Introduction

Human-induced global climate change and associated sea-level rise can have major adverse consequences for coastal ecosystems and societies. The Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) projects an increase in globally averaged surface temperature of 1.4 to 5.8°C over the period 1990 to 2100. Based on these projections of future climate, global mean sea level is projected to rise by 9 to 88 cm between 1990 and 2100, with a central value of 48 cm, which is 2.2 to 4.4 times the observed rate over the 20th century. Even with drastic reductions in greenhouse gas emissions, sea level will continue to rise for centuries beyond 2100 because of the long response time of the global ocean system. An ultimate sea-level rise of 2 to 4 metres is possible for atmospheric carbon dioxide concentrations that are twice and four times pre-industrial levels, respectively (Church et al., 2001).

In the 1990s, a large and concerted effort was made to assess the implications of sea-level rise on coastal countries. Many studies have been carried out on local, national and regional scales. These studies have shown that most coastal areas are vulnerable to the adverse consequences of sea-level rise to some degree, although there is considerable variation in possible impacts (De la Vega-Leinert et al., 2000a, 2000b; see also http://www.survas.mdx.ac.uk/).

Global vulnerability assessments carried out by Hoozemans et al. (1993) and Baarse (1995) suggest that some 189 million people presently live below the once-per-1000-years storm-surge level. They estimate that, under present conditions, an average of 46 million people per year experience storm-surge flooding. This number would double if sea level rises 50 cm (92 million people/year) and almost triple if it rises one metre (118 million people/year). Between 86% and 92% of these people would experience flooding even more than once a year. These projections do not take into account any further population growth, changes in storm frequencies and intensities or adaptive responses.

These global vulnerability assessments have been the key sources of quantitative information on the potential impacts of sea-level rise on regional and global scales. They played an important part in the preparation of the World Coast Conference 1993 and several IPCC reports. They have also been used extensively for further academic analyses, including integrated assessment modelling. However, with the widespread use of the existing global vulnerability assessments, their limitations have become increasingly apparent. These limitations include:

• The obsolescence of underlying data sources;
• The reliance on sea level as the only climate variable that determines coastal vulnerability;
• The static, one-scenario approach of instantaneously raising sea level on today's world;
• Arbitrary assumptions regarding socio-economic development and adaptation.

Whilst Nicholls (2002) has tackled some of these problems, opportunities have arisen to combine data, scenarios and assessment models into a new integrated modelling activity. The project DINAS-COAST (Dynamic and Interactive Assessment of National, Regional and Global Vulnerability of Coastal Zones to Climate Change and Sea-Level Rise; http://www.dinas-coast.net/) builds on methods and expertise developed in a range of scientific-technological disciplines to develop an innovative, interdisciplinary methodology. The aim of the three-year project is to develop a dynamic, interactive and flexible CD-ROM-based tool that will enable its users to produce quantitative information on a range of coastal vulnerability indicators, for user-selected climatic and socio-economic scenarios and adaptation policies, on national, regional and global scales, covering all coastal nations. This tool is called DIVA, which stands for Dynamic and Interactive Vulnerability Assessment.

At the core of the tool is a simulation model that integrates knowledge of both natural and social sciences. Whilst integrated models have been built before, DIVA addresses two new challenges:

• It must be developed in a form that can be made available to a broad community of end-users;
• It must be developed by a geographically distributed group of scientists.

The first challenge requires a powerful yet user-friendly graphical user interface and a computationally efficient (i.e., fast) model. The second challenge calls for an innovative, modular approach to model development. The individual partners in the DINAS-COAST consortium independently develop modules representing processes of natural and social coastal subsystems, which are then "plugged" together to form one integrated model. Efficient means of communication, methods to harmonise concepts and an intuitive way to express knowledge from different disciplines are essential to facilitate this process of model development.