Sea surface temperature clustering based on type-2 fuzzy theory

Spatial data clustering is an effective method to find interesting spatio-temporal clustering patterns. There are many uncertainties in sea surface temperature (SST) clustering, so clustering methods with uncertainty must be used. Type-2 fuzzy theory takes into account the uncertainty of membership grade while fuzzy C means (FCM) not. Based on the analysis of interval type-2 fuzzy C means (IT2FCM), the paper utilizes two normal cloud models to express fuzzifiers m₁ and m₂, and uses two cloud drops to substitute them. The method considers the uncertainty of two fuzzifiers, and avoids many times of repeated tests, which reduces computation cost. The paper applies the improved IT2FCM into global SST clustering, and discovers some interesting climate patterns.