Este trabalho objetiva o desenvolvimento de um pacote no software estatístico R com a implementação de estimadores em modelos de regressão normal univariados com parametrização geral, uma particularidade do modelo definido em Patriota e Lemonte (2011). Essa classe contempla uma ampla gama de modelos conhecidos, tais como modelos de regressão não lineares e heteroscedásticos. São implementadas correções nos estimadores de máxima verossimilhança e na estatística de razão de verossimilhanças. Tais correções são efetivas quando o tamanho amostral é pequeno. Para a correção do estimador de máxima verossimilhança, considerou-se a correção do viés de segunda ordem, enquanto que para a estatística da razão de verossimilhanças aplicou-se a correção desenvolvida em Skovgaard (2001). Todas as funcionalidades do pacote são descritas detalhadamente neste trabalho. Para avaliar a qualidade do algoritmo desenvolvido, realizaram-se simulações de Monte Carlo para diferentes cenários, avaliando taxas de convergência, erros da estimação e eficiência das correções de viés e de Skovgaard. / This work aims to develop a package in R language with the implementation of normal regression models with general parameterization, proposed in Patriota and Lemonte (2011). This model unifies important models, such as nonlinear heteroscedastic models. Corrections are implemented for the MLEs and likelihood-ratio statistics. These corrections are effective in small samples. The algorithm considers the second-order bias of MLEs solution presented in Patriota and Lemonte (2009) and the Skovgaard\'s correction for likelihood-ratio statistics defined in Skovgaard (2001). In addition, a simulation study is developed under different scenarios, where the convergence ratio, relative squared error and the efficiency of bias correction and Skovgaard\'s correction are evaluated.

Water resource planning and management practices in the southeastern United States may be vulnerable to climate change. This vulnerability has not been quantified, and decision makers, although generally concerned, are unable to appreciate the extent of the possible impact of climate change nor formulate and adopt mitigating management strategies. Thus, this dissertation aims to fulfill this need by generating decision worthy data and information using an integrated climate change assessment framework. To begin this work, we develop a new joint variable spatial downscaling technique for statistically downscaling gridded climatic variables to generate high-resolution, gridded datasets for regional watershed modeling and assessment. The approach differs from previous statistical downscaling methods in that multiple climatic variables are downscaled simultaneously and consistently to produce realistic climate projections. In the bias correction step, JVSD uses a differencing process to create stationary joint cumulative frequency statistics of the variables being downscaled. The functional relationship between these statistics and those of the historical observation period is subsequently used to remove GCM bias. The original variables are recovered through summation of bias corrected differenced sequences. In the spatial disaggregation step, JVSD uses a historical analogue approach, with historical analogues identified simultaneously for all atmospheric fields and over all areas of the basin under study. In the second component of the integrated assessment framework, we develop a data-driven, downward hydrological watershed model for transforming the climate variables obtained from the downscaling procedures to hydrological variables. The watershed model includes several water balance elements with nonlinear storage-release functions. The release functions and parameters are data driven and estimated using a recursive identification methodology suitable for multiple, inter-linked modeling components. The model evolves from larger spatial/temporal scales down to smaller spatial/temporal scales with increasing model structure complexity. For ungauged or poorly-gauged watersheds, we developed and applied regionalization hydrologic models based on stepwise regressions to relate the parameters of the hydrological models to observed watershed responses at specific scales. Finally, we present the climate change assessment results for six river basins in the southeastern United States. The historical (baseline) assessment is based on climatic data for the period 1901 through 2009. The future assessment consists of running the assessment models under all IPCC A1B and A2 climate scenarios for the period from 2000 through 2099. The climate assessment includes temperature, precipitation, and potential evapotranspiration; the hydrology assessment includes primary hydrologic variables (i.e., soil moisture, evapotranspiration, and runoff) for each watershed.

