Over the past decades, climate change and urbanization have altered the regional hydro-environments, causing a series of stormwater management problems including urban flood and non-point pollution. Low impact development (LID) has been identified as a sustainable strategy for stormwater management. However, given the complex impacts of climate change and urbanization on hydro-environments, the performance of LID strategy under future changes remains largely unexplored. Accordingly, this research characterized the LID performance under changes in climate and urbanization. To provide an additional reference to sustainable stormwater management, the following specific topics were addressed: (1) Through hydraulic and water quality modeling, the LID performances of flood mitigation and pollution removal were systematically evaluated at the city scale. (2) Through uncertainty analysis, the impact of model parameter uncertainty on the LID performance was taken into account. (3) Through sensitivity analysis, the impact of LID technical parameters (e.g., surface features, soil textures) on the LID performance was quantified. (4) Through scenario analysis, the LID performances under different hydrological patterns were compared. (5) Through spatial analysis, the distribution of LID performance on different land-cover types was determined. (6) Through adopting general circulation model (GCM) projections, the LID performance under future climate scenarios with different representative concentration pathways (RCPs) was investigated. (7) Through adopting integrated assessment model (IAM) projections, the LID performance under future urbanization scenarios with different shared socioeconomic pathways (SSPs) was explored. (8) By coupling climate and urbanization projections with land-cover distribution, the spatiotemporal trends of LID performance in the future were characterized.:Table of Contents List of Abbreviations VII List of Peer-Reviewed Publications on the Ph.D. Topic IX List of Co-authored Peer-Reviewed Publications on the Ph.D. Topic X 1 General Introduction 1 1.1 Background 1 1.2 Objectives 3 1.3 Innovation and Contribution to the Knowledge 3 1.4 Outline of the Dissertation 4 1.5 References 5 2 Literature Review 9 2.1 Hydraulic and Water Quality Modeling 9 2.1.1 Hydraulic Model 9 2.1.2 Water Quality Model 10 2.2 Low Impact Development (LID) 10 2.2.1 LID Practice 10 2.2.2 LID Performance 11 2.3 Performance Evaluation 13 2.3.1 Scenario Analysis 13 2.3.2 Spatial Analysis 13 2.3.3 Uncertainty Analysis 14 2.3.4 Sensitivity Analysis 14 2.4 Future Changes in Climate and Urbanization 15 2.4.1 Climate Change 15 2.4.2 Future Urbanization 16 2.5 References 17 3 Impact of Technical Factors on LID Performance 27 3.1 Introduction 28 3.2 Methods 30 3.2.1 Study Area 30 3.2.2 Model Description 31 3.2.2.1 Model Theory 31 3.2.2.2 Model Construction 31 3.2.2.3 Model Calibration and Validation 32 3.2.2.4 Model Uncertainty Analysis by GLUE Method 34 3.2.3 Hydrological Pattern Design 35 3.2.4 LID Strategy Design 35 3.2.4.1 Implementation of LID Practices 35 3.2.4.2 Sensitivity Analysis by Sobol’s Method 36 3.2.5 Correlation Analysis Using a Self-Organizing Map 37 3.2.6 Description of the RDS Load Components 37 3.3 Results 38 3.3.1 RDS Migration and Distribution in Baseline Strategy 38 3.3.1.1 RDS Migration under Hydrological Scenarios 38 3.3.1.2 RDS Distribution on Land-Cover Types 39 3.3.2 RDS Removal in LID Strategies 40 3.3.2.1 RDS Removal by LID Strategies 40 3.3.2.2 Spatial Distribution of the RDS Removal 42 3.3.2.3 LID Parameter Sensitivity Analysis Result 43 3.4 Discussion 45 3.4.1 Factors Influencing RDS Migration in the Baseline Strategy 45 3.4.2 RDS Removal Performance by LID Strategy 46 3.5 Conclusions 47 3.6 References 47 4 Impact of Hydro-Environmental Factors on LID Performance 53 4.1 Introduction 54 4.2 Methods 56 4.2.1 Study Area 56 4.2.2 Modeling Approach 56 4.2.2.1 Model Theory 56 4.2.2.2 Model Construction 56 4.2.2.3 Model Calibration and Validation 57 4.2.2.4 Model Uncertainty Analysis 57 4.2.3 LID Performance Analysis 58 4.2.3.1 LID Practice Implementation 58 4.2.3.2 LID Performance Evaluation 58 4.2.4 Hydrological Pattern Analysis 59 4.2.4.1 Scenario of ADP Length 59 4.2.4.2 Scenario of Rainfall Magnitude 59 4.2.4.3 Scenario of Long-Term pre-Simulation 60 4.2.5 Sensitivity Analysis of Hydrological Scenarios 60 4.3 Results 61 4.3.1 LID Performance under Different ADP Lengths 61 4.3.2 LID Performance under Different Rainfall