National Institute for Space Research
Biblioteca Digital da Memória Científica do INPENot a member yet
6175 research outputs found
Sort by
Detecção de áreas de desmatamento no Estado do Tocantins, no período 2006/7 a 2010/11
Amazônia and Cerrado are the highest biomes that cover the Brazilian land. They have been affected by deforesting. This study analyzed the deforestation detection in the Tocantins State, Brazil, in both those biomes, from 2006/7 to 2010/2011. It was checked the detections from analogical comparison among CBERS, Landsat e Resourcesat imaging and detections from DETER, a near real time detection conduced by INPE with MODIS data. In 42.783,72 km² monitored by the analogical method, it were delimited and classified like deforestation 60.098,44 ha allocated in 842 polygons. The lowest analogical delimited polygon was 0,34 ha, the highest were 1.631,08 ha and their average was 71,38 ha. It occurred a heavy deviating among the different studied locals. These data represented 1,40 ha deforested on each square kilometer and 19,68 polygon on 1.000 km². In 39.630,37 km² monitored by the DETER system, it were warned 114 deforestation alerts, accounting 12.489,46 ha. The lowest DETER alert was 6,40 ha, the highest was 2.468,51 ha and their average was 109,46 ha. These data represented 0,32 ha deforested on each square kilometer and 2,88 alerts on 1.000 km². Both the methods are essential to the deforestation detection, despite their peculiarities and the regional landcover peculiarities.Pages: 7313-732
Comparação entre índices de vegetação gerados a partir de dados dos sensores MODIS e Vegetation-2
Vegetation indices have been considered as the basis of several empirical models relating radiometric data from orbital images and biophysical properties of vegetation. Frequently these models are defined by a specific sensor data set and their results are assumed as universal or absolute. MODIS sensor, onboard of Terra satellite and Vegetation-2 sensor, onboard of SPOT 5 satellite, generate similar products, but the impact of choosing one or other sensor data set on empirical models performance has not been fully understood. MODIS and Vegetation-2 data from two Eucalyptus spp. farms from FIBRIA S/A located in São Mateus and Aracruz (Espirito Santo state, Brazil) and also from two seasons were compared. The comparison was performed by linear regression tests, including β0=0 and β1=1 hypothesis tests at 5% of significance. Results indicated that MODIS and Vegetation-2 data are significant different even being collected at near period of time and at similar illumination conditions. The Eucalyptus spp phenological changes have not been felt by both set of data, but their relationship has been influenced with higher correlations detected at dry season.Pages: 3023-303
Monitoreo histórico de los desmontes en áreas de bosque nativo de la Provincia de Entre Ríos, Argentina
The objective of this study was the historical monitoring of forest clearance in the region of the "Espinal" (thorn forest) in northeastern of Entre Ríos Province. Monitoring changes in land use and cover is needed for government policies of conservation and management of natural resources. The study area was the northern of Entre Rios province. The typical vegetation is Espinal thorn forest, subjected to selective logging for extensive livestock ranching and now subjected to agriculture too. The methodology use Landsat imagery and change detection technique to form a package with spectral bands and a band of intensity of changes. On this package we applied the technique "Image segmentation by region growing with manual seeding". Then, the thematic aggregation based on visual interpretation was made. Finally, the map of historical clearance was obtained by Boolean operations between annual thematic masks. The segmentation of images combined with successive masking achieved good results for historical monitoring of forest clearing.Pages: 3039-304
Mapeamento do uso da terra do município de Ituiutaba-MG por meio da classificação automática de Bhattacharya (2011)
The land use update is needed to understand the patterns of agricultural space organization, increasingly altered by human action. This study constitutes a subsidy to the action of legislators and planners. The techniques of remote sensing and geoprocessing for automatic classification constitute support for land use mapping, optimizing the time and, when allied with the field evidence, allow more accurate in the obtained results. The main objetive is to map the land use and vegetation cover in Ituiutaba-MG from the use of TM/Landsat5 (2011) and automatic classification by the method of Bhattacharya. The methodology included the use of 221/222 path and 73 row from TM sensor acquired in August 2011, released by INPE, then it was proceeded to their georeference, mosaic, cut, segmentation and mapping of images by the method of Bhattacharya in Spring 5.1.8 software. Field work was made in August 2011, and subsequent corrections of the mapping errors were made in the software ArcGIS 10.0. The results identified seven categories of land use in study area, which occupy the following percentages in relation to the total area of the municipality: Bodies of water (1.04%) Natural Vegetation (27.32%); Reforestation (0.07%) Agriculture (5.57%); Sugar cane (13.14%), Pasture (51.80%) and Urban Area (1.07%).Pages: 8076-808
Simulação das bandas espectrais do sensor QuickBird-2 a partir do sensor WorldView-2
Data comparison between two different sensors must analyze one of their several characteristics. In order to evaluate the potential of the improvements on spectral resolution of WorldView-2 sensor, the reference data must have the same spatial and radiometric resolutions, besides the same geometric and atmospheric conditions of acquisition. For most areas throughout the world, there are no available images acquired by WorldView-2 and QuickBird-2 sensors in the same date or under the same geometric and atmospheric conditions. Therefore, the combination of the bands of given sensor to simulate the bands of a reference sensor represents an attractive possibility for the handling of remote sensing data, searching for its improvement. This work presents a methodology for the simulation of the multispectral bands of QuickBird-2 sensor applying the weighted combination of the spectral bands that possess spectral overlapping with each reference band of WorldView-2 sensor. The analysis of the spectral overlapping was performed through the discretizing of the spectral curves. Visual and statistical tests were accomplished to confirm the effectiveness of the methodology.Pages: 7981-798
Utilização de geotecnologias para inferências populacionais no espaço urbano do município de Guarulhos-SP