In recent years, there has been a surge of research interest in the analysis of time series and spatial data. While on one hand more and more sophisticated models are being developed, on the other hand the resulting theory and estimation process has become more and more involved. This dissertation addresses the development of statistical inference procedures for data exhibiting dependencies of varied form and structure. In the first work, we consider estimation of the mean squared prediction error (MSPE) of the best linear predictor of (possibly) nonlinear functions of finitely many future observations in a stationary time series. We develop a resampling methodology for estimating the MSPE when the unknown parameters in the best linear predictor are estimated. Further, we propose a bias corrected MSPE estimator based on the bootstrap and establish its second order accuracy. Finite sample properties of the method are investigated through a simulation study. The next work considers nonparametric inference on spatial data. In this work the asymptotic distribution of the Discrete Fourier Transformation (DFT) of spatial data under pure and mixed increasing domain spatial asymptotic structures are studied under both deterministic and stochastic spatial sampling designs. The deterministic design is specified by a scaled version of the integer lattice in IRd while the data-sites under the stochastic spatial design are generated by a sequence of independent random vectors, with a possibly nonuniform density. A detailed account of the asymptotic joint distribution of the DFTs of the spatial data is given which, among other things, highlights the effects of the geometry of the sampling region and the spatial sampling density on the limit distribution. Further, it is shown that in both deterministic and stochastic design cases, for "asymptotically distant" frequencies, the DFTs are asymptotically independent, but this property may be destroyed if the frequencies are "asymptotically close". Some important implications of the main results are also given.

Made available in DSpace on 2014-06-12T18:29:15Z (GMT). No. of bitstreams: 2 arquivo4279_1.pdf: 1161561 bytes, checksum: 80aee0b17f88de11dd7d0999ad1594a1 (MD5) license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2008 / Esta tese de doutorado trata da realização de inferências no modelo de regressão linear sob heteroscedasticidade de forma desconhecida. No primeiro capítulo, nós desenvolvemos estimadores intervalares que são robustos à presença de heteroscedasticidade. Esses estimadores são baseados em estimadores consistentes de matrizes de covariâncias propostos na literatura, bem como em esquemas bootstrap. A evidência numérica favorece o estimador intervalar HC4. O Capítulo 2 desenvolve uma seqüência corrigida por viés de estimadores de matrizes de covariâncias sob heteroscedasticidade de forma desconhecida a partir de estimador proposto por Qian eWang (2001). Nós mostramos que o estimador de Qian-Wang pode ser generalizado em uma classe mais ampla de estimadores consistentes para matrizes de covariâncias e que nossos resultados podem ser facilmente estendidos a esta classe de estimadores. Finalmente, no Capítulo 3 nós usamos métodos de integração numérica para calcular as distribuições nulas exatas de diferentes estatísticas de testes quasi-t, sob a suposição de que os erros são normalmente distribuídos. Os resultados favorecem o teste HC4

L’étude de la variabilité du climat est désormais indispensable pour anticiper les conséquences des changements climatiques futurs. Nous disposons pour cela de quantité de données issues de modèles de circulation générale (GCMs). Néanmoins, ces modèles ne permettent qu’une résolution partielle des interactions entre le climat et les activités humaines entre autres parce que ces modèles ont des résolutions spatiales souvent trop faibles. Il existe aujourd’hui toute une variété de modèles répondant à cette problématique et dont l’objectif est de générer des variables climatiques à l’échelle locale àpartir de variables à grande échelle : ce sont les modèles de régionalisation ou encore appelés modèles de réduction d’échelle spatiale ou de downscaling en anglais.Cette thèse a pour objectif d’approfondir les connaissances à propos des modèles de downscaling statistiques (SDMs) parmi lesquels on retrouve plusieurs approches. Le travail s’articule autour de quatre objectifs : (i) comparer des modèles de réduction d’échelle statistiques (et dynamiques), (ii) étudier l’influence des biais des GCMs sur les SDMs au moyen d’une procédure de correction de biais, (iii) développer un modèle de réduction d’échelle qui prenne en compte la non-stationnarité spatiale