京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第20689号 / 工博第4386号 / 新制||工||1682(附属図書館) / 京都大学大学院工学研究科都市環境工学専攻 / (主査)教授 清水 芳久, 教授 田中 宏明, 教授 米田 稔 / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

The world energy needs increasing in a significant number and currently it is mainly sourced from the finite fossil fuel. Other than that, fossil fuel is the main source of CO2 emissions that leads to the increasing rate of global warming that will ultimately change the earth's climate. Thus, researcher began to look for alternative energy that is renewable and has the least impact to the world's climate; one example is wind energy. Sweden has experienced a significant increase in wind energy generation, where the wind production constantly grows in the double-digit rates since 2010. However, developing a wind energy requires a significant research and feasibility study in order to provide an economically sustainable operation. Wind farm needs to be located in areas with a good wind potential, however there are several technical and economic limitation on where the wind farm should be located. On top of that, social rejection might also hinder the development of wind energy. Audio-visual obstruction and disturbance to the natural state of the environment are the main arguments that were used to challenge the development of wind farm. Therefore, a multi-disciplinary study needs to be conducted in order to find the perfect balance; which is exactly what this study is all about. The result of this study shows that despite the threat of climate change, wind farm in Sweden could thrive due to the increasing wind speed across the country. It was also discovered that around 30% of the country is suitable, from the social, technical, and economic point of view, as a new location for wind farms. This study also reveals that most of the canceled / rejected wind farms were probably caused by the social rejection due to their proximity to population center or conservation areas. This study also discussed the concept of place attachment and identity that leads to the NIMBY attitude and reflects the concept of the social acceptance issue that happened in Sweden and on how to localize the wind farm concept to the local residence.

The assessment of regional bioenergy potentials from different types of natural land cover is an integral part of simulation tools that aim to assess local renewable energy systems. This work introduces a new workflow, which evaluates regional bioenergy potentials and its impact on water demand based on geographical information system (GIS)-based land use data, satellite maps on local crop types and soil types, and conversion factors from biomass to bioenergy. The actual annual biomass yield of crops is assessed through an automated process considering the factors of local climate, crop type, soil, and irrigation. The crop biomass yields are validated with historic statistical data, with deviation less than 7% in most cases. Additionally, the resulting bioenergy potentials yield between 10.7 and 12.0 GWh/ha compared with 13.3 GWh/ha from other studies. The potential contribution from bioenergy on the energy demand were investigated in the two case studies, representing the agricultural-dominant rural area in North Germany and suburban region in South Germany: Simulation of the future bioenergy potential for 2050 shows only smaller effects from climate change (less than 4%) and irrigation (below 3%), but the potential to cover up to 21% of the transport fuels demand in scenario supporting biodiesel and bioethanol for transportation.

No / Technology application to agricultural productivity is thought to be the solution to meet food demand of the growing population. In a rapidly changing world, with the prospect of decreasing arable land due to urbanization and industrialization, agricultural output requires a 70 % increase in production levels and efficient growth in the harvesting, distribution and consumption of the resources, to meet demand. There are innovations in Information and Communications Technology that can be applied to the agricultural sector in areas of precision farming, use of farm management software, wireless sensors, and use of agricultural machinery. Remote sensing technology is playing a key role through precision agriculture. This paper highlights ways in which precision agriculture is impacting on agriculture with the use of unmanned aerial vehicles for image capturing, processing and analysis.

