research and development - defra, uk - science...

MINISTRY OF AGRICULTURE, FISHERIES AND FOOD CSG 15Research and Development

Final Project Report(Not to be used for LINK projects)

Two hard copies of this form should be returned to:Research Policy and International Division, Final Reports UnitMAFF, Area 6/011A Page Street, London SW1P 4PQ

An electronic version should be e-mailed to [email protected]

Project title CAMELOT (Coastal Area Modelling for Engineering in the LOng Term)

MAFF project code FD 1001

Contractor organisation and location

HR Wallingford Group LtdHowbery Park, WallingfordOxon OX10 8BA

Total MAFF project costs £ 1,245,146

Project start date 01/04/93 Project end date 31/03/00

Executive summary (maximum 2 sides A4)

CAMELOT 1/1.5Effective shoreline management requires a strategic approach that looks beyond the extent of a particular coastal defence scheme and the boundaries of a particular coastal authority. In order to design environmentally sound and cost-effective defences, coastal engineers must be able to predict long-term coastal morphodynamic behaviour over time spans of up to decades and over lengths of coastline of tens of kilometres corresponding to coastal sediment cells or sub-cells.

In 1993, MAFF Flood and Coastal Defence Division set up an ambitious seven-year programme of research into the development of methods for making predictions of coastal morphology over periods of up to decades. This research programme was given the name CAMELOT (Coastal Area Modelling for Engineering in the LOng Term). CAMELOT was implemented in two consecutive programmes. CAMELOT 1 (April 1993 - March 1998) and CAMELOT 1.5 (April 1998-March 2000). In CAMELOT 1, much effort was put into the development of numerical morphodynamic models that operated at medium-term timescales (from a few hours to several weeks). The work during CAMELOT 1.5 was aimed at consolidating the achievements of CAMELOT 1 through validation of morphodynamic models against field data, development of methodologies of model use, and compilation of a manual of coastal morphology for end-users.

MODELS AND METHODSThe central activity in CAMELOT has been the development and testing of numerical models for long-term morphodynamic predictions. This work has realised four main types of CAMELOT product specified below:

(i) Computer models(ii) Data sets and analysis techniques(iii) Methodologies(iv) Dissemination activities

Brief descriptions of each type of product are provided are provided in the main report, emphasising the role that each can play in engineering applications.

CSG 15 (Rev. 12/99) 1

Projecttitle

CAMELOT (Coastal Area Modelling for Engineering in theLOng Term)

MAFFproject code

FD 1001

Present morphodynamic prediction methods can be broadly divided into “bottom-up” predictors, and “top-down” predictors. Taking the example of coastal morphology, bottom-up predictors are numerical models based on equations describing the detailed physical processes of tidal flows, waves, sediment transport and sediment budget, while top-down predictors are qualitative-descriptive methods, empirical methods for extrapolating past behaviour to the future, and regime methods, that deal with the response of the morphodynamic-system as a whole without considering individual processes.

It became clear during CAMELOT that no one model approach or type of model would be sufficiently general to apply to all coastal situations and engineering applications. Accordingly, different models (and different modules within models) were developed for different timescales, coastal configurations, and sediment types. In general, CAMELOT models are based on the bottom-up approach and mostly, these use well-established theories, but in a few cases some theoretical development of these physical processes was undertaken. In such circumstances, this work was often supported in an EU MAST research project. An example was the development of a shingle transport formula which has since been incorporated into the COSMOS-2D and BEACHPLAN models (brief descriptions of these models are given in the main scientific report). Another example is the theoretical and modeling work of University of Wales at Bangor on boundary layer processes, undertaken within CAMELOT, which has been extended by HR to be applicable inside as well as outside the surf zone. This work has recently been incorporated into COSMOS-2D giving a more detailed representation of wave-driven currents and sediment transport rates, and the new version is known as COSMOS-k/t.

To appraise the predictive ability of these model developments, model validation exercises were also performed. There is a need to achieve a consensus about what is meant by a "validated model" in general, and for morphodynamic models in particular. In the context of CAMELOT, validation relates mainly to the accuracy of model predictions in comparison with detailed data sets with the emphasis on the usefulness or dependability of model predictions in real or simulated engineering design problems.

RESULTSCAMELOT established that there are a number of factors which make long-term predictions of coastal morphology much more difficult than for the medium term. These are (i) a substantial lack of long term morphodynamic data, (ii) the impossibility of knowing the future sequences of wave conditions (and other essentially random forcing factors such as wind), and (iii) the possibility of chaotic-type behaviour of the morphology that would render predictions of time sequences of morphology impossible beyond a certain "prediction horizon" even if the input sequence of forcing conditions were well known. These are analogous to problems faced in weather and climate prediction.

In addition to these gaps in our understanding, the high demand on computing time also limits the application of medium-term process-based (i.e. ‘bottom-up’) models, at least without adaptation, to yearly or decadal timescales. One approach to overcome this problem, attempted in CAMELOT, is by ‘filtering’ the small-scale processes so that the models work directly at longer timescales and use larger timesteps. One such technique, applied successfully in PISCES, was to use tidally averaged sediment transport rates. However, a difficulty with all such filtering approaches is that the range of applicability is restricted; in the above example, for instance, waves have to be assumed constant or negligible.

The validation of morphological models is also a much more difficult problem than for models of, say, wave or tidal conditions. This is because a) there are more variables to consider, b) there is a much larger inherent variability in sediment predictions, and c) predictions are of time evolution rather than steady-state conditions, and errors can be expected to accumulate with time. Many comparisons were undertaken in CAMELOT of medium-term model predictions with data. For long-term models, validation is even more of a problem. As well as the problems that concern medium-term model validation, there is a lack of long-term morphological data of suitable duration and resolution, and the problems of inherent predictability require to be addressed.

CONCLUSIONSPrediction of coastal morphology over periods up to decades is an essential part of many coastal engineering design studies. It is also a scientifically complex problem addressing issues which are at the forefront of research in other environmental scientific disciplines as well as coastal morphology. CAMELOT has been a particularly challenging project through its wide ranging nature, including the need for both a high standard of innovative research while at the same time providing practical tools and methodologies for use in engineering applications. The tasks of understanding the scientific issues, developing and testing models, and providing guidelines for applications are far from complete. Nevertheless, general guidance can be given about the correct models and methodologies for particular applications. At the present stage of development, it is important to recognise that the crucial element in any engineering application is the expertise of the investigators. The models should be regarded as tools that will give useful predictions provided they are used, and the results interpreted, with sufficient skill.

To aid investigators in using models for long term morphodynmaic predictions, this research has produced a manual that translates the findings of CAMELOT into practical steps. The manual is called ‘Coastal Morphology Modelling: A guide to model selection and usage’. In this manual an attempt has been made to describe and classify the coastal evolution models that are available, and their input data requirements, as an aid to choosing and establishing a model for a particular application. In addition, guidance is provided on some of the causes of uncertainty in such modelling and advise on methods for investigating this uncertainty. The importance of validation, and the use of sensitivity testing are discussed, and guidance provided.