Magnitudes 62 4.3.3 Spatial Distribution of LID Performance 63 4.3.4 Sensitivities of LID Performance to ADP Length and Rainfall Magnitude 66 4.4 Discussion 68 4.4.1 Impact of ADP Length and Rainfall Magnitude on LID Performance 68 4.4.2 Spatial Heterogeneity of LID Performance 68 4.4.3 Research Implications 69 4.5 Conclusions 70 4.6 References 71 5 Impact of Future Climate Patterns on LID Performance 77 5.1 Introduction 78 5.2 Methods 80 5.2.1 Study Area 80 5.2.2 Hydraulic and Water Quality Model 80 5.2.2.1 Model Development 80 5.2.2.2 Model Calibration and Validation 81 5.2.3 Climate Change Scenario Analysis 81 5.2.3.1 GCM Evaluation 81 5.2.3.2 Greenhouse Gas (GHG) Concentration Scenario 82 5.2.3.3 GCM Downscaling 83 5.2.4 LID Performance Analysis 83 5.2.4.1 Implementation of LID Practices 83 5.2.4.2 Evaluation of LID Performance 84 5.2.4.3 Sensitivity Analysis on LID Performance 86 5.3 Results 86 5.3.1 Hydrological Characteristics under Future Climate Scenarios 86 5.3.2 LID Performance under Future Climate Scenarios 87 5.3.2.1 LID Short-Term Performance 87 5.3.2.2 LID Long-Term Performance 90 5.3.3 Impact of ADP Length and Rainfall Magnitude on LID Performance 92 5.3.3.1 LID Performance Uncertainty 92 5.3.3.2 Spatial Distribution of LID Performance 93 5.3.3.3 Sensitivity of LID Performance to Climate Change 95 5.4 Discussion 97 5.4.1 LID Performance in Short-Term Extremes and Long-Term Events 97 5.4.2 Impact of Climate Change on LID Performance 97 5.4.3 Research Implications 99 5.5 Conclusions 100 5.6 References 100 6 Impact of Climate and Urbanization Changes on LID Perfor-mance 109 6.1 Introduction 110 6.2 Methods 112 6.2.1 Study Area 112 6.2.2 Modeling Approach 112 6.2.2.1 Model Development 112 6.2.2.2 Model Calibration and Validation 113 6.2.3 Future Scenario Derivation 113 6.2.3.1 Climate Change Scenario 113 6.2.3.2 Urbanization Scenario 115 6.2.4 Flood Exposure Assessment 115 6.2.5 Implementation and Evaluation of LID Strategy 117 6.2.5.1 Implementation Scheme of LID Strategy 117 6.2.5.2 Performance Evaluation of LID Strategy 117 6.3 Results 118 6.3.1 Flood Exposure in Baseline and Future Scenarios 118 6.3.1.1 Hydrological Change in Future Climate Scenarios 118 6.3.1.2 Catchment Change in Future Urbanization Scenarios 118 6.3.1.3 Population and GDP Exposures to Flood in Future 121 6.3.2 Flood Exposure with Consideration of LID Strategy 123 6.3.2.1 Runoff Mitigation Performance of LID Strategy 123 6.3.2.2 Peak Mitigation Performance of LID Strategy 124 6.3.2.3 Population and GDP Exposures to Flood under LID Strategy 125 6.4 Discussion 126 6.4.1 Climate Change and Urbanization Exacerbated Flood Exposure Risk 126 6.4.2 LID Strategy Mitigated Flood Exposure Risk 126 6.5 Conclusions 127 6.6 References 127 7 Discussion and General Conclusions 133 7.1 Stormwater Management Performance of LID Strategy 133 7.2 Impact of Influencing Factors on LID Performance 134 7.3 LID Performance under Future Changes 135 7.4 Research Implications 136 7.5 References 137 8 Outlook of Future Research 139 8.1 Optimization of LID Performance 139 8.2 Cross-regional Study on Future Changes 139 8.3 Macro-scale Flood Risk Management 140 8.4 References 141 9 Appendices 143 9.1 Appendix for Chapter 3 143 9.1.1 The Determination of the GLUE Criteria 143 9.1.2 Model Uncertainty Analysis 143 9.1.3 The LID Installation Location 144 9.1.4 Figures 145 9.1.5 Tables 147 9.2 Appendix for Chapter 4 153 9.2.1 Scenario of Long-term pre-Simulation 153 9.2.2 Figures 153 9.2.3 Tables 158 9.3 Appendix for Chapter 5 164 9.3.1 Tables 164 9.4 Appendix for Chapter 6 169 9.4.1 Figures 169 9.4.2 Tables 170 9.5 Data Source 177 9.6 References 178

Översvämningar till följd av intensiva nederbördstillfällen har de senaste åren ökat i antal och omfattning. Dessa händelser förväntas bli vanligare i framtiden och skapa fler översvämningar. Med anledning av detta är det intressant att undersöka hur översvämningar i framtiden breder ut sig och vilka vattennivåer som bildas med förväntad nederbörd. Att modellera översvämningar kräver data som i vissa fall kan vara både tidskrävande och omständig att införskaffa. Möjliga avgräsningar och antaganden i modellparametrar kan då vara intressanta att göra som fortfarande ger användbara resultat. En skyfallskartering har genomförts med 2D-hydraulisk modellering i Kumla med programvaran MIKE 21 Flow Model FM. De översvämningskartor som skapades användes för att identifiera områden i Kumla som riskerar att drabbas av höga vattennivåer