Journal of Coastal Research

24

917924

West Palm Beach, Florida

July 2008

A New Global Coastal Database for Impact and Vulnerability

Analysis to Sea-Level Rise
Athanasios T. Vafeidis, Robert J. Nicholls, Loraine McFadden, Richard S.J. Tol, Jochen Hinkel,
Tom Spencer, Poul S. Grashoff, Gerben Boot, and Richard J.T. Klein

Department of Geography
University of the Aegean
University Hill
Mytilene 81100
Lesvos, Greece
a.vafeidis@ieee.org

School of Engineering and

the Environment and
Tyndall Centre for Climate
Change Research
University of Southampton
Highfield
Southampton SO17 1BJ
United Kingdom

Potsdam Institute for

Climate Change Research
(PIK)
P.O. Box 601203
D-14412 Potsdam
Germany

Cambridge Coastal Research

Unit
Department of Geography
University of Cambridge
Downing Place
Cambridge CB2 3EN
United Kingdom

WL Delft Hydraulics
P.O. Box 177
2600 MH Delft
The Netherlands

Stockholm Environment
Institute
Box 2142
103 14 Stockholm
Sweden

Flood Hazard Research

Centre
Middlesex University
Enfield EN3 4SA
United Kingdom

Economic and Social

Research Institute
Whitaker Square
Sir John Rogersons Quay
Dublin 2, Ireland

ZMK, University of Hamburg

Bundesstrasse 55
20146 Hamburg
Germany
Demis bv
Kluyverweg 2a
2629 HT Delft
The Netherlands

Institute for Environmental

Studies
Vrije Universiteit
De Boelelaan 1087
1081 HV Amsterdam
The Netherlands
Engineering and Public Policy
Carnegie Mellon University
Baker Hall 129
Pittsburgh, PA 15214
U.S.A.

ABSTRACT
VAFEIDIS, A.T.; NICHOLLS, R.J.; MCFADDEN, L.; TOL, R.S.J.; HINKEL, J.; SPENCER, T.; GRASHOFF, P.S.;
BOOT, G., and KLEIN, R.J.T., 2008. A new global coastal database for impact and vulnerability analysis to sea-level
rise. Journal of Coastal Research, 24(4), 917924. West Palm Beach (Florida), ISSN 0749-0208.
A new global coastal database has been developed within the context of the DINAS-COAST project. The database
covers the worlds coasts, excluding Antarctica, and includes information on more than 80 physical, ecological, and
socioeconomic parameters of the coastal zone. The database provides the base data for the Dynamic Interactive Vulnerability Assessment modelling tool that the DINAS-COAST project has produced. In order to comply with the
requirements of the modelling tool, it is based on a data model in which all information is referenced to more than
12,000 linear coastal segments of variable length. For efficiency of data storage, six other geographic features (administrative units, countries, rivers, tidal basins or estuaries, world heritage sites, and climate grid cells) are used to
reference some data, but all are linked to the linear segment structure. This fundamental linear data structure is
unique for a global database and represents an efficient solution to the problem of representing and storing coastal
data. The database has been specifically designed to support impact and vulnerability analysis to sea-level rise at a
range of scales up to global. Due to the structure, consistency, user-friendliness, and wealth of information in the
database, it has potential wider application to analysis and modelling of the worlds coasts, especially at regional to
global scales.
ADDITIONAL INDEX WORDS: Segmentation, coastal geographic information system (GIS), data model, climate
change, global change.

INTRODUCTION
The expected accelerated rise in global mean sea levels
through the 21st century may cause a number of physical
changes to the worlds coasts and hence can endanger coastal
DOI:10.2112/06-0725.1 received 19 July 2006; accepted in revision 15
February 2007.

populations and infrastructure, as well as threaten many

coastal ecosystems (MCLEAN et al., 2001). The sensitivity of
the coastal zone to sea-level rise, in conjunction with its importance in terms of social, economic, and ecological value
(e.g., AGARDY et al., 2005; COSTANZA et al., 1997; TURNER,
SUBAK, and ADGER, 1996), highlights the need for consistent
national- to global-scale assessments of potential impacts and

918

Vafeidis et al.

possible responses to this global phenomenon and hence vulnerability to these changes. Such evaluations are an important component in the formulation of international climate
change policies (NICHOLLS, TOL, and HALL, 2007), which are
necessary for reducing the magnitude of potential impacts via
mitigation (reducing greenhouse gas emissions) and/or adaptation to sea-level rise and climate change.
Global and regional vulnerability studies have provided
quantitative assessments of potential impacts of sea-level
rise at global and regional scales (e.g., GORNITZ et al., 1994;
HOOZEMANS, MARCHAND, and PENNEKAMP, 1993; MIMURA,
2000; THIELER and HAMMAR-KLOSE, 1999). However, the
scope of these studies has been limited by the available data
in terms of resolution, coverage, parameter availability, and
dated sources: this is a generic problem for broad-scale coastal analysis (NATIONAL RESEARCH COUNCIL STAFF, 2004). In
addition to these limitations, data quality and integration
constitute further problems; even in those cases where data
and tools are available to coastal scientists for the analysis
and modelling of coastal processes, these usually exist in
fragmented forms. This fact compromises the consistency, reliability, and versatility of evaluations based on such sources.
It has long been recognised that appropriate and reliable information within organised, planned, and coherent coastal
databases is an essential prerequisite for coastal zone management (WEYL, 1982), and such systems continue to be developed (e.g., BARTLETT, 2000; BARTLETT et al., 1997; MILLARD and BRADY, 2000; VAFEIDIS et al., 2004). The importance of data quality and integration is also reflected in the
proposal of the U.S. National Research Council for a geospatial framework for the U.S. coastal zone (NATIONAL RESEARCH COUNCIL STAFF, 2004).
In order to address the preceding issues and provide a consistent source of data at global scales, the Dynamic and Interactive Assessment of National, Regional and Global Vulnerability of Coastal Zones to Climate Change and Sea-Level
Rise (DINAS-COAST) project has developed a new global
coastal database termed the Dynamic Interactive Vulnerability Assessment (DIVA) Coastal Database (DINAS-COAST
CONSORTIUM STAFF, 2006a), which is available at http://
www.civil.soton.ac.uk/DIVA and http://www.dinas-coast.net.
This database contains physical, ecological, and socioeconomic parameters and covers the worlds coast, excluding Antarctica. It underpins the DIVA modelling tool that the project
has developed (DINAS-COAST CONSORTIUM STAFF, 2006b;
HINKEL and KLEIN, 2007; NICHOLLS, KLEIN, and TOL, 2007).
This modelling tool enables users to analyse a wide range of
mitigation and adaptation policies in terms of coastal impacts
and vulnerability by producing quantitative data on a range
of coastal and socioeconomic scenarios at national, regional,
and global scales, including adaptation (and mitigation)
strategies. For this reason, the database has been specifically
designed to address the data requirements of the project and
the needs of researchers in the area of vulnerability assessment of coastal zones. It is also expected to be used for wider
assessment of regional and global coastal issues.
This paper discusses the importance and the role of global
databases in assessing global and regional coastal vulnerability to sea-level rise, emphasises the need for a new global

