Processing heterogeneous information is necessary in many domains, but it is specially important in the study of complex systems like geologic phenomena responsible for natural hazards. The developments in sensor, communication and computer technologies allows the monitoring and storage of different geophysical, geochemical and geological variables on a continuous base. In addition, field observation and laboratory analysis incorporate more information, also changing with time. Some of the problems associated with these large masses of information are: (a) the presence of heterogeneous variables (i.e. ratio, interval, nominal, ordinal, and other more complex like images, spectra, etc.), (b) the uncertainty associated with the observed variables in terms of observation or measurement errors, vagueness, subjectivity, etc, and (c) the incompleteness due to monitoring gaps, irregular sampling frequencies and missing data.