Publicações

An increase of geo-spatial data availability due to recent technological advances in data acquisition instruments (GPS, Sensor Networks, Remote Sensing Satellites), has brought Geo-Information Science researchers to the following scenario: a huge amount of data to carry out their research as opposed to a lack of tools to handle it adequately. Thus, the demand for the inclusion of better tools for handling temporal and behavioral dimensions in relation to software such as Geographical Information Systems has increased. Nevertheless, the current GIS, referred to as 20th Century GIS, were not designed in order to meet such a demand. Therefore, the need has come up to design a new GIS generation, the 21st Century GIS, with tools capable of addressing researchers needs. This work aims at proposing the development of software architecture to support a data-intensive Geographical Information Science as part of an effort to establish a 21st Century Geographical Information Engineering.

Algoritmos de segmentação e classificação detectam feições presentes em imagens, produzindo conjuntos de dados vetoriais compostos por polígonos. Um processo de generalização é necessário para gerar mapas com diferentes graus de detalhe em diferentes escalas e reduzir o armazenamento de dados. Este artigo tem o objetivo de revisar e sugerir melhorias em métodos de generalização de polígonos, quando aplicado a estes conjuntos de dados vetoriais, buscando manter a topologia original e minimizar a discrepância de área do mapa de saída. O conjunto de dados usado é parte do mapa de desflorestamento da região Amazônica, medido pelo PRODES. Os resultados mostram que a eficiência da generalização pode ser melhorada e a discrepância de área também pode ser reduzida.

This communication evaluates the mechanisms associated with the inherent difficulty of global coupled ocean-atmosphere models to simulate the thermal vertical structure of the equatorial Atlantic. The study is done with use of the coupled GCM developed at CPTEC, which consists of CPTECs global atmosphere model fully coupled over the global tropics (40S-40N) with GFDLs MOM3 ocean GCM. Coupled and uncoupled simulations were calculated in ensemble mode, with ten members runs, 8 months long each starting at every calendar month for the period 1982-2001. Systematic differences between simulations done with the CGCM and OGCM forced with prescribed atmospheric fluxes, as well as against PIRATA array data over the tropical Atlantic are used to argue that the systematic errors of the coupled model over the equatorial Atlantic can be separable into a surface and under-thermocline ones. Evidences are shown to suggest that surface temperature systematic errors are attributable to insufficiencies on AGCM stratus parameterization, which generated excess solar radiation over the regions of stratus decks over cold waters along the southwestern coast of Africa. For the under-thermocline processes, however, it is shown a remarkable locking to the seasonal cycle, with systematic errors maximum amplitude occurring in April/May, regardless of the month of model initialization. Linkages to AGCM rainfall deficiencies over the equatorial landmasses are speculated.

Carbon Monoxide (CO) is emitted to the atmosphere over the Amazon by biomass combustion as well as being formed by oxidation of hydrocarbons. Carbon dioxide (CO2) is emitted by combustion and respiration and absorbed by photosynthesis. CO and CO2 mixing ratios were measured continuously at Maxaranguape, a coastal site, and Tapajos km67, an inland site, during the BARCA (Balanço Atmosférico Regional de Carbono na Amazônia) campaign. We report here results from November 2008 through May 2009 including both the two intensive flight campaigns, BARCA-A in November-December 2008 and BARCA-B in May 2009 and the period between. The Maxaranguape coastal station is located on a bluff overlooking the Atlantic Ocean about 40 km north of Natal. Steady easterly trade winds deliver tropical marine air to the site most of the time except for occasional periods of nighttime land breezes. The Tapajos km67 site is located in the Tapajos National Forest about 60 km south of Santarem. Trajectory analysis indicates that coastal air off northeast Brazil contributes to the easterly flow entering the Amazon basin. Mixing ratio gradients between this pair of sites is an overall indicator of the net influence of sources in the eastern Amazon basin. During the dry season CO concentrations at km67 exceed the mixing ration in marine air by 100 – 200 ppb, indicating the influence of biomass burning across the region. CO mixing ratios at km67 during the wet season are on average 40 ppb higher than at Maxaranguape, which we interpret as CO production from biogenic hydrocarbons, though there may also be some influence from urban areas in eastern Amazon and along the coast. Throughout the wet season in 2009 leading up to the BARCA-B flights in May, CO2 at km67 was elevated by 15-25 ppm relative to CO2 in the marine-air inflow sampled at Maxaranguape, suggesting a net efflux of CO2 during this entire period and consistent with the analysis of airborne observations.