coastal database, and presents the database that has been

developed within the DINAS-COAST project. Within this context, existing data collections and databases that have been
specifically designed for the coastal zone are reviewed and
the new characteristics and comparative advantages of the
DIVA Coastal Database are assessed. Furthermore, an overview of the database is presented and its potential use in
coastal analysis is discussed.

GLOBAL COASTAL DATABASES: AN OVERVIEW

In the last two decades, the availability of digital spatial data
for the worlds biomes has vastly increased as a result of advancements in data capture and input techniques (PEUQUET,
1988). Nevertheless, collections of data are not necessarily adequate for the needs of global-scale analysis, as merging data
from various sources may raise several questions (FLEWELLING
and EGENHOFER, 1999). As TOMLINSON (1988) observes, the
way in which Earth observations are stored and integrated
largely determines the degree to which spatial processes can be
understood. Therefore, well-organised and designed data systems are needed to underpin our understanding of the processes
taking place over large parts of the globe, including the coast
(BARTLETT, 2000; TOMLINSON, 1988).
The large increase in global data availability has had a
significant impact on coastal science. New and better-quality
data for large parts of the worlds coastal areas on factors
such as waves, water quality, sediment fluxes, and elevation
are becoming available via remote sensing. These data are
extensively utilised to support local-scale studies (BRYAN et
al., 2001; SCOTT LEE and SHAN, 2003). At the same time,
new spatial global datasets on various biophysical and socioeconomic parameters such as population distribution (CIESIN and CIAT STAFFS, 2004; LANDSCAN STAFF, 2003) and
recently gross domestic product density (NORDHAUS, 2006)
are being developed and upgraded in order to satisfy data
needs for interdisciplinary analysis. However, the forms in
which these vast volumes of data exist and the difficulties
involved in the process of integrating them are still a major
barrier to their systematic utilisation within impact and vulnerability analysis. Commonly encountered limitations include (1) different data formats, (2) incompatible projections
between datasets, (3) high costs for integrating nondigital datasets, (4) lack of metadata, and (5) incomplete spatial coverage. Providing linkages among different data types, datasets, and databases, in order to enable their combined use,
and making this information available to nonspecialised users and policy makers can greatly enhance the potential of
existing data and tools in assisting the coastal science community (VAFEIDIS et al., 2004). Some attempts have been
made, especially in recent years, to develop global databases
for the worlds coastal regions. While these databases are an
improvement on existing available datasets, they remain limited in various aspects, such as the type and amount of information they contain or the choice of the data model for the
representation of information. These existing coastal databases are assessed here.

Journal of Coastal Research, Vol. 24, No. 4, 2008

A New Global Coastal Database

Table 1. Main characteristics of existing coastal databases.

Database

LOICZ

Coverage

Global

EUROSION Europe

Data
Format

Resolution

Data Sources

Raster

Half degree Existing public

datasets

Various

1:100,000

Existing datasets,
newly developed
datasets for
EUROSION

SAUP

Coastal
Raster
areas and
oceans

Half degree Existing datasets,

newly developed
datasets for SAUP

CoastBase

Europe

Not specified Existing datasets

Various

LOICZ LandOcean Interactions in the Coastal Zone, SAUP Sea

Around Us Project.

Examples of Coastal Databases

Due to the complex and dynamic nature of coastal zone
systems, their analysis implies the need to simultaneously
model multiple processes and their interactions rather than
single isolated processes. For such analysis to be viable, the
underlying data needs to be harmonised and held in a single,
well-structured database (ROBERTS and MOORE, 1998). Examples of limited sets of global- or regional-scale coastal data
include the Sea Around Us Project (SAUP) database (http://
www.saup.fisheries.ubc.ca) (A LDER, 2003); CoastBase, the
European virtual coastal and marine data warehouse (http:/
/www.coastbase.org); the EUROSION database (http://
www.eurosion.org); and the LandOcean Interactions in the
Coastal Zone (LOICZ) typology (MAXWELL and BUDDEMEIER,
2002). The main characteristics of these databases are summarised in Table 1.
These databases include collections of existing datasets, as
well as additional data generated for the purposes of projects
in which the databases were developed (e.g., EUROSION).
The coastal element of the SAUP database is a subset of the
global ocean database and includes a series of biophysical parameters and some additional attributes, such as the percentage of water area covered by wetlands (CHUENPAGDEE
and PAULY, 2004). The EUROSION database is fully integrated within a geographic information system (GIS) and was
built to support the assessment of coastal erosion in Europe.
It includes a series of parameters that impact coastal erosion
processes and data on protection measures (e.g., coastal defences). The LOICZ database is probably the most comprehensive collection of coastal parameters at global scale, prior
to the DIVA database, as it merges numerous global datasets
and contains a wealth of information on physical parameters.
It was developed to provide a basis for categorising coastal
and marine environments at global and regional scales, including upscaling detailed studies (e.g., BUDDEMEIER et al.,
2002, pp. 4851; CROSSLAND et al., 2005), and underlies
LOICZview, an online interactive software tool for statistically classifying such data and for assisting environmental
scientists in developing coastal typologies. Finally, CoastBase
aimed to link existing databases and new aggregated data in
an attempt to design a European virtual coastal and marine
data warehouse to improve coastal and marine information

