Publicações

A necessidade de informações acerca das dinâmicas de uso da terra tornou estudos relacionados à classificação de uso e cobertura da terra na região Amazônica relevantes no campo científico, com destaque ao uso de classificadores supervisionados. No entanto, o desempenho de um classificador supervisionado é dependente dos dados de entrada utilizados (imagem a ser classificada, definição das classes e coleta de amostras de treinamento), que em muitos estudos não são propriamente avaliados. Para contribuir com o entendimento de como variações nos dados de entrada afetam os resultados da classificação supervisionada, este estudo analisou os resultados de classificações de cobertura da terra em uma região da Amazônia brasileira, considerando diferentes métodos de classificação baseados em pixel, legendas com distintos níveis de detalhe (número de classes) e a sensibilidade dos classificadores à variação das amostras de treinamento (Incerteza) com o uso de uma imagem ótica de média resolução. Para tanto, foram geradas 100 imagens classificadas para três legendas com diferentes níveis de detalhe e diferentes configurações de quatro algoritmos de classificação supervisionados: Máxima Verossimilhança (ML), Máquina de Vetores Suporte (SVM), Árvore de Decisão (J48) e k-Vizinhos mais Próximos (K-NN). No total, foram obtidos 111 (3 níveis de legenda x 37 configurações de classificadores) grupos de 100 imagens classificadas, que foram utilizados para gerar imagens combinadas em que cada pixel recebe o rótulo mais frequente nesses grupos de imagens (moda) e mapas da Incerteza associada. De forma geral, a combinação de imagens retornou uma imagem classificada de igual ou maior exatidão que imagens que a compõem, com diferenças menos expressivas em função dos classificador utilizado que ao se considerar os grupos de 100 imagens classificadas, embora a diferença no valor de exatidão seja pequena. A maioria das imagens classificadas mais acuradas foram obtidas utilizando alguma configuração do algoritmo K-NN. No entanto, esse algoritmo apresentou maiores valores de Incerteza quando comparado a ML e SVM. Essa maior variabilidade (maiores valores de Incerteza), em compensação, colabora para a obtenção de melhores resultados da operação da moda, pois permite reclassificar adequadamente pixels de classificação instável. O algoritmo ML apresentou resultados mais consistentes (variam pouco em relação ao conjunto de amostras de treinamento), adequadamente acurados e possui utilização mais simples, já que a etapa de ajuste de parâmetros não é necessária. Quanto menor o nível de detalhe da legenda, maior a exatidão das imagens classificadas e menor sensibilidade das imagens classificadas à variação das amostras de treinamento (Incerteza). A diminuição do nível de detalhe da legenda resulta também em diferenças menos sutis na exatidão das imagens classificadas por diferentes algoritmos na configuração ótima.

O presente artigo tem por objetivo revelar a importância do monitoramento e da avaliação das políticas públicas de educação ambiental, bem como apresentar um processo participativo de construção de indicadores que dialogam diretamente com o Tratado de Educação Ambiental para Sociedades Sustentáveis e Responsabilidade Global. As políticas públicas de educação ambiental, neste contexto, são compreendidas a partir de uma abordagem plural e de enfrentamento coletivo dos problemas socioambientais. Os primeiros resultados obtidos são oito dimensões articuladas de indicadores de monitoramento e avaliação de políticas públicas de educação ambiental. A partir desses resultados iniciais foram realizadas análises que conectam as referidas dimensões ao Tratado e à Política Nacional de Educação Ambiental, apontando para a criação de um sistema de monitoramento nacional, com o potencial de disparar processos adaptados localmente.