till följd av skyfall motsvarande 100- och 200-årsregn. En stor osäkerhet vid modellering av översvämningar är att validera resultaten som fås fram. Ofta saknas information om tidigare översvämningar. De nederbördstillfällen som används är ofta så stora att det saknas data om liknande händelser tidigare. Vid översvämningsmodellering anväds data som beskriver olika typer av modellparametrar. Dessa kommer med ytterligare osäkerheter som kan göra valideringen problematisk. För att undersöka hur stor effekt olika modellparametrar har på resultatet genomfördes en känslighetsanalys där differenskartor skapades mellan undersökta scenarion och referenskartor. Skyfallskarteringen visade att stora delar i Kumla drabbas av översvämningar för både ett 100- och 200-årsregn. Området Kumlaby identifierades som känsligt och får höga vattennivåer. Detta beror mest troligt på omgivningens topografi och att Kumlaby underlagras av leror med låg infiltrationskapacitet. I känslighetsanalysen identifierades markens råhet och infiltrationskapacitet vara styrande parametrar för översvämningens utbredning och vattennivåer. Dessa påverkar främst hur höga vattenflöden som uppstår och översvämningens utbredningen och vattennivåer. Kunskap om dessa parametrar är viktigt för att undvika över- eller underskattning av en översvämning. Användningen av avrinningskoefficienter istället för markens råhet, infiltrationskapacitet och evaporation undersöktes. Differensen i översvämningens utbredning och vattennivåer blev stor i och utanför Kumla tätort. På mindre områden kan det vara mer lämpligt att använda en avrinningskoefficient när en mer detaljerad klassning kan göras av de markytor som finns. Ett scenario som undersöktes i känslighetsanalysen var installation av gröna tak på alla byggnader i Kumla. Simuleringarna som genomfördes visade att både utbredningen och vattennivåer minskade. Detta till följd av större lagringskapacitet och motstånd mot vattenflöden som kommer med gröna tak. / Urban floods caused by intense rainfall have occurred more frequently the last couple of years. These rainfall events are expected to become more common in the future and create more floods in urban areas. This makes it important to investigate the extent and water levels from urban floods in the future. In order to simulate floods, different types of data is needed. This data can be both time consuming and difficult to obtain. With this in mind, it is interesting to investigate possible simplifications and assumptions of model parameters. A cloud burst mapping was made with 2D hydraulic modelling in Kumla with the software MIKE 21 Flow Model FM. The flood maps created were used to identify areas in Kumla which have a higher risk of being subject to high water levels. One uncertainty while modelling urban floods is the process of validating the results. There is often a lack of data for the used rainfall events or information from previous floods in the area. In flood modelling data is used which describes different model parameters, these comes with additional uncertainties and can make the validation more difficult. A sensitivity analysis was made to be able to examine effects on the results from variations in model parameters. The cloud burst mapping showed that large parts of Kumla will be affected by water levels which goes up to 1 m. The area Kumlaby was identified as being sensitive for high water levels. This is due to placement of Kumlaby below higher ground which causes water to flow toward Kumlaby. The ground below is mostly made up of clay which has low infiltration capacity. In the sensitivity analysis the bed resistance and infiltration capacity were identified as governing parameters regarding the extent and water levels of urban floods. In order to avoid over- or underestimation of floods it is important to have knowledge about these parameters in the model area. The use of a runoff coefficient instead of bed resistance, infiltration and evaporation were examined. The difference of the resulting flood were large in the whole model area. In smaller areas a runoff coefficient could be used with better results when a more detailed description can be made of the surfaces in the area. A scenario where green roofs were assumed to have been installed on all buildings in Kumla were examined. The simulations showed that both the extent and water levels decreased. This due to the fact that green roofs have a capacity to store water and delay flows of water.