<p>The reason for performing an Environmental Impact Assessment (EIA) is to incorporate environmental concern in different kinds of plans and projects. The purpose of such an assessment is to identify and describe direct, indirect and cumulative environmental impacts.</p><p>Geographical Information System (GIS) is a tool that can be used to combine spatial extension of both sensitive areas and different environmental impacts in a quick and easy way. Because of that, descriptions of environmental impacts and motivation of different standpoints on a specific issue can be more correct and easier to make if GIS is used as a tool.</p><p>Hence, GIS can contribute to improve the quality of Environmental Impact Assessments. If the benefits of using GIS are to out weight the costs, geographical information of satisfactory detail, actuality and accuracy need to be available at reasonable prices.</p><p>In this paper, case studies are carried out for three different geographical analyses to investigate the use of GIS as a tool in EIA. From these case studies, more general conclusions about the benefits and limitations of using GIS for Environmental Impact Assessments are also drawn. The criteria, after which benefits and limitations of GIS have been estimated, are data availability, time consumption for performing the analyses and how the results from the performed analyses can make impact assessments easier, motivate different standpoints and increase the comprehensiveness of the EIA-report.</p><p>The three different GIS-analyses includes calculation of pollution load in a catchment area, estimation of the visual impact from planned buildings and generation of alternative locations for underwater pipes. These analyses have been applied on one ongoing EIA for a freight terminal and one ongoing EIA for an underwater wastewater transmission pipe. The analyses in these case studies have been undertaken with ArcGIS software using the extensions “Spatial Analyst” and “3D Analyst”.</p><p>It is shown in this study that due to uncertainty in available model values and the time consuming data manipulation, it is unlikely that calculations of pollution loads with GIS will be used to any larger extent in EIA. To carry out visual assessments with help of GIS to estimate visual impacts is on the other hand assumed to be useful in EIA-work. Both to assess impacts and to estimate how changes in building design can alter those impacts. To use GIS to produce alternative locations for underwater pipes is also considered valuable in EIA-work, even though there is a considerable lack of data to predict the environmental class of marine areas and the connections between available data and real environmental values are weak. However, the risk of damaging vulnerable and high valued marine areas should decrease when applying this type of analyse.</p> / <p>Miljökonsekvensbeskrivningar (MKB) genomförs för att möjliggöra att hänsyn om miljön tas vid olika typer av exploateringsprojekt och exploateringsplaner. Syftet med en miljökonsekvensbeskrivning är att identifiera och beskriva direkta, indirekta och kumulativa miljöeffekter.</p><p>Geografiska informationssystem (GIS) är ett verktyg som kan användas för att snabbt och enkelt kombinera rumslig utsträckning av både känsliga områden och olika miljökonsekvenser. GIS kan därför användas för att underlätta beskrivningar av miljökonsekvenser och motivera olika ställningstaganden. Därigenom kan GIS bidra till bättre grundade och mer rättvisande miljökonsekvensbeskrivningar. Men för att vinsterna av att genomföra GIS-analyser ska överstiga kostnaderna krävs att geografisk information av tillräcklig detaljeringsgrad, aktualitet och säkerhet finns att tillgå till rimliga kostnader.</p><p>Syftet med denna studie har varit att genom fallstudier undersöka om tre olika GIS-analyser kan användas som verktyg i miljöbedömningar med idag tillgänglig data, samt att utifrån fallstudierna dra mer allmänna slutsatser om vinster och begränsningar av att använda GIS i miljökonsekvensbeskrivningar. De kriterier som har använts för att bedöma vinster och begränsningar är tidsåtgång för analyserna, tillgång på data samt hur analyserna kan vara ett stöd i MKB-arbetet genom att underlätta beskrivningar av konsekvenser, motivera ställningstaganden som görs och öka rapportens begriplighet.</p><p>De tre genomförda GIS-analyserna innefattar beräkning av föroreningsbelastning inom ett avrinningsområde, bedömning av byggnaders landskapspåverkan genom synlighetsanalys samt generering av förslag till alternativ ledningsdragning för en undervattensledning. Dessa analyser har tillämpats på en pågående MKB för en detaljplan för en partihall och en pågående MKB för en avloppsvattenledning under vatten. Analyserna i fallstudierna har genomförts i programmet ArcGIS med tilläggen ”Spatial Analyst” och ”3D Analyst”.</p><p>Studien visar att det på grund av osäkerheter i tillgängliga schablonvärden och tidsåtgång för databehandling är osannolikt att beräkning av föroreningsmängd inom ett avrinningsområde med GIS kommer att användas till någon större utsträckning i MKB. Att med GIS genomföra synlighetsanalyser för att utvärdera landskapspåverkan bedöms däremot kunna användas i MKB, både för att bedöma påverkan och för att utvärdera hur en förändring i en byggnads utformning kan förändra den påverkan. Även att med GIS generera förslag till alternativa ledningsdragningar under vatten bedöms kunna fylla en funktion i MKB trots att tillgången på data för att kunna beskriva olika marina områdens naturvärden är klart bristfällig och kopplingen mellan tillgängliga data och verkliga naturvärden är osäker. Denna typ av analys bör ändå minska risken att värdefulla och skyddsvärda marina områden skadas vid lednignsdragning.</p>