CSG 15 (1/00) 2

Projecttitle


MAFFproject code

FD 1001

Scientific report (maximum 20 sides A4)

1. INTRODUCTION

1.1 CAMELOT 1 and CAMELOT 1.5Effective shoreline management requires a strategic approach that looks beyond the extent of a particular coastal defence scheme and the boundaries of a particular coastal authority. In order to design environmentally sound and cost-effective defences, coastal engineers must be able to predict long-term coastal morphodynamic behaviour over time spans of up to decades and over lengths of coastline of tens of kilometres corresponding to coastal sediment cells or sub-cells.

In 1993, MAFF Flood and Coastal Defence Division set up an ambitious seven-year programme of research into the development of methods for making predictions of coastal morphology over periods of up to decades. This research programme was given the name CAMELOT (Coastal Area Modelling for Engineering in the LOng Term). CAMELOT was implemented in two consecutive programmes. CAMELOT 1 (April 1993 - March 1998) and CAMELOT 1.5 (April 1998-March 2000). In CAMELOT 1, much effort was put into the development of numerical morphodynamic models that operated at medium-term timescales (from a few hours to several weeks). These models were based on representing physical processes that take place over wave and tidal periods. A number of such models were developed (for example, the profile model, COSMOS-2D, and the area model PISCES) and their predictions were compared against hydrodynamic and morphological data from a range of laboratory and field measurements. This was largely successful, but it was found that improvements to the representations of some processes, such as those in the swash zone, were needed.

The work during the first year of CAMELOT 1.5 was aimed at consolidating the achievements of CAMELOT 1 through validation of morphodynamic models against field data, development of methodologies of model use, and compilation of a manual of coastal morphology for end-users. During 1999-2000, CAMELOT 1.5 continued this work, following the broad theme of morphological predictions over yearly and decadal timespans, with specific emphasis on the topics described below. Contemporary with CAMELOT, HR has undertaken research projects in coastal morphodynamics at a more scientific level for the European Commission MAST (Marine Science and Technology) research programme. These projects include SASME, COAST3D and PACE.

The project involved HR Wallingford (HR), the Proudman Oceanographic Laboratory (POL), and, in a smaller role, the University of Wales, Bangor (UWB). The programme of activities was wide ranging, some activities involving collaboration between the participating institutes. A feature of this wide scope of CAMELOT is that it has provided a focus for integrating subsequent MAFF research projects concerned with more specific problems in coastal morphology. Examples are projects on the morphological behaviour of offshore sandbanks (HR Wallingford, Posford Duvivier, Delft Hydraulics, University of Twente and IMS Plymouth); coastal spits and nesses (Babtie Group and Birkbeck College); and work on new statistical analysis methods (Halcrow, HR Wallingford and University of Cambridge). This report provides an overview of the main coastal morphology modelling work by HR over the full course of the project up to its completion in March 2000.

1.2 CAMELOT RationaleThe overall aim of CAMELOT has been:

To provide an overall methodology, including the development and validation of a suite of numerical models, to enable practising engineers to predict accurately the evolution of coastal morphology on length scales of up to a coastal sub-cell and timescales of up to several decades.

At the outset of CAMELOT, the best approaches to achieve this overall objective were not fully understood. Work in CAMELOT has therefore followed a number of approaches, some of which have proved more successful than others. A large amount of flexibility has been needed, to allow some approaches to be abandoned if after initial work they are judged unlikely to be useful, while permitting new ideas to be followed which were not appreciated at the outset, but which arose as a result of increased understanding gained during the project.

An example is the use of numerical morphodynamic models based on detailed representations of the main physical processes of waves, tides and sediment transport. These models have proved useful for predictions of morphology over periods of days or weeks. At the start of CAMELOT it was thought that these models could be used for longer timespans, over years and decades, and the main problems would be ones of computational time and efficiency. However, during the project it became clear that there were more fundamental reasons why such models could not be used for these longer time periods, at least without substantial modification. Other approaches for predicting coastal morphology over these longer timescales had therefore to be devised, along with a fresh assessment of what features of coastal morphology can in principle be predicted.

The project management of CAMELOT reflected this need for flexibility. Milestones were set for tasks identifiable in detail at the outset, such as field data collection, laboratory experiments and some model validation exercises. For other tasks, such as development of new

CSG 15 (1/00) 3

Projecttitle


MAFFproject code

FD 1001

models or application of new data analysis techniques, objectives were set at a high level only. Assessment of progress in CAMELOT was made by review by an Advisory Committee, meeting at approximately nine-monthly intervals. The committee consisted of potential users of CAMELOT products from civil engineering consultants, the Environment Agency and academia, in addition to MAFF and the CAMELOT participants. As well as assessing progress, the committee provided advice on user requirements and the future course of the project.

1.3 Coastal engineering applications of CAMELOTCAMELOT products are designed as an aid to appraising any engineering scheme requiring knowledge of the future evolution of coastal morphology up to decadal timescales. Such applications would include:

Evolution of natural coastlines. Evolution of coastlines after installation of hard or soft defences. Effects of long term changes of sea level, wave climate or sediment supply. Movement of nearshore bed features such as bars and banks. Effects of dredging and dumping of sediment Infill and migration of channels. Exposure of submerged objects or natural features, e.g. pipelines or veneer beaches.

2. SCOPE OF CAMELOT

2.1 Geographical extentHR Wallingford’s role in CAMELOT covers the beach and nearshore zone. This typically refers to a coastal sediment cell or sub-cell in the longshore direction, and the zone from the top of the beach out to a closure depth (at which bed level changes are negligible, typically about 10-20m depth) in the cross-shore direction. POL's role covers the coastal zone further offshore, and entire sea areas (Figure 1).

Figure 1 Schematic coastal sub-cell and the nearshore region

2.2 TimescalesMany coastal engineering design studies need to consider morphodynamic responses over relatively short periods typically corresponding to an extreme event lasting a few days, and also the cumulative effects over years or decades covering the design lifetime of a scheme. For scientific reasons, separate modelling approaches are also needed for these two timescales. The following terminology has been used in CAMELOT:

Short Term Timescales from seconds to hoursMedium Term Timescales from hours to several weeksLong Term Timescales from several weeks to decades

Most practical applications of CAMELOT products will be for medium and long term timescales.

2.3 Coastline sediment types and structuresExposed sand or shingle/gravel coastlines have been considered in CAMELOT. Coasts which are predominantly muddy are subject to quite different sediment processes and have not been addressed. Saltmarshes, estuaries and inlets are also not considered. CAMELOT

CSG 15 (1/00) 4

Projecttitle


MAFFproject code

FD 1001

models include some representation of seawalls, groynes and offshore breakwaters, but small-scale morphodynamic processes such as scour around structures were not included.