et temporelle du climat dans un contexte de modélisation dite spatiale et enfin, (iv) établir une définitiondes saisons à partir d’une modélisation des régimes de circulation atmosphérique ou régimes de temps.L’intercomparaison de modèles de downscaling a permis de mettre au point une méthode de sélection de modèles en fonction des besoins de l’utilisateur. L’étude des biais des GCMs révèle une influence indéniable de ces derniers sur les sorties de SDMs et les apports de la correction des biais. Les différentes étapes du développement d’un modèle spatial de réduction d’échelle donnent des résultats très encourageants. La définition des saisons par des régimes de temps se révèle être un outil efficace d’analyse et de modélisation saisonnière.Tous ces travaux de “Climatologie Statistique” ouvrent des perspectives pertinentes, non seulement en termes méthodologiques ou de compréhension de climat à l’échelle locale, mais aussi d’utilisations par les acteurs de la société. / The study of climate variability is vital in order to understand and anticipate the consequences of future climate changes. Large data sets generated by general circulation models (GCMs) are currently available and enable us to conduct studies in that direction. However, these models resolve only partially the interactions between climate and human activities, namely du to their coarse resolution. Nowadays there is a large variety of models coping with this issue and aiming at generating climate variables at local scale from large-scale variables : the downscaling models.The aim of this thesis is to increase the knowledge about statistical downscaling models (SDMs) wherein there is many approaches. The work conducted here pursues four main goals : (i) to discriminate statistical (and dynamical) downscaling models, (ii) to study the influences of GCMs biases on the SDMs through a bias correction scheme, (iii) to develop a statistical downscaling model accounting for climate spatial and temporal non-stationarity in a spatial modelling context and finally, (iv) to define seasons thanks to a weather typing modelling.The intercomparison of downscaling models led to set up a model selection methodology according to the end-users needs. The study of the biases of the GCMs reveals the impacts of those biases on the SDMs simulations and the positive contributions of the bias correction procedure. The different steps of the spatial SDM development bring some interesting and encouraging results. The seasons defined by the weather regimes are relevant for seasonal analyses and modelling.All those works conducted in a “Statistical Climatologie” framework lead to many relevant perspectives, not only in terms of methodology or knowlegde about local-scale climate, but also in terms of use by the society.

Esta tesis se ocupa de los problemas de la estimación econométrica con muestras pequeñas, en los contextos del los VARs monetarios y de la investigación empírica del crecimiento. Primero, demuestra cómo mejorar el análisis con VAR estructural en presencia de muestra pequeña. El primer capítulo adapta la especificación con prior intercambiable (exchangeable prior) al contexto del VAR y obtiene nuevos resultados sobre la transmisión monetaria en nuevos miembros de la Unión Europea. El segundo capítulo propone un prior sobre las tasas de crecimiento iniciales de las variables modeladas. Este prior resulta en la corrección del sesgo clásico de la muestra pequeña en series temporales y reconcilia puntos de vista Bayesiano y clásico sobre la estimación de modelos de series temporales. El tercer capítulo estudia el efecto del error de medición de la renta nacional sobre resultados empíricos de crecimiento económico, y demuestra que los procedimientos econométricos robustos a incertidumbre acerca del modelo son muy sensibles al error de medición en los datos. / This thesis deals with the problems of econometric estimation with small samples, in the contexts of monetary VARs and growth empirics. First, it shows how to improve structural VAR analysis on short datasets. The first chapter adapts the exchangeable prior specification to the VAR context, and obtains new findings about monetary transmission in New Member States. The second chapter proposes a prior on initial growth rates of modeled variables, which tackles the Classical small-sample bias in time series, and reconciles Bayesian and Classical points of view on time series estimation. The third chapter studies the effect of measurement error in income data on growth empirics, and shows that econometric procedures which are robust to model uncertainty are very sensitive to measurement error of the plausible size and properties.