Översvämningar till följd av skyfall har blivit allt vanligare och förväntas att öka i takt med klimatförändringarna. Översvämningar kan ställa till stora skador för ett samhälle, framförallt när de samhällsviktiga verksamheterna blir drabbade. För att undvika att detta sker bör samhället vara byggt för att tåla stora volymer vatten som faller vid ett skyfall. Ett steg för att nå dit är att göra en skyfallskartering där flödesvägar, vattenvolymer och översvämningens utbredning tas fram genom hydraulisk modellering. Utifrån skyfallskarteringen kan sårbara områden upptäckas och förebyggande åtgärder kan utföras för att minska översvämningens negativa påverkan. Akademiska sjukhuset i Uppsala är en samhällsviktig verksamhet och har tidigare haft problem med översvämningar. I den här studien har en skyfallskartering utförts på Akademiska sjukhusets område för att ta reda på översvämningens utbredning vid ett skyfall och vilka åtgärder som lämpar sig för att förhindra översvämningar. Skyfallskarteringen utfördes i det tvådimensionella (2D) hydrauliska modelleringsprogrammet MIKE 21 Flow Model. Eftersom en skyfallskartering baseras på många generaliseringar finns det vissa osäkerheter kring valet av parametrar. Därför har även en känslighetsanalys utförts kring valet av regntyp (Chicago Design Storm (CDS) jämfört med ett blockregn), regnets varaktighet, grönytornas avrinningskoefficient och markens infiltrationshastighet. Indata till modellen baserades på olika kartdata som bearbetades i GIS-programmet ArcMap. Flera olika regn med varierande återkomsttid simulerades. Resultaten visade att det blir översvämning inne på sjukhusområdet vid ett 100-årsregn som förvärras när återkomsttiden ökar. Åtgärdsanalysen utfördes genom att lägga in förändringar i höjdmodellen för att se hur det påverkar översvämningens utbredning. Analysen visade att åtgärder som jordvallar och höjdsättning av marken kan tillämpas på området för att minska översvämningsrisken. Resultatet från känslighetsanalysen visade att ett CDS-regn ger större översvämningskonsekvenser i modelleringen än om ett blockregn av samma återkomsttid och varaktighet används. Känslighetsanalysen av varaktigheterna visade att en lång varaktighet kan leda till låga flödestoppar som inte representerar ett skyfall väl. En avrinningskoefficient på 0,4 beskriver infiltrationen i området väl och när en större avrinningskoefficient används tenderar översvämningen att bli större på grönytorna. Till sist visade resultatet att infiltrationshastigheten är en känslig parameter som bör väljas efter mer noggrann analys av marken i modelleringsområdet. / Flooding as a cause of cloudbursts have become more common and is expected to increase with climate change. Floods can cause substantial damage to a society, especially when the critical societal functions are affected. To avoid this the city should be built to tolerate large volumes of water from cloudbursts. As a step on the way to accomplish this, a cloudburst mapping could be made where flow paths, water volumes and the extent of the flooding are studied through hydraulic modelling. Through the cloudburst mapping, vulnerable areas can be spotted, and flood prevention measures can be taken to lessen the extent of the floods negative impact. Uppsala University Hospital serves a critical societal function and has previously had problems with flooding. In this project a cloudburst mapping has been made in the two dimensinoal (2D) hydraulic modelling program, MIKE 21. This was done to find out the extent of a flood caused by a cloudburst event and what measures that can be taken to prevent floods. A cloudburst mapping is based off many generalized assumptions and there are some uncertainties when selecting the parameters. Because of this, a sensitivity analysis was performed on the selection of rain-type (Chicago Design Storm (CDS) vs. block-rain), rain duration, the runoff coefficient and the soil's infiltration capacity. The inputs of the model were based off different geographic data and then constructed in the GIS-program ArcMap. Several different rain events with varying duration and return periods were simulated. The results showed that there is considerable flooding in the area after a rain with a 100-year return period and it gets worse when the return period increases. The flood prevention analysis was made by editing the terrain to mimic flood prevention measures and study how the extent of the flood responds to the edits. The analysis showed that measures like soil barriers and changes in elevation were effective in lessening the risk of flooding. The results from the sensitivity analysis showed that a CDS-rain causes a more significant flooding compared to a block-rain of the same return period and duration. The sensitivity analysis of the rain duration proved that a long duration can lead to flat flow curves that doesn't resemble a flow curve from a cloudburst event. A runoff coefficient of 0.4 describes the infiltration in the area well and with a larger coefficient the flooding on greenery tend to grow. Lastly, the infiltration capacity proved to be a sensitive parameter that needs to be selected carefully, preferably after a thorough soil analysis.

