Les circulations atmosphériques synoptiques et locales expliquent les variabilités spatiotemporelles des précipitations en Amazonie brésilienne. Ce vaste espace géographique est caractérisé par un réseau pluviométrique peu dense et hétérogène qui limite un suivi exhaustif des pluies. La télédétection est alors l´outil préférentiel afin de pallier cette lacune spatiale. Toutefois, les estimations quantitatives des précipitations (QPE) dérivées des domaines spectraux des infrarouges (IR) et Micro-ondes (MO), ou de techniques multi canaux souffrent d´erreurs. Les sources d´incertitudes et biais des QPE peuvent provenir de la calibration des capteurs, de l´algorithmie et de l´échantillonnage des champs de pluie. Dans cette thèse, l´évaluation des algorithmes TRMM3B42(v6), CMORPH et PERSIANN s´appuie sur une base de données pluviométriques dont la fiabilité est assurée par un contrôle de qualité rigoureux. Sur la période 2003-2009, l’inter comparaison des produits satellitaires montre un faible accord (coefficient de détermination < 0,30) avec la base de référence en Amazonie brésilienne Légale. TRMM3B42(v6) présente la meilleure performance avec un échantillonnage correct des champs de pluie. En outre, un cas d’études montre que les erreurs des QPE par satellite sont plus importantes en milieu forestier, notamment pour le produit MO MWCOMB. Enfin, la correction des grilles satellitaires par PDF Matching permet de réaliser une analyse climatologique approfondie et améliorée par rapport à la base de référence / Synoptic and local atmospheric circulations explain the spatiotemporal variability of rainfall in the Brazilian Amazon. This wide geographical area is characterized by a sparse and heterogeneous rain gauges network, which limits an exhaustive monitoring of rainfall. Remote sensing is then the preferred tool to fill this gap space. However, quantitative precipitation estimates (QPE) derived from spectral Infrared (IR) and Microwave (MW), or multi-channel techniques suffer from errors. Sources of uncertainty and bias of QPE may come from the sensor calibration, the algorithms and sampling of rain fields. In this thesis, the evaluation of algorithms TRMM3B42(v6), CMORPH and PERSIANN relies on a rain gauges database which reliability is assured by rigorous quality control. Over the period 2003-2009, the inter comparison of satellite products show low agreement (coefficient of determination < 0.30) with reference database in the Legal brazilian Amazon. TRMM3B42(v6) has the best performance with a correct sampling of rain fields. In addition, a case study shows that satellite QPE errors are larger in the forest, including the product MW MWCOMB. Finally, the satellite grids correction by PDF Matching allows a detailed and improved climatologically analysis compared with reference database / Circulações atmosféricas sinóticas e locais explicam a variabilidade espaço-temporal das chuvas na Amazônia brasileira. Esta vasta área geográfica é caracterizada por uma rede pluviômetrica heterogênea, o que limita um monitoramento exaustivo da chuva. O sensoriamento remoto é então a ferramenta adequada para preencher este espaço vazio. No entanto, estimativas quantitativas de precipitação (QPE) derivadas do espectro Infravermelho (IR) e Microondas (MO), ou técnicas multi-canais sofrem erros. Fontes de incerteza e de viés das QPE podem vir a partir da calibração de sensores, dos algoritmos e da amostragem de chuva. Nesta tese, a avaliação dos algoritmos TRMM3B42(v6), PERSIANN e CMORPH se baseia em um banco de dados pluviométricos cujo a confiabilidade assegurada por rigoroso controle de qualidade. Durante o período 2003-2009, a comparação dos produtos de satélite mostram baixa concordância (coeficiente de determinação < 0,30) com a base de referência na Amazônia brasileira Legal. TRMM3B42(v6) tem o melhor esempenho com uma amostragem correta da chuva. Além disso, um estudo de caso mostra que os erros das QPE por satélite são maiores na floresta, incluindo o produto MO MWCOMB. Finalmente, a correção das grades de satélite por PDF Matching permite uma análise climatológica detalhada e melhorada em relação a base de referência

The purpose of this work was to present and apply an analytical tool to the flows of goods and income between economic agents and the environment that allows us to calculate the Environmental Gross Domestic Product (EGDP) for Brazilian Legal Amazon. In order to achieve this goal the Environmental Social Accounting Matrix (ESAM) was developed. The model was developed from the traditional Social Accounting Matrix, the Environmental Input-Output Analysis Models, and from the United Nations handbook on the System of Integrated Environmental and Economic Accountings. The EGDP was calculated from the Environmental Input-Output Analysis, since the unavailability of data did not allow the application of the model of ESAM. The flows between the economy and the environment considered were the emissions of green house gases (depreciation of natural capital) and the investment needed to return the air to the same quality it had before being polluted. The results showed that the inclusion of depreciation of natural capital in the Gross Domestic Product (GDP) and in the added value (AV) calculation for each industry of the Brazilian Legal Amazon Region diminished the AV significantly in the industries of cattle (235%), soybean (77%), and other activities of livestock and agricultural (24%). In the Rest of Brazil, the industries with the highest impact of depreciation of natural capital on AV were soybean (30%), cattle (15%) and sugarcane (13%). The total EGDP of the Brazilian Legal Amazon Region was 15% smaller than its total GDP, whereas this difference for the rest of Brazil was 1%. Considering that the GDP is a component of economic wellbeing, the results show a significant reduction in economic wellbeing due to green house gas emissions released into the air by economic activities, mainly due to land use changes. The Environmental Social Accounting Matrix Model is a useful tool to help decision makers since it offers an analytical instrument for economic behavior and the impacts