This thesis applies the runoff process in ostrovská Bystřice river catchment in Ore Mountains. For this purpose we used runoff precipitation model HEC-HMS. Our specific goal is to model the soil runoff, which is observed on eleven runoff precipitation events in the period 2009 - 2013. We analyzed basic physical characteristics of soils, which occur in observed river basin. The results were afterwards used to set up parameters of studied methods. We found out, that the moisture from antecedent precipitation influences modeling of soil runoff. The antecedent precipitation conditions the change of basic physical characteristic. We disregarded the influence of evapotranspiration and effect of vegetation on soil runoff. For the parameterization we have chosen two methods, which describe soil runoff. It was SCS CN and Green Ampt methods. These two methods have been compared on sample of resulting events. The methods were manually and automatically calibrating. The results showed on the insignificant difference between both observed methods. No better significant predicative capacity was manifested for one or other methods. Key worlds: soil runoff modeling, SCS CN, Green Ampt, HEC-HMS, infiltration

Increased urbanization causes pervious greenfields to be converted to impervious areas increasing stormwater runoff. Most of the urban floods occur because existing drainage systems are unable to handle peak flows during rainfall events. During a storm event, flood runoff will carry contaminants to receiving waters such as rivers and creeks. Engineers and scientists have combined their knowledge to introduce innovative thinking to manage the quality of urban runoff and harvest stormwater for productive purposes. The introduction of pervious pavements addresses all the principles in Water Sensitive Urban Design. A pervious pavement is a load bearing pavement structure that is permeable to water. The pervious layer sits on the top of a reservoir storage layer. Pervious pavements reduce the flood peak as well as improve the quality of stormwater at source before it is transported to receiving waters or reused productively. To be accepted as a viable solution, understanding of the influence of design parameters on the infiltration rate (both from the bedding and the sub-base) as well as strength of the pavement requires to be established. The design of a particular pavement will need to be customized for different properties of sub layer materials present in different sites. In addition, the designs will have to meet local government stormwater discharge standards. The design of drainage systems underneath pervious pavements will need to be based on the permeability of the whole pervious system. The objectives of the research project are to: • Understand the factors influencing infiltration capacities and percolation rates through the pervious surface as well as the whole pavement structure including the bedding and the sub-base using a laboratory experimental setup. • Obtain relationships between rainfall intensity, infiltration rate and runoff quantity based on the sub-grade material using a computational model to assist the design of pervious pavements. A laboratory scale pavement was constructed to develop relationships between the surface runoff and the infiltration volume from a pervious pavement with an Eco-Pavement surface. 2 to 5mm crushed gravel and 5 to 20mm open graded gravel were chosen as the bedding and sub-base material. Initial tests such as dry and wet density, crushing values, hydraulic conductivity, California Bearing Ratio tests for aggregate material were conducted before designing and constructing the pavement model. A rainfall simulator with evenly spaced 24 sprays was set up above the pervious pavement surface. The thesis presents design aspects of the laboratory scale pavement and the tests carried out in designing the pavement and the experimental procedure. The Green and Ampt model parameters to calculate infiltration were obtained from the laboratory test results from aggregate properties. Runoff results obtained from rainfall simulator tests were compared with the Green and Ampt infiltration model results to demonstrate that the Green and Ampt parameters could be successfully calculated from aggregate properties. The final infiltration rate and the cumulative infiltration volume of water were independent of the rainfall intensity once the surface is saturated. The model parameters were shown to be insensitive to the final infiltration capacity and to the total amount of infiltrated water. The Green and Ampt infiltration parameters are the most important parameters in designing pervious pavements using the PCSWMMPP model. The PCSWMMPP model is a Canadian model built specially for designing pervious pavements. This is independent of the type of sub-grade (sand or clay) determining whether the water is diverted to the urban drainage system (clay sub-grade) or deep percolation into the groundwater system (sand sub-grade). The percolation parameter in Darcy's law is important only if the infiltrated water recharges the groundwater. However, this parameter is also insensitive to the final discharge through the subgrade to the groundwater. The study concludes by presenting the design characteristics influencing runoff from a pervious pavement depending on the rainfall intensity, pavement structure and sub-grade material and a step-by step actions to follow in the design.

