Presentation on theme: "Overview of Urban Drainage Flood Prediction Methods and Approaches J.Y. Chen1 and B.J. Adams2 1. Water Survey Division, Environment Canada 2. Department."— Presentation transcript:
1 Overview of Urban Drainage Flood Prediction Methods and Approaches J.Y. Chen1 and B.J. Adams Water Survey Division, Environment Canada 2. Department of Civil Engineering, University of TorontoGood afternoon, our topic of the presentation is “overview of urban drainage flood prediction methods and approaches”, the topic looks quite broad, but we will try to stay focused.
2 Presentation Outline Urban drainage modeling approaches Analytical model developmentModel evaluation and comparisonConclusionsWe start with an introduction to urban drainage modeling approaches, or model classification with emphasis in the development of analytical models.Since some of you may be not quite familiar with the methodologies, we try to paint a broad picture for the analytical model development by sharing our thoughts and experience with you. Model evaluation and comparison is based on a case study of the subcatchment located in Don river watershed. From the results of the case study, we attempt to draw conclusions.
3 Methods for Urban Drainage Flood Prediction Statistical/stochastic methodsFlood frequency analysisRegional flood frequency analysisTime series analysisDeterministic methodsConceptual modelsDifferent types of methods or models may be used for urban drainage flood prediction, these methods generally fall into two broad categories, namely, statistical methods and deterministic methods. For the statistical methods, flood frequency analysis may be used at a gauged site to estimate flood quantiles with a specified return period, regional flood frequency analysis can be used for an ungauged site; time series analysis includes auto regression analysis, stepwise regression analysis, etc. An important type of deterministic model is conceptual model, which attempts to simulate major components of rainfall-runoff physical processes
4 Deterministic Conceptual Modeling Methods Water budgetModel inputsModelstructureRunoff routingModel calibrationConceptual models normally consist of two major components, namely water budget and runoff routing components, the water budget component establishes water balance among rainfall, evaporation, soil moisture content, and different runoff components, the routing component reflects the damping effect of catchment storage. This type of model is normally characterized by the water budget component. In model applications, significant involvement is needed in model calibration and verification.Model verificationPrediction
5 Approaches Used for This Study Design storm approachSimulation approachContinuous simulationEvent simulationDerived probability distribution approachDesign storm, simulation and derived probability distribution approaches are used for this study. Depending on the time scale is used, simulation may be further divided into event and continuous simulation. Nevertheless, different approaches may have its advantages and disadvantages. For example design storm approach cannot consider antecedent soil moisture condition, and continuous simulation is usually considered more accurate, but time consuming.
6 Analytical Model Development Closed-form analytical models are developed with derived probability distribution theoryProbability distributions of runoff volumes and peak flow rates can be derived from probability distributions of rainfall characteristicsThe key element in the development of analytical model is the rainfall-runoff transformation, based on which the probability distributions of the rainfall characteristics are mathematically transformed to create the probability distributions of system outputs. such as, runoff volume and peak flow rates.We will give you some highlights of these transformations
7 Rainfall-Runoff Transformation Runoff coefficient basedExtended form
8 Rainfall-Runoff Transformation (Cont’d) Infiltration based
9 Analytical Model Statistic analysis of rainfall records Rainfall characteristics, e.g.,rainfall event volume,duration & interevent timeOverflowStoragefacilityExceedance probabilityof a runoff spill volumeProbability distributionsof rainfall characteristicsRainfall-runofftransformationAverage annualvolume of spillsAverage annualnumber of spillsPDF or CDF of runoffevent volumeExpected value ofrunoff event volumeAverage annual runoffvolumeAverage annual runoffcontrol efficiency
19 ConclusionsPeak flow rates from event simulation models appear to be lower than the results from continuous simulation modelEvent simulation models appear to be more conservative than continuous simulation model for runoff volume estimation
20 Conclusions (Cont’d)The closed-form analytical models developed with derived probability distribution theory, are capable of providing comparable results to continuous simulation resultsDifferent models may vary not only in modeling approach, but also in the level of complexity, it can be challenging to select an appropriate model with a desired level of performance