Page 198 - Introduction to Mineral Exploration

9: MINERAL EXPLORATION DATA 181

Vector 100 m cells 250 m cells
10
km

Legend Granite Mylor beds
N
Sea Quartz porphyry dykes Gramscatho group

FIG. 9.2 Vector and raster representation of lithologies from a geological map. The area in SW Cornwall,
England and the original map was compiled at 1:250,000 scale based largely on a map of the British Geological
Survey. Note the impact of using different cell sizes in a raster map. The lithology of the rasterized cell is
taken from the lithology with the largest area within the cell.

model, polygons are formed by the use of TABLE 9.2 Advantages and disadvantages of raster
software as a mesh of lines, often known as and vector formats. (After Bernhardsen 1992.)
arcs, that meet at nodes. Another variation of
this model often used for height data is that of Raster Vector
the triangular irregular networks (TIN) and is
used to visualize digital elevation surfaces or Data collection Rapid Slow
Small
Data volume
Large
construct digital terrain models (DTM). The Data structure Simple Complex
TIN model is similar to the polygons used in Graphical treatment Average Good
ore resource and reserve calculation (see sec- Geometrical accuracy Low High
tion 10.5). Area analysis Good Average
Generalization Simple Complex
9.1.3 Storage methods
The simplest solution for data storage is that
of the flat file method in which each point has duplicate tuples are allowed; and (iv) tuples can
associated x, y (and z) coordinates, as well as be rearranged without changing the nature of
attributes. This is the familiar style of an the relation (Bonham-Carter 1994). The flat file
accounting ledger or table and implemented method is however an inefficient way to store
electronically in a spreadsheet, such as Lotus data as a minor change, for example, a change
1-2-3 and Microsoft Excel. In this format to the name of a lithology in Table 9.3 (which is
(Table 9.2) attributes are stored in columns or part of the table associated with Figs 9.4 & 9.5),
fields, and rows, that are known as tuples. The requires a global search and change of all exam-
features of this type of storage are that: (i) all ples. Data are more efficiently stored and edited
the data are represented in the table; (ii) any cell in a relational database in which the data are
in the table must have a single value, replic- stored as a series of tables linked by unique
ate samples require additional tuples; (iii) no keys, such as sample numbers. The flat file