These lecture notes are intended to provide third-year mathematics undergraduates who are already familiar with inviscid fluid dynamics with some of the basic facts about the modelling and analysis of viscous flows. Writing the notes has been an interesting task because so many of the phenomena to be described are not only associated with vitally important mechanisms in everyday life but they are also readily observable without any need for instrumentation. More sophisticated realisations are also readily available, for instance in the very valuable collection of photographs “An Album of Fluid Motion”, edited by Van Dyke (2). Thus it is all the more stimulating when the mathematics that emerges when these phenomena are modelled is novel and suggestive of new methodologies.

The notes are strictly not self-contained and should be read in conjunction with standard texts which are referenced. We have concentrated on trying to present some of the salient physical ideas and mathematical ramifications as starkly as possible and, to this end, many of the exercises have been designed to be worked as an integral part of the notes; they are only put at the end of chapters for convenience. The starred exercises cover more advanced material and can be omitted at a first reading.

Experience has shown us that, in the twentieth century, theoretical mechanics generally has been one of the best vehicles for learning about physical applied mathematics.

The beginning

In the spring of 1984 Klaus Hasselmann invited wave modellers to Hamburg to discuss possible joint work. The WAM (Wave Modelling) group emerged. I dutifully reported on progress (Komen, 1985a,b, 1986, 1987a,b, 1990, 1991a,b). Now, ten years later, the WAM group has achieved what it wanted to achieve: a third generation computer model has been developed, which is able to predict the wave conditions at sea; it is used for global and regional applications; the model is a useful tool for interpreting satellite observations of the ocean and increased our understanding of the role of waves in air/sea interaction and the coupling between the atmosphere and the ocean. The results obtained are presented in this book.

WAM was one of a number of international collaborations: JONSWAP, SWAMP, SWIM, WAM, … Observations during the Joint North Sea Wave Project (Hasselmann et al, 1973) and also by Mitsuyasu and collaborators (Mitsuyasu, 1966, 1969, Mitsuyasu et al, 1971) had established the importance of the nonlinear transfer in governing the shape and evolution of a wind sea spectrum. Based on this finding ‘parametric’ wave models had been developed based on the approximation of a self-similar shape of the wind-sea spectrum.

Summary and outlook

The WAM group has realized its objectives: a third generation wave model has been developed; it runs in global and regional modes; it extended our understanding of the underlying physics; and data assimilation schemes have been developed and tested. This book is testimony of what has been achieved. However, careful reading will also reveal open ends, which range from small inconsistencies to major open issues. In this chapter we give an outlook on expected developments.

Compared to the so-called second generation models considerable progress has been made regarding the formulation of the evolution equation for the wave spectrum. The present WAM model is based on an explicit formulation of the physics of generation of waves by wind, nonlinear wave-wave interactions and dissipation due to whitecapping and bottom processes, rather than on the approach of ad hoc modelling which was commonplace with second generation models. The latter approach was shown to be inadequate under extreme circumstances such as hurricanes (SWAMP, 1985), while the WAM model gives for rapidly varying wind fields very satisfactory results (see chapter IV). Nevertheless, under ‘normal’ circumstances both approaches give similar results for the wave height. The reason for this is that although second generation models have inadequate physics they have been tuned to a considerable extent. Thus, the benefits of a third generation model are mainly related to a better representation of the spectrum itself and to a more explicit formulation of the underlying physics of wave evolution.

Despite the progress, we still are not able to make wave predictions that always fall within the error bands of the observations.

Impact of satellite wave measurements on wave modelling

Through the launch of ocean observing satellites, wave modellers are now for the first time receiving detailed wave data on a global, continuous basis. This can be expected to have a profound impact on wave modelling. The first us ocean satellite SEASAT demonstrated in 1978 that wave heights could be accurately measured with a radar altimeter and that a SAR (synthetic aperture radar) was capable of imaging ocean waves. Unfortunately, SEASAT failed after three months, and further satellite wave measurements were not made until the radar altimeter aboard GEOSAT was put into orbit in 1985. This changed with the launch of the first European Remote Sensing Satellite ERS-1 in July 1991. Since then, both radar altimeter and SAR wave data have been produced again globally in a continuous, near-real-time mode (cf. table 1.1).

Even before satellite wave data became available on a quasi-operational basis, the recognized potential of these data had a strong influence on wave modelling. One of the principal motivations for developing the third generation wave model WAM was to provide a state-of-the-art model for the assimilation of global wind and wave data from satellites for improved wind and wave field analysis and forecasting. Prior to the development of the WAM model, wave modellers had available only first and second generation wave models. The former were known to be based on incorrect physics, while the latter contained essentially correct physics but were restricted numerically through an artificial separation of the spectrum into a wind sea component of prescribed spectral shape and a swell spectrum, with rather arbitrary parametrizations of the coupling between the two spectral regimes.

Wave modelling in historical perspective

The third generation ocean wave models considered in this book are numerical models which integrate the dynamical equations that describe the evolution of a wave field. Their development followed progress in understanding ocean wave dynamics and experience with practical forecasting methods.

