No description available.

El texto completo de este trabajo no está disponible en el Repositorio Académico UPC por restricciones de la casa editorial donde ha sido publicado. / Recurrent economic and human losses occur in populated areas caused by the debris flow, known in Peru as “huayco” and for which there is little information. To determine the effectiveness of debris flow barriers installed in 2016 in three high-risk creeks with slopes ranging from 29 to 35%, it was analyzed satellite images with the Geographical Information System (GIS). For that, it is necessary to obtain the volumes of both soil erosion in the upper basin transported by the debris flow and the retained solids by the barriers. Topographic, geological, geomorphological, and hydrological characteristics were evaluated, as well as the evolution of the population in the dejection cone. It was no possible to obtain results for all the destructive events because there are no cartographies, and the available satellite images do not have enough temporal or spatial resolution or present cloudiness greater than 20%. The field investigations after the ENSO 2017 occurrence made it possible to verify that the debris flow barriers allowed to avoid the loss of human life and material damage. The field measured volumes of the solids retained by the barriers and the previously estimated volumes of erosion were compared, and retention efficiency of 80–90% was obtained. It recommends continuing with this research because it is necessary to know debris flow characteristics for analyzing the convenience and technical requirements for the implementation of retention barriers in other dangerous streams that are difficult to access.

A new approach to improve the determination of building boundaries through automatic processing of light detection and ranging (LIDAR) data is presented. The LIDAR data is processed and interpolated into a grayscale image of intensity values corresponding to height measurements. Ground measurements are separated from non-ground measurements by using a progressive morphological filter. With these measurements now distinct, further separation of non-ground measurements into building and non-building measurements is performed by growing regions with similar characteristics. These building areas are then refined, resulting in a ground plan representation of building boundaries, known as building footprints. Several algorithms are then implemented to clean these footprints. A new method is developed to analyze actual known satellite imagery in order to confirm identified building footprints.

A relative, lie-value image normalisation (LVIN) procedure was investigated as a means of estimating surface reflectances from sequences of Landsat TM and ETM+ imagery, and standardising image data for change detection studies when there are uncertainties in sensor calibration and atmospheric parameters over time. The basis of the LVIX procedure is that for an A-date sequence, the digital numbers (DNs) of N-1 overpass images can be mapped to the reflectance values of a reference image for a set of pseudo- invariant targets (PITs) common to all images in the sequence. The robust M-estimator was employed to provide the transformation function that achieved the mapping. The investigation also showed that in some instances the LVIN procedure could incorporate the modelled Path DN-the modelled DN for a target of zero surface reflectance. A lack of surface validation data was a limitation in the investigation. However, a qualitative evaluation of the LVIN procedure was possible by examining the pre- and post-normalisation image histograms. In a comparison with the results of the 6S radiative transfer code, it war observed that when both overpass and reference images were acquired with the same sensor, the LVIK procedure appeared t o correct for atmospheric effects; and when overpass and reference images were with different sensors, the LVIN procedure also corrected for between-sensor differences. Moreover, it was demonstrated for the more "temporally-invariant" PITs that the procedure retrieved surface reflectances that were on average within ±0.02 reflectance units. / The ability of the LVIK procedure to standardise sequences of image data was further demonstrated in the study of vegetation change. The normalised difference vegetation index (NDVI) was calculated from LVIN estimates of surface reflectance for a selection of sites around the township of Mt. Barker, Western Australia. NDVI data had characteristics consistent with data that have been corrected for atmospheric effects. A modification to the LVIN procedure was also proposed based on an investigation of some empirically-derived vegetation reflectance relationships. Research into the robustness of the relationships for a greater range of vegetation types is recommended.