Submitted by Marco Antônio de Ramos Chagas (mchagas@ufv.br) on 2017-03-10T14:51:47Z No. of bitstreams: 1 texto completo.pdf: 3086596 bytes, checksum: e5d76ab9a5012dbfd8871cecf53abe9f (MD5) / Made available in DSpace on 2017-03-10T14:51:47Z (GMT). No. of bitstreams: 1 texto completo.pdf: 3086596 bytes, checksum: e5d76ab9a5012dbfd8871cecf53abe9f (MD5) Previous issue date: 2005-08-29 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / O conhecimento do processo de infiltração da água no solo é de extrema importância para diversas áreas da engenharia de conservação de água e solo. O modelo de Green-Ampt-Mein-Larson (GAML) é um dos mais utilizados para a simular a infiltração por ser relativamente simples e por vir apresentando bons resultados na estimativa do processo. Todavia, a difícil determinação dos parâmetros de entrada de GAML e o fato destes parâmetros não representarem efetivamente as condições reais de ocorrência da infiltração ainda dificultam sua aplicação. O presente trabalho teve o objetivo de melhorar a estimativa da infiltração de água no solo, considerando, para tal, duas vertentes distintas e independentes: uma baseada em uma análise físico- matemática mais detalhada do processo de infiltração e do deslocamento da frente de umedecimento pelo perfil do solo; e outra baseada em simples adequações dos parâmetros de entrada de GAML. Desenvolveu-se um novo modelo para a estimativa da infiltração de água no solo (GAML-c), fundamentado no GAML, baseando-se numa descrição mais aproximada da geometria e do deslocamento da frente de umedecimento. O GAML-c realiza uma discretização do teor de água para o qual o solo é umedecido quando da ocorrência do processo de infiltração, simulando o deslocamento simultâneo de diversas sub-frentes de umedecimento pelo perfil do solo. Paralelamente, propôs-se, também, a adequação dos parâmetros de entrada de GAML, a saber: substituir o teor de água do solo na saturação (θ s ) pelo teor de água na zona de transmissão (θ w ); substituir a condutividade hidráulica do solo saturado (K 0 ) pela taxa de infiltração estável (Tie); e calcular o potencial matricial na frente de umedecimento (ψ f ) por meio de uma equação baseada na condutividade hidráulica relativa aos teores de água inicial do solo (θ i ) e da zona de transmissão, e dos parâmetros pressão de borbulhamento (ψ b ) e índice de distribuição do tamanho de poros (λ), oriundos do ajuste da curva de retenção de água no solo feito pelo o modelo de Brooks & Corey. O conjunto de adequações propostas foi denominado GAML-t, não tendo qualquer relação com o procedimento GAML-c. Testes experimentais foram realizados em laboratório a fim de determinar a infiltração em três diferentes classes de solo: Argissolo Vermelho (PV), Latossolo Vermelho (LV) e Latossolo Vermelho-Amarelo (LVA). Avaliou-se o desempenho de GAML-c e de GAML-t, comparando-os com o modelo de GAML aplicado em sua forma original e também adequado segundo cinco diferentes propostas. O GAML-c foi avaliado considerando quatro diferentes cenários, relativos à consideração dos parâmetros de entrada K 0 e θ w : K 0 igual à Tie e θ w experimental (TW); K 0 determinado pelo permeâmetro de carga constante e θ w experimental (KW); K 0 igual à Tie e θ w igual a θ s (TS); e K 0 determinado pelo permeâmetro de carga constante e θ w igual à θ s (KS). Verificou-se que o GAML-c aplicado utilizando-se os parâmetros da simulação TW foi capaz de estimar o perfil de umidade do solo, provendo estimativas de infiltração aceitáveis em todos os tipos de solo estudados; mas ainda necessitando de ajustes. O GAML-t foi capaz de prover boas estimativas da infiltração de água em todos os solos estudados, sendo recomendado para aplicação na estimativa de infiltração. / Knowledge of water infiltration process is extremely important to many areas of Soil and Water Conservation Engineering. Gren-Ampt-Mein-Larson infiltration model (GAML) is one of the models that are most used to predict infiltration process because of its relative simplicity and encouraging results. However, the major obstacle in using GAML model is not just due to the difficulties in estimating it's parameters, but also due to the fact these parameters do not represent field infiltration conditions. This study have had the objective of give better alternatives in predicting infiltration process, considering two