919

search and exchange. However, it did not reach the point of

developing the actual data warehouse and remains a prototype that contains a very limited set of data sources (ELEVELD, SCHRIMPF, and SIEGERT, 2003).
The EUROSION, LOICZ, and SAUP databases probably
constitute the most comprehensive attempts to effectively
merge coastal data at regional to global scales. Nevertheless,
they are limited by factors including the type of the data model, the forms in which the data can be distributed, and the
limited socioeconomic coverage, which in most cases consists
of population data. Some of these limitations stem from the
fact that the previously mentioned databases, with the exception of EUROSION, were not originally designed or did
not include data intended solely for coastal applications; not
all are focused on the coastal zone but also cover inland or
ocean regions. The data model has been determined primarily
by the form in which the underlying data exist, which in most
cases is a raster data model (with the exception of the EUROSION database, which combines different data types). Despite its advantages in most global applications, the raster
data format can be inefficient for coasts due to their essentially linear nature; raster also gives a fixed database resolution, which does not always address the requirements of
coastal analysis. Finally, the available datasets are in formats that require the use of GIS software and expertise.
These factors can significantly limit the potential use of these
databases in coastal modelling and analyses, including impact and vulnerability analysis.

THE DIVA COASTAL DATABASE

Recognising the need for a new global-scale data system to
produce reliable evaluations of coastal impacts and adaptation,
the DINAS-COAST project set as one of its primary objectives
the compilation of a new global coastal database. This database has been designed specifically for impact and vulnerability analysis under sea-level rise. It is intended to form a consistent basis for producing reliable evaluations of coastal vulnerability. The database is one of the three principal elements
of the DIVA modelling tool, which also includes a range of
interacting impact/adaptation algorithms and a graphical user
interface (HINKEL, 2005; HINKEL and KLEIN, 2007).
Due to its explicit spatial nature, the database was initially
developed within a GIS. A new data model was developed to
represent coastal space, which decomposed the worlds coast
into a series of linear segments. It is noteworthy that most
impact and vulnerability analyses have explicitly or implicitly employed some form of coastal segmentation to create a
data structure. Such an example is the first Global Vulnerability Assessment, where the database structure was essentially based on national boundaries, producing 192 segments
(HOOZEMANS, MARCHAND, and PENNEKAMP, 1993; NICHOLLS and HOOZEMANS, 2005). Segmentation of the coast into
discrete coastal units is also applied at more detailed scales
for the purposes of shoreline or coastal management (e.g.,
BRENNER, JIMENEZ, and SARDIA, 2006; MANGOR, 2001).
In the following sections, the process of compiling the database is discussed. First, the data model and the structure that
underlie the DIVA database are presented and their implemen-

Journal of Coastal Research, Vol. 24, No. 4, 2008

920

Vafeidis et al.

tation is described. Then, the data sources and methods used

for populating the database are examined. Finally, the important tasks of validation of the database and the generation of
supporting metadata on the database are discussed.

Representation of Coastal Space: The Data Model

The decision regarding how to represent information in a
system is an important factor for the organisation and reliability of a database. The data model and structure employed
for describing the system have been identified by B ARTLETT
et al. (1997) as two of the major concerns to the developers
of coastal information systems, the most immediate of which
is the selection of an appropriate model of coastal space. The
representation of the coastline also affects GISs, which are
founded on the assumption that the two horizontal dimensions are essentially equivalent (GOODCHILD, 2000). The
common recognition of the coastline as a linear entity led
SHUPENG (1988) to suggest using a linear model for coastal
databases, a suggestion that many researchers have followed
(e.g., FRICKER and FORBES, 1988; SHERIN, 2000; SHERIN and
EDWARDSON, 1996).
Based on the concept of linear representation of the coastline, the DINAS-COAST project created a global model of
coastal space in which geographic information is represented
as a collection of geographic features (HINKEL, 2005; HINKEL
and KLEIN, 2007). All data are expressed as attributes of seven geographic feature types, namely, linear coastline segments of variable length, subnational administrative units,
countries, major selected rivers, major selected tidal basins
or estuaries, world heritage sites, and relevant climate and
sea-level scenarios based on the grid of the Climate and Biosphere Groupe (CLIMBER) climate model as described by
GANOPOLSKI and RAHMSTORF (2001). The fundamental core
of the database consists of the linear segments, and all attribute data are spatially referenced to the relevant coastal segments, which possess unique identifiers that link the different feature types via the coastal segments (Figure 1). The
segmentation model is designed to define homogeneous units
for vulnerability analysis at the (broad) scales of analysis.
These units are defined based on the longshore variation of
several parameters related to (1) coastal geomorphology (related to the natural system response), (2) population density
(related to human exposure/response and the human dimension), and (3) administrative boundaries (related to political
decision-making entities) (MCFADDEN et al., 2007). While
these variables are not comprehensive, they represent a selective set of boundary conditions across the naturalhuman
system divide that allow objective segments to be defined at
an appropriate scale for the DINAS-COAST project. The spatial variability of the broad-scale parameters becomes the reference unit for the segmentation: the point at which a parameter changes along the coast is plotted against the coastline,
defining a linear stretch or coastal segment. Each segment
reflects a changing response or sequence of responses within
the coastal system to sea-level forcing. All attribute data are
then referenced to these autonomous units, using methods
and metrics discussed in the following sections. An analytical
description of the segmentation rules, the underlying data,