The distance of sound propagation in water is inversely proportional to the frequency; there for the low frequency noise generated by distance shipping contribute most to the low frequency components of the ocean ambient noise. In order to improve the performance of sonar application, understanding and predication shipping noise is important. In this study, noise data was collected by vertical line array (VLA), in South China Sea of Asian Seas International Acoustics Experiment (ASIEX) in 2001. Due to the limited access to the satellite imagery, and based on the assumption that fishing vessels in South China Sea has similar operational pattern and lighting in recent years, the night time satellite image data was acquired by Defense Meteorological Satellite Program DMSP - Operational Linescan System(DMSP-OLS) of 2005. Linear regression was analyzed between fishing light pixel distribution and low frequency noise level variation, fishing light was expressed in terms of total pixel number and total light intensity, and distance from fishing vessel to VLA was assumed by spherical and cylindrical spread. The results show high correlation between total light intensity and noise level, and cylindrical spreading is better assumption in shallow water on the continental shelf at low frequency. As a conclusion, this study successfully prove that by using the night satellite image, the low frequency fishing vessel noise can be predicted with reasonable accuracy.

Grasslands are a significant source of biodiversity in farmed landscapes that is important to monitor. New generation satellites such as Sentinel-2 offer new opportunities for grassland’s monitoring thanks to their combined high spatial and temporal resolutions. Conversely, the new type of data provided by these sensors involves big data and high dimensional issues because of the increasing number of pixels to process and the large number of spectro-temporal variables. This thesis explores the potential of the new generation satellites to monitor biodiversity and factors that influence biodiversity in semi-natural grasslands. Tools suitable for the statistical analysis of grasslands using dense satellite image time series (SITS) with high spatial resolution are provided. First, we show that the spectro-temporal response of grasslands is characterized by its variability within and among the grasslands. Then, for the statistical analysis, grasslands are modeled at the object level to be consistent with ecological models that represent grasslands at the field scale. We propose to model the distribution of pixels in a grassland by a Gaussian distribution. Following this modeling, similarity measures between two Gaussian distributions robust to the high dimension are developed for the lassification of grasslands using dense SITS: the High-Dimensional Kullback-Leibler Divergence and the -Gaussian Mean Kernel. The latter outperforms conventional methods used with Support Vector Machines for the classification of grasslands according to their management practices and to their age. Finally, indicators of grassland biodiversity issued from dense SITS are proposed through spectro-temporal heterogeneity measures derived from the unsupervised clustering of grasslands. Their correlation with the Shannon index is significant but low. The results suggest that the spectro-temporal variations measured from SITS at a spatial resolution of 10 meters covering the period when the practices occur are more related to the intensity of management practices than to the species diversity. Therefore, although the spatial and spectral properties of Sentinel-2 seem limited to assess the species diversity in grasslands directly, this satellite should make possible the continuous monitoring of factors influencing biodiversity in grasslands. In this thesis, we provided methods that account for the heterogeneity within grasslands and enable the use of all the spectral and temporal information provided by new generation satellites.

