Rainfall Runoff Modelling

RAINFALL RUNOFF MODELLING

CHAPTER 1 INTRODUCTION
1.1 DEFINITION OF RUNOFF: Runoff may be defined as that portion of the precipitation as well as any other flow contribution that enters the natural surface stream or channel. Thus it is a flow collected i.e., output from the drainage basin in a given unit of time. It is one of the different phases of hydrological cycle. The runoff from a catchment in any specified period is the total amount of water that flows into the natural stream and is expressed as: (i) Millimeters (or centimeters) of water over the entire catchment area (also called as drainage basin) or (ii) In hectare-meters or sometimes in cubic meters of water per unit area of the catchment or drainage basin.
Rainfall is known as the main contributor to the generation of surface runoff. Therefore there is a significant and unique relationship between rainfall and surface runoff. When precipitation falls towards the earth, a portion of it is retained by the vegetation as interception, stored in the depressions of the ground as depression storage and as soil moisture Part of the precipitation reaches the underground as infiltration. When all these requirements are satisfied the excess precipitation spreads and covers the soil surface with a film of water called as surface detention and flows over the land surface as overland flows and enters the (natural) channel and flows as surface runoff. The water retained as interception, depression, storage and soil moisture (i.e., capillary water) constitutes basin recharge. The surface runoff may consist of the following two portions. (i) Overland flow that flows over the surface (i.e., surface runoff) to join the nearby channel. (ii) Inter-flow which is the portion of the precipitation that infiltrates into the soil and flows laterally in the surface soil to an adjacent channel. This is further
classified as prompt interflow (with minimum time lag) and delayed interflow. The part of the precipitation that percolates into the ground through the soil to join the ground water is called as ground water runoff. The overland flow and interflow are usually combined together to form direct runoff.
1.2 CLASSIFICATION OF RUNOFF: Depending upon the time delay between the precipitation and the runoff, the runoff may be classified into: (i) (ii) Direct runoff Base flow which is delayed flow from the catchment that joins the natural channel as ground water flow 1.3 METHODS OF ESTIMATING RUNOFF: The various methods of indirect measurement of runoff may be classified as follows: (1) Rational Method Rational method is well known as one of the basic approach to compute stormwater flows from rainfall by relating peak runoff to rainfall intensity through a proportionally factor. When the first flow rate or discharge formula was established, the rainfall intensities were not considered as a significant factor.
where, Q = calculated flow rate (m3/s), C = runoff coefficient, I = rainfall intensity (mm/h), A = area of catchment involved (ha).
Although the method can be considered as the most reliable approach in estimating the design storm peak runoff, experience has shown that it only provides satisfactory results on small catchments of up to 80 hectares only. (2) By Unit Hydrograph Method: The Unit Hydrograph (UH) of a drainage basin is defined as a hydrograph of direct runoff resulting from one unit of effective rainfall which is uniformly distributed over the basin at a uniform rate during the specified period of time known as unit time or unit duration. The unit quantity of effective rainfall is generally taken as 1mm or 1cm and the outflow hydrograph is expressed by the discharge ordinates. The unit duration may be 1 hour, 2 hour, 3 hours or so depending upon the size of the catchment and storm characteristics. However, the unit duration cannot be more than the time of concentration, which is the time that is taken by the water from the furthest point of the catchment to reach the outlet. Figure 1 shows a typical unit hydrograph.
Figure 1 Typical Unit Hydrograph
CHAPTER 2 RAINFALL RUNOFF MODELLING

2.1 INTRODUCTION: Rainfall runoff models (RRMs) are standard tools routinely used today for hydrological investigations in engineering and environmental science. They are applied to extend stream flow time-series in space and time to evaluate management strategies and/or catchment response to climate and/or land use variability for the calculation of design floods as load models linked to water quality investigations for real-time flood forecasting or to provide boundary conditions for atmospheric circulation models. While flood modelling is of primary interest, such models allow studying the interactions between surface water and groundwater and the unsaturated zone. The rainfall -runoff process is a highly complex, nonlinear, and dynamic physical mechanism that is extremely difficult to model. This chapter presents an approach that combines data and technique, to develop integrated models of the rainfall runoff process. The generic guidelines outline a procedure for applying a hydrological model. This can be summarised as occurring in four phases: 1. Project management, 2. Problem definition, 3. Option modelling,
4. Compare Options and select the best. This deals only with problem definition and option modelling because the first and last phases are discussed sufficiently for the purpose of rainfall-runoff modelling in the generic guidelines. A further reason is that rainfall-runoff modelling is usually only a part of a larger hydrological modelling project. This section describes the process of developing a rainfall-runoff mode:
2.2 COLLATE AND REVIEW DATA The amount and quality of data available to develop a model should be determined. This can influence the selection of models, the performance criteria, and the approach to calibrate models. A bare minimum data set sufficient to make an approximate estimate of mean annual catchment yield would include catchment area along with spatial and temporal characteristics of rainfall and potential evapotranspiration (PET). A comprehensive data set would include long-term streamflow measurements and rainfall and PET data collected at one or more locations within the catchment along with land use coverage, vegetation cover and impervious area information, including changes over time. The quality of the data should be reviewed prior to using to detect errors, nonstationary if any, and understand uncertainties that may influence estimates.