Figure 1. Entity-relationship diagram of the DINAS-COAST database.

and the rationale underpinning the methodology can be found

in MCFADDEN et al. (2007).
According to this methodological framework, the segmentation of the coastline was implemented within a GIS. The
Digital Chart of the World (DCW) coastline (ENVIRONMENTAL SYSTEMS RESEARCH INSTITUTE STAFF, 2002) was employed as the base layer since its size, scale, and level of detail were the most optimal available, in terms of information
content and computational requirements, for a global database. However, for the purposes of the present study, around
1500 small islands and atolls, which are areas known to be
particularly vulnerable to sea-level rise (NURSE et al., 2001),
were added to the original layer. These additional islands and
atolls, not included in the DCW, were selected from the more
detailed World Vector Shoreline dataset (SOLURI and WOODSON, 1990). Based on the preceding coastline and using global
datasets (Table 2) containing spatial information on the distribution of the segmentation parameters that were described
earlier in this section, the segmentation was implemented in
a series of steps. Each step involved splitting the coastline at
the points where the values of each of the parameters
changed, thus generating segments with uniform values
along their length. This process was performed using standard GIS operations and techniques such as dataset reprojections, format conversions, overlay analyses, and buffering
(LONGLEY et al., 2001). The segmentation process divided the
worlds coast into 12,148 segments (Figure 2), which constitute the fundamental spatial reference units for the database,
the integrated modules, and the graphical user interface of
the DIVA modelling tool.

Populating the Database

Compilation of the database was based on a series of primary and secondary data sources (Table 2). Primary data included datasets derived from remote sensing that offer a glob-

Journal of Coastal Research, Vol. 24, No. 4, 2008

A New Global Coastal Database

921

Table 2. Examples of different types of datasets used for compiling the

DIVA Coastal Database. These datasets were developed by the DINASCOAST consortium, or existing public datasets were used and modified,
based on additional data and expert judgement, according to project needs.
Analytical descriptions of the datasets included in the database can be
found in the DIVA database documentation.
Dataset

Type

Gridded Population of the

World, v. 3
World elevation and bathymetry
Geomorphic type
Landform type
Tidal range
Wetlands database

Raster layer

Second-level administrative boundaries

World heritage sites
Surge heights
Uplift/subsidence
Land use
GDP per capita
Tourist arrivals/departures
World rivers
Tidal basins
Storm surges
Wave climate

Source

CIESIN

Raster layer

National Geophysical Data

Center
Analogue map McGill (1958)
Analogue map Valentin (1952)
Raster layer
IGBP-LOICZ
Tabular data Cambridge Coastal Research Unit
Polygon layer DCW, ESRI
Tabular data
Model output
Point layer
Raster layer
Tabular data
Tabular data
Tabular data
Point layer
Tabular data
Raster layer

UNESCO
Delft Hydraulics
Peltier (2000)
IMAGE Project
CIA, WRI
WRI
Delft Hydraulics
Delft Hydraulics
Delft Hydraulics
IGBP-LOICZ

CIA Central Intelligence Agency, CIESIN Center for International

Earth Science Information Network, DCW Digital Chart of the World,
ESRI Environmental Systems Research Institute, GDP gross domestic product, IGBP International GeosphereBiosphere Programme,
LOICZ LandOcean Interactions in the Coastal Zone, UNESCO
United Nations Educational, Scientific, and Cultural Organization, WRI
World Resources Institute. From VAFEIDIS et al., 2006.

ally consistent view of the Earth, are largely unaffected by

national data collection practices, can be aggregated comparatively easily, and are readily available (RHIND and CLARK,
1988). These data were complemented with analogue cartographic data from various sources, which were converted into
digital form and incorporated in the database. Secondary
data sources (e.g., paper maps and tabular data) were extensively used for the generation of digital datasets when digital
data were not available or when existing databases were not
deemed adequate for the requirements of DINAS-COAST.
The integration of primary and secondary data sources offers
the most promising approach to compiling up-to-date and scientifically valid global databases (R HIND and CLARK, 1988).
It should be noted that integration of multisource data is
prone to numerous limitations (e.g., variable temporal and
spatial scaling, different formats, and different levels of reliability) but nonetheless essential (TOMLINSON, 1988).
The integration of coastal datasets from such a large variety of data sources constituted a significant challenge as the
spatial data inheritance unavoidably produced variations in
quality, timeliness, and coverage. All digital datasets acquired, and all analogue ones converted to digital form, were
processed and customised to a standard form in order to overcome inherent variations in geographic extent, map projections, and data representation. One of the most important
aspects of this work was the geographic integration of the
datasets collected or generated. Georeferencing can consti-

Figure 2. Example of coastline segmentation in Southeast Asia (approximately 375 segments). Small segments, resulting mainly from variations
in population density, can be seen in the coast of Vietnam, while larger
segments can be seen in the northern coast of Borneo.

tute a major problem even in large-scale (i.e., small-area)

studies (MUELLER, MEISSNER, and WEINREBE, 2002). For
this purpose, a series of global projections were used for performing projection-dependent calculations. These projections
were selected depending on the processing needs and the nature of the individual datasets. After preprocessing was completed, all data were referenced to the common Geographic
Reference System WGS1984 Datum.
Using the storage and analytical powers of a GIS, all database parameters were referenced to the coastline segments
(Figure 3) and were exported in the form required by DIVA
(consistent with the seven geographic features). The process-

Figure 3. An example of a coastline segment (highlighted), and a view

of associated data fields, which are analytically described in the metadata
documentation (Vafeidis et al., 2006). The geographic (latitude and latitude) coordinates of the centre point of the segment are 29.706 and
92.941, respectively.