different ways: one based on a most detailed physical-mathematical analysis of the infiltration process and of the wetting front movement through the soil profile; and other one based on simple modifications of GAML model parameters. It was developed a new model to simulate infiltration process (GAML-c) that is based on GAML model, and that provides a better description of the wetting front geometry and displacement through the soil profile. GAML-c model simulates the simultaneous displacement of many sub-wetting fronts through the soil profile. It was also proposed the following modifications in GAML model parameters: replace the saturation moisture content parameter (θ s ) with the "filed saturation" moisture content (θ w ); replace the saturated hydraulic conductivity parameter (K 0 ) with the infiltration rate after long time of wetting (Tie); and calculate wetting front suction head (ψ f ) by the use of an equation that makes use of relative hydraulic conductivity in the initial soil moisture (θ i ) and "field saturation" moisture, and also of Brooks & Corey bubbling pressure and pore-size distribution index parameters. These three modifications were called GAML-t and do not have any relation to GAML-c model. Experimental infiltration tests were conduced at laboratory using three different soil types: Red Ultisol (PV), Red Oxisol (LV) and Red-Yellow Oxisol (LVA). The performances of GAML-c model and GAML-t modifications were evaluated and compared to the performances of GAML model applied with its original and modified parameters. Five different ways of modify GAML parameters were used. GAML-c was evaluated using four different scenarios: considering K 0 equals to Tie and the maximum soil moisture equals to θ w (TW); considering K 0 value determined by constant-head permeameter method and the maximum soil moisture equals to θ w (KW); considering K 0 equals to Tie and the maximum soil moisture equals to θ s (TS); and considering K 0 value determined by constant-head permeameter method and the maximum soil moisture equals to θ s (KS). It was verified that GAML-c model applied using TW scenario was able to simulate soil water profile, giving acceptable prediction of infiltration to the three soil types. However, GAML-c still needs better development. GAML-t was able to give very good predictions of infiltration to the three soil types, being strongly suggested to simulate infiltration process. / Tese importada do Alexandria

Made available in DSpace on 2014-07-29T15:01:47Z (GMT). No. of bitstreams: 1 Dissertacao parte1 Thiago Quintiliano de Castro.pdf: 5430014 bytes, checksum: f3e872648dc1cc1c6c038172a0c1138e (MD5) Previous issue date: 2011-09-22 / This study aimed to evaluate the experimental performance of three types of permeable pavements constructed in three different structural conditions in order to control runoff in urbanized areas of the city of Goiania, Goias, adapting to the legal requirements and using local manufactured materials. It was evaluated nine experimental plots of 3.2 m² as the following types: PAV - concrete block "paver", PCP - porous concrete plate and GCC - concrete block "concregrama" and the following structural conditions: I - base of sand and natural subgrade, II - base of sand and compacted subgrade and III - base of sand, gravel subbase and compacted subgrade. Using an artificial rain simulator, 18 tests were performed with two pre-defined rain, a medium intensity (69 mm / h) and a high intensity (180 mm / h). The subgrade soil and building materials were characterized, measured the surface and subsurface runoff, and moisture of the layers of the pavement. The parameters of the models of Horton and Green-Ampt were obtained by adjusting the calculated data infiltration. Combinations PCP-II, PCP-III, II-CCG, CCG-III and PAV-III showed little or no runoff. The delay and persistence in critical times of the runoff hydrograph also secured good results to the PCP, which generally showed higher soaking times (7,2 to 30,4 minutes) compared to the other types of pavement. The structure that showed the best hydrological performance was the III, whose results ranged from 0 to 0,19 for the runoff coefficient. The results showed that porous pavements evaluated in this study contribute to the reduction of runoff, because of low runoff coefficients (0 to 0,36) obtained. Constructive guidelines