2.3 SETTING UP AND BUILDING A MODEL The catchment characteristics are considered along with the knowledge on data available and any other information available to the modeller. The rainfall-runoff model chosen is conceptualised and an initial parameter set is identified. When the model is first set up consideration should be given to all constraints which are limiting and their effects on the modelling. Section 5 provides more details associated with this step.
2.4 CALIBRATION AND VALIDATION
Model calibration is a process of systematically adjusting model parameter values to get a set of parameters which provides the best estimate of the observed streamflow (in the case of rainfall-runoff models). The term validation, as applied to models, typically means confirmation to some degree that the calibration of the model is acceptable for the intended purpose. In the context of rainfall runoff modelling, validation is a process of using the calibrated model parameters to simulate runoff over an independent period outside the calibration period (if enough data is available) to determine the suitability of the calibrated model for predicting runoff over any period outside the calibration period. It is a very common situation in a project that involves rainfall runoff modelling for flow time series to be required for several catchments or sub catchments within the model domain and for data to be available from two or more stream flow gauges to facilitate calibration and validation. At locations where gauged flows are available and flow estimates are required, two options are available to the modeller: (1) The rainfall runoff models may be calibrated independently for each gauged catchment. In this case, independent parameter sets will be derived for the rainfall runoff models of each catchment; or (2) A joint calibration may be performed, with rainfall runoff models calibrated with consistent parameters to fit to the gauge records from two or more gauges. In this case, a single set of rainfall runoff model parameters will be produced for all of the catchments that represent a compromise to fit the flows at all of the gauges within that group. Calibration of a rainfall runoff model normally involves running the model may times, trialling different values of parameters, with the aim of improving the fit of the model to the calibration data. Calibration can be facilitated: Manually, with the modeller setting the parameter values, running the model to
inspect the results and then repeating this process many times; Using automated optimisation, with an optimiser algorithm running the model
hundreds or thousands of times with different parameter values; or
Using a hybrid approach of automated optimisation phases, interspersed with
manually implemented trials of parameter sets. 2.5 CLIMATIC DATA: Climatic data is the most important driver of any rainfall runoff modelling process. The calibration and validation of models also involves comparison to observed streamflow data. Checks should therefore be performed on the input data and the comparison data set for calibration and validation to be used in rainfall runoff modelling before any attempt is made to apply or calibrate the models. Investigations into data to be used for rainfall runoff modelling include checks of: Stationarity of the data time series , i.e. has there been any systematic or step change
in the statistical properties over the time of data collection, and if so why; Spatial coherence of data, i.e., is the data consistent with regional spatial and temporal
patterns and trends; Accuracy of the spatial location of the gauging site; Consistency in the approach used to date and time stamp the data, particularly for data
provided by different agencies; Procedures use for spatially interpolation of point observations to gridded data
estimates or estimated series across catchment areas e.g., time series plots at different levels of temporal aggregation, ranked plots, residual mass curves, double mass curves, etc. One major factor which will apply across all types of time series data used is that the time base must be kept consistent so that the data applies to the same time period.