In Central Brazil every year large areas of forest, cerrado and pasture land are burned emitting primary carbonaceous aerosols into the atmosphere. These interfere with the radiative budget, affecting directly the amount of solar radiation that is reflected and scattered back to space, and partly absorbed by the atmosphere. The long wave terrestrial radiation is also affected. Aerosols also affect cloud microphysics acting as cloud condensation nuclei (CCN), modifying the precipitation pattern and the cloud albedo. During the South American burning season, aersosols have a significant impact on local and regional air quality affecting visibility and human health by particle inhalation causing pulmonary diseases. The recognition of atmospheric aerosols importance led to the necessity of quantifying and optimizing its sources, distributions and sinks in the troposphere to improve simulations with atmospheric chemical models. An improvement of the aerosol representation in such models can be achieved through data assimilation methods including available aerosol observations. In this study an assimilation system was used based on the three dimensional variational data assimilation method (3D-VAR) using the Coupled Chemical-Aerosol-Tracer Transport – Brazilian Regional Modeling System (CCATT-BRAMS). The assimilated observations are mass concentration (MC) of particular matter smaller than 2.5 µm (PM2.5). The observations were collected in October 2007, during the Cloud-Aerosol Interaction Measurements (CLAIM) campaign, which took place in the region of the city Alta Floresta in the state of Mato Grosso, Brazil. The measurements of PM2.5 were collected by a DATARAM instrument aboard the aircraft that collected information with 10s or 30s frequency. During the campaign 17 flights were carried out, 13 of which could be used in this study. In this paper we present the data and its assimilation results, evaluate its improvement of the model simulation and discuss its general impact on the chemical state of atmosphere.

Brazil ranks in third place within the top 10 renewable electricity producers with almost 80% of the electricity generated from renewable sources. This work describes the status of solar and wind energy resource assessment at Centre for Earth System Science (CCST) of Brazilian Institute for Space Research (INPE) and also the major efforts of the to provide reliable information. It describes the recent studies in development by Renewable Energy team at CCST/INPE in numeric modeling in order to provide reliable resource assessment in order to provide the information demands of Brazilian energy sector.

In 1997 scientists proposed an agenda for biogeochemical research in river corridors of the Amazon basin to both identify gaps in knowledge and to outline a possible approach to filling those gaps. Most previous research in the Amazon had focused on natural processes with a majority of effort on the main stem and its largest tributaries; much less was known about the small river corridors at both the micro (<100 ha) and meso (<100,000 ha) scale or the riparian forests or floodplains around these rivers. In addition, although many potential impacts on rivers and floodplains from anthropogenic land cover or land use conversion were proposed little empirical data were available. A nested sampling design from micro- to macro-scale was proposed. Over the intervening 13 years of research as part of the Largescale Biosphere Atmosphere project in Brazil much new information has been gained. Many studies at the small watershed scale demonstrated clear impacts on stream water flow paths from groundwater to surface water due to forest loss. In addition, processes controlling chemical composition were advanced including surface interflow and in-stream processing. At the meso-scale distributed hydrologic models successfully captured human influence on river process. At the macro-scale carbon processing from land to water was further elucidated, particularly with the extent and activity of flood inundated areas. The lag in response from small watershed to the macro-basin remains a pressing question. Similarly, the role of impoundments or climate change on flood stage and floodplain function needs further consideration.

The study of atmospheric trace gases is increasing over the years. One of these gases is sulfur dioxide (SO2), which is found in the troposphere and stratosphere, as a result of both natural and anthropogenic emissions. A campaign was made in Cubatao-SP, from 04/12/2007 to 08/07/2007, to measure the total column of this gas with a Brewer Spectrophotometer. With this instrument the total column of SO2 was calculated. Since this region is highly industrialized, high total columns of SO2 were expected, which indeed occurred, with a daily average of 6.4 DU during the whole period.