Publicações

É apresentada a capacidade dos modelos numéricos em reproduzir o clima presente sobre a América do Sul. Características e desenvolvimentos do modelo global atmosférico do CPTEC, do modelo global acoplado oceano-atmosfera do CPTEC e modelos regionais climáticos são descritos. No modo climático, os modelos globais utilizam resolução de cerca de 200km, enquanto os modelos regionais geralmente utilizam a resolução de cerca de 50 km. Métodos estatísticos de downscaling e resultados sobre o Brasil são apresentados. Processos de retroalimentação oceano-atmosfera, radiação-nuvem, biosfera-atmosfera são discutidos e resultados de simulações numéricas são apresentados. Avaliações dos modelos globais atmosférico e acoplado e regionais climáticos em representar alguns fenômenos meteorológicos que atuam na região: El Niño-Oscilação Sul, Zona de Convergência do Atlântico Sul, Zona de Convergência Intertropical, Ciclones extratropicais, Modo Anular do Hemisfério Sul e Jato de Baixos Níveis, são apresentadas. Em geral, os modelos numéricos simulam satisfatoriamente estes fenômenos. Alguns erros sistemáticos são identificados. Problemas de poluição atmosférica e do efeito da ilha de calor são simulados para a megacidade de São Paulo e consequências para mudanças climáticas são discutidas. A problemática da elevação do nível do mar é discutida. Incertezas em simulações do clima presente são exploradas a partir de variações de configuração de um mesmo modelo, o HadCM3, ou a partir de diferentes modelos que possuem diferentes configurações, sejam globais atmosféricos, globais acoplados oceano-atmosfera ou regionais climáticos.

Lightning is one of the major causes of weather related mortality in several countries around the world. However, in many countries, no reliable statistics are available about the number of annual lightning fatalities, in part due to the inexistence of specific institutions responsible for such statistics and the inherent difficulties to obtain such information. Therefore, several assumptions are needed in order to obtain an estimate of the annual number of global lightning fatalities. In the past, the level of development of different countries and the available statistics of some developed and non-developed countries were used to estimate for the first time an annual number of 24,000 fatalities globally. This paper presents a new approach to estimate the number of global fatalities. It considers the country population and lightning activity obtained from the Lightning Imaging Sensor (LIS) in the countries where lightning statistics of fatalities are available to estimate a vulnerability index. This index is then extrapolated to all countries by continent in order to estimate the annual global number of fatalities. With this new method, an estimation of approximately 6000 annual fatalities globally is obtained.

A 10-year study about lightning fatalities in Brazil is presented. It is the more complete study already done in South America. The study indicates the death rates in the country sorted by season, state, city, type of region, age, season, gender and circumstances. Ten years of data were collected from the Federal Civil Defense Agency, the Ministry of Health and the media press. It also includes an analysis about the differences involved in lightning deaths in each region of the country. The analysis suggests that specific actions should be adopted in terms of lightning protection to minimize risks in Brazil. Also, it provides useful information in order to identify the most important lightning safety rules for the country hat can be used in a national lightning safety campaign.

Neste projeto foram realizadas atividades que levaram ao acesso e uso de dados que auxiliam no desenvolvimento e aprimoramento de modelos globais que representam a ocorrência e os efeitos do fogo na vegetação. O principal conjunto de dados encontrado está relacionado à atividade de fogo na superfície terrestre, de janeiro de 2001 até dezembro de 2010, baseado em dados de satélites. Estes dados foram agregados mensalmente a uma resolução espacial de 0,5° de latitude e longitude, e convertidos em formato comum a outros dados utilizados no Centro de Ciência do Sistema Terrestre (CCST). Estes dados foram então relacionados a outros tipos de variáveis ambientais, sendo elas: distribuição de biomas naturais, população global, precipitação, umidade do solo e água precipitável na atmosfera, originados em outros institutos e projetos no próprio CCST. Para síntese, os dados foram também organizados em climatologias trimestrais (média climatológica dos meses de janeiro a abril, março a junho, agosto a setembro e outubro a dezembro). A partir dos mapas criados, observamos que as ocorrências de fogo aparecem sobre os continentes e fora do oceano, e não existem em regiões de gelo ou de deserto, dando confiabilidade na construção do banco de dados. As regiões que apresentam seca em determinada época do ano apresentam também maiores ocorrências de fogo, e o bioma natural mais atingido é em geral a savana. Onde há uma quantidade maior de umidade no solo, precipitação, água precipitável e população, a atividade de fogo em geral é baixa ou nula. Os dados foram também analisados através de gráficos que já neste ponto indicam as formas funcionais das equações que podem ser usadas para representar as relações entre a atividade de fogo e as outras variáveis, e que poderão ser usadas futuramente no desenvolvimento e aprimoramento de novos modelos. Na maioria dos casos, estes gráficos indicam relações não lineares entre as variáveis. Por exemplo, os valores de atividade de fogo em geral decrescem exponencialmente com a precipitação, umidade no solo e população global. Os desenvolvimentos deste projeto alcançaram os objetivos esperados, destacando-se principalmente as contribuições para a construção de bancos de dados que são de grande importância para o desenvolvimento e testes de modelos em escala global da dinâmica do fogo nos ecossistemas terrestres.