2.6 CATCHMENT DETAILS: (1) Location of gauges (streamflow, rainfall and evaporation): The streamflow recorded at the catchment outlet is a combined response to the spatial distribution of rainfall and evaporation across the catchment. There are uncertainties
associated with the streamflow measurements due to rating curve errors as well as due to extrapolation outside the limits of the rating curve. There is spatial variability in rainfall (and to smaller extent evaporation) across a catchment which is not captured when undertaking lumped catchment modelling using a single rain gauge. There might be problems with the location of the rain gauge in terms of capturing a representative rainfall for all the rainfall events especially for catchments with high rainfall gradients. (2) Topography and Catchment Areas: Size of catchment may have an effect to the runoff generation in terms of the runoff efficiency (volume of runoff per unit area). The larger the size of the catchment, the larger is the time of concentration and the smaller the runoff efficiency. The land use characteristics also contributed for the surface flow process whereby the infiltration excess flow is the main process in terms of runoff generation on degraded land while saturation excess overland flow is more relevant for agricultural land. The catchment area for a catchment represents the contributing area to the catchment outlet where the streamflow is measured. The catchment boundaries (and the corresponding catchment area) can either be derived from topographic map layers or using the catchment digital elevation model (DEM) and a standard package. It is usually easier to determine catchment area for the catchments located in steeper terrain compared to those located in very flat areas (especially when using DEM). (3) Soil types: A catchments rainfall-runoff response is related to the soil types in the catchment. The surface soil characteristics determine the infiltration rates and so the contributions from different flow components (surface runoff, through flow and base flow). Soils information can be obtained from any soils field work that has been undertaken in the catchment or from large scale soil properties maps. In most practical applications of conceptual rainfall-runoff models, soil information is seldom directly used as input in the calibration process because the inherent spatial variability in soil properties within a catchment is typically sufficiently large that it has been difficult to demonstrate statistically robust relationships between conceptual model parameters and soil properties. (4) Vegetation: Land cover/vegetation cover in a catchment can often be correlated with the amount of interception storage/loss and actual evapotranspiration in a catchment. The land cover
across the catchment can be derived from large scale vegetation mapping based on satellite imagery or remotely sensed data. Vegetation cover data has not typically been used explicitly in directly determining rainfall runoff model parameters, although there have been some recent studies which have demonstrated the importance of catchment land cover in rainfallrunoff model calibration and for predictions in ungauged basins. (5) Water Management Infrastructure: Water management infrastructure within a catchment can allow humans to make very substantial modifications to flows within a catchment. Water management infrastructure may include large dams, farm dams and off stream storages, extractions, man-made canals or diversion pipelines and discharges from sewage treatment plants. Locations of these infrastructures should be identified where they exist within the catchment so that their potential impact on stream flows may be assessed. Recorded flows at the catchment outlet may require adjustment to allow for the influence of water management infrastructure located upstream of each of the flow gauging locations.
2.7 FLOW DATA: Reliable measurements of streamflow data are critical for successfully calibrating a rainfall-runoff model to a catchment. The streamflow data for all the gauged locations can be obtained from the respective state government water management agencies. Considerations in checking streamflow data include: The agency collecting the data and the quality assurance procedures (if any)
implemented by that organisation during data collection; Reliability of the rating of levels to flows for the gauge; The accuracy, extent and currency of cross sections surveyed at the gauge site.
(Surveyed cross sections may only extent to the top of the stream bank and gauging for flows extending onto the floodplain may use a cross section that is inaccurate); for; How hydraulically stable (variable over time) the rating site is; Vegetation and substrate material for the channel bed, channel banks and floodplain
and the influence of assumptions made about these on gauged flows; The ratio of the highest flow outputs to the highest flow that the gauge has been rated
Examination of potential backwater effects for the site from influences that are
downstream of the site, such as stream confluences, bridge crossings, culverts, dams or weirs; Hysteresis effects leading to different flow rates for the same recorded level on rising
and falling limbs of hydrographs; How well maintained the gauging site and instrumentation used for measuring water
levels has been; Any changes to the gauging instrumentation over time; The length of time since the last rating at high flows; Length of record at the site; Availability of quality codes with the flow data record; Proportion of missing data; Trends in when data is missing from the record and how this might influence any If there are a number of gauges closely located that basically represent the same Assessment of the above factors will inform whether the data is useful in calibration of the model, independent validation of the model or whether the data should be ignored.
infilling procedures; and catchment the data sets may be able to be combined to give a longer record for the site.
2.8 RAINFALL DATA: Rainfall-runoff modelling still depends heavily on the records from point rain gauges, both recording rain gauges giving estimates of rainfall intensities at time steps of one hour or better and daily rain gauges. Recording rain gauges become more important, but they are more expensive to operate and much fewer in number. Thus it may still be necessary to use daily gauges to get an estimate of the total volume of rainfall over a catchment, using the nearest recording rain gauge to give an appropriate idea of the distribution of rainfall in time, the storm profile.