Les travaux réalisés au cours de cette thèse s’inscrivent dans les problématiques de localisation d’un véhicule par vision. Nous nous plaçons en particulier dans le cas de parcours sur de longues distances, c’est à dire plusieurs kilomètres. Les méthodes actuelles de localisation et cartographie simultanées souffrent de problèmes de dérives qui les rendent difficilement exploitables après plusieurs centaines de mètres. Nous proposons dans ce mémoire de pallier ces limites en exploitant une connaissance à priori sur la géométrie de l’environnement parcouru.Cette information est extraite d’un Système d’Information Géographique. En particulier, les travaux réalisés se basent sur les modèles 3D des bâtiments des villes et sur une carte de la route.Dans la première partie de ce mémoire, nous proposons une approche permettant de corriger hors ligne une reconstruction SLAM en exploitant la connaissance d’un modèle 3D simple de l’environnement. Cette correction s’applique en deux étapes. En premier lieu, un recalage non-rigide entre le nuage de points reconstruit et le modèle 3D est effectué de sorte à retrouver la cohérence globale de la reconstruction. Dans le but de raffiner le nuage de points obtenu, un ajustement de faisceaux contraint par le SIG est alors effectué sur l’ensemble de la reconstruction.La particularité de cet ajustement de faisceaux est qu’il prend implicitement en compte les contraintes géométriques apportées par le modèle 3D. La reconstruction ainsi corrigée est alors utilisée en tant que base de données pour la relocalisation en ligne d’une caméra mobile. La précision de relocalisation obtenue est en particulier suffisante pour les applications de réalité augmentée.Dans la deuxième partie de ce mémoire, nous détaillons une solution permettant de corriger en ligne la reconstruction SLAM. Pour cela, les contraintes géométriques apportées par le SIG sont exploitées au fur et à mesure de la trajectoire du véhicule. Nous montrons tout d’abord que la connaissance de la position relative de la caméra par rapport à la route permet de corriger de façon robuste la dérive de facteur d’échelle. De plus, lorsque les contraintes géométriques sont suffisantes, la reconstruction SLAM réalisée jusqu’à l’instant courant est recalée sur le SIG.Cela permet de corriger ponctuellement la dérive observée sur la position courante de la caméra.Le processus complet permet dès lors de localiser le véhicule avec une précision semblable à celle d’un système GPS sur des trajectoires de plusieurs kilomètres.Les deux méthodes proposées ont été testées à la fois sur des séquences de synthèse et réelles. Des résultats qualitatifs et quantitatifs sont présentés tout au long de ce mémoire. / This thesis deals with the vision based geolocalisation of a vehicle. In particular, the problem of localisation on large sequences, i.e. several kilometers, is studied. In this context, state of the art Simultaneous Localisation and Mapping systems suffer from drift. In consequence,existing SLAM methods can not provide accurate localisation of the camera after several hundred meters. Thus, we propose in this thesis to avoid the drift phenomenon by exploiting a simple knowledge about the geometry of the environment. This information is provided by a Geographical Information System. In particular, our work is based on coarse 3D city models and road maps.In the first part, we propose an offline two steps correction of SLAM reconstructions based on a 3D city model of the area. First, the reconstructed 3D point cloud and this 3D city model are aligned through a non-rigid transformation. This step allows the SLAM reconstruction to regain its global consistency. Then, a bundle adjustment constrained with the GIS is applied on the entire reconstruction to refine its geometry. The innovation of this bundle adjustment is that it takes into account the geometrical constraints provided by the 3D city model in a single term.The obtained 3D point cloud can then be considered as a feature landmark database. Finally, this database is used to localise a moving camera in real-time. In pratice, the precision of the obtained localisation is sufficient for augmented reality applications. In the second part of this manuscript, we present a solution which makes possible the online correction of a SLAM reconstruction. The GIS geometrical constraints are exploited over the vehicle trajectory. First, we show that the scale factor drift can be robustly corrected thanks to the knowledge of the ground plane equation. Furthermore, the current SLAM reconstruction is fitted onto the GIS when the geometrical constraints are sufficient. It punctually ensures the correction of the current camera position. The entire process allows the geolocalisation of a vehicle on several kilometers. The obtained precision is close to GPS.The two proposed solutions have been validated of both synthetic and real sequences. Quantitative and qualitative experiments are presented over this manuscript.

