This paper appears in


Marine and Coastal Geographical Information Systems
edited by D.J. Wright and D.J. Bartlett
pages 295-315, 1999.


Copyright © reserved by Dawn Wright and Taylor & Francis publishers.
May be freely distributed electronically in whole or in part, but please keep this notice attached and do not alter the text.

CHAPTER TWENTY-TWO

Epilogue

Darius J. Bartlett and Dawn J. Wright

22.1 INTRODUCTION

As we write this chapter, a mere year and a half remains of the 20th century. The closing of a century (and a millennium) is a natural point at which to take stock of past achievements, assess current practices, and make plans for the future. The authors and editors of this book have attempted to provide just such an assessment with regard to research and recent developments in marine and coastal geographical information systems.

For practitioners in these fields, these are exciting times. Born in the 1960s (though conceived out of theories, concepts, technologies and visions laid in earlier times), the earliest GISs were designed to help solve land-based problems. Over the ensuing decades they have thrived and multiplied in that environment, growing ever more powerful and sophisticated in the process, so that in the late 1990s, their use for most terrestrial applications is so widely accepted as to frequently go unremarked.

We have seen earlier in this book, that GISs started their migration seawards in the mid-1970s (Bartlett, 1999). For much of the subsequent decade, the effective limit for most GIS applications lay at the coast. Here, there were challenges enough: the conceptual and technical challenges of representing a highly dynamic, multidimensional, fuzzy-bounded environment in a digital framework; and the institutional challenges (some would say these were much the harder to resolve) of reconciling multiple demands and frequently conflicting interests, both human and non-human, at the shore. Gradually, technical innovations, patient research, and political developments in equal measure addressed these different concerns, and prevailed to the point where coastal GIS is now starting to come of age.

With emerging success at the shore, it was only a matter of time before the first incursions of GIS to the deep ocean realm occurred. Wright (1999) has charted the key stages of this important evolutionary progression. Some of the challenges encountered in coastal and littoral environments are also met in the deep ocean, where they combine with others unique to the benthic and, particularly, the abyssal domains. Probably the most taxing problems in these latter categories arise from the sheer lack of knowledge regarding much of what happens below the surface of the world’s oceans. With few exceptions, human engagements with marine environments have literally only scratched the surface, and we enter the 21st century still knowing remarkably and embarrassingly little about the two-thirds of the surface (and what happens below the surface) of this planet we live on. Thus, for deep ocean GIS, one of the biggest and most pressing challenges is simply that of acquiring more and more reliable data to work with.

We should also consider the changing milieu in which this evolution has occurred. In the early days of GIS, particularly in the 1960s and 1970s, most of the pioneering research and development in GIS arose within the academic sector. By the start of the 1980s, the commercialisation of mainstream GIS was well underway and, for the next two decades, much GIS development work was vendor-led. For specialist applications however (and this would include both the coastal and the marine sectors) it is still more usual, and perhaps more appropriate, for the core research to remain located within the academic sector. History shows that in disciplines that are undergoing rapid expansion and evolution (a situation which applies to marine and coastal science, as well as to GIS itself), many of the greatest advances arise from the work of radicals and visionaries, operating in academic environments that are comparatively free from the conservatism imposed by the commercial constraints of market viability. It is no accident that most of the chapters of the current volume dealing with technical and conceptual aspects of marine and coastal GIS are written by authors working in universities or in government-funded research settings.

Nevertheless, after many years of focus on terrestrial (environmental and socio-economic) applications, there are signs that the commercial sector is increasingly paying heed to the specialist needs of marine and coastal GIS users. Within the industry, many of the leading GIS developers and vendors are involved in collaborative research with marine scientists, to the certain benefit of both sides of the partnership. Recent industry-wide moves towards open, extendable GIS will certainly play an important role in this process, and it is likely that the future will see an increasing range and quantity of sector-specific extensions being developed for and by specialist end-user communities. These will "plug into" and add to a core set of generic system capabilities provided by the GIS vendors. Such developments are already evident in the growing desktop GIS sector: for example in the expanding suite of coastal, marine and other extensions available or under development for the ArcView and MapInfo systems respectively.