Journal of Coastal Research, Vol. 24, No. 4, 2008

922

Vafeidis et al.

ing undertaken included GIS operations such as resampling,

proximity analyses, buffering, and map algebra zonal functions (LONGLEY et al., 2001), which were employed for calculating some function or statistic of the input data values located in zones defined by the coastline segments. These zonal
functions were applied to all coastline segments for various
tasks, such as for calculating spatial statistics (e.g., average
population density along a segment) or for quantifying the geometric characteristics of the segments (e.g., coastline length).
Although the coastline segment is a linear unit of reference,
the values of the individual parameters also often include an
implicit depth, as they have been calculated for a coastal zone
that extends inland. The width of the zone varies depending
on the nature of the parameter and the form of the original
data. Therefore, parameters such as coastal slope were calculated directly on the linear representation of the coast (e.g.,
based on the values of pixels intersected by the vector line of
the coast), while others, such as the area of mangroves, were
calculated at variable widths following the extent of the borders of their polygon representation. Analytical descriptions
for how the values of each parameter were calculated can be
found in the DIVA metadata documentation (VAFEIDIS et al.,
2006). The values resulting from these operations were then
attributed to the respective coastline segments, thus compiling
the DIVA database. The database in its final form includes
information on more than 80 physical and socioeconomic parameters at the coast (e.g., Table 2).

Validation
The compilation of the database was followed by extensive
testing which aimed to identify errors or omissions in the
data, ensure that the data had been correctly transferred to
the coastline segments, and examine the internal consistency
of the database. To identify gaps in the global coverage or
potential errors in the data, the analytical and processing
capabilities of GIS were employed and the standard spatial
analysis functions already mentioned in previous sections
were used for exploring the data and comparing them against
other datasets. The methodologies employed varied depending on the type of the dataset. Errors in the original data or
gaps in coverage were identified and corrected utilising data
from additional sources (e.g., local-scale datasets) or on the
basis of the DINAS-COAST consortiums expert knowledge
when data were not readily available. It has to be noted that
when external data sources (e.g., population and elevation datasets) were employed for populating the database, the validation process did not involve a detailed evaluation of the
accuracy of the original datasets used as it was assumed that
this process had been performed by the creators of the datasets. Instead, the validation process could be better described in those cases as a general assessment of the quality
of the employed data and of their suitability in being transferred to the data model used in the DIVA database.
In the second and third steps of the validation process, a
series of algorithms was generated for evaluating the transfer
of data to the unit segments and the internal consistency of
the database. The process, which was fully automated, did
not detect errors but identified areas where improvements in

the spatial referencing of the input data could be made. These

areas contained 2.5% of the total number of segments (approximately 300 segments), and updates were implemented
automatically based on the logging outputs of the algorithms.
The updates primarily included corrections in the process of
referencing attribute data to the geographic features using
additional information from the results of the validation algorithms (e.g., an attribute value based on the proportional
length of a coastal sediment situated within a zone of a specific value rather than based on the mean value of the zones
intersected by the segment). The results of the validation process are described in detail in the associated DIVA metadata
document (VAFEIDIS et al., 2006).

Metadata
Recognising the fundamental importance of effective metadata to describe all constituent datasets and, hence, to support
widespread use of the database by user groups both within and
outside the DINAS-COAST consortium, a detailed metadata information document was created (VAFEIDIS et al., 2006). This
document contains information for all parameters of the DIVA
database, including all geographic features of DIVA (coastline
segments, administrative units, countries, tidal basins, rivers,
world heritage sites, and CLIMBER grid cells). Information is
listed according to the system properties of the DIVA model
(parameters, initialised variables, and drivers) (cf. HINKEL and
KLEIN, 2007). The metadata include detailed descriptions of the
individual parameters and initialised variables, information on
the data properties and on the underlying data sources, a presentation of the methodologies employed for generating datasets, and methods and techniques used for attributing the data
to the coastline segments.

DISCUSSION AND CONCLUSIONS

A new global coastal database has been compiled within
the DINAS-COAST project. This database was specifically
designed to address the needs of vulnerability assessment of
coastal zones under sea-level-rise scenarios. It merges data
of different types and from various sources to create a consistent and coherent source of information based on physical
and socioeconomic parameters within the coastal zone, and it
covers the entire globe, excluding Antarctica. Rather than being a static collection of georeferenced digital data, the DINAS-COAST database provides a flexible data system relevant to current assessment and understanding of coastal vulnerability to sea-level rise. It is in a form amendable to subsequent analysis and progressive upgrade as new and more
detailed datasets emerge or the data requirements of the
DIVA model expand. Importantly, this new database has a
fundamentally different structure from that of all other global datasets, which are primarily based on raster data models.
The representation of the coast in DINAS-COAST is based
on the concept of a linear model. This type of representation
constitutes an improvement compared to earlier linear data
models of the coast, as it does not divide the coast in an arbitrary way but is instead based on a series of scientific criteria (MCFADDEN et al., 2007).
The structure and contents of the database reflect the priorities, aims, and perspectives of the DINAS-COAST project