Flood forecasting maps are essential for rapid disaster response and risk management, yet the computational complexity of physically-based simulations hinders their application for efficient high-resolution spatial flood forecasting. To address the problems of high computational cost and long prediction time, this thesis proposes to develop deep learning neural networks based on a flood simulation dataset, and explore their potential use for flood prediction without learning hydrological modelling knowledge from scratch.  A Fully Convolutional Network (FCN), FCN with multiple outputs (Multioutput FCN), UNet, Graph-based model and their Recurrent Neural Network (RNN) variants are trained on a catchment area with twelve rainfall events, and evaluated on two cases of a specific rainfall event both quantitatively and qualitatively. Among them, Convolution-based models (FCN, Multioutput FCN and UNet) are commonly used to solve problems related to spatial data but do not encode the position and orientation of objects, and Graph-based models can capture the structure of the problem but require higher time and space complexity. RNN-based models are effective for modelling time-series data, however, the computation is slow due to its recurrent nature. The results show that Multioutput FCN and the Graph-based model have significant advantages in predicting deep water depths (>50 cm), and the application of recurrent training greatly improves the long-term flood prediction accuracy of the base deep learning models. In addition, the proposed recurrent training FCN model performs the best and can provide flood predictions with high accuracy.

La crue de 1910 de la Seine a eu une incidence directe sur le fonctionnement des différents réseaux (réseau électrique, assainissement des eaux usées, transport, eau potable). Le réseau RATP a été particulièrement atteint dans son fonctionnement. Les dommages qu’une crue centennale pourrait engendrer aujourd’hui risquent d’être plus importants encore car le réseau actuel est plus vulnérable du fait des nombreux équipements électriques et informatiques qu’il comporte. La majorité des émergences (les entrées d’eau) de la RATP est située en zone inondable. Lors d’une crue majeure de la Seine, les écoulements dus aux inondations se propagent directement dans la partie souterraine et centrale du réseau (Métro et RER) par le biais de ces émergences. Cette thèse s'intéresse à la simulation hydrodynamique des écoulements dans le réseau RATP en utilisant le logiciel MIKE URBAN dédié à la modélisation des réseaux d’assainissement. Cette modélisation nécessite une bonne connaissance de l’origine des écoulements pour mieux les prendre en compte. En effet, le réseau RATP est inondé par les eaux superficielles et les eaux d’infiltration. Afin de mieux quantifier les volumes entrants dans le réseau, un modèle physique d’une bouche de métro type a été réalisé. Les résultats des essais physiques ont permis de valider un modèle numérique qui caractérise les écoulements autour d’une bouche de métro et quantifie les volumes entrants. Cela a permis également de proposer une formule théorique de débit tenant compte de la géométrie d’une bouche de métro. Les écoulements par infiltration sont quant à eux modélisés en fonction de la charge de la nappe et validés avec des mesures in situ. Ce travail de recherche a comme objectif d’améliorer et valider un modèle de simulation. Il s’agit de mettre en œuvre un outil opérationnel d’aide à la décision qui permettra à la cellule inondation de la RATP de bien comprendre le fonctionnement de son réseau afin d’améliorer son plan de protection contre le risque inondation. / The 1910 flood of the Seine had a direct impact on the functioning of the different networks (Electricity network, sewerage, transport, water distribution). The RATP network was particularly affected in its functioning. The damage that centennial flood could cause today may be even greater because the current network is more vulnerable because of the numerous electrical and computer equipment that it comprises. The majority of the emergences (The water ingress) of the RATP is located in flood areas. During a major flooding of the Seine, the flows due to the floods propagate directly into the underground and central part of the network (Metro and RER) through these emergences. This thesis is interested in a hydrodynamic simulation by MIKE URBAN, Model used to model the RATP network due to its MOUSE engine developed by DHI for the sewerage networks. This work also presents the results obtained on a physical model of a subway station. The experimental data were used to model water ingress within the RATP network from the subway station. Network protection against infiltration requires a thorough knowledge of underground flow conditions. Infiltrations through the tunnels are estimated numerically. The aim of this research is to improve and validate a simulation model. It is a question of implementing an operational decision support tool which will allow the flood cell of the RATP to understand the functioning of its network in order to improve its flood risk protection plan.