As has been indicated repeatedly in the various contributions to this book, much research remains to be done into improving, extending and optimising the capabilities of spatial information systems in the marine and coastal realms. Essential areas ripe for investigation include:

22.1.1 Strategies and Techniques for Marine and Coastal Data Collection

Data are the raw materials that fuel GIS use. Without adequate data, even the most sophisticated analytical techniques are rendered useless. The current paucity of reliable, relevant and usable data for the marine and coastal domain has been a recurring thread in the chapters of this book. Arising from this need, there is a clear imperative to expand significantly the range of data collecting technologies available to the scientist and manager. Even around our shores, routine collection of relevant environmental and other data is far from established, and huge gaps occur in the sampling density of such data, in both the spatial and the temporal dimensions. Offshore, the situation is even worse. Recent generations of satellite remote sensors have greatly helped routine monitoring and data collection for the ocean surface, but current sensors are unable to penetrate the water column. At present, remote sensing of the ocean bed, or the water columns overlying the seafloor, is essentially restricted to the use of sonar and occasionally moored buoys or drifters bearing data loggers.

22.1.2 Marine and Coastal Data Standards

The need for data standards is now recognised in all GIS application areas, and no less with regard to marine and coastal data. It may even be argued that the need for clear and unambiguous standards for data definition, quality control and data exchange is all the more important in marine and coastal environments, given the multiplicity of users and interest groups acquiring and using these data and also, frequently, the international dimension of the applications concerned. Although, as was indicated above, the global marine and coastal database is still limited in both size and content, the growing number of national and international research initiatives planned or under way make it certain that sizes of data holdings, and the flow of data and information between institutions, will grow rapidly and substantially. This mobility of data carries with it many benefits, but also a number of possible hazards: among the latter, the most pressing and potentially serious is the risk of data quality loss and creeping error or uncertainty in the results obtained from using such data; plus the concomitant danger that the wrong data will be applied inadvertently to the wrong situation. It is essential that the whole issue of creating and applying sound, internationally agreed data standards is investigated thoroughly, so as to maximise benefits and minimise any concomitant disadvantages.

22.1.3 Automating and Optimising Shoreline Definition and Maritime Boundaries

As was indicated in Chapter 21 (Palmer and Pruett, 1999), there have been numerous recent advances in the definition of shorelines and maritime boundaries. Nevertheless, much research remains to be done in this area. In particular, while WGS-84 provides a global horizontal datum for shoreline determination, the development of a global vertical datum is an area of ongoing investigation and discussion. Additionally, since shorelines are constantly undergoing change due to the forces of nature (and these changes can frequently be very rapid, or even catastrophic, in response to extreme weather events and other intense disturbances), new and more automated collection techniques are being explored to locate and extract rapidly the land/sea interface from multi-spectral and other imagery. Thus, key areas of research and development that need support from the international community include: the definition and adoption of a global vertical datum to improve the accuracy of future shoreline determination; and the development and adoption of an objective process for the automated or semi-automated extraction of shoreline from imagery to provide timely updates for shoreline changes.

Improvements in the definition of maritime boundaries are contingent upon the decisions of the international hydrographic community, the United Nations, unilateral and bilateral agreements of and between coastal states, and adjudication of disputes by authoritative bodies. Palmer and Pruett (1999) provides a summary of an approach for creating a comprehensive source of global maritime boundaries. The utility of a global maritime boundary database (GMBD) is a function of the currency of information. As subsequent claims, counter- claims, disputes and adjudicated boundaries are announced, they will be incorporated in the GMBD. This is a challenging task, since the dynamic nature of geo-political declarations ensures the need for frequent updates: this, too, poses many important questions worthy of research, including the development and testing of consensus-building techniques, for ensuring wider acceptance of international decisions on boundary definition; and investigation of the role of GIS as a tool for communicating such decisions to governments, other maritime interests, and the wider public.

22.1.4 Working with Three-dimensional (3-D) Data

The present generation of GIS evolved largely out of a translation of conventional cartographic metaphors into a digital environment. As a result, many of the inherent limitations of traditional mapping are embedded within current geoinformation systems. Foremost among these is the continuing emphasis on working in two, or (in the case of surface modelling) two and a half, spatial dimensions (Raper, 1999). Much geography, particularly in the coastal and marine realms, requires all three spatial dimensions to be considered equally, whether simply for visualisation, or else for more demanding spatial analyses. In fact visualisation and analysis are two separate, but complementary, aspects of working in three spatial dimensions: the former is concerned with the graphic presentation of data to the viewer in a volumetric (3-D) form, either in hard copy or on a computer display; while the latter involves actual manipulation and processing of the data in order to derive information or further data as output, which may or may not then be visualised.

