Mining diatom algae fossil data for discovering past lake salinity

Climate changes around a large body of water have an intertwined relationship with the salinity of the water and diatom algae growing within it. One may use the diatom algae fossils obtained from bottom of an inland lake to conclude the historical climate changes around the lake and by extension the historical salinity of the water. The discovery of the historical quantified salinity of inland lakes is extremely important to understanding climate change, carbon dioxide levels, and global warming. In this research effort, the past salinity levels for Santa Fe Lake located in New Mexico, USA, were discovered by mining the data of diatom algae fossils. Modified Rough Sets as the first component of the proposed hybrid system were used to establish the relationships between diatom algae data, expressed in linguistic values, and the climate changes. The established relationships were extended to embrace the linguistic values of water salinity. The outcome was a set of fuzzy patterns. Fuzzy Logic as the second component of the proposed hybrid system was employed to: (i) provide the membership functions for the different linguistic values of the salinity and (ii) produce a crisp value for the salinity of the water related to each slice of diatom fossil using the crisp values of algae abundance indices in each slice. The validity of the findings was tested which revealed 72% of accuracy for the produced results.