Ramon Morais de Freitas

Cropland expansion changes deforestation dynamics in the southern Brazilian Amazon

Intensive mechanized agriculture in the Brazilian Amazon grew by >3.6 million hectares (ha) during 2001–2004. Whether this cropland expansion resulted from intensified use of land previously cleared for cattle ranching or new deforestation has not been quantified and has major implications for future deforestation dynamics, carbon fluxes, forest fragmentation, and other ecosystem services. We combine deforestation maps, field surveys, and satellite-based information on vegetation phenology to characterize the fate of large (>25-ha) clearings as cropland, cattle pasture, or regrowing forest in the years after initial clearing in Mato Grosso, the Brazilian state with the highest deforestation rate and soybean production since 2001. Statewide, direct conversion of forest to cropland totaled >540,000 ha during 2001–2004, peaking at 23% of 2003 annual deforestation. Cropland deforestation averaged twice the size of clearings for pasture (mean sizes, 333 and 143 ha, respectively), and conversion occurred rapidly; >90% of clearings for cropland were planted in the first year after deforestation. Area deforested for cropland and mean annual soybean price in the year of forest clearing were directly correlated (R2 = 0.72), suggesting that deforestation rates could return to higher levels seen in 2003–2004 with a rebound of crop prices in international markets. Pasture remains the dominant land use after forest clearing in Mato Grosso, but the growing importance of larger and faster conversion of forest to cropland defines a new paradigm of forest loss in Amazonia and refutes the claim that agricultural intensification does not lead to new deforestation.

Fraction images derived from Terra Modis data for mapping burnt areas in Brazilian Amazonia

The objective of this paper is to present a method for mapping burnt areas in Brazilian Amazonia using Terra MODIS data. The proposed approach is based on image segmentation of the shade fraction images derived from MODIS, using a non‐supervised classification algorithm followed by an image editing procedure for minimizing misclassifications. Acre State, the focus of this study, is located in the western region of Brazilian Amazonia and undergoing tropical deforestation. The extended dry season in 2005 affected this region creating conditions for extensive forest fires in addition to fires associated with deforestation and land management. The high temporal resolution of MODIS provides information for studying the resulting burnt areas. Landsat 5 TM images and field observations were also used as ground data for supporting and validating the MODIS results. Multitemporal analysis with MODIS showed that about 6500 km2 of land surface were burnt in Acre State. Of this, 3700 km2 corresponded to the previously deforested areas and 2800 km2 corresponded to areas of standing forests. This type of information and its timely availability are critical for regional and global environmental studies. The results showed that daily MODIS sensor data are useful sources of information for mapping burnt areas, and the proposed method can be used in an operational project in Brazilian Amazonia.

Monitoring land cover in Acre State, western Brazilian Amazonia, using multitemporal remote sensing data

This article presents the use of multitemporal remote sensing data for monitoring land cover changes in Acre State, western Brazilian Amazonia. For this study, the 2000 Landsat ETM+ (ETM+, Enhanced Thematic Mapper Plus), the 1990 Landsat TM (TM, Thematic Mapper) and 1980 Landsat MSS (MSS, Multispectral Scanner System) were used. The 2005 and 2007 Terra MODIS (MODIS, Moderate Resolution Imaging Spectroradiometer) images were also used to map deforestation that had occurred during the recent years and to map burned areas that occurred in the 2005 dry year. The ETM+, TM, MSS and MODIS images were converted to vegetation, soil and shade fraction images, using linear spectral mixing model, in order to reduce the data volume for digital processing. Then land cover maps were produced by digital classification of these fraction images. The results showed that deforestation increased by 7100 km2 from 1980 to 1990, by 5100 km2 from 1990 to 2000 and by 3300 km2 from 2000 to 2007. About 2100 km2 of vegetation regrowth areas were observed in the 2000 ETM+ images. The analysis of 2005 MODIS images showed that 3700 km2 of deforested areas and 2800 km2 of forested areas were burned in Acre State in 2005. This kind of information is important for regional and global environmental studies and for efforts to control such burning and deforestation activities in the future.

Temporal series of EVI/MODIS to identify land converted to sugarcane

Time series of remote sensing EVI/MODIS data are an important source of information to evaluate the direct land conversion to sugarcane in response to increased ethanol production in Brazil. The present work uses a time series of EVI/MODIS data to identify the land use prior to the conversion to sugarcane in a dense cultivated region in Parana¿ State, Brazil. Some sugarcane fields were selected over MODIS images acquired during the period of 2000 to 2008 in order to obtain reference time series to perform the land use classification prior to 2005. Using the temporal behavior of the EVI curves it was possible to distinguish among pasture land, annual crops and sugarcane in order to identify the land use prior to sugarcane conversion. It was noticed that in 2000 several of the annual crop fields in 2005, were pasture land in 2000 which were gradually converted to annual crops until 2005 and then to sugarcane.

Mapeamento das áreas de cana-de-açúcar na região norte fluminense - RJ por uso de técnicas de sensoriamento remoto

This study aimed to evaluate sugarcane growth dynamics occupation in land cultivated at six major cities of the North Fluminense, Rio de Janeiro, from 1984 to 2007. Eighteen images of Landsat TM sensor, visual interpretation and linear spectral mixing model (MLME) were used to generate thematic maps of sugar cane plantation spatial distribution and quantification. Based on these maps was possible to analyze the sugarcane field spatial distribution quantifying them in each municipality. The results indicated that between the years 2004 to 2007, there was a decrement in the total area occupied by the culture in 43,308.33 ha and, from 2000 to 2007 and increase of 24,422.72 ha, mainly in the Campos dos Goytacazes, Sao Francisco de Itabapoana and Cardoso Moreira municipalities. The MLME used allowed a real live mapping of the sugarcane areas.