While GIS supports two-dimensional spatial analysis very well, it does not provide easy-to-use 3-D visualisation or volumetric analysis, as is found with more specific software such as Fledermaus, Dynamic Graphics, IBM Visualization Data Explorer or Spyglass. Some attempts have been made to extend capability into the third dimension, most notably with Intergraph’s Voxel Analyst, but so far such developments, while welcome, are still limited in the functions they provide. For example, a 3-D Analyst extension is now available for ESRI’s ArcView, but despite the name this tool is suitable only for 2.5-D (surface) modelling and viewing, since the underlying data structure is not truly 3-D. Likewise, current database management systems (DBMSs) do not directly provide this capability either, although the prevailing view from industry is that it is probably easier and more appropriate to provide data to scientific visualisation applications from a DBMS than from a system such as ArcView.

22.1.5 Time Series Data

Most current GISs have only rudimentary support for analysis of the time series data needed in many marine and coastal studies. While considerable advances have been made in recent times, this weakness in GIS is still rooted in imperfect conceptual understandings of temporal data generally and, more specifically, of the spatial dynamics of many marine and coastal environments. Where temporality has been incorporated into marine or coastal GIS, it has mostly arisen through extension of the DBMS underpinning the software, where support for maintaining and viewing data as time series is better understood, primarily because of heavy use of time series by banks and stock exchanges who have lobbied DBMS companies for this functionality. The current fix for most GIS specialists is to provide links from the GIS to special purpose programs that provide time series analysis (e.g., Fourier analysis, power spectra, stochastic simulation, etc.). This is not an ideal solution, however, and the quest for truly temporal GIS remains one of the "holy grails" of much current GIS research. Any benefits accruing from such research would be felt across almost the entire span of the marine and coastal science, and would have important roles in many other application domains.

22.1.6 Computational Models and Experimental Flow

GIS have excellent capabilities for connecting maps to computational models or for linking programs running on one platform to those running on another. Linkages have been successfully made between GIS and a wide variety of process models drawn from the terrestrial domain: in the natural environment, these include groundwater contamination models, climate models, soil loss equations, surface hydrological models and others; while in the socio-economic domain they include modelling spheres of influence and sales pitches, modelling epidemics and flows of commodities, etc. In contrast, many of the techniques involved in coupling marine and coastal models to GIS are still poorly investigated or understood, and thus the benefits and synergy that can arise from bringing these different tools together are rarely seen.

Even where such coupling of models and GIS can be achieved, linking and scheduling these to run as a logical sequence for scientific investigations is no easy task. Just running a single model from within (or alongside) a GIS can be non-trivial. Running a series of computational (numerical) experiments with a GIS is even more problematic. Thus, an important direction for future research would appear to lie in developing and providing templates and computational tools for extracting data or importing results to, from, and between multiple process models and GIS. At present, such operations are best done via scientific notebook systems (e.g., Cuny et al., 1997; Skidmore et al., 1998) specifically designed to support scientific collaboration and simultaneous computational experimentation (i.e., several scientists working on the same data set, at the same time via a collaboration, even though they may be physically separated by great distances). It would be a significant advantage for marine and coastal scientists alike, to be able to schedule and conduct such experiments within a unified and integrated spatial information system environment.

22.1.7 Human Factors

While some of the research issues outlined above are technological in focus, it should be emphasised that, in common with recent concerns in many other areas of information system application, a growing degree of attention is now being devoted to the human and institutional contexts within which coastal/marine GIS use takes place. The past two decades have seen a huge, rapid uptake of GIS in academic, commercial, governmental, scientific, and other sectors, for a range of marine and coastal applications, and this awareness is translating into a gradually-expanding user base for the technology. However, using GIS, and using GIS well, are not necessarily one and the same thing. There is a need to ensure that end-users remain aware of both the strengths and the limitations of their tools, and this in turn requires that training of end-users keeps pace with advances in concepts and technology. It is important, also, to remember that GIS should always be a means to an end, and should never become the end in itself.