on the type of permeable pavement that performed better were drawn and described. / O presente trabalho teve como objetivo avaliar o desempenho experimental de três tipos de pavimentos permeáveis construídos em três condições estruturais diferentes com vistas ao controle do escoamento superficial em áreas urbanizadas do município de Goiânia, Goiás, adequando-se às exigências legais e utilizando-se de materiais provenientes de fabricantes locais. Foram avaliadas nove parcelas experimentais de 3,2 m² conforme os seguintes tipos de revestimento: PAV - Bloco de concreto maciço tipo paver , PCP Placa de concreto poroso e CCG Bloco de concreto vazado tipo concregrama , e as seguintes condições estruturais: I - base de areia e subleito natural, II - base de areia e subleito compactado e III - base de areia, sub-base de brita tipo 1 e subleito compactado. Utilizando-se de um simulador de chuva artificial, foram realizados 18 ensaios com duas chuvas pré-definidas, uma de média intensidade (69 mm/h) e outra de alta intensidade (180 mm/h). Foram caracterizados o solo do subleito e os materiais de construção, medidos os escoamentos superficial e subsuperficial, bem como a umidade das camadas dos pavimentos. Os parâmetros dos modelos de Horton e de Green-Ampt foram obtidos por meio do ajuste dos dados calculados de taxa de infiltração. Os pavimentos PCP-II, PCP-III, CCG-II, CCG-III e PAV-III apresentaram pouco ou nenhum escoamento superficial. O retardo e o prolongamento nos tempos críticos do hidrograma de escoamento superficial também garantiram bons resultados ao revestimento PCP, que no geral apresentou os maiores tempos de embebição (7,2 a 30,4 minutos) em comparação aos demais tipos de revestimento. A estrutura que apresentou o melhor desempenho hidrológico foi o Trecho III, cujos resultados variaram de 0 a 0,19 para o coeficiente de escoamento. A análise dos resultados mostrou que os pavimentos permeáveis avaliados neste trabalho contribuem à redução do escoamento superficial, em razão dos baixos coeficientes de escoamento (0 a 0,36) obtidos, assim como ao armazenamento temporário de águas de chuva e ao incremento da infiltração de água no solo urbano. As diretrizes construtivas do tipo de pavimento permeável que obteve melhor desempenho foram traçadas e descritas.

Submitted by Marco Antônio de Ramos Chagas (mchagas@ufv.br) on 2017-03-09T14:21:34Z No. of bitstreams: 1 texto completo.pdf: 768607 bytes, checksum: 53e3ec1372f6119b7d22f36cd39f49bd (MD5) / Made available in DSpace on 2017-03-09T14:21:34Z (GMT). No. of bitstreams: 1 texto completo.pdf: 768607 bytes, checksum: 53e3ec1372f6119b7d22f36cd39f49bd (MD5) Previous issue date: 2005-08-29 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Infiltração é o processo pelo qual a água passa da superfície para o interior do solo, sendo dependente da carga hidráulica na superfície, do teor de água inicial, da textura, da estrutura, da condutividade hidráulica e da porosidade do solo. Diversos modelos se propõem a estimar este processo, dentre os quais se destaca o de Green-Ampt modificado por Mein e Larson (GAML), baseado numa análise física do processo. Embora seja bastante utilizado, este modelo apresenta como desvantagem a dificuldade de obtenção de seus parâmetros de entrada, principalmente do potencial matricial na frente de umedecimento ( ψ m ). Assim sendo, este trabalho teve como objetivos: obter as variáveis necessárias à estimativa do potencial matricial na frente de umedecimento, utilizando-se para tal o modelo de Green-Ampt modificado por Mein e Larson, e estabelecer modelo matemático que relacione o potencial matricial na frente de umedecimento com as características físicas e a condutividade hidráulica do solo saturado, para sua aplicação no modelo de Green-Ampt modificado por Mein e Larson. Foram utilizadas três classes de solos (Latossolo Vermelho - Amarelo – LVA; Latossolo Vermelho – LV; e Argissolo Vermelho – PV). Estes solos foram secos ao ar, destorroados e peneirados em peneira de malha 10 mm, sendo posteriormente acondicionados em colunas de PVC de 200 mm de diâmetro e 800 mm de altura, de modo que a densidade do solo ficasse próxima daquela observada em campo. Na lateral das colunas, foram instaladas horizontalmente sondas de TDR para a determinação do teor de água e para o acompanhamento da frente de umedecimento. Aplicou-se água sobre