A FUZZY MODEL FOR ESTIMATING HEAVY METALS IN AN ANTARTIC SOIL

MAURO A.E. CHAPARRO; MARCOS A.E. CHAPARRO; ANA M. SINITO

Congreso; Segunda Reunión Bienal de la Asociación Latinoamericana de Paleomagnetismo y Geomagnetismo (LATINMAG); 2011

LATINMAG

Magnetic monitoring is well-known tool and constitutes an alternative for assessing pollutants in different environments. In this contribution, a methodology is developed for building a model of Takagi-Sugeno-Kant type and their confidence intervals. The model yields values and confidence intervals of heavy metals (Pb, Cr, Cu and Zn) from soil samples collected in Base Marambio (Antarctica) In this work, the input variables are: specific magnetic susceptibility (χ), anhysteretic remanent magnetisation (ARM), anhysteretic susceptibility to volumetric susceptibility - ratio (κARM/κ), and remanent coercivity (Hcr). Fuzzy c-means clustering (FCM) was used to partition data, hence the membership functions and rules are built. In addition, a model of fuzzy lineal regression was used to construct the (output) consequent functions. The results show not only satisfactory agreement between the model and data, but also provide valued information from the rules analysis that allows us to understand the behaviour of the magnetic and heavy metal variables.