Osmar Abílio de Carvalho Júnior

A New Approach to Change Vector Analysis Using Distance and Similarity Measures

The need to monitor the Earth’s surface over a range of spatial and temporal scales is fundamental in ecosystems planning and management. Change-Vector Analysis (CVA) is a bi-temporal method of change detection that considers the magnitude and direction of change vector. However, many multispectral applications do not make use of the direction component. The procedure most used to calculate the direction component using multiband data is the direction cosine, but the number of output direction cosine images is equal to the number of original bands and has a complex interpretation. This paper proposes a new approach to calculate the spectral direction of change, using the Spectral Angle Mapper and Spectral Correlation Mapper spectral-similarity measures. The chief advantage of this approach is that it generates a single image of change information insensitive to illumination variation. In this paper the magnitude component of the spectral similarity was calculated in two ways: as the standard Euclidean distance and as the Mahalanobis distance. In this test the best magnitude measure was the Euclidean distance and the best similarity measure was Spectral Angle Mapper. The results show that the distance and similarity measures are complementary and need to be applied together.

Identificação regional da Floresta Estacional Decidual na bacia do Rio Paranã a partir da análise multitemporal de imagens MODIS

Parana river basin has one of the major fragments of Decidual Seasonal Forest in Brazil. This vegetation is widely fragmented due to the selective wood exploitation and the growth of pasture areas, what justifies the development of studies in order to understand its dynamics and preserve its diversity. Thus, the present study aimed at defining a method for regional identification of the Deciduous Forest in the Parana river basin. The deciduous forest has a typical phenological cycle in comparison with other savanna physiognomies. Due these characteristics, a temporal series of normalized difference vegetation index (NDVI) images of the MODIS sensor was used for its detection. The adopted methodology may be subdivided into the following steps: (a) elaboration of the 3D cube of NDVI images, where the z profile corresponding to temporal signature or NDVI spectrum, (b) noise elimination using the Minimum Noise Fraction (MNF) transformation, and (c) NDVI temporal variability examination of deciduous forest vegetation, with the establishment of the best NDVI band applied in the vegetation index differencing method. The Deciduous Forest presents a typical NDVI spectral behaviour, with higher values in the raining season and lower values in the dry season, what makes this kind of vegetation different from others. The employment of a changing detection algorithm between two images: one for the dry season and the other for the raining season enhances the localization of the Decidual Seasonal Forest. So, the methodology has proved to be effective for regional delimitation of Deciduous Forests considering the MODIS sensor. Considering the changing detection method, Deciduous Forest region is characterized by presenting NDVI alteration values.

Spatialization of climate, physical and socioeconomic factors that affect the dairy goat production in Brazil and their impact on animal breeding decisions

Brazil has high climate, soil and environmental diversity, as well as distinct socioeconomic and political realities, what results in differences among the political administrative regions of the country. The objective of this study was to determine spatial distribution of the physical, climatic and socioeconomic aspects that best characterize the production of dairy goats in Brazil. Production indices of milk per goat, goat production, milk production, as well as temperature range, mean temperature, precipitation, normalized difference vegetation index, relative humidity, altitude, agricultural farms; farms with native pasture, farms with good quality pasture, farms with water resources, farms that receive technical guidance, family farming properties, non-familiar farms and the human development index were evaluated. The multivariate analyses were carried out to spatialize climatic, physical and socioeconomic variables and so differenciate the Brazilian States and Regions. The highest yields of milk and goat production were observed in the Northeast. The Southeast Region had the second highest production of milk, followed by the South, Midwest and North. Multivariate analysis revealed distinctions between clusters of political-administrative regions of Brazil. The climatic variables were most important to discriminate between regions of Brazil. Therefore, it is necessary to implement animal breeding programs to meet the needs of each region.