2.4 Coastal parametersThe main focus of CAMELOT is the prediction of future coastal morphology. However, in order to do this, predictions of many other parameters are sometimes made, some of which will be useful in design studies. These parameters include nearshore tidal water levels and currents, nearshore wave conditions and wave-induced phenomena (such as set-up, longshore currents and undertow), and sediment transport rates (longshore and cross-shore).

2.5 CAMELOT productsThere are four main types of product delivered by the CAMELOT project:

(i) Computer models(ii) Data sets and analysis techniques(iii) Methodologies(iv) Dissemination activities

Brief descriptions of each type of product are provided in the folowing section, emphasising the role that each can play in engineering applications.

3. COMPUTER MODELS (CAMELOT PRODUCT 1)It became clear during CAMELOT that no one type of model would be sufficiently general to apply to all coastal situations and engineering applications. Accordingly, different models (and different modules within models) have been developed for different timescales, coastal configurations, and sediment types. There have been three stages in the development of these CAMELOT models; Theoretical development, model development and model validation:

3.1 Theoretical developmentPresent prediction morphodynamic prediction methods can be broadly divided into “bottom-up” predictors, and “top-down” predictors. Taking the example of Coastal Morphology, bottom-up predictors are numerical models based on equations describing the detailed physical processes of tidal flows, waves, sediment transport and sediment budget, while top-down predictors are qualitative-descriptive methods, empirical methods for extrapolating past behaviour to the future, and regime methods, that deal with the response of the morphodynamic-system as a whole without considering individual processes.

The bottom-up predictors have the problem that computer-time can become prohibitive for direct long-term simulations. Even then it is not known whether continuous integration of short-term processes will give correct predictions of long-term behaviour, because of incomplete representation of processes and accumulation of errors. bottom-up morphodynamic models have the same basic structure involving calculations of waves, currents, sediment transport and bed level changes, and then repeating these calculations at time intervals using updated bed levels, until the required duration of the simulation is reached. The procedure is shown in the flowchart in Figure 2.

The top-down predictors are simple and quick to use but have the problem that many are based on unproven hypotheses and they do not as yet yield the time-scales of changes. Also past performance of morphodynamic is not necessarily a guide to the future, especially if major changes (e.g. beach nourishment) are planned. The shortcomings of the bottom-up and top-down predictors have suggested the development (still in its infancy) of hybrid models in which a top-down method is essentially calibrated (either at intervals in time, or probabilistically) by a bottom-up method. The hybrid approach appears to offer the optimum way forward for improved prediction methods.

CAMELOT models are based, to a greater or lesser extent, on the bottom-up approach. Mostly, these use well-established theories, but in a few cases some theoretical development of these physical processes was undertaken. In such circumstances, this work was often supported in an EC MAST research project. An example was the development of a shingle transport formula which has since been incorporated into the COSMOS-2D and BEACHPLAN models (brief descriptions of these models are given in the following section). Another example is the theoretical and modelling work of UWB on boundary layer processes, undertaken within CAMELOT, which has been extended by HR to be applicable inside as well as outside the surf zone. This work has recently been incorporated into COSMOS-2D giving a more detailed representation of wave-driven currents and sediment transport rates, and the new version is known as COSMOS-k/t.

CSG 15 (1/00) 5

Projecttitle


MAFFproject code

FD 1001

Figure 2 General procedure for “bottom-up” morphodynamic models

3.2 Model developmentA central activity in CAMELOT has been the development and testing of numerical models. In some cases this has involved extensions and complete overhauls of previously existing models, and in other cases new models have been developed from scratch. In all cases the aim has been to develop these models to a "validated research" level, in which the software code has been written, tested and compared against data sets. The tasks of formal software testing and installation of user interfaces such as GUIs (Graphical User Interface) were outside the scope of CAMELOT, and would be major undertakings for models of this sort. Nevertheless, the models incorporate some user interface features such as keyworded input and graphical post-processing.

The table below lists the principal morphodynamic models and their main characteristics: Items in the table are explained in subsequent paragraphs:

MODEL SCALE TYPE SPEED DEVELOPMENT

COSMOS-2D M-T Profile Fast Commercial Use

COSMOS-k/t M-T Profile Medium Research

COSMOS-3D M-T Multiple Profile Medium Research

PISCES M-T Area Slow Commercial Use

TURBO-PISCES L-T Area Medium Research

BEACHPLAN L-T Planshape Fast Commercial Use

Model. Name of the model developd at HR Wallingford

Scale. Medium term (M-T) or long term (L-T).

Type. Profile models consider a cross-shore slice of the coast, and assume the coastline and depth contours are uniform longshore. This effectively removes the longshore dimension from the calculations, thereby permitting a more detailed representation of cross-shore processes than Area models, while retaining high computational speed. Multiple profile models allow for coastlines to be gently curving, and operate by stacking several profiles along the coast. The profiles are linked with each other via morphodynamic changes, but are independent of each other in calculations of waves and currents. Area models have no restrictions in cross-shore and longshore directions, and consequently can be applied to any coastline geometry. However, they use relatively large amounts of computing time. Planshape models focus on longshore processes and determine the evolution of a representative coastline contour.

Speed. As a rough guide to computing speed, a few hours of processing time on a modern workstation or PC is typically sufficient for a simulation of several years for a planshape model, but only a few days for an area model. However, the speed of each model depends very

CSG 15 (1/00) 6

Projecttitle


MAFFproject code

FD 1001

strongly on factors such as the number of model grid points, the duration of the simulation, the mobility of the bed, and the computer hardware. An approximate indication of their relative speeds is given in the table: "Fast" is about ten times quicker than "Medium" which in turn is about ten times quicker than "Slow".

Development. COSMOS-2D, PISCES and BEACHPLAN have been used in commercial studies by HR and other organisations. COSMOS-k/t, COSMOS-3D and TURBO-PISCES are at the stage of research models undergoing development and testing.

An important task for the coastal engineer is to predict the medium-term response of the beach profile to a given sequence of environmental forcing (such as a storm, change in sea level etc.) or to human interference. The best approach at present is the use of computational coastal profile models of the waves, currents and sediment dynamics. The physical processes involved are complex, and a number of models have been developed by various organisations which incorporate these processes, to varying degrees of completeness.

The profile model most frequently used at HR Wallingford is COSMOS-2D which is a process based model, meaning that many of the physical processes involved in the beach profile evolution are modelled in some manner. The present sediment transport module used within COSMOS-2D is based on Bailard’s energetics sediment transport formula, which was founded on the assumption that the sediment transport rate is proportional to the instantaneous rate of energy dissipation. Due to its accessibility the formula has been widely used. However, it does not provide a good physical representation of sediment transport mechanics and it has been shown that the formulation in some cases performs very badly.