Raingauge-measured volumes may be subject to error. In particular, they depend on the design of the raingauge in relation to wind conditions at the site and rainfall intensities. The best design is thought to be a raingauge with the orifice set at ground level and
10
surrounded by an anti-splash grid but this is not always practical, particularly in environments with frequent snow. A variety of designs of wind shield have been used in different countries to try to mitigate this wind effect. The wind effect can be large; estimates of reductions of up to 20% have been reported at windy sites for gauges only 30cm above the ground compared to ground-level gauges.
2.9 EVAPOTRANSPIRATION: The measured pan evaporation data can be obtained for all the locations with the evaporation gauges installed. Modelling requires potential evapotranspiration (PET). There are a number of methods to convert pan evaporation to PET including. These use climatic variables in the conversion calculation including solar radiation, temperature, vapour pressure, and wind speed which are recorded at some pan recording stations but not all. This further limits the network available to draw data from. When all the required data is available the conversion calculations will use the records but often some variable is missing and estimates of that variable are made and used. Where there is no data for the climatic variables, calculated pan to PET conversion factors from a nearby station can be used to derive PET from pan evaporation. Commonly the spatial products have interpolated layers for a range of climatic factors and the spatial PET layer is calculated from data in these layers rather than interpolating PET calculated at recording stations. 2.10 RAINFALL RUNOFF PROCESSES Apart from recording and/or forecasting rainfall itself, the next most important problem is understanding and forecasting the runoff generated by the rainfall. This difficult problem has attracted enormous amounts of attention and effort around the world. 2.10.1 Factors affecting runoff (1) Rainfall Precipitation, whether it occurs as rain or snow, is the potential source of water that may run off the surface of small watersheds. The extent of the storm and the distribution of rainfall during the storm are two major factors, which affect the peak rate of runoff. The
11
storm distribution can be thought of as a measure of how the rate of rainfall (intensity) varies within a given time interval. (2) Antecedent Moisture Condition The runoff from a given storm is affected by the existing soil moisture content resulting from the amount of precipitation occurring during the preceding five days (antecedent moisture condition). (3) Watershed Area The watershed area or area draining water to the point of interest is usually determined from a topographic map or scaled aerial photograph accompanied by a field review locating manmade features that have diverted the flow of water. (4) Soils Apart from rainfall characteristics such as intensity, duration and distribution, there are other specific factors which have a direct bearing on the occurrence and volume of runoff. The most common factor is the soil type. Due to the variation of runoff production, different studies have been conducted according to particular soil conditions. In general, the higher the rate of infiltration, the lower the quantity of stormwater runoff. Fine-textured soils such as clay produce a higher rate of runoff than do coursetextured soils such as sand. Sites having clay soils may require the construction of more elaborate drainage systems than sites having sandy soils. (5) Surface Cover The type of cover and its condition affects runoff volume through its influence on the infiltration rate of the soil. Fallow land yields more runoff than forested or grass land for a given soil type. Some of the intercepted moisture is so long draining from the plant down to the soil that it is withheld from the initial period of runoff. Vegetation has a significant effect on the infiltration capacity of the soil. A dense vegetation cover shields the soil from the intense raindrop impact which eventually will cause a breakdown of the soil aggregate as well as soil dispersion with the consequence of driving fine soil particles into the upper soil pores. Vegetation, including ground litter, forms numerous barriers along the path of the water flowing over the surface of the land, which slows the water down and reduces its peak rate of runoff. Covering areas with impervious material reduces surface storage and infiltration and thus increases the amount of runoff. (6) Time Parameters
12
Time is the parameter that is used to distribute the runoff into a hydrograph. The time is based on the velocities of flow through segments of the watershed. The slope of the land in the watershed is a major factor in determining the velocity. Two major parameters are time of concentration (Tc) and travel time of flow through the segments (Tt). (7) Storage in the Watershed On very flat surfaces where ponding or swampy areas occur throughout the watershed, a considerable amount of the surface runoff may be retained in temporary storage, thus reducing the rate at which runoff will occur. Storage areas may be created to reduce the rate of runoff in an urbanizing area. These can be effective sediment traps as well as flood detention structures if left permanently in the watershed. Generally the following processes are usually identified as taking place: Evapotranspiration at the surface Surface infiltration Overland flow Unsaturated zone flow Saturated zone flow (groundwater) Consequently, artificial conduits augmenting any natural channels are constructed to convey excess rainfall away from critical areas quickly and efficiently. Such water can of course be stored effectively in detention areas depending on the capacity of the (downstream) conduits and natural channels. Infiltration of rainfall in an urban area is by definition limited, though engineers now recognise that there is considerable value in artificially maximising the infiltration of rainwater in order to limit the cumulative surface runoff, while ensuring that consequent groundwater levels do not adversely affect the foundations of urban structures.