of economic activities on the environment. / The purpose of this work was to present and apply an analytical tool to the flows of goods and income between economic agents and the environment that allows us to calculate the Environmental Gross Domestic Product (EGDP) for Brazilian Legal Amazon. In order to achieve this goal the Environmental Social Accounting Matrix (ESAM) was developed. The model was developed from the traditional Social Accounting Matrix, the Environmental Input-Output Analysis Models, and from the United Nations handbook on the System of Integrated Environmental and Economic Accountings. The EGDP was calculated from the Environmental Input-Output Analysis, since the unavailability of data did not allow the application of the model of ESAM. The flows between the economy and the environment considered were the emissions of green house gases (depreciation of natural capital) and the investment needed to return the air to the same quality it had before being polluted. The results showed that the inclusion of depreciation of natural capital in the Gross Domestic Product (GDP) and in the added value (AV) calculation for each industry of the Brazilian Legal Amazon Region diminished the AV significantly in the industries of cattle (235%), soybean (77%), and other activities of livestock and agricultural (24%). In the Rest of Brazil, the industries with the highest impact of depreciation of natural capital on AV were soybean (30%), cattle (15%) and sugarcane (13%). The total EGDP of the Brazilian Legal Amazon Region was 15% smaller than its total GDP, whereas this difference for the rest of Brazil was 1%. Considering that the GDP is a component of economic wellbeing, the results show a significant reduction in economic wellbeing due to green house gas emissions released into the air by economic activities, mainly due to land use changes. The Environmental Social Accounting Matrix Model is a useful tool to help decision makers since it offers an analytical instrument for economic behavior and the impacts of economic activities on the environment.

The purpose of this work was to present and apply an analytical tool to the flows of goods and income between economic agents and the environment that allows us to calculate the Environmental Gross Domestic Product (EGDP) for Brazilian Legal Amazon. In order to achieve this goal the Environmental Social Accounting Matrix (ESAM) was developed. The model was developed from the traditional Social Accounting Matrix, the Environmental Input-Output Analysis Models, and from the United Nations handbook on the System of Integrated Environmental and Economic Accountings. The EGDP was calculated from the Environmental Input-Output Analysis, since the unavailability of data did not allow the application of the model of ESAM. The flows between the economy and the environment considered were the emissions of green house gases (depreciation of natural capital) and the investment needed to return the air to the same quality it had before being polluted. The results showed that the inclusion of depreciation of natural capital in the Gross Domestic Product (GDP) and in the added value (AV) calculation for each industry of the Brazilian Legal Amazon Region diminished the AV significantly in the industries of cattle (235%), soybean (77%), and other activities of livestock and agricultural (24%). In the Rest of Brazil, the industries with the highest impact of depreciation of natural capital on AV were soybean (30%), cattle (15%) and sugarcane (13%). The total EGDP of the Brazilian Legal Amazon Region was 15% smaller than its total GDP, whereas this difference for the rest of Brazil was 1%. Considering that the GDP is a component of economic wellbeing, the results show a significant reduction in economic wellbeing due to green house gas emissions released into the air by economic activities, mainly due to land use changes. The Environmental Social Accounting Matrix Model is a useful tool to help decision makers since it offers an analytical instrument for economic behavior and the impacts of economic activities on the environment. / The purpose of this work was to present and apply an analytical tool to the flows of goods and income between economic agents and the environment that allows us to calculate the Environmental Gross Domestic Product (EGDP) for Brazilian Legal Amazon. In order to achieve this goal the Environmental Social Accounting Matrix (ESAM) was developed. The model was developed from the traditional Social Accounting Matrix, the Environmental Input-Output Analysis Models, and from the United Nations handbook on the System of Integrated Environmental and Economic Accountings. The EGDP was calculated from the Environmental Input-Output Analysis, since the unavailability of data did not allow the application of the model of ESAM. The flows between the economy and the environment considered were the emissions of green house gases (depreciation of natural capital) and the investment needed to return the air to the same quality it had before being polluted. The results showed that the inclusion of depreciation of natural capital in the Gross Domestic Product (GDP) and in the added value (AV) calculation for each industry of the Brazilian Legal Amazon Region diminished the AV significantly in the industries of cattle (235%), soybean (77%), and other activities of livestock and agricultural (24%). In the Rest of Brazil, the industries with the highest impact of depreciation of natural capital on AV were soybean (30%), cattle (15%) and sugarcane (13%). The total EGDP of the Brazilian Legal Amazon Region was 15% smaller than its total GDP, whereas this difference for the rest of Brazil was 1%. Considering that the GDP is a component of economic wellbeing, the results show a significant reduction in economic wellbeing due to green house gas emissions released into the air by economic activities, mainly due to land use changes. The Environmental Social Accounting Matrix Model is a useful tool to help decision makers since it offers an analytical instrument for economic behavior and the impacts of economic activities on the environment.