To address some of these shortcomings a new shingle transport formula has been incorporated into COSMOS-2D. They found that the new formula worked well for longshore shingle transport, but not for cross-shore transport. This is because the hydrodynamic process description, within the sediment transport module of COSMOS-2D, is not sufficiently detailed to drive the new formula in the cross-shore direction.

In CAMELOT 1.5 a new sediment transport module was added to COSMOS-2D. This module was based on a more detailed hydrodynamical description, which, in turn, enables a more physics-based approach to be used for the sediment transport calculations within the surf-zone. The work in the module undertaken at HR extends the non-breaking wave, sediment transport model supplied by CAMELOT partners Dr A G Davies and Dr Z Li of the University of Wales, Bangor (UWB). The model name, COSMOS-k/t, arises from its origin within the COSMOS-2D framework. The ‘k’ refers to the use of the turbulent kinetic energy (k) resolving turbulence closure model. The ‘t’ refers to the time-varying nature of solution. Calculations are made throughout a wave cycle, rather than only using cycle averaged processes [43].

Box 1 Development of beach profile models in CAMELOT

3.3 Model validationModel results of morphodynamic evolution have been compared against a range of data sets, with particular emphasis on field data and laboratory data at prototype scale. These comparisons with data have been incorporated into the model development process so that strengths and weaknesses can be identified and further development undertaken to address the weak points. Tests with profile models have been carried out for both strong wave conditions, for which beaches generally show a degree of erosion, and more moderate wave conditions, for which there is often a gradual onshore movement of sediment and accretion on beaches. Model predictions are generally better for erosive conditions. Tests with area models have been carried out for more complex coastal configurations, such as a river outflow into a straight coastline, and for a beach protected by one and two offshore breakwaters. The simulation periods were a few days, and some characteristic morphological features were predicted, such as oblique longshore migrating bars and tombolo formation behind offshore breakwaters. Long term simulations with planshape models show evolution towards typical equilibrium planshapes, and good predictions have been made of bulk longshore transport of sediment. All these types of models have been used in commercial applications.

A different type of model comparison exercise concerned data collected off the Holderness coastline in Yorkshire in conjunction with the NERC LOIS (Land-Ocean Interaction Study) programme. In this case the aim was to demonstrate how the PISCES area model could be driven with wave, current and water level data supplied at the model's offshore and lateral boundaries from POL's larger-scale models. Comparisons were then made of PISCES predictions of waves and currents with measurements made by POL inside the modelled region. During the CAMELOT programme significant advances were made in numerical modelling of coastal sediment transport and useful models were developed.

The complexity of the processes and models makes it difficult to assess to what extent morphodynamic models can be validated. This is especially the case since the word "validation" has different meanings within the software industry, academic research and engineering consultancy. In the software industry, validation refers to well-defined procedures for model development, testing and documentation. In academic research, validation mainly relates to the accuracy of model predictions of physical parameters in comparison with detailed data sets. In engineering consultancy, the emphasis is on the usefulness of model predictions in real or simulated engineering design problems. There is a need to achieve a consensus about what is meant by a "validated model" in general, and for morphodynamic models in particular

CSG 15 (1/00) 7

Projecttitle


MAFFproject code

FD 1001

One of the validation studies in CAMELOT considered the 2DH coastal area model PISCES to the stretch of coastline between Happisburgh and Winterton on the East Anglian coast following the construction of artificial reefs (see Figure opposite)

PISCES is a state-of-the-art, fully-interactive coastal area modelling framework, capable of simulating the various processes of wave propagation, tidal and wave-driven current distribution and the associated sediment transport, and the corresponding seabed evolution in complex coastal areas. The aim of the validation was to investigate the ability of a model such as PISCES to accurately predict the actual evolution of the coastline at Happisburgh that occurred prior to the construction of the artificial reefs.

From the morphodynamic model comparisons with data, it is apparent that the accuracy of model predictions (of time sequences of bed levels) is substantially less than, for example, wave or current parameters. This arises from a combination of the intrinsic variability of sediment processes, the complexity of combining a range of separate hydrodynamic, sediment and morphodynamic parameters, and the unavoidable accumulation of prediction errors as calculations proceed through time [42].

Box 2 Area Model Validation in CAMELOT

4. DATA SETS AND ANALYSIS TECHNIQUES (CAMELOT PRODUCT 2)During CAMELOT, measurements have been made of hydrodynamic, sediment and morphodynamic parameters in laboratory and field experiments. These include measurements by POL of waves, currents and sediment concentrations off Holderness and in the Taw estuary in Devon, and laboratory experiments by HR in the UK Coastal Research Facility to investigate the factors affecting the initiation of wave breaking. These data sets provide information to improve theoretical formulations used in models, to validate numerical models, and as scientific knowledge for direct use in practical applications and as a basis for further research.

In addition to making specific measurements to obtain data sets, many pre-existing data sets have been used directly or have been reanalysed. Several laboratory flume data sets, especially those at prototype scale, have been used to validate morphodynamic profile models. In some cases, this has been done as an intercomparison exercise with other European institutes as part of an EU MAST programme (such as the G6M and G8M programmes). A number of field data sets from the UK and overseas have also been used for validation over medium timescales. For yearly and decadal timescales, however, there exist very few data sets with adequate temporal resolution (one month or less) as well as long term coverage. The most useful data sets have been a 30-year set of beach elevation measurements from the central Lincolnshire coastline between 1959 and 1990 with measurements at monthly intervals, and a 10-year set of beach and nearshore bed level measurements from Duck, North Carolina, USA, sampled at 14-day intervals. These data sets have been used to deduce generic modes of morphodynamic behaviour, including limits of predictability for process-based models. In particular, they have revealed information about long term trends on different timescales, and shorter-term variations about these trends.

Some new methods of analysing morphodynamic data have been used in addition to established techniques of trend line fitting, spectral analysis etc. These new methods include empirical orthogonal eigenfunction techniques and fractal analysis.

5. METHODOLOGIES (CAMELOT PRODUCT 3)An important lesson of CAMELOT is that morphodynamic models should not be used naively in a "black box" mode. A good understanding of the underlying scientific concepts and a clear idea of the type of information required for a particular engineering application are needed, although knowledge of details of model coding is not necessary. Much emphasis has therefore been put on the methodologies of using models and tackling morphodynamic problems in general.

5.1 How should morphodynamic models be used?There are a range of questions that need to be addressed before morphodynamic modelling is undertaken.

Should morphodynamic modelling be used at all? A number of coastal engineering morphodynamic problems do not require any morphodynamic modelling. This may be because the scale of the problem does not merit a morphodynamic approach (e.g. an assessment of the sediment budget of a coastal sub-cell as a whole needs a knowledge of sediment fluxes through the boundaries of the sub-cell, but not how the morphology evolves within it). Alternatively, the type of information from non-morphodynamic modelling may be sufficient (e.g. it may be possible to estimate medium term trends of morphology over a sufficiently large area by analysis of past behaviour, or to use modelling solely to give an initial calculation of erosion or accretion rates). Another reason is that although morphodynamic modelling may be appropriate and would supply added value, the benefit is sufficiently marginal not to justify the extra time and/or expense.