Using Gradient Pattern Analysis for land use and land cover change detection

In this work, the computational operation based on Gradient Pattern Analysis - GPA was applied for the first time in MODIS spatial-temporal images over the Amazon region. The study area is located in the Para State, eastern Brazilian Amazonia. Using MOD09 8-day composite product from 2000 to 2009 was elaborated the EVI2 spatial-temporal series of the study area. For each pixel we performed smooth time-series applying wavelets transform method for noise reduction. The GPA objective was characterizing small symmetry breaking, amplitude and phase disorder due to spatial-temporal fluctuations driven by the deforestation and flooded changes detected by MODIS images. For the characterization of spatial-temporal series the Gradient Pattern Analysis showed a new approach to understand LULC changes in the remote sensing images.

Fraction images derived from EO-1 Hyperion multitemporal data for dry season green up analysis in Tapajós National Forest, Brazilian Amazonia

In this study, we present an approach for phenology analysis of Amazon green-up using Linear Spectral Mixing Model applied to Hyperion multitemporal data. The study area was selected in the Tapajós National Forest located in Pará State, Brazilian Amazonia. The region has well-defined dry and wet seasons with yearly rain about 2, 100 mm a dry season occurring from June to October. The study area is primarily covered by dense tropical rain forest (“Floresta Ombrófila Densa”) with a high number of emergent tree species. The EO-1 Hyperion data were acquired in July, August and September 2001, corresponding to the dry season in this region. The Linear Spectral Mixing Model was applied on each calibrated surface reflectance data, generating vegetation, soil, and shade fraction images. Then fundamental statistical analyses were carried out to evaluate the differences within the vegetation and shade fraction images derived from medium spatial resolution Hyperion images for rainforest phenology analysis.

Mapping and monitoring land cover in Acre State, Brazilian Amazônia, using multitemporal remote sensing data

This paper presents the use of multitemporal remote sensing data for monitoring land cover changes in Acre State, Brazilian Amazonia. The 2000 Landsat ETM+, the 1990 Landsat TM, and 1980 Landsat MSS were used. The 2005 and 2007 MODIS images were also used to map deforestation occurred during the recent years and to map burned areas occurred in the 2005 dry year. The Landsat and MODIS images were converted to vegetation, soil, and shade fraction images. Then land cover maps were obtained by digital classification of these fraction images. The deforestation increased 7, 114 km2 from 1980 to 1990, 4, 900 km2 from 1990 to 2000, and 3, 258 km2 from 2000 to 2007. It also showed that about 2, 815 km2 of regrowth areas were observed in the 2000 ETM+ images. The analysis of MODIS images showed that 3, 700 km2 of deforested areas and 2, 800 km2 of forested areas were burned in Acre State in 2005. These information are critical for regional and global environmental studies and for efforts to control such burning and deforestation in the future.

Seasonality of vegetation types of South America depicted by moderate resolution imaging spectroradiometer (MODIS) time series

The development, implementation and enforcement of policies involving the rational use of the land and the conservation of natural resources depend on an adequate characterization and understanding of the land cover, including its dynamics. This paper presents an approach for monitoring vegetation dynamics using high-quality time series of MODIS surface reflectance data by generating fraction images using Linear Spectral Mixing Model (LSMM) over South America continent. The approach uses physically-based fraction images, which highlight target information and reduce data dimensionality. Further dimensionality was also reduced by using the vegetation fraction images as input to a Principal Component Analysis (PCA). The RGB composite of the first three PCA components, accounting for 92.9% of the dataset variability, showed good agreement with the main ecological regions of South America continent. The analysis of 21 temporal profiles of vegetation fraction values and precipitation data over South America showed the ability of vegetation fractions to represent phenological cycles over a variety of environments. Comparisons between vegetation fractions and precipitation data indicated the close relationship between water availability and leaf mass/chlorophyll content for several vegetation types. In addition, phenological changes and disturbance resulting from anthropogenic pressure were identified, particularly those associated with agricultural practices and forest removal. Therefore the proposed method supports the management of natural and non-natural ecosystems, and can contribute to the understanding of key conservation issues in South America, including deforestation, disturbance and fire occurrence and management.

Avaliação da Acurácia do Produto de Altimetria do Laboratório de Agricultura e Floresta (Evaluation of the Accuracy of the Product of Altimetry of the laboratory of Agriculture and Forest)

Since Shuttle Radar Thematic Map (SRTM) data became available, many studies utilized them for applications in geomorphology, topography, vegetation cover studies, tsunami assessment, and urban studies, among others. Recently, a scientific project (http://www.dsr.inpe.br/laf/canasat) engendered on Brazilian National Institute for Space Research, developed a free and friendly tool for instantaneous visualization of local elevation within the concept of a virtual laboratory framework. This tool, named Laboratory of Agriculture and Forest, was build to assess the elevation anisotropy around selected points. The anisotropy visualization is a simple polar plot of elevation around two simple circles, allowing a rapid view of the topography around the selected point. This tool allows interactivity and provides a range of distance between the center of the selected coordinate and the sampled circles. However, overall assessment of the accuracy of this product requires additional regional studies involving accuracy verification methods with higher level of precision, such as the Differential Global Positioning System (DGPS). The study presented in this paper is based on ground truth control collected with DGPS system, with differential base station data, in a mountain in the Itajubá city. The results pointed out to the efficiency of the tool but indicate some necessary care for their accurate use.