The shallow subsurface controls the partitioning of available energy between sensible and latent heat of the land surface, and the partitioning of available water among evaporation, infiltration, and runoff. It is a key component of both the hydrometeorological system and the terrestrial water cycle. A critical part of any hydrological or hydrometeorological forecast model is therefore the algorithms used to represent the shallow soil processes, which include infiltration, evaporation, runoff, and interflow. For climate models, coupled algorithms called “Land Surface Schemes” (LSSs) are developed to represent the lower boundary conditions that deal with the land-to-atmosphere energy and moisture fluxes. Similar algorithms are implemented in regional watershed models and day-to-day operational water resources forecasting models. It is the primary objective of this thesis to provide improved methods for simulating coupled land surface processes, which can be used as components of LSSs or within existing operational hydrology models. These new methods address a number of specific issues inadequately handled by current models, including the presence of shallow boundary conditions, heterogeneity in infiltration, and infiltration and interflow coupling processes. The main objective of the proposed research is to provide consistent physically-based approach for simulating near surface soil moisture processes, so as to complete the parameterization of the interflow/infiltration algorithm in a Hydrology-Land-Surface scheme MESH. The work mainly focuses on the investigation and development of more physically-based infiltration and interflow algorithms. The hope is to determine appropriate relationships between internal state variables (specifically bulk soil moisture) and system boundary fluxes, while simultaneously reducing the number of nonphysical or unknown model parameters. Fewer parameters lead to reduced calibration requirements for distributed hydrological models and consequently accelerate the transfer of such models to engineering practice. Multiple approaches were taken to provide improved relationships between infiltration and lateral drainage, fluxes and storage. These algorithms were tested by a specialized Richards' equation for sloping soils and Monte Carlo simulations. These tests demonstrated reasonable accuracy and improved representation for the hydrological processes.

The shallow subsurface controls the partitioning of available energy between sensible and latent heat of the land surface, and the partitioning of available water among evaporation, infiltration, and runoff. It is a key component of both the hydrometeorological system and the terrestrial water cycle. A critical part of any hydrological or hydrometeorological forecast model is therefore the algorithms used to represent the shallow soil processes, which include infiltration, evaporation, runoff, and interflow. For climate models, coupled algorithms called “Land Surface Schemes” (LSSs) are developed to represent the lower boundary conditions that deal with the land-to-atmosphere energy and moisture fluxes. Similar algorithms are implemented in regional watershed models and day-to-day operational water resources forecasting models. It is the primary objective of this thesis to provide improved methods for simulating coupled land surface processes, which can be used as components of LSSs or within existing operational hydrology models. These new methods address a number of specific issues inadequately handled by current models, including the presence of shallow boundary conditions, heterogeneity in infiltration, and infiltration and interflow coupling processes. The main objective of the proposed research is to provide consistent physically-based approach for simulating near surface soil moisture processes, so as to complete the parameterization of the interflow/infiltration algorithm in a Hydrology-Land-Surface scheme MESH. The work mainly focuses on the investigation and development of more physically-based infiltration and interflow algorithms. The hope is to determine appropriate relationships between internal state variables (specifically bulk soil moisture) and system boundary fluxes, while simultaneously reducing the number of nonphysical or unknown model parameters. Fewer parameters lead to reduced calibration requirements for distributed hydrological models and consequently accelerate the transfer of such models to engineering practice. Multiple approaches were taken to provide improved relationships between infiltration and lateral drainage, fluxes and storage. These algorithms were tested by a specialized Richards' equation for sloping soils and Monte Carlo simulations. These tests demonstrated reasonable accuracy and improved representation for the hydrological processes.

Submitted by Erika Demachki (erikademachki@gmail.com) on 2014-08-29T15:51:26Z No. of bitstreams: 2 Dissertação_pós_defesa.pdf: 13484145 bytes, checksum: d45c9d8ffd66532d7cea17f178e2fe9c (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Made available in DSpace on 2014-08-29T15:51:26Z (GMT). No. of bitstreams: 2 Dissertação_pós_defesa.pdf: 13484145 bytes, checksum: d45c9d8ffd66532d7cea17f178e2fe9c (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Previous issue date: 2014-04-30 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / This study aimed to develop and calibrate a distributed hydrological model used for a one-dimensional drainage of a peri-urban catchment using the optimization multi-objective method Non -dominated Sorting Genetic Algorithm (NSGA - II) for model calibration. Computational algorithms developed in MATLAB environment were adopted to make this study possible. This model allows a precipitation event to set the surface runoff hydrograph at any position of the watershed (planes or channels) regarding infiltration effect and soil physical characteristics. Objective functions were defined and used simultaneously to calibrate the model. From sensitivity analysis performed, it was found that the model is more affected by the parameters related to permeable areas. The model fit was very good, illustrating the applicability of multi-objective calibration in exploring ideal area and to obtain ideal solutions. Validation proved the efficiency of the model used for other different rainfall events in Samambaia stream basin, generating outputs with good accuracy and optimal theoretical value results for Nash & Sutcliffe coefficients of efficiency near the area region. / O presente trabalho propôs desenvolver e calibrar um modelo hidrológico distribuído unidimensional aplicado a drenagem de uma bacia hidrográfica peri-urbana utilizando o método de otimização multi-objetivo Non-dominated Sorting Genetic Algorithm (NSGA-II) para a calibração do modelo. Para tornar possível a realização deste trabalho foram adotadas rotinas computacionais desenvolvidas em ambiente MATLAB. O modelo desenvolvido permite, para um evento de precipitação, determinar o hidrograma de escoamento superficial em qualquer posição da bacia hidrográfica (planos ou canais) considerando o efeito da infiltração e das características físicas do solo. Foram definidas funções objetivo e aplicadas simultaneamente na calibração do modelo. Da análise de sensibilidade realizada, verificou-se que o modelo é mais impactado pelos parâmetros relativos às áreas permeáveis. O ajuste do modelo foi muito bom, ilustrando a aplicabilidade da calibração multi-objetivo em explorar a região ideal e obter soluções ideais. A validação comprovou a eficiência do modelo, aplicada a outros eventos chuvosos diferentes ocorridos na bacia do córrego Samamabaia, gerando saídas com acurácia satisfatória e resultados para os coeficientes de eficiência Nash & Sutcliffe próximos à região do valor ótimo teórico.