The study of ocean wave dynamics has a very long history indeed. Khandekar (1989) quotes Aristotle, Pliny the Elder, Leonardo da Vinci and Benjamin Franklin. In Phillips (1977) it is recalled how Lagrange, Airy, Stokes and Rayleigh, the ‘nineteenth century pioneers of modern theoretical fluid dynamics’, sought to account for the properties of surface waves. Subsequent progress in the twentieth century has been enormous, as any reader of, for instance, the Journal of Fluid Mechanics will realize. The subject has grown so large that it is nearly impossible for one person to have a comprehensive knowledge of all aspects. There are a number of useful textbooks, devoted totally or in part to water waves (Krauss, 1973, Whitham, 1974, Phillips, 1977 and LeBlond and Mysak, 1978, for example). Subjects that are of special relevance to wave modelling are the propagation of waves, their generation by wind, their nonlinear properties, their dissipation and their statistical description. We will discuss all these aspects in later sections where we will also indicate the historical perspective. Here it is enough to recall the milestones formed by the theoretical work of Phillips and Miles on wave generation and by Hasselmann's theory of four wave interactions.

Introduction

Any numerical model of a physical process necessarily represents that process by a set of mathematical relations that approximate the underlying physical laws. The level of approximation is determined principally by two factors: (1) knowledge of the physical processes; (2) computational limitations. The latter forces one to view the problem of the evolution of waves on an oceanic scale as a statistical one. The process of global wave evolution encompasses scales as large as the basin (107 metres) involving ocean currents, topography and wind variations, and as small as the smallest waves (10-3 metres) that modify the wind stress – it could be argued that even smaller scales associated with turbulent interfacial couplings play a role in the process. The statistical approach essentially treats the bottom 4/5 of this physical range of ten orders of magnitude as though those scales respond to well-defined physical laws imposed at the nodes of a numerical grid of typical size of one degree of latitude. The task of defining appropriate mathematical expressions that reflect the essential physics of the process is one of synthesizing the results of theoretical calculations and observational programs. As discussed in chapter I, the mathematical framework is based on a statistical description of waves having a range of scales of about one metre to one kilometre. These waves evolve in response to an action balance equation in which the ‘physics’ is embodied in a set of source functions. In this chapter, we first discuss the source functions individually and then examine the observational evidence of spectral characteristics and wave growth in fetch-limited and in shallow water situations and directional adjustment to turning winds.

Introduction

The principles of wave prediction were already well known at the beginning of the sixties (§ I.I). Yet, none of the wave models developed in the 1960s and 1970s computed the wave spectrum from the full energy balance equation. Additional ad hoc assumptions have always been introduced to ensure that the wave spectrum complies with some preconceived notions of wave development that were in some cases not consistent with the source functions. Reasons for introducing simplifications in the energy balance equation were twofold. On the one hand, the important role of the wave–wave interactions in wave evolution was not recognized. On the other hand, the limited computer power in those days precluded the use of the nonlinear transfer in the energy balance equation.

The first wave models, which were developed in the 1960s and 1970s, assumed that the wave components suddenly stopped growing as soon as they reached a universal saturation level (Phillips, 1958). The saturation spectrum, represented by Phillips' one-dimensional f-5 frequency spectrum and an empirical equilibrium directional distribution, was prescribed. Nowadays it is generally recognized that a universal high-frequency spectrum (in the region between 1.5 and 3 times the peak frequency) does not exist because the high-frequency region of the spectrum not only depends on whitecapping but also on wind input and on the low-frequency regions of the spectrum through nonlinear transfer. Furthermore, from the physics point of view it has now become clear that these so-called first generation wave models exhibit basic shortcomings by overestimating the wind input and disregarding nonlinear transfer.

In the previous chapters we have been dealing with theoretical, physical and numerical aspects of wave modelling, and the WAM model in particular. It is now time to face practical problems and to turn to applications. The WAM model is one of the most widely used wave models in the world, both for forecasting and hindcasting on global and regional scales. There is therefore a wealth of applications and experience which can be used to discuss and to judge the behaviour of the model.

The purpose of applications

Model applications serve two purposes: a practical one and a scientific one. In practical applications the model is accepted as a reliable tool. As such it can be used either in real-time, to forecast sea conditions for ship routing, offshore operations and for coastal protection, or in a hindcasting mode for computation of the sea state during a particular event and to determine wave climatology and extremal statistics. For scientific purposes, once properly verified, the model is, within the limits derived from its formulation, a representation of the real world. It can therefore be used for numerical experiments to simulate field experiments under conditions that would be difficult to find or expensive to organize. The model is a tool for the better understanding of the relevant phenomena and how they interact. It can also be used to study the sensitivity of the results to a change in the input conditions.

General features of wave data assimilation and inverse modelling

Background

In previous chapters we have discussed the use of a mathematical model, the third generation wave prediction model, to compute the state of the sea. We can distinguish between hindcasts, nowcasts and forecasts, the difference being the time for which the sea state is computed relative to the clock time. A forecast field can be computed only by a model. However, the model estimation of hindcast and nowcast fields can be improved using observations, which have been considered in the previous chapters of this book only for validation of the underlying physics or for verification of model results. This combination of using model results and observations to create an optimal estimate of the sea state is called data assimilation or analysis. The word analysis originates in early meteorological applications, for which meteorologists would subjectively draw isobaric patterns on the basis of isolated pressure observations. They analysed the weather. Later, this work was carried out with numerical models.

Observed data can also be used to validate and improve models. When the model improvement is carried out using numerical automatic model fitting techniques, one speaks of inverse modelling: instead of using a given model to compute data, which are then compared with observations, the observations are used in an inverse mode to construct an optimally fitted model.