Radiative transfer models (RTM) have been successfully used to simulate the effect of forest structural and biochemical characteristics, such as tree sizes and shapes, leaf area index (LAI), leaf angle distribution (LAD), on the canopy radiative budget. One particular use of RTM is the analysis of the reflected light by the canopy, which can be measured by remote sensing techniques. RTM allows a physically based interpretation of the reflectance quantity measured by satellite and can help disentangling the multiple source of variation of the reflectance signal. The DART model - Discrete Anisotropic Radiative Transfer - is one of the most complex three-dimensional RTM, since it uses an accurate mathematical approach of physical processes and a great realism of the landscapes under simulation. Its main simulation outputs are the reflectance of the scene (e.g. a forest stand) at particular spectral wavelength from bottom and top of the atmosphere, the simulation of satellite images, and the simulation of localized radiative budget. Despite the DART potential in analyzing biophysical parameters from remote sensing data, few studies report its application in forest plantations in Brazil, which can have a large number of important field measurements to parameterize the model. The main objective of this project is to evaluate if the DART RTM can help understand the satellite-measured canopy reflectance of Eucalyptus plantations and in particular if DART RTM can improve LAI estimation rather than use only empirical models, as spectral vegetation indices. DART model was parameterized using extensive in situ data obtained from a clonal test, part of the EucFlux project. The specific objectives were: i) parameterize the DART model at different growth stages and for different clonal materials of Eucalyptus plantations and compare simulated reflectance with high resolution satellite images acquired on the same site; ii) analyze the relationship between the Leaf Area Index (LAI) and Spectral Vegetation Indices (SVI) based on empirical relationships, and then use the DART model; iii) analyze the advantage and drawbacks of using a generic relationship or a clone-specific relationship between LAI and SVI, and find other criteria for grouping the genotypes in the same. In Chapter 2, we demonstrated the good performance of DART to simulate canopy reflectance of Eucalyptus forest plantations. The simulated reflectance was similar to those measured by very high resolution images from satellite, despite some discrepancies found in the near infrared region. Then, in Chapter 3, we showed that empirical relationships between LAI and SVI were able to give a reasonable precision for generic relationships; however, genotype-scale relationships gave even better results. The same methodology applied on a DART simulated dataset lead to the same conclusions. An intermediate possibility of grouping the genotypes regarding their litter or leaf optical properties gave intermediate performance. We finally concluded about the superiority of NDVI to estimate LAI using a genotype-specific calibration. Overall, DART simulated datasets created in this work enable to calibrate different LAI -SVI relationships in terms of genotypes, sensors and acquisition characteristics. / Modelos de transferência de radiação (MTR) têm sido utilizados com sucesso para simular o efeito das características estruturais e bioquímicas florestais, como tamanhos de árvores e formas, índice de área foliar (IAF), distribuição angular das folhas (DAF) e sobre o balanço de radiação. Um uso particular do MTR é a análise da radiação refletida pela copa, o que pode ser medido através de técnicas de sensoriamento remoto. O MTR pode permitir a interpretação física da quantidade de reflectância medido por satélite, e pode ajudar a diferenciar as múltiplas fontes de variação do sinal de reflectância. O modelo DART - Discrete Anisotropic Radiative Transfer - é um dos modelos tridimensionais de transferência de radiação mais complexos, uma vez que utiliza uma abordagem matemática precisa e um grande realismo na simulação das paisagens. Seus principais resultados de simulação são a reflectância da cena (por exemplo, um povoamento florestal) em determinado comprimento de onda espectral em relação ao topo e à base da atmosfera, a simulação de imagens de satélite e a simulação do balanço de radiação. Apesar do potencial do DART na análise de parâmetros biofísicos de paisagens florestais a partir de dados de sensoriamento remoto, existem poucos estudos sobre sua aplicação em povoamentos florestais no Brasil; que podem dispor de um elevado número de medições de campo importantes para a parametrização do modelo. O principal objetivo deste estudo foi avaliar se o DART pode ajudar a compreender o comportamento da reflectância do dossel das plantações de eucalipto oriunda de dados de imagens de satélite e, em particular, se DART pode melhorar a estimativa do IAF ao invés do uso somente de modelos empíricos como índices espectrais da vegetação. O DART foi parametrizado com extensos dados de campo adquiridos em um experimento com testes clonais do Projeto Eucflux. Os objetivos específicos foram: i) parametrizar o modelo DART em diferentes idades e com diferentes materiais genéticos de plantações de eucalipto e comparar a refletância simulada com imagens de satélite de alta resolução adquiridas no mesmo local; ii) analisar a relação entre o Índice de Área Foliar (IAF) e Índices Espectrais de Vegetação (IEV\'s) com base em relações empíricas, e, em seguida, usando o modelo DART; iii) analisar as vantagens e as limitações do uso de uma relação genérica ou uma relação específica do genótipo entre IAF e IV e encontrar outros critérios para agrupar os genótipos. No Capítulo 2 foi demonstrado bom desempenho do DART para simular a reflectância do dossel das florestas plantadas de eucalipto. As refletâncias simuladas foram semelhantes com as obtidas pelas imagens de satélite de alta resolução, apesar de algumas discrepâncias encontradas na região do infravermelho próximo. No Capítulo 3, foi mostrado que as relações empíricas entre os IEV\'s e os IAF\'s foram capazes de estimar com razoável precisão para as relações genéricas dos plantios. Contudo, as estimativas por genótipo deram resultados superiores. A mesma metodologia foi aplicada em um conjunto de dados simulados pelo DART com as mesmas conclusões. Uma possibilidade intermediária de agrupar os genótipos foi em função das propriedades ópticas da serapilheira ou das folhas, com desempenhos intermediários. Nós concluímos sobre a superioridade do NDVI para estimar o LAI usando uma calibração específica para cada genótipo. Em termos mais gerais, os dados simulados com o modelo DART utilizados neste trabalho permitiram calibrar diferentes relações IAF-IEV em função dos genótipos, sensores e características de aquisição.

