International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B4. Istanbul 2004
therefore stored in a unified table structure, a form which has
been found to facilitate the use of databases by
multidisciplinary scientists (see also Mueller er a/., 2002).
PU -
(
Pme ) H
7 x.
i Ae od
f SET A
§ ANT ee
> CSN ar
EN V t
e: A eR
cA = a pron
MP MA Su b I
P ep le $6 SLTRWGT 0.3
RE SL A WAVECLIM 2
S Mo Si
y uS v T, TIDALRNG 1
CUN eo er UPLIFT 0.02
MNA oe Padi, = SY" "779. |RIVERID 5999
AS C A erit AN oet BASINID -9999
€ ; AREAL 143
Ex j res an
Ty pd C 2 4
mm ue AREA3 $
AREA4 3
AREAS $
AREAS 12 40
: ; |asiD 723
T E» / LENGTHY 317.041
€ E / RED D d
(ss / UNTRIE 9;
y i VAT T ir MAJORITY 16
/ nd ORDER. ID2 -9999
| CLSFID 722
A COUNTRID_! 188
A 4 | TOURIST AR 7083000
* fi |pcouvrer 187
(4/4 |BasinF 9999 Y
odas T »
S J^ [sl | d
Figure 2: Example of coastline segment and associated attribute
data
4. DISCUSSION
The representation of the coast in DINAS-COAST constitutes
an improvement compared to other linear data models that have
been used in the past for representing the coast. Firstly, the
decomposition of space is not performed in an arbitrary way but
is instead based on a series of scientific criteria. Secondly, this
type of representation produces an organized and coherent data
structure that reduces effectively the complexity of the reality
to the minimum required. Moreover, it achieves all the above in
a way that is compatible with the specific demands of the
product of the project, the DIVA tool. One of the limitations of
the employed methodology is the fact that the linear segments
represent the coast in a static way. This limitation stems from
the reliance of the technique on a particular linear feature.
However, this is a general problem which is common to all
linear data reference models of the coast, including dynamic
segmentation (Sherin, 2000). It is noteworthy that most impact
and vulnerability analyses have employed some form of coastal
segmentation at an early stage of analysis. Such an example is
the first Global Vulnerability Assessment where segmentation
was essentially based on national boundaries, producing 192
segments (Hoozemans et al., 1993).
The data model that has been used has played a fundamental
role in achieving two of the main targets of DINAS-COAST,
firstly the inclusion of the complete database within the DIVA
tool that will be supplied to the end users and secondly the aim
to address the needs of a wider community of researchers and
policy makers. The first aim has been achieved without
compromising the level of detail of the information contained in
it. This is demonstrated by the fact that the coastline
segmentation has produced two orders of magnitude more
segments than Hoozemans ef al. (1993). Nevertheless, due to its
resolution, the information contained in the database is not
designed for use at local scales and should be evaluated
considering all the limitations that are associated with global
804
datasets. The second aim has been achieved by making the
database available in an easily-accessible form as its use does
not require specialized software and hardware. This makes the
information included in the DIVA tool accessible to a wide
range of coastal scientists who have been unable to utilize such
information in the past. Within this context, the data model
employed by DIVA and the methods used for transferring the
GIS data to this model can be viewed as a methodological tool
for making geodata available to the broader coastal science
community and to a wider range of disciplines.
S. CONCLUSIONS
The methodology developed within the context of DINAS-
COAST has led to the generation of a global coastal database
for the purpose of vulnerability assessment. The structure and
contents of the database reflect the priorities and perspectives
of the project (Bartlett ef al., 1997) and in that sense these
factors have been defined to a large extent by project-specific
needs. However, the methodological and structural advances
achieved can impact a wider range of applications in a more
general context.
The DIVA database has a fundamentally different structure to
all other global datasets, which are primarily based on raster
data models. This includes the IGBP-LOICZ coastal typology
which is raster based (IGBP-LOICZ, 2003). It is expected that
the cross-disciplinary nature of the database developed and its
consistency and user-friendliness will satisfy, to a large extent,
current information needs of the wider coastal-research
community and will assist coastal scientists in the modelling
and analysis of coastal processes at regional and global scales.
It is proposed to make the DIVA database publicly available via
the International Geosphere Biosphere Program Land Ocean
Interaction in the Coastal Zone (IGBP-LOICZ).
REFERENCES
Bartlett, D., R. Devoy, S. McCall, and I. O'Connor, 1997. A
Dynamically Segmented Linear Data Model of the Coast,
Marine Geodesy, 20:137-151.
Bartlett, D.J., 1993. GIS and the Coastal Zone: An Overview,
Explorations in Geographic Information Systems Technology:
Applications in Coastal Zones Research and Management, (D.
J. Bartlett, ed), vol. 3, Clark Labs for Cartographic Technology
and Analysis, Worcester, Massachussets, pp. 11-24.
Bartlett, D.J., 2000. Working on the Frontiers of Science:
Applying GIS to the Coastal Zone, Marine and Coastal
Geographical Information Systems, (D. J. Bartlett, ed), Taylor
and Francis, London, pp. 11-24.
Debusschere, K.S., S. Penland, K.A. Westphal, P.D. Reimer,
and R.A. McBride, 1992. Aerial Videotape Mapping of Coastal
Geomorphic Changes, Seventh Symposium on Coastal and
Ocean Management (Coastal Zone '89), July 8-12, Long Beach,
CA, pp. 370-390.
Eberhart, R.C., and T.J. Dolan, 1980. Chesapeake Bay
Development Pressures: Rams Database Analysis, Second
Symposium on Coastal and Ocean Management (Coastal Zone
80), New York. American Society of Civil Engineers.
Inter:
Frick
Descı
137.
Good
Geog
and F
Hink:
Tool
LOIC
Hooz
A Gl
Popu
Scale
IGBF
Prog!
http:/
Leatt
video
apprc
Rese
Lock
and
Cont:
Mari
Bartl
McF:
Datal
Coas
23M
Mout
for G
Muel
- Ti
Geog
ESRI
Sheri
Dyna
Data,
(D. J
pp.
Sheri
syste
Tech
Shup
A pil
(R. T
Wey|
Zone
Q1.