CSG 15 (1/00) 8

Projecttitle


MAFFproject code

FD 1001

Which type of model should be used? The choice of model type (or more than one model type used in conjunction) depends on several factors. These include the type and accuracy of information required, the time and length scales of the problem, coastal geometry (including structures), sediment types, the nature of hydrodynamic forcing conditions, and the data available to drive and calibrate the model. The table of the main properties of morphodynamic models shown earlier provides general guidance for the choice of models. However, each application will have its unique factors, and the most appropriate approach, whether or not modelling is suitable, should be assessed individually taking all factors into account and not just on the basis of general guidelines.

What mode of operation of the model should be used? Once a model is chosen, there are usually a number of ways in which it can be used. Some models can produce results in different forms, such as time sequences of bed levels, or statistical information such as mean, maximum and minimum envelopes of bed levels. Similarly, input may be in the form of time sequences of wave conditions or as a scatter diagram of wave climate data. In some cases calibration of variables can be carried out if sufficient suitable data are available. Another possibility is to carry out sensitivity analyses to see how sensitive final results are to variations in the input conditions or internal parameter settings. Special considerations are needed for long term (monthly, yearly and decadal) predictions. These are considered in the next section.

5.2 Long term predictionsThere are a number of factors which make long-term predictions of coastal morphology much more difficult than for the medium term. These are (i) a substantial lack of long term morphodynamic data, (ii) the impossibility of knowing the future sequences of wave conditions (and other essentially random forcing factors such as wind), and (iii) the possibility of chaotic-type behaviour of the morphology that would render predictions of time sequences of morphology impossible beyond a certain "prediction horizon" even if the input sequence of forcing conditions were well known.

Long term trends of coastal morphology. In some cases the large-scale properties of coastal morphology can be predicted over long timespans (to a certain level of accuracy depending on the application). For example, the change in the mean position of the coastline following the construction of groynes, or responding to long-term systematic changes in wave climate and water level, can be predicted using planshape models. Another example is the use of medium-term models in which the shorter term processes have been filtered so that the models are able to operate over longer timespans. An illustration of this is an adaptation of the PISCES model to use tidally averaged sediment transport fluxes. Simulations lasting many years have been carried out using this filtered version of PISCES. In all such filtering methods, however, the range of applications becomes more restricted; in the above example wave conditions are assumed constant or negligible, and cumulative intra-tidal morphodynamic effects are ignored.

Medium term variations about long term trends. The medium term variations about these average trends are much more difficult to predict, but they are, however, of considerable practical importance because they are often substantially larger than the trend changes. Such applications would include calculations of the proportion of time that buried objects or veneer beaches are exposed, the probabilities of occurrence of beach levels immediately before severe events, or the proportion of time that sensitive areas at the back of a beach are eroded below a certain level. The data set from Lincolnshire, in which beach levels at many locations were sampled over a 30-year period at monthly intervals, illustrates the difficulties posed by these medium term variations of beach levels. Typically the data showed vertical accretion/erosion rates between +1 and -3cm/year over the full 30 year period. However, over periods of a few months or less, beach levels often varied by a metre or more. Furthermore, accretion/erosion trends over periods shorter than 30 years showed no relation to each other, so one could not, for example, assume trends over a previous five or ten year period would continue for the next five or ten years.

Referring back to Items (ii) and (iii) in Section 5.1, predictions of time sequences of medium term variations about long term trends will not be possible with any type of model beyond a certain time limit. This is analogous to the time limits of predictability in weather forecasting. However, it may be possible to predict certain types of statistical information, such as probabilities of occurrence of beach levels (in the sense of medium term variations about long term trends).

Chaotic behaviour and limits to predictability. A detailed analysis of the Lincolnshire and Duck data sets was carried out to determine if there was any evidence for the chaotic mechanism, and if so, what the predictability limits were for morphodynamic models. Both data sets revealed that there was evidence for such a mechanism, and the predictability limits were about 12-24 months at Duck, and somewhat less (6-10 months) at Lincolnshire. These limits apply to the surf and intertidal zones, and there is evidence that different limits apply further offshore or on the upper parts of a beach. No models presently exist which can demonstrate this type of chaotic behaviour over these timescales. This is an issue which is at the cutting edge of a number of environmental sciences and is encountered, for example, in climate prediction.

Sequencing of wave conditions. In contrast, however, a relatively straightforward methodology can be adopted, using existing models, to address the problem that future sequences of wave conditions are unknown (although the future wave climate can in many cases be predicted with reasonable accuracy). The approach involves running a morphodynamic model many times (typically about 30 times), each run using the same starting conditions, except that the sequence of input wave data has been reordered. The final results (in terms, for example, of beach levels or coastline positions) are processed to give mean, maximum, minimum, and other probability-of-exceedance values, across the set of model runs. This provides a calculation of the statistics of the medium term variability of beach levels or coastline positions as well as the long term trends. Existing models can be used, and the only penalty is the additional computing time for the set of runs. The methodology has been used with profile models for timespans of several weeks, and for planshape models for several years. In

CSG 15 (1/00) 9

Projecttitle


MAFFproject code

FD 1001

the latter case, the chaotic mechanisms are ignored, but this remains the best presently available method for addressing the problem of prediction of medium term morphodynamic variability about long term trends.

Wave chronology can have a significant effect on longt-term model predictions. This figure shows the envelope of changes in a straight beach following construction of a groyne, showing the effects of wave chronology using a planshape model. In this example the same beach was forced with the same five year wave sequence, but the sequence of events was re-ordered for each of the forty model runs. For each wave sequence, the maximum (most seaward) position of the beach during each of the 40 5-year sequences. One can then draw 40 separate envelopes of maximum beach position. The shaded area shows the region covered by these 40 envelopes. The continuous line shows the envelope of maximum beach positions from one of the wave sequences. Note that the wave conditions were distributed approximately evenly about the groyne direction, accounting for similar accretion patterns on both sides of the groyne [31].

Box 3 Effects of Wave Chronology

5.3 Model OutputWhat types of model output are possible? The foregoing discussion illustrates the need for a careful assessment of what type of information is needed for a particular engineering application. This should take account of both the nature of the application and what is possible to achieve scientifically. In particular, the use of medium term models for long term simulations (without filtering of medium term processes) is strongly discouraged. The use of multiple model runs of long-term models is recommended. Some generic types of model output that are possible are summarised below:

Time sequences of bed levels for medium term predictions Time sequences of bed levels of mean trends for long term predictions Probabilities of exceedance of bed levels for medium-term variations about long term trends at a future time Probabilities of exceedance of bed levels for medium-term variations about long term trends over a future period of time The time taken for a specified morphological state to be reached, with a specified probability of occurrence (e.g. for a channel to infill

to a given level, or a coastline contour to retreat by a given amount)

How should model results be interpreted? Decisions about how model results should be interpreted need to be made at the start of a design study, and should relate to a clear understanding of the type of information needed from the model. These decisions are then reflected in the choice of model and its mode of use. Examples of the types of information that are possible from morphodynamic models are listed above.