<p>  </p> <p>The management of water resources in East Africa is inherently challenged by rainfall variability and the uneven spatial distribution of freshwater resources. In addition to these issues, meteorological and water data collection has been inconsistent over the past decades, and unclearly defined purposes or end goals for collected data have left many datasets ineffectively curated. In light of the data intensiveness of current modelling and planning methods, data scarcity and inaccessibility have become substantial impediments to informed decision-making. Among the outputs of this research are 1) a revised technique for evaluating bias correction performance on reanalysis data for use in regions where precipitation data is temporally discontinuous which can potentially be applied to other types of climate data as well, 2) a new methodology for quantifying qualitative information contained in legislation and official documents and websites for the assessment of relationships between documented meteorological and water data policies and resulting outcomes in terms of data availability and accessibility, and 3) a fresh look at data needs and the value data holds with respect to water resources decision-making and management in the region.</p>

Wind power presently plays a minor role in Western Canada as compared to hydroelectric power in British Columbia and coal and natural gas thermal power generation in Alberta. However, ongoing reductions in the cost of wind power generation facilities and the increasing costs of conventional power generation, particularly if the cost to the environment is included, suggest that assessment of the present and future wind field in Western Canada is of some importance. To assess present wind power, raw hourly wind speeds and homogenized monthly mean wind speeds from 30 stations in Western Canada were analyzed over the period 1971-2000 (past). The hourly data were adjusted using the homogenized monthly means to attempt to compensate for differences in anemometer height from the standard height of 10m and changes in observing equipment at stations. A regional reanalysis product, the North American Regional Reanalysis (NARR), and simulations conducted with the Canadian Regional Climate Model (CRCM) driven with global reanalysis boundary forcing, were compared to the adjusted station wind-speed time-series and probability distributions. The NARR had a better temporal correlation with the observations, than the CRCM. We posit this is due to the NARR assimilating regional observations, whereas the CRCM did not. The NARR was generally worse than the CRCM in reproducing the observed speed distribution, possibly due to the crude representation of the regional topography in NARR. While the CRCM was run at both standard (45 km) and fine (15 km) resolution, the fine grid spacing does not always provide better results: the character of the surrounding topography appears to be an important factor for determining the level of agreement. Multiple simulations of the CRCM at the 45 km resolution were also driven by two global climate models (GCMs) over the periods 1971-2000 (using only historic emissions) and 2031-2060 (using the A2 emissions scenario). In light of the CRCM biases relative to the observations, these simulations were calibrated using quantile-quantile matching to the adjusted station observations to obtain ensembles of 9 and 25 projected wind speed distributions for the 2031-2060 period (future) at the station locations. Both bias correction and change factor techniques were used for calibration. At most station locations modest increases in mean wind speed were found for most of the projected distributions, but with a large variance. Estimates of wind power density for the projected speed distributions were made using a relationship between wind speed and power from a CRCM simulation for both time periods using the 15km grid. As would be expected from the wind speed results and the proportionality of wind power to the cube of wind speed, wind power at the station locations is more likely than not to increase in the 2031-2060 period from the 1971-2000 period. Relative changes in mean wind speeds at station locations were found to be insensitive to the station observations and choice of calibration technique, suggesting that we estimate relative change at all 45km grid points using all pairs of past/future mean wind speeds from the CRCM simulations. Overall, our results suggest that wind energy resources in Western Canada are reasonably likely to increase at least modestly in the future. / Graduate / 0725 / 0608 / jtdaines@uvic.ca