Increasing global mean temperature or global warming has the potential to affect the hydrologic cycle. In the 21st century, according to the UN Intergovernmental Panel on Climate Change (IPCC), alterations in the frequency and magnitude of high intensity rainfall events are very likely. Increasing trend of urbanization across the globe is also noticeable, simultaneously. These changes will have a great impact on water infrastructure as well as environment in urban areas. One of the impacts may be the increase in frequency and extent of flooding. India, in the recent years, has witnessed a number of urban floods that have resulted in huge economic losses, an instance being the flooding of Mumbai in July, 2005. To prevent catastrophic damages due to floods, it has become increasingly important to understand the likely changes in extreme rainfall in future, its effect on the urban drainage system, and the measures that can be taken to prevent or reduce the damage due to floods. Reliable estimation of future design rainfall intensity accounting for uncertainties due to climate change is an important research issue. In this context, rainfall intensity-duration-frequency (IDF) relationships are one of the most extensively used hydrologic tools in planning, design and operation of various drainage related infrastructures in urban areas. There is, thus, a need for a study that investigates the potential effects of climate change on IDF relationships. The main aim of the research reported in this thesis is to investigate the effect of climate change on Intensity-Duration-Frequency relationship in an urban area. The rainfall in Bangalore City is used as a case study to demonstrate the applications of the methodologies developed in the research Ahead of studying the future changes, it is essential to investigate the signature of changes in the observed hydrological and climatological data series. Initially, the yearly mean temperature records are studied to find out the signature of global warming. It is observed that the temperature of Bangalore City shows an evidence of warming trend at a statistical confidence level of 99.9 %, and that warming effect is visible in terms of increase of minimum temperature at a rate higher than that of maximum temperature. Interdependence studies between temperature and extreme rainfall reveal that up to a certain range, increase in temperature intensifies short term rainfall intensities at a rate more than the average rainfall. From these two findings, it is clear that short duration rainfall intensities may intensify in the future due to global warming and urban heat island effect. The possible urbanization signatures in the extreme rainfall in terms of intensification in the evening and weekends are also inferred, although inconclusively. The IDF relationships are developed with historical data and changes in the long term daily rainfall extreme characteristics are studied. Multidecedal oscillations in the daily rainfall extreme series are also examined. Further, non-parametric trend analyses of various indices of extreme rainfall are carried out to confirm that there is a trend of increase in extreme rainfall amount and frequency, and therefore it is essential to the study the effects of climate change on the IDF relationships of the Bangalore City. Estimation of future changes in rainfall at hydrological scale generally relies on simulations of future climate provided by Global Climate Models (GCMs). Due to spatial and temporal resolution mismatch, GCM results need to be downscaled to get the information at station scale and at time resolutions necessary in the context of urban flooding. The downscaling of extreme rainfall characteristics in an urban station scale pose the following challenges: (1) downscaling methodology should be efficient enough to simulate rainfall at the tail of rainfall distribution (e.g., annual maximum rainfall), (2) downscaling at hourly or up to a few minutes temporal resolution is required, and (3) various uncertainties such as GCM uncertainties, future scenario uncertainties and uncertainties due to various statistical methodologies need to be addressed. For overcoming the first challenge, a stochastic rainfall generator is developed for spatial downscaling of GCM precipitation flux information to station scale to get the daily annual maximum rainfall series (AMRS). Although Regional Climate Models (RCMs) are meant to simulate precipitation at regional scales, they fail to simulate extreme events accurately. Transfer function based methods and weather typing techniques are also generally inefficient in simulating the extreme events. Due to its stochastic nature, rainfall generator is better suited for extreme event generation. An algorithm for stochastic simulation of rainfall, which simulates both the mean and extreme rainfall satisfactorily, is developed in the thesis and used for future projection of rainfall by perturbing the parameters of the rainfall generator