@Article{AndradesFoRoBeMeVaCrOl:2014:MaNeQu,
author = "Andrades Filho, Cl{\'o}dis de Oliveira and Rossetti, Dilce de
F{\'a}tima and Bezerra, Francisco Hil{\'a}rio and Medeiros,
Walter Eug{\^e}nio and Valeriano, M{\'a}rcio de Morisson and
Cremon, {\'E}dipo Henrique and Oliveira, Roberto Gusm{\~a}o de",
affiliation = "{} and {Instituto Nacional de Pesquisas Espaciais (INPE)} and {}
and {} and {Instituto Nacional de Pesquisas Espaciais (INPE)} and
{Instituto Nacional de Pesquisas Espaciais (INPE)}",
title = "Mapping Neogene and Quaternary sedimentary deposits in
northeastern Brazil by integrating geophysics, remote sensing and
geological field data",
journal = "Journal of South American Earth Sciences",
year = "2014",
volume = "56",
pages = "316--327",
month = "Dec.",
keywords = "Barreiras Formation, Post-Barreiras sediments, Gamma-ray,
Morphometry.",
abstract = "Neogene and late Quaternary sedimentary deposits corresponding
respectively to the Barreiras Formation and Post-Barreiras
Sediments are abundant along the Brazilian coast. Such deposits
are valuable for reconstructing sea level fluctuations and
recording tectonic reactivation along the passive margin of South
America. Despite this relevance, much effort remains to be
invested in discriminating these units in their various areas of
occurrence. The main objective of this work is to develop and test
a new methodology for semi-automated mapping of Neogene and late
Quaternary sedimentary deposits in northeastern Brazil integrating
geophysical and remote sensing data. The central onshore
Para{\'{\i}}ba Basin was selected due to the recent availability
of a detailed map based on the integration of surface and
subsurface geological data. We used airborne gamma-ray
spectrometry (i.e., potassium-K and thorium-Th concentration) and
morphometric data (i.e., reliefdissection, slope and elevation)
extracted from the digital elevation model (DEM) generated by the
Shuttle Radar Topography Mission (SRTM). The procedures included:
(a) data integration using geographic information systems (GIS);
(b) exploratory statistical analyses, including the definition of
parameters and thresholds for class discrimination for a set of
sample plots; and (c) development and application of a
decision-tree classification. Data validation was based on: (i)
statistical analysis of geochemical and airborne gamma-ray
spectrometry data consisting of K and Th concentrations; and (ii)
map validation with the support of a confusion matrix, overall
accuracy, as well as quantity disagreement and allocation
disagreement for accuracy assessment based on field points. The
concentration of K successfully separated the sedimentary units of
the basin from Precambrian basement rocks. The reliefdissection
morphometric variable allowed the discrimination between the
Barreiras Formation and the Post-Barreiras Sediments. In addition,
two units of the latter (i.e., PB1 and PB2) previously mapped in
the field were promptly separated based on Th concentration. A
regression analysis indicated that the relationship between
geophysical and geochemical values obtained for the PB1, PB2 and
Barreiras Formation is significant (R-squared = 0.91; p-value
<0.05). Map validation presented a high overall accuracy of 84%,
with a coefficient of quantity disagreement of 12% and a
coefficient of allocation disagreement of 8%. These results
indicate that the methodology applied in the central onshore
Para{\'{\i}}ba Basin can be successfully used for mapping the
Barreiras Formation and Post-Barreiras Sediments in other areas of
the Brazilian coast. The ability to rapidly and precisely map
these units using such methodology could reveal their geographic
distribution along the northeastern coast of Brazil.",
doi = "10.1016/j.jsames.2014.09.018",
url = "http://dx.doi.org/10.1016/j.jsames.2014.09.018",
issn = "0895-9811",
label = "lattes: 0307721738107549 2 AndradesFoRoBeMeVaCrOl:2014:MaNeQu",
language = "en",
targetfile = "1-s2.0-S0895981114001308-clodis.pdf",
urlaccessdate = "27 abr. 2024"
}