a superfície das colunas sob taxa constante, por meio de um simulador de chuvas, sendo que o excesso escoado superficialmente foi conduzido para uma caixa de coleta na qual foi instalado um medidor de nível denominado Thalimedes. O volume infiltrado foi determinado por diferença entre o volume precipitado e o escoado. Após a realização dos testes, foram coletadas amostras de solo para determinação da granulometria, densidade do solo, porosidade, macroporosidade e microporosidade. Foram determinados, ainda, os teores volumétricos de água na zona de transmissão e no início do processo e a condutividade hidráulica do solo saturado. O potencial matricial foi calculado por intermédio de um rearranjo no modelo de GAML, considerando a taxa de infiltração estável (Tie) como representativa da condutividade hidráulica do solo saturado e o teor volumétrico de água na zona de transmissão em lugar do teor volumétrico de água correspondente à saturação. De posse dos valores de ψ m e das características do solo, foram ajustados modelos de regressão, sendo a escolha do melhor modelo feita por meio dos coeficientes de determinação (R 2 ), pela significância dos parâmetros no modelo, pelo menor número de variáveis envolvidas e pela facilidade de obtenção dessas variáveis (características do solo). Os valores de condutividade hidráulica do solo saturado, de porcentagem de silte e de macroporosidade foram significativos em todos os modelos nos quais foram utilizados. A Tie não foi significativa em nenhum dos modelos. Todas as equações obtidas utilizando-se a porcentagem de silte em combinação com outras variáveis apresentaram bons resultados na modelagem do ψ m , sendo que, os melhores resultados foram obtidos para o modelo que relaciona o ψ m com a porosidade, a condutividade hidráulica do solo saturado e a porcentagem de silte. / Infiltration is the process by which water passes through the soil surface down to the soil, being dependent on hydraulic load in the surface, on the antecedent moisture, on texture, structure, soil hydraulic conductivity and soil porosity. Several models aim to estimate this process, among which the Green-Ampt modified by Mein and Larson (GAML) model is based on a physical analysis of the process. Although it is frequently used, this model has a limitation for obtaining input parameters, such as the potential head at the wetting front ( ψ m ). This work aimed to obtain the ψ m for GAML model for soils typical of tropical areas, using three classes of soils (Red-Yellow Oxisol, Red Ultisol and Red Oxisol). These soils were air-dried, ground and sieved through 10 mm mesh sieve, placed in PVC columns of 200 mm diameter and 800 mm height, so that the soil density was similar to the field. In the lateral of the columns, horizontal TDR probes were installed for the moisture determination and for the monitoring of the wetting front. The water was applied to the surface of the columns at a constant rate by means of a rain simulator device and the excess drained superficially, was collected and measured in a box using a “Thalimedes” device. The infiltrated volume was calculated by the difference among the precipitate volume and the runoff. After the tests, soil samples were collected for determination of the texture, soil density, porosity, macroporosity, microporosity, volumetric moisture (before and after the tests) and soil hydraulic conductivity. The pressure head at the wetting front was calculated through an adjustment in the GAML model, considering the rate of stable infiltration (Tie) as representative of the hydraulic conductivity, and the transmission zone moisture instead of the moisture saturation. Once having the ψ m values and soil characteristics, regression models were adjusted, choosing the best fit by R 2 , significance of parameters, less number of variables and easy of obtaining soil variables. The soil hydraulic conductivity values, percentage of silt and macroporosity were all significant for all equations. Tie was not significant in none of the equations. All the obtained equations using the amount of silt in combination with another variable showed good results in the ψ m modeling. The best fitted equation was the one which related the ψ m with the porosity, the soil hydraulic conductivity and the percentage of silt. / Dissertação importada do Alexandria

No description available.

