**6. Conclusions**

 In summary, the comparison of these classifications: band ratio, PCA, Crosta, and Fraser techniques illustrates the difficulties and variants inherent in these classification methods. No single procedure appears to optimize all the parameters of interest. This set exemplifies potential variants in field explorations and shows the path for choosing the best alternative, or combination thereof, of interpretation for the problem at hand. For instance, the Fraser technique succeeded in identifying basalt flows covered by vegetation in the north half of the island but failed to locate the submerged basalts that were identified with the Crosta technique. In turn, the latter failed to identify the basalts covered by vegetation. Hydrothermal alterations are identified with different degrees of precision, indicating that if such is the target, one should evaluate various alternatives to optimize results. The analysis of the lineaments in the island, easily obtained by remote sensing, is associated with surface geological features and deep geophysical properties. They should be incorporated as frequently as possible into this type of studies.

The presence of hydrothermal activity in Isla Isabel extends almost fully throughout the island. Detection of such an activity under vegetated areas shows now that the northern half of the island is also the site of extended alterations. The possibility of basalt flooding in this region is high, and it should be the subject of detailed mapping. The high heat flow in the area, the existence of a laccolith under Isla Isabel, the extensive phreatomagmatic activity, and the presence of hydrothermal alterations in the island point to the relevance of this region as a potential geothermal source of major importance.
