A ground characterisation exercise was conducted in order to identify
sites suitable for a blast furnace at the Redcar steelworks in northea
st England. The purpose of this exercise, conducted some 20 years afte
r the investigations were carried out and a decade after the steelwork
s had been constructed, was to develop techniques for managing spatial
data using Geographical Information Systems (GIS) and geostatistics.
Borehole details, from the factual site investigation reports, were en
tered into a database using a relational structure with linked files.
The borehole location and lithology files were keyed in using codes fo
r geological material, weathering grade, strength, borehole type and v
erification status. Laboratory and field test data were input by scann
ing tables in the ground investigation reports and processing the scan
ned images using optical character recognition software. The XBase pro
gramming language was used to execute graphing, GIS and geostatistics
programs; thereby enabling repetitive tasks to be carried out by the c
omputer. This both saved time and reduced errors and inconsistencies t
hat could result from repeated manual execution of the same program. S
ome simple spatial operations could also be carried out entirely from
within dBase, an example being the use of moving-window statistics emp
loying rectangular windows aligned parallel to the axes of the coordin
ate system to study local stationarity. GIS have been found to offer a
n effective tool for analysing, querying and presenting spatial site i
nvestigation information. Geostatistical kriging and simulation techni
ques offer powerful spatial modelling tools for creating visualisation
s of the spatial variability of parameters measured at points during a
ground investigation. The Redcar case history demonstrates how a data
base management system (DBMS) can be used to manipulate site investiga
tion data as part of a modelling and interpretation exercise. DBase so
urce code allows others to follow the same path or build upon the foun
dations herein. (C) 1998 Elsevier Science B.V. All rights reserved.