Sandro Nunes de Oliveira

Comparative analysis of MODIS time-series classification using support vector machines and methods based upon distance and similarity measures in the Brazilian cerrado-caatinga boundary

We have mapped the primary native and exotic vegetation that occurs in the Cerrado-Caatinga transition zone in Central Brazil using MODIS-NDVI time series (product MOD09Q1) data over a two-year period (2011–2013). Our methodology consists of the following steps: (a) the development of a three-dimensional cube composed of the NDVI-MODIS time series; (b) the removal of noise; (c) the selection of reference temporal curves and classification using similarity and distance measures; and (d) classification using support vector machines (SVMs). We evaluated different temporal classifications using similarity and distance measures of land use and land cover considering several combinations of attributes. Among the classification using distance and similarity measures, the best result employed the Euclidean distance with the NDVI-MODIS data by considering more than one reference temporal curve per class and adopting six mapping classes. In the majority of tests, the SVM classifications yielded better results than other methods. The best result among all the tested methods was obtained using the SVM classifier with a fourth-degree polynomial kernel; an overall accuracy of 80.75% and a Kappa coefficient of 0.76 were obtained. Our results demonstrate the potential of vegetation studies in semiarid ecosystems using time-series data.

Mapeamento da vegetação na floresta atlântica usando o classificador de árvore de decisão para integrar dados de sensoriamento remoto e modelo digital de terreno

The management and ecological monitoring of national parks and other protected areas requires a detailed description of the vegetation distribution patterns. This paper aims to produce a vegetation map for the Serra dos Orgaos National Park (PARNASO). This conservation unit is localized in Atlantic Forest within a topographic variation from sea level to 2,263 meters. The vegetation classification based on the ASTER satellite data, high-resolution aerial photographs and Digital Elevation Model (DEM). The DEM indicates vegetation structures in landscape with high spatial variability because it correlates with environmental factors, such as microclimate, moisture, soil and geomorphological processes. Decision tree classifier was used to extract information of DEM and remote sensing data. Seven classes were identified: Agropecuaria (1.29% of total Park area), Campos de Altitude (24.27%), Floresta Ombrofila Densa Alto-Montana (37.47%), Floresta Ombrofila Densa Montana (21.54%), Floresta Ombrofila Densa Sub-Montana (5.22%), Floresta Secundaria (4.13%), and no vegetation area (6.08%). The three highest physiognomies were associated with altitude higher than 1,000 m and represented 55.5% of the total area. The construction of decision trees combining the DEM and remote sensing information can improve the result on the forest tropical distribution.

Modelagem de espectros temporais NDVI-MODIS, no período de 2000 a 2008, na bacia do rio Paracatu, Brasil

The objectives of this work were to model the distribution of NDVI-MODIS data, in six distinct targets: Crop Land Areas, Gallery Forest, Cerrado, Pastureland, Urban Areas and Semideciduous Seasonal Forest; and also to analyze the physiognomies changes of the index through of waveform models, period 2000-2008, at the Paracatu River Basin (Sao Francisco Sub-River Basin). The materials and methods were detached in the steps: (a) getting MODIS images; (b) noise treatment; (c) math modeling of the NDVI temporal signature; and (d) analysis of statistical relation between NDVI and open canopy. The nonlinear regression, waveform model, to spread NDVI by temporal series, got high significance of its parameters (R2 and p) and tolerable error. With this could identify the physiognomies to be delimited, and to simulate future temporal series by land use, vegetation cover and area extension of each physiognomy. The standard of distribution of the NDVI data showed significant relation with field data of open canopy, this could to be a precise indicator of land use change between the temporal series.

Identification of areas prone to shallow landslide in Parque Nacional da Serra dos Órgãos (Brazil) considering seasonal rainfall

Mathematical modeling is being increasingly used to predict events occurring in nature. Within the diverse existing models, one which stands out is the SHALSTAB. This model of prediction of shallow landslide occurrence was applied in Parque Nacional da Serra dos Órgãos (PARNASO) by using data for average monthly pluviosity, aiming to identify, within the landscape, the spatial variability at places prone to shallow landslide throughout the year. The methodology is sectioned into the following stages: a) elaboration of the digital elevation model (DEM) and its derived maps, such as slope and contribution area, b) application of the SHALSTAB model, considering the various events of rainfall throughout the year, and c) quantification of areas prone to landslide for each rainfall event occurred. The model results indicate the dynamics of the locations which present instability due to the seasonality of rainfall intensity.