lowland Alnus glutinosa woodland were assessed in relation to their optimal environmental conditions (Ellenberg indicator values) and survival strategies (Grime CSR-Strategy) to determine the characteristics (Characters of a Habitat; CoaHs) and range of intra-site conditions (Niches of a Habitat; NoaH). The methodologies, using CSR and Ellenberg indicator values in combination, were developed to determine NoaHs and were tested both quantitatively and qualitatively at different lowland Alnus glutinosa sites. The existence of CoaHs and NoaHs in actual sites was verified by detailed quadrat data gathered at three Alnus glutinosa woodlands at Stonebridge Meadows, Warwickshire, UK and analysed using TWINSPAN and DCA ordination. The CoaHs and NoaHs and their component species were confirmed to have the potential to occur in a particular woodland. Following a literature search relating to the management of small wet woodlands within the UK, in conjunction with the current research, broad principles and strategies were identified for the management of lowland Alnus glutinosa woodland. Using the groundflora composition, an innovative procedure is developed and described for identifying the potential variation within a particular site and determining its appropriate management. Case studies were undertaken on distinct woodlands and the methodology proved effective.

Probability distribution modelling to improve stability in nonlinear MIMO control

We consider the direct adaptive inverse control of nonlinear multivariable systems with different delays between every input-output pair. In direct adaptive inverse control, the inverse mapping is learned from examples of input-output pairs. This makes the obtained controller sub optimal, since the network may have to learn the response of the plant over a larger operational range than necessary. Moreover, in certain applications, the control problem can be redundant, implying that the inverse problem is ill posed. In this paper we propose a new algorithm which allows estimating and exploiting uncertainty in nonlinear multivariable control systems. This approach allows us to model strongly non-Gaussian distribution of control signals as well as processes with hysteresis. The proposed algorithm circumvents the dynamic programming problem by using the predicted neural network uncertainty to localise the possible control solutions to consider.

Gas distribution in shallow packed beds

Packed beds have many industrial applications and are increasingly used in the process industries due to their low pressure drop. With the introduction of more efficient packings, novel packing materials (i.e. adsorbents) and new applications (i.e. flue gas desulphurisation); the aspect ratio (height to diameter) of such beds is decreasing. Obtaining uniform gas distribution in such beds is of crucial importance in minimising operating costs and optimising plant performance. Since to some extent a packed bed acts as its own distributor the importance of obtaining uniform gas distribution has increased as aspect ratios (bed height to diameter) decrease. There is no rigorous design method for distributors due to a limited understanding of the fluid flow phenomena and in particular of the effect of the bed base / free fluid interface. This study is based on a combined theoretical and modelling approach. The starting point is the Ergun Equation which is used to determine the pressure drop over a bed where the flow is uni-directional. This equation has been applied in a vectorial form so it can be applied to maldistributed and multi-directional flows and has been realised in the Computational Fluid Dynamics code PHOENICS. The use of this equation and its application has been verified by modelling experimental measurements of maldistributed gas flows, where there is no free fluid / bed base interface. A novel, two-dimensional experiment has been designed to investigate the fluid mechanics of maldistributed gas flows in shallow packed beds. The flow through the outlet of the duct below the bed can be controlled, permitting a rigorous investigation. The results from this apparatus provide useful insights into the fluid mechanics of flow in and around a shallow packed bed and show the critical effect of the bed base. The PHOENICS/vectorial Ergun Equation model has been adapted to model this situation. The model has been improved by the inclusion of spatial voidage variations in the bed and the prescription of a novel bed base boundary condition. This boundary condition is based on the logarithmic law for velocities near walls without restricting the velocity at the bed base to zero and is applied within a turbulence model. The flow in a curved bed section, which is three-dimensional in nature, is examined experimentally. The effect of the walls and the changes in gas direction on the gas flow are shown to be particularly significant. As before, the relative amounts of gas flowing through the bed and duct outlet can be controlled. The model and improved understanding of the underlying physical phenomena form the basis for the development of new distributors and rigorous design methods for them.

Development of a conceptual model of the soil-moisture-plant sub-system of the hydrological cycle

The soil-plant-moisture subsystem is an important component of the hydrological cycle. Over the last 20 or so years a number of computer models of varying complexity have represented this subsystem with differing degrees of success. The aim of this present work has been to improve and extend an existing model. The new model is less site specific thus allowing for the simulation of a wide range of soil types and profiles. Several processes, not included in the original model, are simulated by the inclusion of new algorithms, including: macropore flow; hysteresis and plant growth. Changes have also been made to the infiltration, water uptake and water flow algorithms. Using field data from various sources, regression equations have been derived which relate parameters in the suction-conductivity-moisture content relationships to easily measured soil properties such as particle-size distribution data. Independent tests have been performed on laboratory data produced by Hedges (1989). The parameters found by regression for the suction relationships were then used in equations describing the infiltration and macropore processes. An extensive literature review produced a new model for calculating plant growth from actual transpiration, which was itself partly determined by the root densities and leaf area indices derived by the plant growth model. The new infiltration model uses intensity/duration curves to disaggregate daily rainfall inputs into hourly amounts. The final model has been calibrated and tested against field data, and its performance compared to that of the original model. Simulations have also been carried out to investigate the effects of various parameters on infiltration, macropore flow, actual transpiration and plant growth. Qualitatively comparisons have been made between these results and data given in the literature.