Ce travail vise à caractériser et comparer les fonctionnements hydrodynamiques à plusieurs échelles spatiales en zones cristalline et sédimentaire de l’Ouest nigérien et par suite à proposer un modèle simple de fonctionnement hydrodynamique des bassins expérimentaux cristallins qui soit potentiellement transposable aux échelles supérieures.L’analyse qualitative du paysage montre que ces deux domaines ont des états de surface communs et spécifiques.L’étude expérimentale effectuée sur les états de surface communs (surface biologique, BIOL, d’érosion, ERO, structurale, ST-jac, et cultivée, C) montre, à l’échelle ponctuelle, que la surface ERO a la même valeur de conductivité hydraulique K dans les deux contextes. En revanche, les surfaces C et surtout ST ont des conductivités plus faibles en zone cristalline. Le suivi temporel effectué sur ces deux états de surface en zone cristalline a mis en évidence la stationnarité de la conductivité sur ST (jachère de 5 ans) durant la saison des pluies tandis qu'elle varie fortement sur la surface cultivée en fonction du calendrier cultural et du cumul de pluie. K décroit d’une valeur initiale de 170 mm/h après le sarclage à 20 mm/h (soit la conductivité mesurée sur ST) lorsque la surface sarclée reçoit un cumul de pluie de l’ordre de 70 mm. La conductivité atteint même la valeur de 10 mm/h après un cumul de pluie de 180 mm. Cette variation de la conductivité montre l’avantage à court terme du sarclage qui améliore l’infiltration mais qui à long terme tend à dégrader la surface.Les résultats obtenus à l’échelle de la surface élémentaire (10 m²) valident bien les mesures ponctuelles. La surface ERO a le même coefficient du ruissellement (Kr) en zones cristalline et sédimentaire tandis que les surfaces ST et cultivée ont un Kr plus élevé en zone cristalline. L’analyse de l’évolution temporelle du Kr de la surface cultivée du socle a montré que sur 63 événements pluvieux observés entre 2011 et 2013, 22 ont un Kr supérieur à la moyenne (qui est de 0.25) dont près de 2/3 sont observés après plus de 70 mm de pluie qui suivent le sarclage. Les résultats obtenus à ces deux échelles (ponctuelle et surface élémentaire) sont donc cohérents et montrent que la surface cultivée s’encroûte et peut ruisseler plus que la surface ST et autant que la surface ERO. A l’échelle du bassin versant (5 ha), les Kr sont plus élevés sur les bassins cristallins à cause de ces fortes valeurs de Kr des surfaces élémentaires mais aussi parce qu’ils sont composés d’autres surfaces à forte capacité ruisselante que sont la surface d’affleurement du socle altéré et la surface gravillonnaire ayant un Kr de 0.58. A ces trois échelles (ponctuelle, élémentaire, petit bassin), on note la non-dépendance du fonctionnement hydrodynamique à l'état hydrique initial.L’exploitation des résultats ponctuels obtenus sur le site cristallin (conductivité moyenne de BIOL, ERO, et ST et conductivité variable sur la surface cultivée) dans le modèle de Green et Ampt a permis de caler le potentiel de front par état de surface et décrire de manière très satisfaisante les ruissellements mesurés sur les surfaces élémentaires. Sur la base des ruissellements ainsi calculés, nous avons simulé les hydrogrammes à l’exutoire des bassins expérimentaux en assimilant le fonctionnement de ces derniers à celui des surfaces élémentaires en spatialisant leur infiltrabilité. En faisant l’hypothèse d’une ré-infiltration nulle sur les versants, nous avons appliqué une fonction de transfert simple prenant en compte la distance de chaque surface élémentaire par rapport au réseau hydrographique, une vitesse d'écoulement constante et une pluie imbibante de 3 à 4 mm devant saturer la couverture sableuse de ravine, ce qui est beaucoup moins qu'en zone sédimentaire. Finalement, les hydrogrammes simulés reproduisent assez bien les caractéristiques des hydrogrammes mesurés, ce qui offre une perspective d’application de certains principes du modèle sur de plus grands bassins. / This work aims at characterizing and comparing the hydrodynamical functioning at several spatial scales within the granitic-basement and sedimentary zones of Western Niger. Then, a simple hydrological model that could be suitable for use at larger scales is proposed and tested.Qualitatively, the two geological domains have common and specific surface features.The experimental work carried out onto common surface features (biological crust, BIOL; erosion crust, ERO; fallow structural surface, ST and cultivated, C) shows that, at the point scale, ERO has the same hydraulic conductivity K value in both contexts. On the other hand, surfaces features C and especially ST have lower K values in granitic context.Monitoring of the ST and C sites along the rainy season proved the stationarity of the ST conductivity value. On the contrary, K varies widely with the amount of rain received from an initial value of 170 mm/h after weeding down to 20 mm/h (i.e. the ST measured value) after 70 mm of rain and even 10 mm/h after 180 mm of rain. This variation shows the short-term benefit of weeding onto infiltration but a degradation of the soil surface on the long term.At the plot scale (10 m2), runoff measurements are consistent with point measurements. ERO has the same runoff coefficient (Kr) in granitic and sedimentary zones while ST and C surfaces have a higher Kr in granitic context.Runoff monitoring of the granitic site cultivated plots showed that from a total of 63 rain events between 2011 and 2013, 22 had a Kr value higher than the