Brazil has nearly 5 million hectares of forest planted with species of the genus Eucalyptus and Pinnus and nearly 70% of these forests are for the manufacture of cellulose pulp. Eucalyptus planted forests are being implemented in different regions of the country, developing at sites with different topographies and rainfall patterns. But then some factors limit the geographic distribution of plantations. Remote sensing has surged ahead in the forest sector, and with the use of techniques of image interpretation, we can identify and map the Eucalyptus and differentiate them from other vegetation types, and quantify them. The aim of this study was to map the Eucalyptus spp. using satellite images and to evaluate the geographic distribution, including soil types and slope of the land where these plantations are being developed. Images from the Landsat / TM for the year 2008 and the plantings were mapped by visual interpretation of images. The areas of plantations were compared with those found in the work areas of Kronka et al. (2002), which showed an increase of 75,123 ha in the entire state. The species of this genus have some limitations of climate, soil and topography. Understanding these factors and the choice of sites that meet the minimum requirements of Eucalyptus facilitates their growth, thus ensuring its productivity.

The aim of this investigation is to analyze the phenomenology of positive and negative (stepped-, dart- and recoil) leaders observed in natural lightning from simultaneous high-speed video recordings and electric field measurements. For that intent we have used two different high-speed cameras (obtained by two different cameras, Red Lake Motion Scope 8000S and Photron Fastcam 512 PCI, operating at frame rates ranging from 1000 or 8000 frames per second) and two different types of electric field sensors (slow and fast) in addition to data from a lightning locating system (BrasilDat). All the instruments were GPS time synchronized in order to avoid ambiguities in the analysis and allow us to estimate the peak current of and the distance to each flash that was detected by BrasilDat. From these data it is possible to calculate the two-dimensional speed of each observed leader, allowing us to obtain its statistical distribution along with its correlation to other characteristics of the associated flash, such as return stroke peak current, interstroke interval and presence and duration of continuing current after the return stroke initiated by the leader. Also, the availability of electric field data makes it possible to correlate it to the optical characteristics of each leader type, providing us new insights on the microphysics of these phenomena. In the analyzed dataset, the speeds of positive leaders and negative dart-leaders follow a lognormal distribution at the 0.05 level (according to the Shapiro-Wilk test), while the negative steppedleaders have a similar distribution even though according to a relatively weaker test (KolmogorovSmirnov). Unfortunately, due to the range of frame rates reached by the used cameras, only the minimum propagation speeds could be estimated in the recoil leaders that were observed.

Schonland et al. (1938), in their seminal streak camera studies of lightning, have identified a peculiar type of negative stepped leader that they termed β, a rather rare variant of the type β leader and in it the second and slower stage of the leader is associated with the appearance of one or more fast dart streamers, which travel rapidly down from the cloud along the previously formed track and cease when they have caught up with the slower leader-tip. They reported four events, two were accompanied by a dart streamer (a nomenclature we will keep for historical reasons), one event contained two examples, and one event that was not described in detail. During a July-August 2007 campaign in Tucson, Arizona, USA, we recorded four downward leaders that fit in the type-β22 description given by Schonland et al. (1938). All cases occurred between about 12 and 32 km from a high-speed camera that was operating at 4000 frames per second. Two cases were produced by the first stroke in a flash, and one case was associated with a subsequent stroke that produced a new ground termination. Our fourth case was a subsequent stroke that began with a dart leader and then changed to a type- stepped leader approximately 1350 m above the ground; in this case, we observed a dart streamer (with the leader tip at about 720 m from the ground) about 3.25 ms after the stepped progression began and the return stroke occurred 2 ms later. All the dart streamers that we observed had speeds between 10 6 and 10 7 m s -1 -2 , in agreement with previous observations of recoil leaders (Saba et al., 2008); therefore, we suggest that the type- negative leaders are the visible manifestation of the development of recoil leaders that were initiated inside the cloud and propagate below the cloud-base, during the development of a bipolar, bidirectional leader that precedes a lightning flash to ground. The recoil leaders develop in previously ionized channels that are produced by the in-cloud 2 positive portion of a bidirectional leader, and when they reach the tip of the negative downward portion, they produce an intense pulse of luminosity that optically appears as a dart streamer.

Lightning data obtained by the Brazilian Lightning Detection Network (BrasilDat) in the southeastern region of Brazil (latitudes from 22S to 24S and longitudes from 43W to 47.4W) from 1999 to 2009 were studied in terms of spatial and temporal variations of the flash counts. The spatial variation shows clear effects of the altitude and urban activity on the flash counts. These aspects modulated the flash counts in a way to produce two large lightning spots, one located over the city of São Paulo and the other with almost a same area located near the city of Volta Redonda in the south part of the state of Rio de Janeiro. While the first is related to the heat island produced by the urban activity, the other is related to a particular feature of the relief characterized by the approximation of the two mountain ridges. For both cities (São Paulo and Volta Redonda) three different comparisons were done involving the flash density (in terms of hours of flashes) for a grid cell of 2x2km, the daily (local time) flash variation, the annual percentage of minutes with more than three (typical value for isolated thunderstorms) flashes with respect to the total number of minutes with flashes and the annual variation of the number of hours of flashes. The results are discussed in the context of the two different physical mechanisms responsible for most convection in the two cities.