Radiative transfer models (RTM) have been successfully used to simulate the effect of forest structural and biochemical characteristics, such as tree sizes and shapes, leaf area index (LAI), leaf angle distribution (LAD), on the canopy radiative budget. One particular use of RTM is the analysis of the reflected light by the canopy, which can be measured by remote sensing techniques. RTM allows a physically based interpretation of the reflectance quantity measured by satellite and can help disentangling the multiple source of variation of the reflectance signal. The DART model - Discrete Anisotropic Radiative Transfer - is one of the most complex three-dimensional RTM, since it uses an accurate mathematical approach of physical processes and a great realism of the landscapes under simulation. Its main simulation outputs are the reflectance of the scene (e.g. a forest stand) at particular spectral wavelength from bottom and top of the atmosphere, the simulation of satellite images, and the simulation of localized radiative budget. Despite the DART potential in analyzing biophysical parameters from remote sensing data, few studies report its application in forest plantations in Brazil, which can have a large number of important field measurements to parameterize the model. The main objective of this project is to evaluate if the DART RTM can help understand the satellite-measured canopy reflectance of Eucalyptus plantations and in particular if DART RTM can improve LAI estimation rather than use only empirical models, as spectral vegetation indices. DART model was parameterized using extensive in situ data obtained from a clonal test, part of the EucFlux project. The specific objectives were: i) parameterize the DART model at different growth stages and for different clonal materials of Eucalyptus plantations and compare simulated reflectance with high resolution satellite images acquired on the same site; ii) analyze the relationship between the Leaf Area Index (LAI) and Spectral Vegetation Indices (SVI) based on empirical relationships, and then use the DART model; iii) analyze the advantage and drawbacks of using a generic relationship or a clone-specific relationship between LAI and SVI, and find other criteria for grouping the genotypes in the same. In Chapter 2, we demonstrated the good performance of DART to simulate canopy reflectance of Eucalyptus forest plantations. The simulated reflectance was similar to those measured by very high resolution images from satellite, despite some discrepancies found in the near infrared region. Then, in Chapter 3, we showed that empirical relationships between LAI and SVI were able to give a reasonable precision for generic relationships; however, genotype-scale relationships gave even better results. The same methodology applied on a DART simulated dataset lead to the same conclusions. An intermediate possibility of grouping the genotypes regarding their litter or leaf optical properties gave intermediate performance. We finally concluded about the superiority of NDVI to estimate LAI using a genotype-specific calibration. Overall, DART simulated datasets created in this work enable to calibrate different LAI -SVI relationships in terms of genotypes, sensors and acquisition characteristics. / Modelos de transferência de radiação (MTR) têm sido utilizados com sucesso para simular o efeito das características estruturais e bioquímicas florestais, como tamanhos de árvores e formas, índice de área foliar (IAF), distribuição angular das folhas (DAF) e sobre o balanço de radiação. Um uso particular do MTR é a análise da radiação refletida pela copa, o que pode ser medido através de técnicas de sensoriamento remoto. O MTR pode permitir a interpretação física da quantidade de reflectância medido por satélite, e pode ajudar a diferenciar as múltiplas fontes de variação do sinal de reflectância. O modelo DART - Discrete Anisotropic Radiative Transfer - é um dos modelos tridimensionais de transferência de radiação mais complexos, uma vez que utiliza uma abordagem matemática precisa e um grande realismo na simulação das paisagens. Seus principais resultados de simulação são a reflectância da cena (por exemplo, um povoamento florestal) em determinado comprimento de onda espectral em relação ao topo e à base da atmosfera, a simulação de imagens de satélite e a simulação do balanço de radiação. Apesar do potencial do DART na análise de parâmetros biofísicos de paisagens florestais a partir de dados de sensoriamento remoto, existem poucos estudos sobre sua aplicação em povoamentos florestais no Brasil; que podem dispor de um elevado número de medições de campo importantes para a parametrização do modelo. O principal objetivo deste estudo foi avaliar se o DART pode ajudar a compreender o comportamento da reflectância do dossel das plantações de eucalipto oriunda de dados de imagens de satélite e, em particular, se DART pode melhorar a estimativa do IAF ao invés do uso somente de modelos empíricos como índices espectrais da vegetação. O DART foi parametrizado com extensos dados de campo adquiridos em um experimento com testes clonais do Projeto Eucflux. Os objetivos específicos foram: i) parametrizar o modelo DART em diferentes idades e com diferentes materiais genéticos de plantações de eucalipto e comparar a refletância simulada com imagens de satélite de alta resolução adquiridas no mesmo local; ii) analisar a relação entre o Índice de Área Foliar (IAF) e Índices Espectrais de Vegetação (IEV\'s) com base em relações empíricas, e, em seguida, usando o modelo DART; iii) analisar as vantagens e as limitações do uso de uma relação genérica ou uma relação específica do genótipo entre IAF e IV e encontrar outros critérios para agrupar os genótipos. No Capítulo 2 foi demonstrado bom desempenho do DART para simular a reflectância do dossel das florestas plantadas de eucalipto. As refletâncias simuladas foram semelhantes com as obtidas pelas imagens de satélite de alta resolução, apesar de algumas discrepâncias encontradas na região do infravermelho próximo. No Capítulo 3, foi mostrado que as relações empíricas entre os IEV\'s e os IAF\'s foram capazes de estimar com razoável precisão para as relações genéricas dos plantios. Contudo, as estimativas por genótipo deram resultados superiores. A mesma metodologia foi aplicada em um conjunto de dados simulados pelo DART com as mesmas conclusões. Uma possibilidade intermediária de agrupar os genótipos foi em função das propriedades ópticas da serapilheira ou das folhas, com desempenhos intermediários. Nós concluímos sobre a superioridade do NDVI para estimar o LAI usando uma calibração específica para cada genótipo. Em termos mais gerais, os dados simulados com o modelo DART utilizados neste trabalho permitiram calibrar diferentes relações IAF-IEV em função dos genótipos, sensores e características de aquisição.