6. DISSEMINATION ACTIVITIES (CAMELOT PRODUCT 4)Throughout the project, progress has been disseminated through technical papers and reports, presentations of work at conferences and seminars, publicity leaflets and articles in the MAFF Flood and Coastal Defence newsletter, and a one-day workshop at HR in June 1997 with about 20 invited potential users of CAMELOT products. A full list of dissemination activities is provided at the end of this report. The manual “Coastal Morphology Modelling – a guide to model selection and usage” is a particularly important item of dissemination. It adds detail to many of the issues highlighted in this summary report.

The use of numerical methods to predict future shoreline evolution is an increasingly important aspect of coastal defence strategy in the UK. In recent years, the number and complexity of such methods has increased considerably, although many are still scientific research tools rather than being used routinely for practical studies.

In this manual an attempt has been made to describe and classify the coastal evolution models that are available, and their input data requirements, as an aid to choosing and establishing a model for a particular application. In addition, the manual provides guidance on some of the causes of uncertainty in such modelling and advised on methods for investigating this uncertainty. The importance of validation, and the use of sensitivity testing are discussed, and guidance provided. Increasingly, probabilistic methods are being used in coastal management studies, for example in formal assessments of risk, vulnerability and resilience. This report describes how such techniques can also be used in the prediction of coastal evolution, and results presented in a probabilistic format. Some examples from recent research and site-specific studies have been presented to demonstrate some of the above points.

It is concluded that the intelligent use of these models in a practical context still requires a high level of specialised knowledge. We are a long way from having “expert systems” in which coastal morphological models can be incorporated into wider management modelling suites and be used by non-specialists. At the present stage of development of the subject, therefore, it seems that the appropriate expertise of the investigators is a very important factor when undertaking coastal engineering or management applications of such models [44].

Box 4 Manual of Coastal Morphology Modelling

CSG 15 (1/00) 10

Projecttitle


MAFFproject code

FD 1001

7. FUTURE DEVELOPMENTThere are still many outstanding issues requiring future research and development. Among the most important are:

7.1 Morphodynamic data and analysis techniquesProgress in understanding and modelling nearshore morphodynamic behaviour is hampered by a lack of data of sufficient spatial and temporal coverage and resolution. For medium term applications, the most useful data sets are from large intensive purpose-designed field campaigns in which the morphodynamic behaviour is monitored in detail at one coastal site over a period of a few weeks, and in which measurements of the main physical processes are simultaneously made over the same period at many locations at the site. Recent and planned campaigns reflect these requirements; examples are SANDYDUCK at the Duck site in the USA, COAST3D at sites in the Netherlands and UK, and INDIA at a site in Portugal.

More important, however, is the lack of suitable long term data. Most presently available data sets do not have sufficient temporal coverage (several years or decades) or resolution (monthly or less). The Lincolnshire and Duck data sets are notable exceptions. In view of the costs and practical difficulties of regular large-scale bathymetric monitoring by in situ methods, there will be an increasing role for remote sensing techniques from the ground, aircraft or satellite. Experience is being gained with technologies for measuring nearshore bathymetry, such as LIDAR or the ARGUS photogrammetric system. Remote sensing has the potential for large spatial coverage with high resolution, which would not be practicable with in situ methods.

7.2 Physical processes and model developmentMedium term physical processes are relatively well understood, although there are specific aspects that require further investigation, for example, for swash zone processes and for onshore sediment movement during moderate wave conditions. Much less is known about the morphodynamic mechanisms at yearly and decadal timescales, and models are at a correspondingly earlier stage of development. More data is needed, and generic issues such as limits of predictability require investigation.

7.3 MethodologiesFor medium term predictions it is recommended that, in any application, tests are carried out to investigate the sensitivity of morphodynamic predictions with respect to input data and internal parameter settings. Further systematic studies of sensitivity are needed. Most long term predictions are envisaged to be in the form of probabilities of, for example, beach levels or coastline contour positions. There needs to be an assessment of the most useful parameters and methods of presentation of this type of information. A potential methodology for assessing predictability limits is by "ensemble" modelling, in which several model runs are made with slightly varying inputs. However, this approach will require models which can reliably demonstrate chaotic-type behaviour. A common theme in all the above methodologies is the use of multiple model runs.

7.4 Model validationComparisons of medium term model predictions need to be carried out with field data, and if possible at sites where there have been recent engineering schemes with subsequent monitoring. However, a consensus is needed amongst model developers and users about what constitutes model validation.

7.5 Integration into management toolsThe probabilistic nature of long term predictions lends itself to incorporation into a wider modelling framework involving probabilistic predictions of other engineering design parameters. An example would be to incorporate the probabilistic predictions of beach levels with those of extreme waves and water levels. Morphodynamic models can also be seen as forming an element of, or interfacing with, software- such as GIS, site-specific databases or decision support tools.

8. CONCLUSIONSPrediction of coastal morphology over periods up to decades is an essential part of many coastal engineering design studies. It is also a scientifically complex problem addressing issues which are at the forefront of research in other environmental scientific disciplines as well as coastal morphology. CAMELOT has been a particularly challenging project through its wide ranging nature, including the need for both a high standard of innovative research while at the same time providing practical tools and methodologies for use in engineering applications. The tasks of understanding the scientific issues, developing and testing models, and providing guidelines for applications are far from complete. Nevertheless, general guidance can be given about the correct models and methodologies for particular applications. At the present stage of development, however, it is important to recognise that the most important element in any engineering application is the expertise of the investigators. The models should be regarded as tools that will give useful predictions provided they are used, and the results interpreted, with sufficient skill.

