Caracterización de zonas contaminadas por derrames recientes de hidrocarburos mediante métodos geofísicos

multifrequency electromagnetic induction systems, composed by two small coils separated by a fixed distance (emi or slem, emi stands for frequency-domain electromagnetic induction and slem stands for small-loop frequency-domain electromagnetic induction), are very useful for the prospection of the first tens of meters of the subsoil, and can be employed to detect quickly and efficiently a great variety of buried structures. one coil generates a known inductive magnetic field and another detects the variations of the induced field as the whole system is displaced over the ground, along acquisition lines which are generally parallel to each other. usually, these data are interpreted by direct visualization of the plan-views of the responses measured at each frequency. besides, to obtain quantitative information, 1d inversion methods are applied and then, their results are stitched together to generate 2d and 3d images. however, due to the lack of correlation between the 1d inversion results corresponding to neighboring points, these images usually present numerous undesirable lateral effects that deteriorate the visualization of the target structures and even can prevent their detection. the importance of this problem is higher for studies of resistive electrical anomalies and with poor signal-to-noise ratio. in this work, we present the results of two emi studies carried out in a site contaminated by a recent hydrocarbon spill. the first one was performed two weeks after the spill and the second three months later, once the remediation works had started. the second time, three dipole-dipole geoelectric lines were also acquired, for comparison. we applied to the emi data a conventional 1d inversion method, in combination with novel lateral filtering techniques that allowed for strongly reducing the spurious lateral variations previously mentioned. besides, in this work we propose a method to estimate the real investigation depth achieved in emi surveys, through the ca