Submitted by Fabio Sobreira Campos da Costa (fabio.sobreira@ufpe.br) on 2017-04-25T14:46:06Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) TESE VERSÃO FINAL (CD).pdf: 688894 bytes, checksum: fc5c0291423dc50d4989c1c2d8d4af65 (MD5) / Made available in DSpace on 2017-04-25T14:46:06Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) TESE VERSÃO FINAL (CD).pdf: 688894 bytes, checksum: fc5c0291423dc50d4989c1c2d8d4af65 (MD5) Previous issue date: 2016-11-04 / Young e Bakir (1987) propôs a classe de Modelos Lineares Log-Gama Generalizados (MLLGG) para analisar dados de sobrevivência. No nosso trabalho, estendemos a classe de modelos propostapor Young e Bakir (1987) permitindo uma estrutura não linear para os parâmetros de regressão. A nova classe de modelos é denominada como Modelos Não Lineares Log-Gama Generalizados (MNLLGG). Com o objetivo de obter a correção de viés de segunda ordem dos estimadores de máxima verossimilhança (EMV) na classe dos MNLLGG, desenvolvemos uma expressão matricial fechada para o estimador de viés de Cox e Snell (1968). Analisamos, via simulação de Monte Carlo, os desempenhos dos EMV e suas versões corrigidas via Cox e Snell (1968) e através da metodologia bootstrap (Efron, 1979). Propomos também resíduos e técnicas de diagnóstico para os MNLLGG, tais como: alavancagem generalizada, influência local e influência global. Obtivemos, em forma matricial, uma expressão para o fator de correção de Bartlett à estatística da razão de verossimilhanças nesta classe de modelos e desenvolvemos estudos de simulação para avaliar e comparar numericamente o desempenho dos testes da razão de verossimilhanças e suas versões corrigidas em relação ao tamanho e poder em amostras finitas. Além disso, derivamos expressões matriciais para os fatores de correção tipo-Bartlett às estatísticas escore e gradiente. Estudos de simulação foram feitos para avaliar o desempenho dos testes escore, gradiente e suas versões corrigidas no que tange ao tamanho e poder em amostras finitas. / Young e Bakir (1987) proposed the class of generalized log-gamma linear regression models (GLGLM) to analyze survival data. In our work, we extended the class of models proposed by Young e Bakir (1987) considering a nonlinear structure for the regression parameters. The new class of models is called generalized log-gamma nonlinear regression models (GLGNLM). We also propose matrix formula for the second-order bias of the maximum likelihood estimate of the regression parameter vector in the GLGNLM class. We use the results by Cox and Snell (1968) and bootstrap technique [Efron (1979)] to obtain the bias-corrected maximum likelihood estimate. Residuals and diagnostic techniques were proposed for the GLGNLM, such as generalized leverage, local and global influence. An general matrix notation was obtained for the Bartlett correction factor to the likelihood ratio statistic in this class of models. Simulation studies were developed to evaluate and compare numerically the performance of likelihood ratio tests and their corrected versions regarding size and power in finite samples. Furthermore, general matrix expressions were obtained for the Bartlett-type correction factor for the score and gradient statistics. Simulation studies were conducted to evaluate the performance of the score and gradient tests with their corrected versions regarding to the size and power in finite samples.

[ES] La gestión de los recursos hídricos es de vital importancia para la comprensión de las sequias a largo plazo. En la actualidad, se presentan problemas debido a la disponibilidad y manejo del recurso hídrico. Además, el cambio climático afecta de manera negativa las variables climáticas y la disponibilidad del recurso hídrico. El tomar decisiones en base a información confiable y precisa conlleva un arduo trabajo y es necesario contar con diferentes herramientas que permitan llegar a la gestión de los recursos hídricos. La modelización de las variables climáticas es parte fundamental para determinar la disponibilidad del recurso hídrico. Las más importantes son la precipitación y temperatura o precipitación y evapotranspiración. Los modelos estocásticos se encuentran en un proceso de evolución que permiten reducir la escala de análisis. En esta investigación se ha abordado la modelación de variables climáticas con detalle diario. Se ha planteado una metodología para la generación de series sintéticas de precipitación y temperatura mediante modelización estocástica continua multivariada a escala diaria. Esta metodología también incorpora la corrección del sesgo para precipitación y temperatura de los escenarios de cambio climático con detalle diario. Los resultados de la presente tesis indican que los modelos estocásticos multivariados pueden representar las condiciones espaciales y temporales de las diferentes variables climáticas (precipitación y temperatura). Además, se plantea una metodología para la determinación de la evapotranspiración en función de los datos climáticos disponibles. Por otro lado, los modelos estocásticos multivariados permiten la corrección del sesgo con resultados diarios, mensuales y anuales más realistas que otros métodos de corrección de