Figure 7. Factor-pattern curves derived from analyzing the Minolta data set with a reduced number of samples designed to be equivalent to the shore-based Perkin-Elmer data set. This reduced data set contains the first-derivative values from 1190 samples. Note that even though the number of samples per section was reduced from about 5.5 per section in the full data set to about 1 per section in this data set, the factors are nearly identical. This suggests that closely spaced spectral sampling is unnecessary if the primary goal is to characterize sediment composition.