for the future time periods. In this study, instead of using the customary two states (rain/dry) Markov chain, a three state hybrid Markov chain is developed. The three states used in the Markov chain are: dry day, moderate rain day and heavy rain day. The model first decides whether a day is dry or rainy, like the traditional weather generator (WGEN) using two transition probabilities, probabilities of a rain day following a dry day (P01), and a rain day following a rain day (P11). Then, the state of a rain day is further classified as a moderate rain day or a heavy rain day. For this purpose, rainfall above 90th percentile value of the non-zero precipitation distribution is termed as a heavy rain day. The state of a day is assigned based on transition probabilities (probabilities of a rain day following a dry day (P01), and a rain day following a rain day (P11)) and a uniform random number. The rainfall amount is generated by Monte Carlo method for the moderate and heavy rain days separately. Two different gamma distributions are fitted for the moderate and heavy rain days. Segregating the rain days into two different classes improves the process of generation of extreme rainfall. For overcoming the second challenge, i.e. requirement of temporal scales, the daily scale IDF ordinates are disaggregated into hourly and sub-hourly durations. Disaggregating continuous rainfall time series at sub-hourly scale requires continuous rainfall data at a fine scale (15 minute), which is not available for most of the Indian rain gauge stations. Hence, scale invariance properties of extreme rainfall time series over various rainfall durations are investigated through scaling behavior of the non-central moments (NCMs) of generalized extreme value (GEV) distribution. The scale invariance properties of extreme rainfall time series are then used to disaggregate the distributional properties of daily rainfall to hourly and sub-hourly scale. Assuming the scaling relationships as stationary, future sub-hourly and hourly IDF relationships are developed. Uncertainties associated with the climate change impacts arise due to existence of several GCMs developed by different institutes across the globe, climate simulations available for different representative concentration pathway (RCP) scenarios, and the diverse statistical techniques available for downscaling. Downscaled output from a single GCM with a single emission scenario represents only a single trajectory of all possible future climate realizations and cannot be representative of the full extent of climate change. Therefore, a comprehensive assessment of future projections should use the collective information from an ensemble of GCM simulations. In this study, 26 different GCMs and 4 RCP scenarios are taken into account to come up with a range of IDF curves at different future time periods. Reliability ensemble averaging (REA) method is used for obtaining weighted average from the ensemble of projections. Scenario uncertainty is not addressed in this study. Two different downscaling techniques (viz., delta change and stochastic rainfall generator) are used to assess the uncertainty due to downscaling techniques. From the results, it can be concluded that the delta change method under-estimated the extreme rainfall compared to the rainfall generator approach. This study also confirms that the delta change method is not suitable for impact studies related to changes in extreme events, similar to some earlier studies. Thus, mean IDF relationships for three different future extreme events, similar to some earlier studies. Thus, mean IDF relationships for three different future periods and four RCP scenarios are simulated using rainfall generator, scaling GEV method, and REA method. The results suggest that the shorter duration rainfall will invigorate more due to climate change. The change is likely to be in the range of 20% to 80%, in the rainfall intensities across all durations. Finally, future projected rainfall intensities are used to investigate the possible impact of climate change in the existing drainage system of the Challaghatta valley in the Bangalore City by running the Storm Water Management Model (SWMM) for historical period, and the best and the worst case scenario for three future time period of 2021–2050, 2051–2080 and 2071–2100. The results indicate that the existing drainage is inadequate for current condition as well as for future scenarios. The number of nodes flooded will increase as the time period increases, and a huge change in runoff volume is projected. The modifications of the drainage system are suggested by providing storage pond for storing the excess high speed runoff in order to restrict the width of the drain The main research contribution of this thesis thus comes from an analysis of trends of extreme rainfall in an urban area followed by projecting changes in the IDF relationships under climate change scenarios and quantifying uncertainties in the projections.