average value (0.25) from which 2/3 are observed after the surface had received more than 70 mm rain after weeding.Results obtained at the two scales (point and 10-m2 plot) are thus consistent and show that the cultivated surface gets crusted and may produce runoff more than fallow ST sites and as much as ERO features.At the basin scale (5 ha), Kr values are higher in the granitic site, not only because of the higher Kr value for a given surface feature but also because of the specific low-infiltrating surfaces which are granite outcrops and gravel crusts (Kr = 0.58).At the three previous scales (point, plot and small basin), runoff volume was found independent of soil initial moisture.Using the previous point-scale results in a Green-Ampt infiltration model led to calibrate the wetting front pressure head for each surface feature and to satisfactorily describe runoff volumes obtained at the plot scale.By estimating runoff with the Green-Ampt infiltration model at any given point, basin-scale hydrograms were obtained by adding the contribution of all elementary surfaces. Assuming no re-infiltration of runoff water within the basins, a simple transfer function was chosen accounting for the distance of each surface to the hydrological network, a constant water velocity of 0.05 m.s-1 and a volume of 3-4 mm of water necessary to fill the kori sand cover, which is much less than that in the sedimentary context. Finally, simulated hydrograms reproduce nicely the measured ones, which offers the perspective of applying some principles of the model to larger basins.

Nous montrons que deux processus habituellement négligés dans les modèles simples de bilan hydrique journalier des cultures, ont pourtant un effet non négligeable sur le bilan hydrique, et en particulier diminuent le drainage vers les nappes phréatiques: la captation de l'eau de pluie par le couvert végétal, et le ruissellement hortonien. Sur une culture de maïs, nous calculons ainsi en valeur moyenne annuelle une perte nette pour le sol de 26.7 7.8 mm et 23.7 14.4 mm respectivement pour la captation et le ruissellement. En vue d'études pluriannuelles, nous développons des modèles simples et à base physique de captation et d'infiltration/ruissellement, que nous couplons au modèle de bilan hydrique agropédoclimatique journalier d'un sol multicouches, BILHYNA (TUZET et al., 1992). La captation et le ruissellement sont calculés en continu selon un pas de temps de 0.01 h et abaissent la quantité de pluie journalière incorporée au sol. Une image réaliste des pluies est calculée à partir de mesures instantanées, ou par un modèle original restituant une courbe d'intensité de forme gaussienne à partir des mesures horaires classiques et d'un paramètre climatique. Les modèles sont testés face aux résultats de 3 années continues de mesures sur une parcelle expérimentale de Thiverval-Grignon, alternant culture de maïs et labour d'hiver. Nous calculons la captation avec le modèle simple de MERRIAM (1960). En supposant que le couvert ne transpire pas lorsqu'il est mouillé, nous multiplions la transpiration journalière par la fraction de la journée pendant laquelle le couvert est sec. L'infiltration de l'eau de pluie est décrite par le modèle de GREEN-AMPT (1911), appliqué à un profil de sol hétérogène en humidité initiale (BOUWER, 1969), en succion effective au front d'humectation (YOUNGS, 1974) et en conductivité hydraulique à saturation (selon HILLEL et GARDNER, 1970). Nous fixons les paramètres d'infiltration à partir de relations de VAN GENUCHTEN (COQUET et al., 2002), à -10 cm de potentiel hydrique pour la saturation partielle du sol lors de l'infiltration (BRAKENSIEK & RAWLS, 1983; FOX et al., 1998). Trois couches sont distinguées: lit de semence, labour, et sol non travaillé. Nous traduisons les processus dynamiques essentiels à l'aide de relations empiriques simples et de mesures: développement et sénescence du couvert (LAI), évolution de la rugosité et de la capacité de flaquage de la surface du sol en fonction de l'énergie cinétique des pluies (ONSTAD et al., 1984, ONSTAD, 1984 / KAMPHORST, 2000), évolution de la résistance hydraulique de la croûte de battance (BRAKENSIEK & RAWLS, 1983), modification des paramètres d'infiltration par le travail du sol. Le test du modèle sur la culture du maïs 1999 montre de très bons résultats face aux ruissellement mesuré sur placettes de 1m², sauf en fin de saison où la négligence de la fissuration de la surface du sol amène le modèle à surestimer le ruissellement (37.1% d'erreur, soit 9.8mm, sur toute la période d'étude). Face aux mesures d'humidité du sol (TDR, gravimétrie), le modèle BILHYNA intégrant captation et ruissellement montre une erreur ponctuelle de 20mm au plus du stock d'eau du sol sur 1.10m de sol, traduisant un besoin d'amélioration du drainage à proximité de la capacité au champ, et d'ajustement des paramètres de transpiration du couvert. En moyenne annuelle sur les trois années, sous une pluie de 773.7 59.3 mm, la captation et le ruissellement constituent ensemble une perte nette pour le sol de 65.1 11.9 mm. Leur effet est de 20mm maximum sur le stock d'eau du sol, de l'ordre de l'incertitude sur la mesure de l'humidité. Captation et ruissellement abaissent néanmoins de manière nette la transpiration du couvert et le drainage profond, respectivement de 28.5 3.1 mm et de 45.4 13.9 mm. Ce dernier résultat montre l'importance potentielle d'intégrer la captation et le ruissellement pour améliorer le bilan hydrique du sol et des nappes phréatiques.