Comparing model output and observed data is an important step for assessing model performance and quality of simulation results. However, such comparisons are often hampered by differences in spatial scales between local point observations and large-scale simulations of grid cells or pixels. In this study, we propose a generic approach for a pixel-to-point comparison and provide statistical measures accounting for the uncertainty resulting from landscape variability and measurement errors in ecosystem variables. The basic concept of our approach is to determine the statistical properties of small-scale (within-pixel) variability and observational errors, and to use this information to correct for their effect when large-scale area averages (pixel) are compared to small-scale point estimates. We demonstrate our approach by comparing simulated values of aboveground biomass, woody productivity (woody net primary productivity, NPP) and residence time of woody biomass from four dynamic global vegetation models (DGVMs) with measured inventory data from permanent plots in the Amazon rainforest, a region with the typical problem of low data availability, potential scale mismatch and thus high model uncertainty. We find that the DGVMs under- and overestimate aboveground biomass by 25 % and up to 60 %, respectively. Our comparison metrics provide a quantitative measure for modeldata agreement and show moderate to good agreement with the region-wide spatial biomass pattern detected by plot observations. However, all four DGVMs overestimate woody productivity and underestimate residence time of woody biomass even when accounting for the large uncertainty range of the observational data. This is because DGVMs do not represent the relation between productivity and residence time of woody biomass correctly. Thus, the DGVMs may simulate the correct large-scale patterns of biomass but for the wrong reasons. We conclude that more information about the underlying processes driving biomass distribution are necessary to improve DGVMs. Our approach provides robust statistical measures for any pixel-to-point comparison, which is applicable for evaluation of models and remote-sensing products.

Height Above the Nearest Drainage (HAND), a state-of-the-art topo-hydrological index, has been increasingly used in geo-environmental studies. It describes the local normalized drainage potential of a large region. To date, a trial and error and cumbersome multistep process has been used to obtain the HAND index which does not result in an optimal threshold for contributing area. This study aims at developing a user-friendly geographic information system (GIS) tool, HANDTool, using the Python programming language. We successfully applied the tool for the Chehel-Chai watershed, Iran and the random forest algorithm was used to model groundwater potential. Results revealed that the HAND index made a great contribution to groundwater potential modeling among the other conditioning factors. This tool gives valuable insights on the soil-topography-moisture shared interactions and vegetation condition, especially in ungauged watersheds.

Este trabalho apresenta uma análise da sustentabilidade de 16 municípios que compõem a mesorregião do Marajó, Pará. A pesquisa utilizou o Barômetro da Sustentabilidade (BS) como ferramenta de análise e empregou 20 indicadores para o bem-estar humano (dimensão socioeconômica) e 06 para o bemestar ecológico (dimensão ambiental). A abordagem do Barômetro da Sustentabilidade indicou que os municípios do Marajó variam de potencialmente insustentável a intermediário. O Marajó mostrou um quadro de desequilíbrio em relação ao nível de sustentabilidade dos municípios, e a isso associamos a ineficácia dos resultados de políticas públicas integradas aliada à discrepância dos diferentes modelos de gestão municipal. ABSTRACT: This paper presents an analysis of the sustainability of 16 cities from the Marajó region, State of Pará. We used the Barometer of Sustainability as an analytical tool and applied 20 indicators for human well-being (human dimension) and 06 for ecological health (environmental dimension). The Barometer approach shows a range of sustainability of Marajó municipalities between potentially unsustainable and intermediate. The Marajó displays a disequilibrium in relation to the level of sustainability of the municipalities, and to this we associate the inefficiency of the results of integrated public policies combined with the discrepancy of the different municipal management models.

Climate change and biodiversity loss are expected to simultaneously affect ecosystems, however research on how each driver mediates the effect of the other has been limited in scope. The multiple stressor framework emphasizes non-additive effects, but biodiversity may also buffer the effects of climate change, and climate change may alter which mechanisms underlie biodiversityfunction relationships. Here, we performed an experiment using tank bromeliad ecosystems to test the various ways that rainfall changes and litter diversity may jointly determine ecological processes. Litter diversity and rainfall changes interactively affected multiple functions, but how depends on the process measured. High litter diversity buffered the effects of altered rainfall on detritivore communities, evidence of insurance against impacts of climate change. Altered rainfall affected the mechanisms by which litter diversity influenced decomposition, reducing the importance of complementary attributes of species (complementarity effects), and resulting in an increasing dependence on the maintenance of specific species (dominance effects). Finally, altered rainfall conditions prevented litter diversity from fueling methanogenesis, because such changes in rainfall reduced microbial activity by 58%. Together, these results demonstrate that the effects of climate change and biodiversity loss on ecosystems cannot be understood in isolation and interactions between these stressors can be multifaceted.