Orientador: Sérgio Luis de Carvalho / Banca: Edson Pereira Tangerino / Banca: Maria José Alencar Vilela / Resumo: O estudo foi realizado na microbacia do Córrego Campestre que está localizada no Município de Lins, Estado de São Paulo, compreendendo uma área de 80 km2. Nessa área está inserida a Cidade de Lins com uma área urbana de 20 km2, perfazendo 25% da microbacia em estudo. Através das coletas e análises de amostras de água, no período de seca e chuva, dos principais córregos da microbacia do Córrego Campestre, foi possível obter resultados do Índice de Qualidade da Água - ÍQA. Com bases nos resultados combinados, com as informações obtidas através das imagens de satélite e fotos dos locais, constatou-se que os efeitos das atividades antrópicas principais desenvolvidas na região urbana, a urbanização e lançamentos de resíduos nos cursos d'água sem nenhuma forma de tratamento, provocaram elevados teores de matéria orgânica e coliformes, comprovado através da analise das características físico-químicas e bacteriológicas. Constatou-se também, a partir das observações feitas "in loco", fortes evidências de degradação da microbacia, principalmente por um antigo lixão que está localizado às margens do Córrego Campestre. Através da analise das imagens de satélite, utilizando um Sistema de Informações Geográficas - SIG, foi possível identificar a inexistência de vegetação ciliar, o que contribuiu para o carregamento de resíduos poluentes do solo aos cursos d'água após as chuvas intensas. / Abstract: The study it was carried through in the microwatershed of the Campestre Stream that is located in the region of the Lins city, São Paulo State, understanding an area of 80 km 2. In this area the Lins city with an urban area of 20 is inserted km 2, with about 25% of the microwatershed in study. Through the collections and analyze chemistry of samples of water, in the drought periods and rain, of the main streams in the microwatershed of the Rural Stream, it was possible like this to obtain results of the Index of Quality of the Water - ÍQA and analyses of the parameters physical, chemical and biological. Those combined results with the information obtained through the satellite images and pictures of the places, it was verified that the effects of the activities main human actions developed in the urban area, the urbanization and releases of residues in the courses of water without any handling form, they provoked high tenors of organic matter, coliformes and nutritious, proven through it analyzes it of the physiochemical and bacteriological characteristics. It was also verified, starting from the observations done " in loco ", strong evidences of degradation of the microwatershed mainly for an old deposit garbage that is located to the margins of the Rural Stream. Through it analyzes her/it of the satellite images, being used... (Complete abstract click electronic access below) / Mestre

Tendo-se em conta a importância da lavoura tritícola para o país, e que o aumento da produtividade depende também da correta execução das operações de implantação e condução da cultura, este trabalho tem como objetivo avaliar a estabilidade dos processos de produção e ao atendimento aos padrões especificados de um sistema de plantio direto na cultura do trigo. Os dados foram coletados em um talhão da Fazenda Paiquerê, do Grupo Lúcio Miranda, no município de Piraí do Sul, Estado do Paraná. Foram consideradas seis operações do processo produtivo do trigo sob sistema plantio direto: colheita da cultura antecessora, pulverização pré-semeadura, semeadura, adubação de cobertura, pulverização para tratamento fitossanitário e colheita do trigo. Para a avaliação do sistema produtivo em questão, foram determinados, para cada operação considerada, indicadores de desempenho e limites de especificação para cada indicador. Foram utilizados para a avaliação dos resultados, ferramentas de análise estatística descritiva, de controle da qualidade e do SIG. Pode-se concluir