Sixteenth International Congress of the Brazilian Geophysical Society Visualization of Euler deconvolution estimates on traditional and plateau plots Felipe F. Melo* and Valéria C. F. Barbosa, Observatório Nacional Copyright 2019, SBGf - Sociedade Brasileira de Geofísica This paper was prepared for presentation during the 16 th International Congress of the Brazilian Geophysical Society held in Rio de Janeiro, Brazil, 19-22 August 2019. Contents of this paper were reviewed by the Technical Committee of the 16 th International Congress of the Brazilian Geophysical Society and do not necessarily represent any position of the SBGf, its officers or members. Electronic reproduction or storage of any part of this paper for commercial purposes without the written consent of the Brazilian Geophysical Society is prohibited. ____________________________________________________________________ Abstract Euler deconvolution is a popular method for interpretation of potential field data. The method estimates the position and the base level for a given nature (i.e., structural index – SI) of a geologic source based on the magnetic dada. These estimates are computed inside a moving data window spanning the whole dataset. The selection of the best estimates is related to the plot of these estimates. Traditionally, for each SI, the estimates of source horizontal coordinates obtained for each position of a moving data window are plotted in the plan view of the study area and the estimates of source depth are represented by different color or size. This traditional plot requires a criterion to select the best estimates and deal with the spread of solutions. In this case, mostly, the base- level estimates are computed but they are not used or shown. We compare the traditional plot of the estimates from magnetic fields with an alternative plot where, for each SI, every estimate is plotted against the central position of the moving data window. At this plot, when the correct SI is used the solutions form a plateau of constant values. We perform the analysis for synthetic data sets generated by different SI and for a real data set, beyond that; we plot base-level estimates in the traditional way. We run Euler deconvolution over the Diorama anomaly, located in the Goiás Alkaline Province, Brazil, and analyze the estimates over different plots. From the results of both plots we may infer that Diorama anomaly is generated by at least two vertical plugs. Introduction Euler deconvolution (Reid et al., 1990; Barbosa et al., 1999) has become the most popular technique for interpretation of potential-field data because it can process data sets fast. Euler deconvolution works with potential- field measurements, their gradients, and a given integer number called the structural index (SI), which in turn depends on the nature (type) of the geologic source (Reid et al., 2014; Reid and Thurston, 2014). Mathematically, Euler deconvolution assumes a tentative SI, solves a system of equations in a moving data window scheme and estimates the four parameters: base level, horizontal and vertical positions of a geologic source. Traditionally, for each tentative SI, one map is generated where the estimates of the horizontal and vertical coordinates obtained by Euler deconvolution are plotted. Specifically, the Euler deconvolution estimates of the horizontal coordinates are plotted in the plan view (− plane) of the study area (Reid et al., 1990), and usually vertical estimates are plotted in different color or size. In this case, base-level estimates are usually neglected. When the correct SI is used the estimates cluster over the source (Thompson, 1982). The cumbersome with this plot is the large amount of solutions (Barbosa and Silva, 2011). In order to deal with this spray of solutions and define the correct SI some authors developed some criteria, e.g., Thompson (1982) and Reid et al. (1990). FitzGerald et al. (2004) provided an overview on best practices to select the solutions of Euler deconvolution. On the other hand, to define the correct SI, Barbosa et al. (1999) used the minimum correlation between base-level estimates and the observed anomaly. Alternatively, to the traditional plot, Silva and Barbosa (2003) proposed to plot each coordinate estimate against the central position of the moving data window. When the correct SI is assumed, depth estimates define a plateau over the source; however horizontal estimates always define a plateau over the source (Silva et al., 2001) regardless the assumed SI. Melo et al. (2013) took advantage of this behavior to locate automatically the horizontal plateaus. Melo and Barbosa (2018) worked with base-level estimates on this plateau plot and showed that over the source the depth and base-level estimates define plateaus if the correct SI is assumed, and no interfering anomalies are present. In case of interfering anomalies, the minimum standard deviation of base-level estimates over a plateau defines the correct SI. The plateau is delineated either by the depth or horizontal estimates. In this work, we analyze Euler deconvolution solutions displayed in the traditional (Reid et al., 1990) and in an alternative way, the plateau plot (Silva and Barbosa, 2003; Melo and Barbosa, 2018). Furthermore, we analyze base- level estimates plotted in the traditional way. To this end, we generate magnetic anomalies from single-point sources related to different SI. We simulate two scenarios: a single source and two sources with distinct SI. Application on real data from part of the Goiás Alkaline Province, Brazil, shows that both plots lead to similar results and we infer that at least two vertical plugs generate the Diorama anomaly. Method Reid et al. (1990) defined the Euler deconvolution as  −      +  −      +  −      =  − ℎ, (1) where  is the SI,  is a base level or background value,   ,   and   are the source positions, ,  and  are the observation positions, ℎ = ℎ, ,  is the total-field anomaly and ℎ  ⁄ , ℎ  ⁄ and ℎ  ⁄ are the gradients with respect to the variables ,  and , respectively.