There is, therefore, great need for research into these human aspects: how can the technology be fine-tuned and adjusted to best serve the working environments and information needs of its marine or coastal users? And what are the training requirements of system end users, and how can these needs be best met? Specific areas where such human-oriented research appears to be most needed include the design and development of decision support systems to assist coastal or marine sector managers. What decisions require to be made, and what data and rules apply in the making of such decisions? How can these rules be translated into expert or other systems to provide necessary technological support, while still keeping sufficient control in the hands of the human operator? The design of appropriate user interfaces (e.g., Su, 1999), including, interfaces that will allow efficient use of GIS on the bridge of a pitching ship in a storm; within the confines of a submersible on the deep ocean floor; or on hand-held computers being operated by scuba divers; etc.; and the use of GIS as a teaching, training and awareness-building tool for promoting sustainable ocean use, integrated coastal zone management and other long- and short-term strategies, and for assisting in community participation in (especially coastal) planning and decision-making.

22.2 CONCLUSION

We live in a very rapidly changing world. Space exploration and travel, now taken almost for granted, are the product of a mere 30 years; the Internet which, as indicated in the Preface, was instrumental in helping this book see the light of day, is only some 20 years old; and the cell phones that are so ubiquitous in modern society are even younger. GISs are still also comparatively recent innovations and, against a plethora of potential user communities and application areas, continue to evolve and develop with sometimes bewildering rapidity.

The nature of science in general and, specifically, marine and coastal science, has also undergone radical change and evolution in the past three decades. In common with many other branches of, particularly the environmental sciences, marine and coastal science has largely rejected the reductionism and the hard- engineering-dominated ethos of "command and control" that so characterised much of the science applications in the first half of the 20th century. Gone is the idea of the seas as a sink, unlimited in its capacity to soak up the pollutants and waste products of human society. Gone, too, is the notion of the ocean as a boundless source of fish and other resources, without need for control or rationalisation of exploitation. And gone is the concept of armouring the shore and waging war between humanity and the raging ocean. Instead, in both the marine and the coastal domains, the new ethos preaches an holistic message, with sustainability of resource use, and the needs or rights of nature and of future human generations embodied firmly within its texts.

In the closing years of the 20th century, many concerns have been expressed by scientists, politicians and the public alike, regarding the health of the world’s oceans and coastal zones. There are also indications that these warnings are being heard and heeded by decision-makers at all levels from the international to the strictly local. Whether in the natural sciences or the humanities, whether pitched at the conceptual, the technical or the application end, or, as is increasingly being demanded and attempted, in the integration of data and applications from all of these interest groups and professions, the role of GIS in this new thinking is clear. Applying GISs to marine and coastal environments presents taxing, but particularly satisfying challenges to end users and system developers alike. The chapters of this book demonstrate eloquently the many advances that have been achieved in recent years and the steps currently underway to address remaining issues. There is much research and development yet to be done, and we, the editors and authors of this volume, very much hope that the lead and examples demonstrated in these pages will encourage other scientists, investigators and end-users to join the ranks of the marine and coastal GIS communities.

22.3 REFERENCES

Bartlett, D.J., 1999. Working on the frontiers of science: applying GIS to the coastal zone, in this volume, Chapter 2.

Cuny, J.E., Dunn, R., Hackstadt, S.T., Harrop, C., Hersey, H., Malony A.D., and Toomey, D.R., 1997, Building domain-specific environments for computational science: A case study in seismic tomography. International Journal of Supercomputer Applications, 11, pp. 1-21, http://www.cs.uoregon.edu/~harrop/papers/ETPSC96/.

Raper, J., 1999. 2.5- and 3-D GIS for coastal geomorphology, in this volume, Chapter 9.

Palmer, H. and Pruett, L., 1999, GIS applications to maritime boundary delimitation, in this volume, Chapter 21.

Skidmore, J., Sottile, M., Cuny, J. and Malony, A., 1998, A prototype notebook-based environment for computational tools. In Proceedings, Supercomputing 98, Orlando, Florida (Piscataway, New Jersey: IEEE), pp. 1-13, http://www.csi.uoregon/edu/nacse/vine/pub/sc98.html.

Su, Y., 1999, A user-friendly marine GIS for multi-dimensional visualisation, in this volume, Chapter 16.

Wright, D.J., 1999. Down to the sea in ships: The emergence of marine GIS, in this volume, Chapter 1.