que o sistema de plantio direto de trigo considerado neste estudo não atendeu aos padrões especificados, necessitando de melhorias em todas as operações envolvidas no processo. A utilização das cartas de controle, do histograma e do SIG identificou os pontos que apresentaram problemas e que não se encontravam dentro das especificações agronômicas em todas as operações avaliadas. O uso de técnicas geoestatísticas aliadas à tecnologia SIG permitiu o adequado zoneamento dos indicadores analisados, identificando áreas com diferentes quantidades. / Considering the relevance of wheat crop in Brazil, and that the yield increase depends also on the correct execution of implementation and management of the crop, this work aimed to evaluate how stable the production processes are and to the specified patterns attending of a no-tillage wheat crop. The data were collected at Paiquerê farm, belonged to Lúcio Miranda Group, in of Piraí do Sul city, Paraná state. Six operations of the production process of the no-tillage wheat were considered: harvesting of the former crop, spraying before drilling, drilling, fertilizer spreading, spraying and wheat harvesting. To evaluate the studied production system, it were determined to each considered operation, performance indicators and specification limits to each indicator. Tools of descriptive statistics analysis, of quality control and GIS were used. It was concluded that the evaluated system did not attend the specified patterns, needing improvements in all operations involved in the process. The utilization of control charts, histogram and GIS identified where there were problems and that were not in the agronomic specification range in all the evaluated operations. The use of geostatistical techniques together GIS technology allowed the adequate zoning of the studied indicators, identifying areas with distinct quantities.

O objetivo deste trabalho é apresentar os conceitos da logística, em especial a logística de serviços, e algumas de suas ferramentas, como a roteirização de veículos e previsão de demanda por serviços, aplicadas aos serviços urbanos, neste caso o serviço de remendos em pavimentos, visando aumentar a eficiência desse serviço. O serviço de remendos, muitas vezes chamado de tapa-buracos, é uma atividade de manutenção comum nas cidades. Para observar a aplicação das ferramentas foi elaborado um estudo de caso na cidade de São Carlos. Dados sobre o serviço de remendos em pavimentos foram coletados e, com a ajuda de um sistema de informações geográficas – SIG, foram gerados roteiros que foram comparados com os dados originais. As rotas simuladas pelo SIG foram mais eficientes do que as praticadas na realidade, mostrando a utilidade dos conceitos da logística e também a utilidade do SIG na gerência da infra-estrutura urbana. A previsão de demanda por serviços de remendos não pôde ser observada devido à falta de dados históricos, fundamentais a essa etapa do trabalho. / The objective of this work is to introduce the concepts of logistics, especially the service logistics and some of its tools as the vehicle routing and the demand forecast for services, applied to the urban services, in this case the patching service in pavements in order to increase the efficiency of this service. The patching service, many times called tapa-buracos (in Brazil), is a common activity of maintenance in the cities. To observe the application of the tools one case study was elaborated in the city of São Carlos. Data about the patching service in pavements were collected and, helped by the geographic information system – GIS, routes were created and compared to the original data. The paths simulated by the GIS were more efficient than the real ones, showing the utility of the logistics concepts and also the utility of the GIS on the management of the urban infrastructure. The demand forecast for services of patching could not be observed due of the lack of historical data, essential to this part of the work.