The interest of urban environment studies has raised among researchers as well as remote sensing and Geographic Information System (GIS) are commonly used to acquire detailed information on urban land cover. Thus, the purpose of this paper is to define a methodology for urban analysis that could provide information of the population characteristics through land cover mapping using a orbital image with high spatial resolution. The study area corresponds a section of the municipality of Guarulhos (São Paulo State, Brazil), a important town in the metropolitan area of São Paulo, where were selected 25 polygons from digital mesh of 2010 Census from Brazilian Institute for Geography and Statistics (IBGE). The methodology applied is based on three main steps: 1- Classification of IKONOS imagery, obtained in 2007, pan-sharpened, using software ENVI EX and Object-Based Image Analysis (OBIA) methods and techniques resulting 7 cover classes: asphalt, vegetation cover, shadow and different types of roofs (light, dark and ceramic). In this step were applied image segmentation, customized new attributes and defined classification rules. 2- Choice of variables about population characteristics and households in the 2010 Census. 3- Calculating bivariate correlation between results of land cover and population variables. The results were presented, discussed and the methodology proposed sounds very promising to be applied to update urban environment information for urban planning and territorial organization.Pages: 794-80
Mapeamento Fotogramétrico Digital: Um Estudo Comparativo da Bacia Hidrográfica do Rio Piabanha nos Ambientes E-Foto e LPS
The use of photogrammetry has been potentiated due to the great need for rigorous planning, in line with the demands of new times. In order to have students trained and other stakeholders in the area of photogrammetry has been developed the software E-FOTO, in order to supply the lack of free softwares in this area of knowledge, since the other existing ones are expensive. The project E-FOTO is based on the philosophy of self-learning and a program components in free, open source and, tending to the basic principles of the free software community. As a contribution to the development of the project E-FOTO, this article technical, opted to choose an area where there was a busy relief, where the deformations are larger and the correlations Between the images are not so simple to achieve. In order to test the ability of E-FOTO to work with this type of region we used the software ERDAS IMAGINE, specifically the module LPS, a commercial software already established in the labor market as a parameter for comparison of results obtained. It is important to remember that this project was done when the E-foto software has not worked in an integrated way, in 2011. Currently, in 2013, there have been numerous implementations, and the software has integrated all its functions and presenting their increasingly satisfactory results. Recently, the E-foto was awarded the third place in the 6th CATCON, a competition called educational software for photogrammetry, which occurred in Australia during the Congress of the International Society for Photogrammetry and Remote Sensing, which signals to us all the frank evolution of this software to the academic community.Pages: 2134-214
Análise histórica da adequabilidade do uso das terras em uma microbacia hidrográfica no Município de Piracicaba
Old aerial photographs (1940) were used for a temporal and spatial analysis of land use suitability of a watershed in Piracicaba County. Aerial photographs of 1962, 1978 and 1995 were used too. The analyses were performed on Idrisi. Despite the antiquity, the 1940aerial photographs were interpreted in a satisfactory manner and the control points selected allowed the overall georeferencing and analyses. The use of old aerial photographs showed that already in 1940 a considerable part of the watershed, 27,66%, was inadequately used. This land use pattern repeated in 1962, 1978 and 1995. While 10,25% of the watershed was inadequately used in all dates, only 56,51% of the land suitable for crops was used with sugarcane in 1995.Pages: 4353-435
Geotecnologias aplicadas na identificação de áreas suscetíveis ao fogo no Parque Nacional da Chapada dos Veadeiros
The research aimed to use geotechnologies to identify susceptible areas to fire in the Chapada dos Veadeiros National Park and the surrounding area. The study area presents 2376,73km² and its located in UTM zone 23S, between the coordinates 177422 248096 mE and 8420578 8473325 mN, in the state of Goiás. We created the following thematic maps: land use and vegetation cover, hypsometry, slope, orientation of slopes, proximity to rods and urban areas. For each thematic map, we created a susceptibility map to fire, with classes ranging from high, to moderate and low. The areas considered as having a high susceptibility to fire presented the following conditions: occupied with Park Savanna, Grassy-Woody Savanna and Cultivated Pasture; lower altitudes than 1200m; greater slopes than 40%; strands with orientation toward the north, northeast and northwest; areas near roads, trails outside the parks and urban areas. The results showed that in relation to the land use and vegetation cover, the park and the surrounding area have presented 87,79% of its area in the upper class of susceptibility to fire. The areas with altitudes up to 1200m represent 65,22% of the area and indicate a high susceptibility to fire and the slopes greater than 40% are 6,10% and also indicate high susceptibility to fire. Regarding the orientation of the slopes, 40,20% of the area has a high susceptibility to fire. Regarding the proximity to roads and urban areas, the study area has 4,05% in the upper class of susceptibility to fire.Pages: 5169-517
Análise comparativa dos modelos de elevação SRTM, ASTER GDEM e TOPODATA para estimar o fator topográfico (LS) da USLE
Given the great variety of MDE from several sources and with different spatial resolutions, the question arises: which data are most suitable for the objective and scale of work. The LS factor integrates the gradient of slopes and its length, is considered the most critical factor to estimate soil loss. This study evaluated the performance of SRTM, ASTER GDEM and TOPODATA to estimate the USLE in the Environmental Protection Area Gama e Cabeça de Veado in the Distrito Federal. Based on these research results, we observed that there was no statistically substantial difference in the USLE of the three models. The average of the three factors LS derived from three different MDE presented similar values as well as the average of three USLE generated. This result seems to be associate with the terrain characteristic of the APA, which is flat and gently ondulated, since the influence of spatial resolution in the LS factor is more pronounced in MDE with high spatial resolution and on sloping land. However, the choice of the appropriated DEM for each study is not a great concern, since the three tested DEM showed similar results.Pages: 4435-444