sesgo. Estos modelos climáticos son una herramienta para pronosticar eventos o escenarios futuros que permiten tomar mejores decisiones de manera anticipada. Estos modelos se programaron en el entorno de MatLab con el objetivo de aplicarlos a diferentes zonas de estudio de manera eficiente y automatizada. Los análisis realizados en la presente tesis se realizaron para la cuenca del Júcar con un buen desempeño para las condiciones de la cuenca. / [CA] La gestió dels recursos hídrics és de vital importància per a la comprensió de les sequeres a llarg termini. En l'actualitat, es presenten problemes a causa de la disponibilitat i maneig del recurs hídric. A més, el canvi climàtic afecta de manera negativa les variables climàtiques i la disponibilitat del recurs hídric. El prendre decisions sobre la base informació de confiança i precisa comporta un ardu treball i és necessari comptar amb diferents eines que permeten arribar a la gestió dels recursos hídrics. La modelització de les variables climàtiques és part fonamental per a determinar la disponibilitat del recurs hídric. Les més importants són la precipitació i temperatura o precipitació i evapotranspiració. Els models estocàstics es troben en un procés d'evolució que permet la incorporació de més detalls reduint l'escala d'anàlisi. En aquesta investigació s'ha abordat el modelatge de variables climàtiques amb detall diari. S'ha plantejat una metodologia per a la generació de sèries sintètiques de precipitació i temperatura mitjançant modelització estocàstica contínua multivariada a escala diària. Aquesta metodologia també incorpora la correcció del biaix per a precipitació i temperatura dels escenaris de canvi climàtic amb detall diari. Els resultats de la present tesi indiquen que els models estocàstics multivariats poden representar les condicions espacials i temporals de les diferents variables climàtiques (precipitació i temperatura). A més es planteja una metodologia per a la determinació de l'evapotranspiració en funció de les dades climàtiques disponibles. D'altra banda, els models estocàstics multivariats permeten la correcció del biaix amb resultats diaris, mensuals i anuals més realistes que altres mètodes de correcció de biaix. Aquests models climàtics són una eina per a pronosticar esdeveniments o escenaris futurs que permeten prendre millors decisions de manera anticipada. Aquests models es van programar a l'entorn de Matlab amb l'objectiu d'aplicar-los a diferents zones d'estudi de manera eficient i automatitzada. Les anàlisis realitzades en la present tesi es van realitzar per a la conca del Xúquer amb un bon acompliment per a les condicions de la conca. / [EN] Management of the water resources is important for understanding long-term droughts. Currently, there are problems due to the availability and management of water resources. Furthermore, climate change negatively affecting climate variables and the availability of water resources. Making decisions based on reliable and accurate information involves hard work and it is necessary to have different tools to achieve the management of water resources. The modeling of the climatic variables is a fundamental part to determine the availability of the water resource. The most important are precipitation and temperature or precipitation and evapotranspiration. Stochastic models are in a process of evolution that allows the incorporation of more details by reducing the scale of analysis. In this research, the modeling of climatic variables has been approached in daily detail. A methodology has been proposed for the generation of synthetic series of precipitation and temperature by means of multivariate continuous stochastic modeling on a daily scale. This methodology also incorporates the bias correction for precipitation and temperature of the climate change scenarios with daily detail. The results of this thesis indicate that multivariate stochastic models can represent the spatial and temporal conditions of the different climatic variables (precipitation and temperature). In addition, a methodology is proposed for the determination of evapotranspiration based on the available climatic data. On the other hand, multivariate stochastic models allow bias correction with more realistic daily, monthly and annual results than other bias correction methods. These climate models are a tool to forecast future events or scenarios that allow better decisions to be made in advance. These models were programmed in the MatLab software with the aim of applying them to different study areas in an efficient and automatically. The work in this thesis was carried out for the Júcar basin with a good performance for the conditions of the basin / Hernández Bedolla, J. (2022). Análisis estocástico de datos climáticos como predictor para la gestión anticipada de sequías en recursos hídricos [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/182095 / TESIS