Journal of Coastal Research, Vol. 24, No. 4, 2008

A New Global Coastal Database

(cf. BARTLETT et al., 1997), and in that sense they have been
defined largely by project-specific needs. However, the methodological and structural advances achieved extend beyond
the boundaries of the DINAS-COAST project. It is expected
that the cross-disciplinary nature of the DIVA database will
support other types of coastal analysis on global and regional
scales, such as global change studies within the International
GeosphereBiosphere Programme and IHDP LOICZ project
(CROSSLAND et al., 2005; KREMER et al., 2004). The global
coverage of the database ensures accessibility to broad-scale
data in areas where detailed information is not readily available. The greatly improved global wetland component of the
database, for example, has underpinned a new and improved
broad-scale model of wetland vulnerability. The model provides a dynamic and integrated assessment of wetland loss
and, drawing on parameters from the database, estimates
transitions among different vegetated types and open water
under a range of scenarios of sea-level rise and changes in
accommodation space through human intervention (MCFADDEN, SPENCER, and NICHOLLS, 2007). In addition to its
potential as a source of input data, the database can be employed as an independent source for the validation of model
results. Finally, it must be noted that due to its global coverage, the information contained in the database is not appropriate for analysis at local scales and should be evaluated
considering all limitations associated with global datasets.
Particular attention must be given to the fact that the scope
and scale of the DIVA database render it inappropriate for
coastal management decisions due to data uncertainties associated with the databases intended use in global and regional climate policy analyses. Nevertheless, the database
can provide a valuable tool for highlighting areas where further in-depth analysis is necessary, thus assisting in the
prioritisation of regions where coastal management at local
scales is more urgently required.
Despite the previously mentioned restrictions, the interdisciplinary nature, the consistency, and the user-friendliness of
the database render it a significant advance in the field of
global coastal databases and can provide an effective basis for
the modelling and analysis of coastal processes at regional and
global scales. Future work may include updating the database
with new data as these become available, upgrading of the
database with larger-scale and higher-resolution data for specific regions, refining the coastline segmentation for regional
to local case studies, and updating the database with localscale data, thus rendering the database suitable for coastal
management issues in those regions. This might require nested databases, as at finer scales a linear data model may be
less appropriate: this is a topic for further research. Such work
can form the basis for creating a database that links data
across a wider range of scales and can therefore constitute an
innovative tool for coastal analysis across different scales.

ACKNOWLEDGMENTS
This work was supported by the DINAS-COAST project,
which received funding from the European Union under contract number EVK2-2000-22024. The contributions of the entire DINAS-COAST consortium, without whom this work

923

would not have been possible, are gratefully acknowledged.

They comprise the Potsdam Institute for Climate Research,
Middlesex University, Hamburg University, Delft Hydraulics, Vrije Universiteit Amsterdam, University of Cambridge,
and DEMIS bv.

LITERATURE CITED
AGARDY, T.; ALDER, J.; DAYTON, P.; CURRAN, S.; KITCHINGMAN, A.;
WILSON, M.; CATENAZZI, A.; RESTREPO, J.; BIRKELAND, C.; BLABER, S.; SAIFULLAH, S.; BRANCH, G.; BOERSMA, D.; NIXON, S.;
DUGAN, P.; DAVIDSON, N., and VOROSMARTY, C. (eds.), 2005. Ecosystems and Human Well-Being: Current State and Trends. Millennium Assessment Report Series: Global Assessment Reports.
Washington, DC: Island Press, pp. 513549.
ALDER, J., 2003. Putting the coast in the Sea Around Us project. The
Sea Around Us Project Newsletter, 15, 12.
BARTLETT, D.; DEVOY, R.; MCCALL, S., and OCONNOR, I., 1997. A
dynamically segmented linear data model of the coast. Marine Geodesy, 20, 137151.
BARTLETT, D.J., 2000. Working on the frontiers of science: applying
GIS to the coastal zone. In: WRIGHT, D.J., and BARTLETT, D.J.
(eds.), Marine and Coastal Geographical Information Systems.
London: Taylor and Francis, pp. 1124.
BRENNER, J.; JIMENEZ, J.A., and SARDIA, R., 2006. Definition of homogeneous environmental management units for the Catalan
coast. Environmental Management, 38(6), 9931005.
BRYAN, B.; HARVEY, N.; BELPERIO, T., and BOURPAN, B., 2001. Distributed process modeling for regional assessment of vulnerability
to sea-level rise. Environmental Modeling and Assessment, 6, 57
65.
BUDDEMEIER, R.W.; SMITH, S.V.; SWANEY, D.P., and CROSSLAND,
C.J., 2002. The Role of the Coastal Ocean in the Undisturbed Nutrient and Carbon Cycles. LOICZ Reports and Studies Series No.
24, pp. 4851.
CHUENPAGDEE, R. and PAULY, D., 2004. Improving the state of
coastal areas in the AsiaPacific region. Coastal Management, 32,
315.
CIESIN AND CIAT STAFFS, 2004. Gridded Population of the World
(GPW), Version 3. Center for International Earth Science Information Network, Columbia University and Centro Internacional
de Agricultura Tropical. Palisades, New York: CIESIN, Columbia
University.
COSTANZA, R.; DARGE, R.; DE GROOT, R.; FARBER, S.; GRASSO, M.;
HANNON, B.; LIMBURG, K.; NAEEM, S.; ONEILL, R.V.; PARUELO,
J.; RASKIN, R.G.; SUTTON, P., and VAN DEN BELT, M., 1997. The
value of the worlds ecosystem services and natural capital. Nature, 387, 253260.
CROSSLAND, C.J.; KREMER, H.H.; LINDEBOOM, H.J.; MARSHALL
CROSSLAND, J.I., and LE TISSIER, M.D.A. (eds.), 2005. Coastal
Fluxes in the Anthropocene. The LandOcean Interactions in the
Coastal Zone Project of the International GeosphereBiosphere
Programme Series: Global Change. Berlin: Springer, 232p.
DINAS-COAST CONSORTIUM STAFF, 2006a. DIVA Coastal Database
Version 1.0, Potsdam Institute for Climate Impact Research (PIK),
Potsdam, Germany. http://www.civil.soton.ac.uk/DIVA and http://
www.dinas-coast.net (accessed July 4, 2006).
DINAS-COAST CONSORTIUM STAFF, 2006b. DIVA 1.0, Potsdam Institute for Climate Impact Research, Potsdam, Germany, CD-ROM.
ELEVELD, M.A.; SCHRIMPF, B.H.W., and SIEGERT, A.G., 2003. User
requirements and information definition for a virtual coastal and
marine data warehouse. Ocean & Coastal Management, 46, 487
505.
ENVIRONMENTAL SYSTEMS RESEARCH INSTITUTE STAFF, 2002. Digital Chart of the World. Redlands, California: ESRI.
FLEWELLING, D.M. and EGENHOFER, M.J., 1999. Using digital spatial archives effectively. International Journal of Geographical Information Science, 13(1), 18.
FRICKER, A. and FORBES, D.L., 1988. A system for coastal description and classification. Coastal Management, 16, 111137.

