S Y S D E C O I T A L I A
P A G I N A | 0 2
Classification was performed using 163 sea truth training samples, while classification accuracy
was evaluated using an additional 76-98 validation samples. The user's accuracy, the producer's
accuracy and the K-index were also determined.
The collected mapping data were processed in a GIS
environment by which the multitemporal variations
in the distribution of the transplanted P. oceanica
prairie were monitored. Specifically, an overall
classification accuracy ranging from 93.67% to
99.02% was obtained, while the K-index ranged from
0.90 to 0.98.