A cana-de-açúcar é uma cultura de grande expressão em vários estados brasileiros. Para que a mesma apresente bons níveis de produtividade, quer seja nas áreas tradicionais de cultivo ou nas áreas em expansão, é de extrema importância que haja a seleção e alocação das diferentes variedades de acordo com os ambientes de produção, os quais envolvem aspectos relacionados à qualidade dos solos e aos níveis esperados de produtividade em função disso. Apesar da importância do ambiente de produção para o manejo varietal e operacional do canavial, este não considera aspectos relacionados ao clima dos locais de cultivo. Sendo assim, o presente estudo buscou desenvolver um procedimento para a obtenção de classes de ambientes de produção para a cultura da cana-de-açúcar sob o enfoque climático, por meio do uso de modelos agrometeorológicos de estimativa das produtividades potencial e atingível, e da eficiência climática resultante dessas. A partir da utilização do modelo da Zona Agroecológica (modelo FAO), associado à penalização da produtividade pelo déficit hídrico, foi calculada a produtividade potencial e a produtividade atingível da cultura da cana-de-açúcar em 178 localidades do estado de São Paulo. Por meio do quociente entre essas produtividades, estimou-se a eficiência climática (). Tanto as variáveis climáticas como as produtividades foram espacializadas por meio da técnica de regressão em que foram utilizadas coordenadas geográficas integradas. Foi calculado o balanço hídrico climatológico normal, na escala mensal, para a caracterização climática do Estado, e o balanço hídrico seqüencial, na escala decendial, para a penalização da produtividade potencial pelo déficit hídrico e obtenção da produtividade atingível. O balanço hídrico foi determinado de acordo com o método proposto por Thornthwaite e Mather (1955) e a evapotranspiração de referência foi calculada utilizando-se a equação de Priestley e Taylor (1972). Essas simulações foram realizadas considerando-se períodos de 30 anos de dados, de 1973 a 2003, disponibilizados pelo SIGRH/DAEE-SP. Foram consideradas cinco épocas de plantio para a cana planta (fevereiro, março, julho, setembro e outubro) e três ciclos de maturação para a cana-soca (precoce, média e tardia). Foi inserido no modelo agrometeorológico empregado um fator de penalização das produtividades potencial e atingível pela ocorrência de baixas temperaturas do ar. De acordo com os mapas, foi possível propor uma classificação dos ambientes com base na produtividade atingível, a qual variou de 50 até 100 t ha-1 e com base nos valores obtidos de eficiência climática, que variaram de 0,35 a 0,65. Conclui-se com isso, que o modelo agrometeorológico apresentou desempenho satisfatório, possibilitando juntamente com o emprego de um sistema de informações geográficas, a obtenção das classes climáticas dos ambientes de produção para a cultura da cana-de-açúcar, que poderão subsidiar o planejamento das usinas quanto aos manejos varietal e operacional dos canaviais. / Sugarcane is a crop of great expression in several Brazilian states. For obtaining success, with high yield levels, in both traditional and expanding areas, it is very important that the selection and allocation of different varieties be done according to the production environments, which normally involve aspects related to the soil quality and the expected yield levels associated to that. Even considering the importance of the production environment concept for the varietal and operational management in the sugarcane mills, it does not consider aspects related to the climate conditions of the production areas. Based on that, the present study aimed to develop a procedure for obtaining the classification of the sugarcane production environments according to the climatic conditions, by the use of agrometeorological models for estimating potential (PP) and attainable (PA) yields, and the resulting climatic efficiency (). This study used the association of the FAO Agro-ecological Zone model and the water deficit depletion model respectively for estimating PP and PA for sugarcane crop in 178 locations in the state of São Paulo. The was calculated by the ratio between PA and PP. All the variables related to the climate and to the yield were spacialized by the multiple linear regression technique, which uses the geographic coordinates, altitude and their integration as independent variables. The monthly normal water balance was used to characterize the climate and the water availability conditions of the state, while the 10-day serial water balance was used to characterize the water deficit during the crop cycle for each year of the historical series. The water balance was determined by the Thornthwaite and Mather (1955) method and the reference evapotranspiration was estimated by the Priestley and Taylor (1972) equation. The sugarcane crop yield were simulated for a 31-year period, from 1973 to 2003, with data obtained from SIGRH/DAEE-SP. For such simulations five planting dates were considered for plant crop (February, March, July, September and October) and three cycles of maturation for ratoon crop (early, medium and late). A low temperature depletion factor was introduced in the yield models for considering the negative effect of frosts on sugarcane crop. According to the maps generated, it was possible to propose a production environment classification based on PA, which ranged from 50 to 100 t ha-1, and also based on , which varied between 0.35 and 0.65. These data when applied to a geographical information system (GIS) allowed to classify the production environments not only by using soil data but also using the classes defined by the climatic conditions, giving subsides for the crop planning, variety management and operational activities.