Classificação de padrões de savana usando assinaturas temporais NDVI do sensor MODLS no Parque Nacional Chapada dos Veadeiros

Savannas are the main vegetation type in Central Brazil, covering approximately 23% of the national territory. Locally known as Cerrado, Brazilian Savannas are formed by amosaic of different physiognomies such as grassland, shrubland and woodland that have atypical phenological cycle. ln this context, the MODIS data provide daily measurements well suited to monitor the seasonal phenology of vegetation. The present work aims to evaluate the advantages of the temporal signatures to detect Brazilian Savanna vegetation types in the Chapada dos Veadeiros National Park, Brazil. The adopted methodology may be subdivided into the following steps: (a) elaboration of the 3D cube of NDVI from temporal MODIS images, where the z profile corresponding to temporal signature, (b) noise elimination by combining Median Filter and Minimum Noise Fraction techniques, (c) endmember detection, and (d) spectral classification using Spectral Correlation Mapper method. The results demonstrate that the savanna physiognomies present typical temporal signatures. The endmembers correspond to the three major physiognomic domains: (a) Cerrado grassland, herbaceous dominated region; (b) Cerrado, mostly amixture of grasses and shrubs; and (c) Cerrado woodland, densely covered by trees. Comparison with Landsat 7/ETM+ image demonstrates the classification efficiency of the temporal series. The study concluded that the NDVI series is useful in differentiating the amount of vegetation types The methodology efficiency has been proved for regional delimitation of savanna physiognomies even considering the low spatial resolution of the 250m MODIS sensor and the high spectral mixture.

Tratamento de ruídos e caracterização de fisionomias do Cerrado utilizando séries temporais do sensor MODIS

Cerrado is formed by a mosaic of grassland, shrubland and woodland physiognomies with a typical phenological cycle. Thus, MODIS data provide daily measurements which allows to monitor the seasonal phenology of the vegetation. The objective of this work was to characterize savanna formations, forest formations and cerrado areas converted by anthropic actions, by using temporal series of MODIS NDVI and EVI after noise reduction. The adopted methodology should be divided into the following steps: (a) elaboration of the temporal cube with NDVI and EVI, in which the z profile corresponds to temporal signature, (b) noise elimination, (c) detection of temporal signature. The Minimum Noise Fraction Transformation (MNF) method was applied to reduce noise in temporal signature. The results showed that the NDVI values were higher than the EVI; and there was a relationship with the seasons of the year. The forest formations presented the highest values of NDVI and EVI, showing the lowest variations among the seasons. The converted areas of Cerrado presented the lowest values in both indices, and their values decreased in the beginning of the dry season, probably because it was the harvesting season. The study concluded that the NDVI and EVI temporal series are useful in differentiation among vegetation types.

Aplicação do método de identificação espectral para imagens do sensor ASTER em ambiente de cerrado

The spectral classifiers allow a good estimate for the mapping of the materials from the similarity between the reference curve and the image. Initially the spectral classifiers had been developed for hyperspectral images analysis. However, some works demonstrate good results for the application of these techniques in multispectral images. The present work aims to evaluate the spectral classifier Spectral Identification Method (SIM) in ASTER image. The Spectral Identification Method (SIM) is proposed to establish a new similarity index and three estimates according to the significance levels (5%, 10% and 15%) of the materials. This method is based on two statistical procedures: ANOVA and Spectral Correlation Mapper (SCM) coefficient. This information can be used to evaluate the degree of correlation among the materials in analysis. The advantage of this method is to validate according to levels of significance of the most probable areas of the sought material. The method was applied to ASTER image at the Military Instruction Field located Formosa (GO) close to Federal District. The images were acquired with atmosphere correction. The pixels size from the SWIR image were duplicated in order to join the VNIR and SWIR images. Endmembers were detected in three steps: a) spectral reduction by the Minimum Noise Fraction (MNF) transformation, b) spatial reduction by the Pixel Purity Index (PPI) and c) manual identification of the endmembers using the N-dimensional visualizer. The classification was made from the endmembers of nonphotosynthetic vegetation (NPV), photosynthetic vegetation (PV) and soil. These procedures allowed to identify the main scenarios in the study area.