CHAPTER 3 RAINFALL RUNOFF MODELS

3.1 DEFINITION OF RAINFALL-RUNOFF MODEL: The best model, is always that which achieves the greatest realism with the least parameter and model complexity.
13
Model can also be understood as a system of inter-related components and the relationships between them. The system analysis involves the breaking down its complexity into simple manageable subsystems connected by flows of causality, matter, energy or information. The purpose of systems analysis is to make complex systems more easily understood. Watershed models simulate natural processes of the flow of water, sediment, chemicals, nutrients, and microbial organisms within watersheds, as well as quantify the impact of human activities on these processes. Simulation of these processes plays a fundamental role in addressing a range of water resources, environmental, and social problems.
3.2 CLASSIFICATION OF RAINFALL-RUNOFF MODELS: Models are normally characterized or classified to help describe and discuss their capabilities, strengths, and limitations. There is no universal method to characterize rainfallrunoff models, and models have been classified in several ways depending on the criteria of interest. From the above references five categories are chosen and are presented below: 1) Event and Continuous Simulation Models, 2) Conceptual and Hydrodynamic Models, 3) Lumped and Distributed Parameter Models, and 4) Models with Fitted, Physically Determined, or Empirically Derived Parameters. 5) Channel flow routing models One common classification scheme not included below is the differentiation between deterministic and stochastic models. Deterministic modeling is simply a category of stochastic modeling that disregards the uncertainties in the model, its parameters, and its inputs.
1) Event and Continuous Simulation Models Rainfall-runoff models are either event models or continuous simulation (CS) models. Event models typically estimate the runoff from an individual storm event, i.e., describing a relatively short period within the hydrologic record. Event models ordinarily evaluate a partial set of the hydrologic processes that affect the watershed: infiltration, overland and channel flow, and possibly interception and detention storage. Most event
14
models use a constant time interval, whose value may typically range from minutes to several hours. Continuous simulation models operate for a sustained period that includes both rainfall events and interstorm conditions. To legitimately evaluate the stream flow during interstorm periods, continuous simulation models should include additional hydrologic properties such as evapotranspiration, shallow subsurface flow, and ground-water flow. Also crucial to these models is an accounting of the soil moisture and how it relates to changes in infiltration. The CS time interval can be daily, hourly, subhourly, or variable. Models that provide only daily simulation are not ordinarily useful for stormwater applications. For an event model, the initial conditions (antecedent soil moisture, stream and reservoir levels, etc.) must either be subjectively assigned by the user, calibrated with some type of error-reduction procedure, or approximated by an external procedure. When an explicit accounting procedure and the past climatological record are used to estimate the initial conditions, the function of the event model can approach that of a CS model. In modeling a long period that contains a number of floods of various magnitudes, the application of CS models provides more opportunities to compare model results with observed runoff. A long calibration period with a variety of hydrological conditions increases confidence in model results. Event models are typically applied to fewer storms, but increased confidence can be gained by calibrating the model to as many storms as possible. 2) Conceptual and Hydrodynamic Models The categorization describes the types of equations used in the model to describe the hydrologic processes. These categories of models are identified as: 1) "Black-box" or transfer functions, 2) Conceptual models, and 3) Hydrodynamic models. Black-box models rely upon a statistical correspondence between the model input (rainfall) and model output (runoff) without relation to any underlying physical processes. Conceptual models are described as "models which are formulated on the basis of a simple arrangement of a relatively small number of elements, each of which is itself a simple representation of a physical relationship." This definition is sufficiently broad enough to include hydrodynamic models, but conceptual models usually represent an intermediate level of component sophistication. Hydrodynamic models -- sometimes also termed
15