A two-dimensional numerical model (RiverFLO-2D) has been enhanced to simulate flooding of urban areas by developing an innovative wet and dry surface algorithm, accounting for variable rainfall, and recoding the model computer program for parallel computing. The model formulation is based on the shallow water equations solved with an explicit time-stepping element-by-element finite element method. The dry-wet surface algorithm is based on a local approximation of the continuity and momentum equations for elements that are completely dry. This algorithm achieves global volume conservation in the finite element, even for flows over complex topographic surfaces. A new module was implemented to account for variable rainfall in space and time using NEXRAD precipitation estimates. The resulting computer code was parallelized using OpenMP Application Program Interface, which allows the model to run up to 5 times faster on multiple core computers. The model was verified with analytical solutions and validated with laboratory and field data. Model application to the Malpasset dam break and Sumacarcel flooding event show that the model accurately predicts flood wave travel times and water depths for these numerically demanding real cases. To illustrate the predictive capability of the enhanced model, an application was made of the city of Sweetwater flooding in Miami-Dade County, FL caused by the Hurricane Irene. The simulation starts with dry bed and rainfall is provided by NEXRAD estimates. Integrating NEXRAD rainfall estimates, developing a novel dry-wet area algorithm and parallelizing RiverFLO-2D code, this dissertation presents a proof of concept to accurately and efficiently predict floods in urban areas, identifying future improvements along this line of research.

The overall purpose of this study was to define the water balance of feedlot pens in a Saskatchewan cattle feeding operation for a one year period. Although the initial intention of the study was focused upon an active feedlot, cattle were removed from the pens in July 2003. Therefore, the year of analysis was conducted on the manured surface of an inactive feedlot. The water balance was also performed on a scraped soil surface, since manure is removed from the pens and spread on agricultural land, leaving the pen surfaces bare for a short period of time each year. <p>During the monitoring period (Sept. 2003 to Aug. 2004), 313 mm of precipitation was received at the feedlot, but only 84 mm of that total was received before June 2004. Winter precipitation was very low (33 mm) and there was no observed runoff from it. Runoff collection weirs in operation for only part of the summer recorded no runoff. The Green-Ampt and USDA SCS runoff models, as well as a snowmelt runoff equation, were used to predict runoff from both the manure pack, as well as the scraped soil surface. Using manure and soil hydraulic parameters determined in the laboratory (from falling head permeameter measurements) and the field (from rainfall simulations), as well as incorporating the greatest 24 hour rainfall amounts and 30 minute intensities experienced at the feedlot, the USDA model found that 29 mm of runoff would occur from the scraped soil surface. Additionally, snowmelt runoff was estimated to be 19 mm for the winter precipitation received. Drainage beneath the 0.6 m soil depth was negligible and the top 0.6 m of soil experienced an increase in moisture of 54 mm. Finally, 211 mm was lost as evaporation. For the manure pack, no runoff was predicted using the Green-Ampt and USDA SCS models and snowmelt runoff equation, which corresponded well to the lack of runoff measured both from the weir and rainfall simulations. Drainage beneath 0.6 m soil depth was negligible. Of the 313 mm of precipitation that fell during the study year, 42 mm was stored within the manure pack and the rest was lost as evaporation (271 mm).