Detection of karst depression by aster image in the Bambui Group, Brazil

Karst is a characteristic geological feature of areas comprised of limestone. Due to the solubility of these rocks in water, exhibit an extreme heterogeneity of hydraulic conductivities. The characterizing features of karst aquifers are the open conduits, which provide low resistance pathways for ground water flow. Overall cave orientation is largely controlled by hydraulic gradient, joint patterns and other tectonic features, such as faulting and folding. The karst depressions may form on the surface by subsurface actions (dissolution and collapse). Thus, the depressions often show regularity of pattern or alignments, frequently in association with structurally guided cave systems below. The present work aims at to detect depressions zone, as dolines and uvalas in the limestone of the Bambui Group (Central Brazil) using ASTER and ASTERDEM images. A photogeological study, carried out on aster image allowed us to elaborate geomorphological map of dolines. Some guidance to detect dolines can be associated with fracture permeability dominated by nearly vertical joints and joint swarm is provided by fracture trace mapping from remote sensing. Commonly, dolines can be identified on the image and DEM as topographic depressions, which very often contain water or moist vegetation. The methodology allowed determining a doline distribution pattern what is important to environmental planning.

Integração de dados de sensoriamento remoto multi resoluções para a representação da cobertura da terra utilizando campos contínuos de vegetação e classificação por árvores de decisão

This paper aims to develop a methodology of multisensor integration for an Amazon monitoring system. The proposed system employs the Vegetation Continuous Fields (VCF) method that uses the decision tree algorithm. The algorithm uses a set of independent variables, in this case MODIS multi-temporal metrics, to recursively split a dependent variable, in this case training data from class memberships, into subsets, which maximize the reduction of squares of sum of the residuals. The training data are obtained by high-resolution imagery classification (Landsat/TM, ETM+ and CBERS 2/CCD). In this study, an automated algorithm was developed from IDL language in the ENVI software and the statistical procedure of the S-PLUS software. The study area is Mato Grosso State with an extensive area of Amazon forest. The scenes are classified in three classes: forest, non-forest, and water. Comparisons of the final product with regional land cover maps derived from PRODES revel general agreement. Therefore, the results of this study suggest that the methodology is appropriate for land cover determination in the Amazon forest.

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.

Spatial dependence of malaria epidemics in municipalities of the Brazilian Amazon

Introduction: In 2010, there were 305 (37.8%) municipalities with malaria epidemics in the Brazilian Amazon. The epidemics spread can be explained by the spatial distribution pattern. Objective: To analyze the spatial dependence, autocorrelation, of the malaria epidemics in the municipalities of this region. Methods: An automated algorithm was used for the detection of epidemic municipalities in 2003, 2007 and 2010. Spatial dependence was analyzed by applying the global and local Moran index on the epidemic months proportion variable. The epidemic municipalities clusters were identified using the TerraView software. Results: The global Moran index values were 0.4 in 2003; 0.6 in 2007; and 0.5 in 2010 (p = 0.01), confirming the spatial dependence among the epidemic municipalities. Box Map and Moran Map identified inter-municipal, interstate and borders clusters with spatial autocorrelation (p < 0.05). There were 10 epidemic municipalities clusters in 2003; 9 in 2007 and 8 in 2010. Discussion: The epidemic municipalities clusters may be linked to the health facilities difficulties on acting together. The structural limitations of the health services can be overcome by territorial integration to support planning and control activities, strengthening the interventions. Conclusion: The routine analysis of the epidemic municipalities clusters with spatial and temporal persistence may provide a new indicator of planning and integrated control prioritization, contributing to malaria epidemics reducing in inter-municipal, interstate and borders areas.ABSTRACT: Introduction: In 2010, there were 305 (37.8%) municipalities with malaria epidemics in the Brazilian Amazon. The epidemics spread can be explained by the spatial distribution pattern. Objective: To analyze the spatial dependence, autocorrelation, of the malaria epidemics in the municipalities of this region. Methods: An automated algorithm was used for the detection of epidemic municipalities in 2003, 2007 and 2010. Spatial dependence was analyzed by applying the global and local Moran index on the epidemic months proportion variable. The epidemic municipalities clusters were identified using the TerraView software. Results: The global Moran index values were 0.4 in 2003; 0.6 in 2007; and 0.5 in 2010 (p = 0.01), confirming the spatial dependence among the epidemic municipalities. Box Map and Moran Map identified inter-municipal, interstate and borders clusters with spatial autocorrelation (p < 0.05). There were 10 epidemic municipalities clusters in 2003; 9 in 2007 and 8 in 2010.