CSG 15 (1/00) 11

9. CAMELOT: PAPERS, REPORTS AND PUBLICITY ARTICLES1 Southgate H N and Nairn R B (1993), "Deterministic Profile Modelling of Nearshore Processes. Part 1. Waves and Currents",

Coastal Engineering, Vol 19, pp27-56

2 Nairn R B and Southgate H N (1993), "Deterministic Profile Modelling of Nearshore Processes. Part 2. Sediment Transport and Beach Profile Development", Coastal Engineering, Vol 19, pp57-96

3 Chesher T J, Wallace H M, Meadowcroft I C and Southgate H N (1993), "PISCES. A Morphodynamic Coastal Area Model. First Annual Report", HR Report SR 337 (Prepared under MAFF Contract CSA 2090 but closely allied to work in CAMELOT)

4 Southgate H N (1993), "European Medium-Scale Coastal Field Measurements. Report on a Workshop at HR Wallingford on 27 and 28 May 1993", HR Report SR 363 (Prepared under MAFF Commission FD0801 but closely allied to work in CAMELOT)

5 Southgate H N (1993), "The Effects of Wave Event Sequencing on Long-Term Beach Response", Extended abstract for Large Scale Coastal Behaviour '93, Florida, USA

6 Chesher T J (1993), "Tidal Schematisation in Morphodynamic Area Models", Extended abstract for Large Scale Coastal Behaviour '93, Florida, USA

7 Wallace H M (1993), "Coastal Sand Transport and Morphodynamics. A Review of Field Data", HR Report SR 355 (Prepared under MAFF Commission FD0801 but closely allied to work in CAMELOT)

8 HR Wallingford and Proudman Oceanographic Laboratory (1993), "Predicting Long Term Coastal Morphology", MAFF Flood and Coastal Defence Newsletter, No 4

9 Southgate H N and Wallace H M (1994), "Breaking Wave Persistence in Parametric Surf Zone Models", Proc. Coastal Dynamics '94, Barcelona, Spain

10 Wallace H M and Chesher T J (1994), "Validation of Coastal Morphodynamic Models with Field Data", Proc. Coastal Dynamics '94, Barcelona, Spain

11 Soulsby R L, Southgate H N, Thorne P D and Flather R A (1994). "Modelling Long Term Coastal Morphology: The CAMELOT project", Proc. MAFF Conference of River and Coastal Engineers, Loughborough University of Technology

12 Chesher T J (1995), "Numerical Morphodynamic Modelling of Keta Lagoon', Proc. Coastal Dynamics '95, Gdansk, Poland

13 Chesher T J, Price D M and Southgate H N (1995), "The Application of Medium-Term Morphodynamic Models to the Long-Term", Seminar Paper for 'Morphology of Rivers, Estuaries and Coasts', HYDRA 2000, IAHR, London

14 Southgate H N and Beltran L M (1995), "Time Series Analysis of Long-Term Beach Level Data from Lincolnshire, UK", Proc. Coastal Dynamics '95, Gdansk, Poland

15 Southgate H N (1995), "The Effects of Wave Chronology on Medium and Long Term Coastal Morphology", Coastal Engineering, Vol 26, No 3/4, pp251-270

16 Southgate H N (1995), "Prediction of Wave Breaking Processes at the Coastline", Book article in Advances in Fluid Mechanics, Ed. M. Rahman, Pub. Computational Mechanics Publications.

17 Damgaard J S and Soulsby R L (1996), "Longshore Bed-Load Transport", Proc. 25th ICCE, Orlando, USA

18 Damgaard J S, Stripling S and Soulsby R L (1996), "Numerical Modelling of Coastal Shingle Transport", HR Report TR 4

19 Gillott P W K and Southgate H N (1996), "A One Component Model for Low Frequency Waves", HR Report TR 1. (Prepared under MAFF Commission FD0202 but closely allied to work in CAMELOT)

20 Stripling S and Damgaard J S (1996), "COSMOS-2D: Numerical Model Evaluation using NOURTEC (Terschelling) Data", HR Report TR 25. (Prepared under MAFF Commission FD08 but closely allied to work in CAMELOT)

21 Southgate H N, Beresford P J, Channell A R and Adams R (1996), "Wave Breaking Experiments in the UK Coastal Research Facility", HR Report TR 2

22 Southgate H N and Beltran L M (1996), "Self-Organisational Processes in Beach Morphology", Proc. Conference on Physics of Estuarine and Coastal Seas, The Hague, The Netherlands

Projecttitle


MAFFproject code FD 1001

23 Southgate H N and Stripling S (1996) "Measurements of Wave Breaking in the UK Coastal Research Facility", Proc. 25th ICCE, Orlando, USA.

24 Weare T J (1996), "Computational Hydraulics in 1997", Proc. 2nd Int. Conf. on Hydrodynamics, Hong Kong (contains contributions on Modelling of Nearshore Morphology by H N Southgate and Coastal Physics by R J S Whitehouse)

25 Cesare M and Bunn N P (1997), "The Development of COSMOS-3D", HR Report TR 28.

26 Beltran L M and Southgate H N (1997), "TURBO-PISCES. An Efficient Coastal Area Morphodynamic Model", HR Report TR 11

27 Nicholson J, Broker I, Roelvink J A, Price D M, Tanguy J-M, and Mory M (1997), "Intercomparison of Coastal Area Morphodynamic Models", Coastal Engineering, Vol 31, pp97-124

28 Pechon P, Rivero F, Johnson H, Chesher T J, O'Connor B A, Tanguy J-M, Karambas T, Mory M, Hamm L (1997), "Intercomparison of Wave-Driven Current Models", Coastal Engineering, Vol 31, pp199-216

29 HR Wallingford and Proudman Oceanographic Laboratory (1997), "CAMELOT - Long Term Coastline Evolution", MAFF Flood and Coastal Defence Newsletter, No 10

30 Southgate H N (1997), "Non-Linear Methods for Analysis of Long-term Beach and Nearshore Morphology", 27th IAHR Congress, San Francisco, USA

31 Southgate H N and Capobianco M (1997), "The Role of Chronology in Long-Term Morphodynamics. Theory, Practice and Evidence", Proc. Coastal Dynamics '97, Plymouth, UK

32 Stripling S and Damgaard J S (1997), "Improvement of Morphodynamic Predictions in COSMOS-2D", HR Report IT 443

33 Stripling S (1997), "Numerical Modelling of Q3D Processes", HR Report IT 445

34 Stripling S (1997), "PISCES Numerical Model Evaluation Using SCAWVEX (Holderness) Data", HR Report TR 30

35 Moeller I (1997), “Statistical Methods for Characterising Long-Term Beach Profile Morphodynamics: Theory and Applications to Beach Profile Data from Duck, North Carolina”, HR Report IT 454.

36 Millard T K and Eguiagaray M (1998), "Enhancements to the Beach Planshape Mathematical Model (BEACHPLAN)", HR Report IT 454.

