430
N.J. Brown et al.
does degrade response times compared to transforming the data into some Arc specific
format: however, the increase in response time is out-weighed by the benefit of having only
one image of the data. The system has now reached a point where external decision makers
could be productively involved.
ITE has created similar products for the analysis of various environmental subjects
at a range of scales; these have included:
the DOE funded climate change Core model (Parr and Eatherall, 1994)
the MAFF funded ITE Wetlands GIS (Brown et al., 1997)
the EC funded LANDECONET demonstrator (Brown and Firbank, 1997)
In one important respect the DSS has adopted an alternative approach to that used in these
projects. These three products have used various software including ArcInfo and ArcView
GIS, but in each case the majority of datasets used have been converted into the format
preferred by the host software. In the DSS, data created for LOIS/BIOTA are maintained in
their original form.
While the system is designed primarily to stimulate scientific cooperation and
information exchange between researchers, it also provides secure and convenient access to
a wide variety of data and models and will provide an aid for effective coastal management.
The conceptual model of the flow and storage of information within the system is
shown in Fig. 3.
The DSS is being developed in the Arc Macro Language (AML) of the ArcInfo
GIS system on UNIX workstations. The decision to implement the system in AMLs was
mainly prompted by the speed and ease with which it can be developed and modified in
response to the needs of its users, the BIOTA research community.
Within the DSS a series of menu interfaces have been developed, which
incorporate some of the analysis methods appropriate for the areas of study. Some of these
GIS facilities are standard operations such as raster overlay, whereas more complex facilities
have also been implemented.
Some of the main facilities offered by the GUI are LOIS specific maps, physical
data (such as profiles), biological data, national maps (such as Ordnance Survey), and GIS
and modelling functions.
A central feature of the design is that data are stored in their "native" format.
Within every directory a "readme" file is kept which provides a brief description (source,
date, accuracy, resolution etc.) of all the data, files, programmes and sub-directories which
are present. UNIX shell programmes are started at regular intervals by the "cron" facility to
ensure that "readme" files are kept up to date. Users of the system only have "read" access to
the data. Use of "permissible" combinations or analysis of data is controlled by the menu
system. Because of this design approach the system can be distributed across different
machines on the local area network (LAN).
More complex analysis includes the comparison of the estimated distribution of a
plant species derived from: remotely sensed data (spectral response), quadrat samples
(botanical survey) and modelled from exposure and inundation rates.
Précédent

- 435/596

Suivant