Understanding the links between biodiversity, ecosystem services (ES) and human well-being (HWB) is a prerequisite for furthering the agenda of several multilateral environmental agreements and global goals. We performed a systematic review to discuss the extent to which biodiversity research has addressed the interface between ES and HWB and we focused on Brazil as a case study of global relevance. We found that biodiversity research in Brazil remains focused predominantly on biological processes and that research on the links with ES and HWB is in its early phases, exhibiting scarce connections with provisioning and cultural services. This pattern reveals the nature of existing funding policies and scientific gaps in the country. Given the global relevance of Brazil’s stock of biodiversity and ES, we argue that research on their links with HWB will be a crucial element of the national and global process of achieving Sustainable Development Goals by 2030.

Coastal water resources are a worldwide key socio-environmental issue considering the increasing concentration of population in these areas. Here, we propose an integrative transdisciplinary approach of water resource, water management and water access in Recife (NE Brazil). The present-day water situation is conceptualized as an imbricated multi-layered system: a multi-layered water resource, managed by a multi-layered governance system and used by a multi-layered social population. This allows identifying processes of quantitative, qualitative, and sanitary conflicts between governance and population strategies regarding water supply, as well as the institutional and individual denials of these conflicts. Based on this model, we anticipate future water-related problematic fates. Concerning the water resource system, the rapid groundwater level decrease due to unsustainable water predatory strategies, and the very low recharge rate have drastically modified the aquifer system functioning, inducing hydraulic connection between shallow groundwater (contaminated and locally salty) and deep ones (mostly fresh, with local inherited salinity), threatening the deep strategic water resource. Concerning the water governance system, the investments to increase the capacity storage of surface water, the water regulation agencies and the public/private partnership should shortly improve the water supply and wastewater issue. Nevertheless, the water situation will remain highly fragile due to the expected water demand increase, the precipitation decrease and the sea-level increase. Concerning the water access system, the population variably perceives these current and further effects and the possible mitigation policies, and develops alternative individual strategies. Authorities, policymakers and water managers will have to implement a well-balanced water governance, taking into account the specificities of the PPP, public and private groundwater users, and with a strong political willingness for a sustainable water management to ensure water supply for all the population. In other words, an anticipatory and integrated vision is necessary to reduce the discrepancies in this complex system.

The aviation industry is committed to reducing its environmental impact and has established ambitious goals to decrease CO2 emissions by 50% by to 2050. The replacement of fossil fuels by jet-biofuels is one of the main strategies to attain the emission targets. Therefore, the aim of this study was to provide a detailed survey on land availability for sugarcane production, one of the most promising feedstock options for jet-biofuels, in the Brazilian state of São Paulo, where the technological resources are concentrated. This analysis was carried out by integrating georeferenced information on land use, protected areas, soil fertility, terrain slope and climatic conditions into a conditional decision support procedure, based on Boolean inference techniques. Our results showed that 3,501,590 ha would be potentially available for sugarcane expansion in the year 2013. Almost 80% of the mapped lands have high economic potential once they are located in a distance lower than or equal to 25 km from the processing units. If properly included in the productive sector, the available lands could increase sugarcane production by 73% in relation to current levels. That could lead to a growth of 147% (20.6 billion liters) in the ethanol production compared to the production obtained in 2015. Nevertheless, long-term policies, logistic improvements and environmental standard definitions on jet-biofuels still remain as major challenges to boost feedstock production, as well as, to implement financial and regulatory measures necessary to promote jet-biofuels production and use.