Journal of Coastal Research, Vol. 24, No. 4, 2008

924

Vafeidis et al.

GANOPOLSKI, A. and RAHMSTORF, S., 2001. Rapid changes of glacial

climate simulated in a coupled climate model. Nature, 409(6817),
153158.
GOODCHILD, M.F., 2000. Foreword. In: WRIGHT, D.J., and BARTLETT, D.J. (eds.), Marine and Coastal Geographical Information
Systems. London: Taylor and Francis, pp. xiiixv.
GORNITZ, V.M.; DANIELS, R.C.; WHITE, T.W., and BIRDWELL, K.R.,
1994. Vulnerability to sea-level rise in the U.S. southeast. Journal
of Coastal Research, Special Issue No. 12, 327338.
HINKEL, J., 2005. DIVA: an iterative method for building modular
integrated models. Advances in Geosciences, 4, 4550.
HINKEL, J. and KLEIN, R.J.T., 2007. Integrating knowledge for assessing coastal vulnerability to climate change. In: MCFADDEN, L.,
NICHOLLS, R.J., and PENNING-ROWSELL, E.C. (eds.), Managing
Coastal Vulnerability: An Integrated Approach. Amsterdam: Elsevier Science, pp. 6178.
HOOZEMANS, F.M.J.; MARCHAND, M., and PENNEKAMP, H.A., 1993.
A Global Vulnerability Analysis: Vulnerability Assessment for Population, Coastal Wetlands and Rice Production on a Global Scale.
Delft, The Netherlands: Delft Hydraulics.
KREMER, H.H.; LE TISSIER, M.D.A.; BURBRIDGE, P.R.; TALAUEMCMANUS, L.; RABALAIS, N.N.; PARSLOW, J.; CROSSLAND, C.J.,
and YOUNG, W., 2004. LandOcean Interactions in the Coastal
Zone: Science Plan and Implementation Strategy. International
GeosphereBiosphere Programme Report 51/International Human Dimensions Programme on Global Environmental Change
Report 18. Stockholm: IGBP Secretariat. 60p.
LANDSCAN STAFF, 2003. Global Population Database. Oakridge National Laboratory, Oakridge, Tennessee.
LONGLEY, P.A.; GOODCHILD, M.F.; MAGUIRE, D.J., and RHIND,
D.W., 2001. Geographic Information Systems and Science. Chichester, United Kingdom: John Wiley & Sons.
MANGOR, K., 2001. Shoreline Management Guidelines. Horsholm,
Denmark: DHI Waste & Environment.
MAXWELL, B.A. and BUDDEMEIER, R.W., 2002. Coastal typology development with heterogeneous data sets. Regional Environmental
Change, 3, 7787.
MCFADDEN, L.; NICHOLLS, R.J.; VAFEIDIS, A.T., and TOL, R.S.J.,
2007. A methodology for modelling coastal space for global assessment. Journal of Coastal Research, 23(4), 911920.
MCFADDEN, L.; SPENCER, T., and NICHOLLS, R.J., 2007. Broad-scale
modelling of coastal wetlands: what is required? Hydrobiologia,
577(1), 515.
MCGILL, J.T., 1958. Map of coastal landforms of the world. Geographical Review, 48(3), pp. 402405.
MCLEAN, R.F.; TSYBAN, A.; BURKETT, A.; CODIGNOTTO, J.O.;
FORBES, D.L.; MIMURA, N.; BEAMISH, R.J., and ITTEKKOT, V.,
2001. Coastal zones and marine ecosystems. In: MCCARTHY, J.J.,
CANZIANI, O.F., LEARY, N.A., DOKKEN, D.J., and WHITE, K.S.
(eds.), Climate Change 2001: Impacts, Adaptation and Vulnerability. Cambridge: Cambridge University Press, pp. 343380.
MILLARD, K. and BRADY, A., 2000. The contribution of technology in
supporting data exchange in integrated coastal zone management.
In: FLEMING, C.A. (ed.), Coastal Management: Integrating Science,
Engineering and Management. London: Thomas Telford, pp. 97
106.
MIMURA, N., 2000. Distribution of vulnerability and adaptation in
the Asia and Pacific region. In: MIMURA, N., and YOKOKI, H.
(eds.), Global Change and AsiaPacific Coasts. Proceedings of
APN/SURVAS/LOICZ Joint Conference on Coastal Impacts of Climate Change and Adaptation in the AsiaPacific Region (Asian
Pacific Network for Global Change Research, Kobe and the Center
for Water Environment Studies, Ibaraki University, Kobe, Japan),
pp. 2125. http://www.survas.mdx.ac.uk (accessed June 20, 2006).
MUELLER, M.; MEISSNER, B., and WEINREBE, W., 2002. TerraMarISTerrestrial and Marine Information System. In: BREMAN,
J. (ed.), Marine Geography: GIS for the Oceans and the Seas. Redlands, California: ESRI Press, pp. 97102.
NATIONAL RESEARCH COUNCIL STAFF, 2004. A Geospatial Framework for the Coastal Zone, National Needs for Coastal Mapping
and Charting Committee, National Research Council of the National Academies. Washington, DC: National Academies Press.
NICHOLLS, R.J. and HOOZEMANS, F.M.J., 2005. Global vulnerability

analysis. In: SCHWARTZ, M. (ed.), Encyclopedia of Coastal Science.