37 Damgaard J S (1998), "Numerical Schemes for Morphodynamic Simulations", HR Report IT 459

38 Peet A H and Damgaard J S (1998), "COSMOS-k/t. Modelling Cross-Shore Sediment Transport within the Surf-Zone" HR Report TR 38

39 Damgaard J S and Peet A H (1998), "Numerical Modelling of Coastal Morphology. A Study of Model Sensitivity and Other Aspects of Engineering Importance", HR Report TR 55

40 Southgate H N (1998), "The CAMELOT Programme for Predicting Long Term Coastal Morphology. An Overview from HR Wallingford", 33rd MAFF Conference of River and Coastal Engineers, Keele University

41 Southgate H N and Moeller I (1998), "Fractal Properties of Coastal Profile Evolution at Duck, North Carolina, USA", submitted to Journal of Geophysical Research

42 Damgaard J, Chesher J, Hall L, (1999). “Simulation of the sediment budget for the Happisburgh to Winterton Reefs”. PISCES application study, HR Wallingford Report TR 83, May 1999

43 Peet A and Damgaard J, (1998). “COSMOS-k/t, Modelling cross-shore sediment transport within the surf zone”, HR Wallingford Report TR 38, March 1998

44 Southgate H, Brampton, A (2000). “Coastal Morphology Modelling – A guide to model selection and usage”, HR Wallingford Report SR 570, April 2000

13

Projecttitle



45 Southgate H, Lopez de San Roman Blanco (1999), Predicting Long-term Coastal Morphology, in “Coastal Management: Integrating science, engineering and management”, Proceedings of the international conference organised by the Institution of Civil Engineers and held in Bristol, UK, September 1999, edited by C. A. Fleming

46 Southgate H N and Moeller I, “Fractal properties of coastal profile evolution at Duck, North Carolina”, J. Geophys. Res., Vol. 105, pp11489-11507, May 2000

47 Southgate H N “Guidelines for error assessment in models of long term coastal morphology”, Proc. IAHR Conference on River, Coastal and Estuarine Morphodynamics, Genoa, Italy, Sept 1999 (paper and presentation)

48 Lopez de San Roman Blanco, B and HN Southgate, 1999. "Efectos de la cronologia del oleaje en la modelizacion de la forma en planta de las playas." V Jornadas Espanolas de Puertos y Costas. La Coruna, 1999. pp.87-89 (paper and presentation)

49 Capobianco M, de Vriend H J, Stive M J F, Southgate H N and Jimenez J, (1999) “Large-scale coastal evolution. Toward a robust model-based methodology”, submitted to J. Coastal Research

50 Hanson H, de Vriend H J, Steetzel H, Nicholls R, Capobianco M, Jimenez J, Plant N, Larson M, Stive M J F, Southgate H N, Hinton C, Roszynski G, (1999) “Long-Term Modelling of Coastal Evolution”, submitted to J. Coastal Research

51 Lopez de San Roman Blanco B and Southgate H N, (1998) “The effects of wave chronology in beach planshape modelling”, HR Report TR 67

52 Lopez de San Roman Blanco B and Southgate H N, (1999) “Comparison of BEACHPLAN predictions with shoreline data from Kolobrzeg, Poland”, HR Report, to appear

53 Peet A H, (1999) “COSMOS-2D. Model validation”, HR Report TR 84

54 Southgate H N (1999) “Guidelines for error assessment in models of long term coastal morphology”, submitted to IAHR Symposium on River, Coastal and Estuarine Morphology, Genoa, Sept. 1999

10.CAMELOT: PRESENTATIONS OF SEMINARS, CONFERENCES AND WORKSHOPSLarge Scale Coastal Behaviour '93, Florida, USA. Presentation on "The Effects of Wave Event Sequencing on Long-Term Beach response" by H N Southgate (including poster presentation)

Large Scale Coastal Behaviour '93, Florida, USA. Presentation on "Tidal Schematisation in Morphodynamic Area Models" by T J Chesher

Coastal Dynamics '94, Barcelona, Spain. Presentation on "Validation of Coastal Morphodynamic Models with Field Data" by H M Wallace

Coastal Dynamics '94, Barcelona, Spain. Presentation on "Breaking Wave Persistence in Parametric Surf Zone Models" by H N Southgate

HR Computational Hydraulics Course (1994). Presentation on "Coastal Sediment Transport and Morphodynamics" by H N Southgate

MAFF River and Coastal Engineering Group Course at HR (1994). Presentation on "Modelling Coastal Morphodynamics" by H N Southgate

MAFF River and Coastal Engineering Group Course at HR (1994). Demonstration of PISCES model by T J Chesher

MAFF River and Coastal Engineering Conference (1994). Presentation on "Modelling Long-Term Coastal Morphology. The CAMELOT Project" by R L Soulsby

Workshop on Tidal Inlet Modelling, Vicksburg, US (1994). Presentation on "Development of the PISCES Model and Applications to Inlets" by H N Southgate Seminar at University of Cantabria, Santander, Spain (1995). Presentation on "Modelling Coastal Morphodynamics and Applications of the PISCES Model" by H N Southgate

14

Projecttitle



Coastal Dynamics '95, Gdansk, Poland. Presentation on "Numerical Modelling of Keta Lagoon", by T J Chesher

Coastal Dynamics '95, Gdansk, Poland. Presentation on "Time Series Analysis of Long-Term Beach Level Data from Lincolnshire, UK" by H N Southgate

Seminar on 'Morphology of Rivers, Estuaries and Coasts', HYDRA 200, IAHR, London (1995). Presentation on "The Application of Medium-Term Morphodynamic Models to the Long-Term" by T J Chesher

IAHR Meeting on the UK Coastal Research Facility, London (1995). Presentation on "Wave Breaking Experiments in the UKCRF" by H N Southgate

ICCE, Orlando, Florida (1996). Presentation on "Longshore Bedload Transport" by J S Damgaard

ICCE, Orlando, Florida (1996). Presentation on "Measurements of Wave Breaking in the UK Coastal Research Facility" by H N Southgate

Seminars at Delft Hydraulics and Utrecht University, The Netherlands. (1996). Presentations on "Fractals and Coastal Morphology" by H N Southgate

MAFF Workshop on Long-Term Coastal Morphodynamics, HR Wallingford (1997). Presentation on "CAMELOT Products and Knowledge from HR" by H N Southgate

Coastal Dynamics '97, Plymouth, UK (1997). Presentation on "The Role of Chronology in Long-Term Morphodynamics. Theory, Practice and Evidence" by H N Southgate

Coastal Dynamics '97, Plymouth, UK (1997). Presentation on "Quantifying the Uncertainty in Beach Planshape Modelling to Wave Chronology" by T K Millard

27th IAHR Congress, San Francisco, California (1997). Presentation on "Non-Linear Methods for Analysis of Long-Term Beach and Nearshore Morphology" by H N Southgate

Seminar at the University of Twente, The Netherlands (1998). Presentation on "Fractals, Chaos and Coastal Morphology" by H N Southgate

MAFF River and Coastal Engineering Conference (1998). Presentation on "The CAMELOT Programme for Predicting Long Term Coastal Morphology. An Overview from HR Wallingford" by H N Southgate

Southgate H N, “Coastal morphology: analogies with fluid turbulence and discrete modelling”, Euromech 395 Symposium on Coastal, Estuarine and River forms, Univ. of Twente, The Netherlands, June 2000 (presentation)

All HR Wallingford reports are available at www.hrwallingford.co.uk

Please press enter

15

research and development - defra, uk - science...

Documents