In this dissertation, the mapping of outdoor environments and localization of a mobile robot in that setting is considered. It is well known that in the absence of a map or precise pose estimates, localization and mapping is a coupled problem. However, in this dissertation this problem is decoupled in to two disjoint steps / mapping and localization on the acquired map. First the images of the outdoor environment is downloaded from a website such as Google Earth and then these images are processed by utilizing several artificial neural network topologies to create maps. Once these maps are obtained, the localization is done by using Monte Carlo localization. This dissertation addresses a solution for the information which is most of the time taken for granted in most studies / a prior map of environment. Mapping is solved by using a novel approach / the map of the environment is created by processing satellite images. Several global localization techniques are developed and evaluated to be used with these map so as to localize a mobile robot globally. The outcome of this novel approach presented here may serve as a virtual GPS. Mobile phone applications can localize a user within a circle of uncertainty without GPS. This crude localization may be used to download relevant satellite images of the local environment. Once the mobile robot is localized on the map created from the satellite images by using available techniques in the literature i.e. Monte Carlo localization, it may be claimed that it is localized on Earth.

First part of the thesis introduces possibilities of depicting of the data provided by the MSG (Meteosat Second Generation) satellite. Different color combinations (RGB - red, green, blue) description made by the basic channels combination follows. The main part of this thesis is focused on T-re profiles (T - temperature, re - effective cloud particle radius). The analysis of profiles of temperature versus the effective radius of cloud particles shows the vertical distribution of the effective particle size in clouds. The profiles are computed and constructed from the satellite retrieved data and they show graphically the distribution of cloud particles and indirectly represent the severity of the developing storms. Forecasting and predicting dangerous phenomena such as hailstorms or tornados that occur in severe storms is the main objective of the technique. This nowcasting tool was tested in Central Europe and the end of this thesis presents the results of the testing.