Dordrecht, The Netherlands: Kluwer Academic Publishers, pp.
486491.
NICHOLLS, R.J.; TOL, R.S.J., and HALL, J.W., 2007. Assessing impacts and responses to global-mean sea-level rise. In: SCHLESINGER, M. (ed.), Human-Induced Climate Change: An Interdisciplinary Assessment. Cambridge, U.K.: Cambridge University Press,
pp. 119134.
NICHOLLS, R.J.; KLEIN, R.J.T., and TOL, R.S.J., 2007. Managing
coastal vulnerability and climate change: a national to global perspective. In: MCFADDEN, L., NICHOLLS, R.J., and PENNING-ROWSELL, E.C. (eds.), Managing Coastal Vulnerability: An Integrated
Approach. Amsterdam: Elsevier Science, pp. 221240.
NORDHAUS, W.D., 2006. Geography and macroeconomics: new data
and new findings. Proceedings of the National Academy of Sciences,
103, 35103517.
NURSE, L.A.; SEM, G.J.H.; SUAREZ, A.; POH POH, W.; BRIGUGLIO,
L., and RAGOONADEN, S., 2001. Small island states. In: MCCARTHY, J.J., CANZIANI, O.F., LEARY, N.A., DOKKEN, D.J., and
WHITE, K.S. (eds.), Climate Change 2001: Impacts, Adaptation and
Vulnerability. Cambridge, United Kingdom: Cambridge University
Press, pp. 843875.
PELTIER, W.R., 2000. Glacial Isostatic Adjustment Corrections. In:
DOUGLAS, B.C.; KEARNEY, M.S., and LEATHERMAN, S.P. (eds.),
Sea Level Rise; History and Consequences. San Diego, California:
Academic Press, pp. 6595.
PEUQUET, D.J., 1988. Issues involved in selecting appropriate data
models for global databases. In: MOUNSEY, H., and TOMLINSON,
R. (eds.), Building Databases for Global Science. London: Taylor
and Francis, pp. 6677.
RHIND, D. and CLARK, P., 1988. Cartographic data inputs to global
databases. In: MOUNSEY, H., and TOMLINSON, R. (eds.), Building
Databases for Global Science. London: Taylor and Francis, pp. 79
104.
ROBERTS, A.M. and MOORE, R.V., 1998. Data and databases for decision support. Hydrological Processes, 12, 835842.
SCOTT LEE, D. and SHAN, J., 2003. Combining Lidar elevation data
and IKONOS multispectral imagery for coastal classification mapping. Marine Geodesy, 26, 117127.
SHERIN, A.G., 2000. Linear reference data models and dynamic segmentation: application to coastal and marine data. In: WRIGHT,
D.J., and BARTLETT, D.J. (eds.), Marine and Coastal Geographical
Information Systems, Volume 95116. London: Taylor and Francis.
SHERIN, A.G. and EDWARDSON, K.A., 1996. A coastal information
system for the Atlantic Provinces of Canada. Marine Technology
Society Journal, 30(4), 2027.
SHUPENG, C., 1988. The coastline as a base for global databases: a
pilot study in China. In: MOUNSEY, H., and TOMLINSON, R. (eds.),
Building Databases for Global Science. London: Taylor and Francis, pp. 202215.
SOLURI, E.A. and WOODSON, V.A., 1990. World vector shoreline. Hydrographic Review, 67(1), 2736.
THIELER, E.R. and HAMMAR-KLOSE, E.S., 1999. National Assessment of Coastal Vulnerability to Sea-Level Rise: Preliminary Results for the U.S. Atlantic Coast. U.S. Geological Survey Open File
Report 99-593. http://pubs.usgs.gov/of/of99593 (accessed June 26,
2006).
TOMLINSON, R., 1988. Opening remarks. In: MOUNSEY, H., and
TOMLINSON, R. (eds.), Building Databases for Global Science. London: Taylor and Francis, pp. 19.
TURNER, R.K.; SUBAK, S., and ADGER, W.N., 1996. Pressures, trends
and impacts in coastal zones: interactions between socio-economic
and natural systems. Environmental Management, 20, 159173.
VAFEIDIS, A.; NICHOLLS, R.J.; MCFADDEN, L.; HINKEL, J., and
GRASHOFF, P.S., 2004. Developing a global database for coastal
vulnerability analysis: design issues and challenges. International
Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 35(B), 801805.
VAFEIDIS, A.T.; BOOT, G.; COX, J.; MAATENS, R.; MCFADDEN, L.;
NICHOLLS, R.J.; SPENCER, T., and TOL, R.S.J., 2006. The DIVA
database documentation. http://www.civil.soton.ac.uk/DIVA and
http://www.dinas-coast.net (accessed July 6, 2006).

Journal of Coastal Research, Vol. 24, No. 4, 2008

A New Global Coastal Database

VALENTIN, H., 1952. Die Kusten der Erde. Petermanns Geographische Mitteilungen Erganzungsheft, 246, 118 p.
WEYL, P.K., 1982. Simple information systems for coastal zone management. Coastal Zone Management Journal, 9(2), 155182.

Journal of Coastal Research, Vol. 24, No. 4, 2008

925