physically based models -- are also simplifications of reality and have a certain amount of empiricism. However, these models are generally based on the most recent physics-based understanding of the hydrologic processes that control the runoff response in the watershed In reality, the boundaries between conceptual and hydrodynamic models are fuzzy. Individual models will normally combine both conceptual and hydrodynamic components. Not all hydrologic properties can be represented by hydrodynamic components, which force all models to have some conceptual and empirical aspects. The predominant manner in which the components are modelled results in the overall classification. 3) Lumped and Distributed Parameter Models In lumped conceptual models the parameters and variables represent average values over the entire catchment. Therefore, the description of the hydrological processes cannot be based directly on the equations that are supposed to be valid for the individual soil columns. As a result the equations are semi empirical, but still with a physical basis. The model parameters cannot usually be assessed from field data alone, but a have to be obtained through the help of calibration. These models operate with different but mutually interrelated storages representing physical elements in a catchment. The mode of operation may be characterized as a bookkeeping system that is continuously accounting for the moisture contents in the storages. The hydrologic parameters used in the rainfall-runoff models can be represented in either a lumped or distributed manner. The lumping method averages the total rainfall, its distribution over space, soil characteristics, overland flow conditions, etc. for the entire watershed, ignoring all flow-routing mechanisms that exist within it. The expectation is that any minor details of the rainfall-runoff process will be inconsequential, resulting in an "average" flood condition. Although certain lumped parameters may implicitly represent physical attributes of the hydrologic system, they cannot be expected to have any direct physical interpretation. Lumped models can be made to behave more like distributed parameter models by adopting a detailed database and dividing a watershed into very small sub watersheds. Distributed parameters both describe the geographical variation of parameters across the watershed and discriminate between changes in the hydrologic processes that occur throughout the watershed. In a fully distributed model, the hydrology of each small element of the watershed is distinctly simulated to include the hydrologic interactions with bordering elements. In reality, parameters in the distributed models have to be lumped to a small degree
16
to match the grid scale used for computations. In addition, the fitting of distributed, hydrodynamic models to observed streamflow at present is usually accomplished through the simplification and calibration of certain parameters. Therefore, without a sufficiently detailed database, a distributed model effectively may deteriorate into a lumped system model. A third approach simulates the hydrologic processes for a discrete number of land use and soil types. A land use and soil type combination, termed a hydrologic response unit (HRU), may occur in numerous locations in the same watershed; however, the hydrologic response is modelled for this combination only once, and this response is assumed to be homogeneous for all locations having that HRU. The HRU parameter approach is used in many rainfall-runoff models. Depending on how the watershed is partitioned, either the hydrologic response from each HRU is assigned to individual elements throughout the watershed, or the responses from several HRUs are prorated and aggregated to represent the lumped response from a sub watershed. Within the framework of any individual model, the level of distribution can be user-controlled. It has been stated that the appropriate extent to which a modeller chooses to distribute the parameters should depend upon the objectives of the study and the available data, time, and money. Many studies suggest that distributed parameter models are desirable because they have the greatest potential for use in describing land use change, water-quality modelling, and forecasting on ungagged watersheds. These advantages assume that the parameters of distributed models are more physically realistic than the lumped model parameters, which should be the case when the model is well designed. Distributed parameters have the potential to be physically interpreted and, when this is the case, greater confidence can be placed in the use of the model. One reason that distributed parameter models have not seen widespread use is the availability of detailed databases. Future improvements in data acquisition, including the application of geographical information systems (GIS), will likely lead to more extensive use of distributed and HRU parameter models. 4) Models with Fitted, Physically Determined, or Empirically Derived Parameters: Parameters for rainfall-runoff models can be 1) Fitted through calibration, 2) Determined from field measurements, or 3) Empirically fixed.
17
One group of empirical models are statistically based using statistical methods such as ARIMA (Autoregressive Integrated Moving Average). Another group of empirical models are based on the unit hydrograph model (or applying the principles of unit hydrograph). The third group of empirical models are data-driven models using methods such as artificial neural networks, model trees, nearest neighbour method, evolutionary algorithm, support vector machines, etc. Fitted parameters, set in the calibration process, typically have no little or no physical interpretation. Physically determined parameters are derived from measurable watershed characteristics such as slope, channel width, hydraulic conductivity of soils, etc. Measured values may not always produce the best results when used directly in a model. Thus, some physically determined parameters may be adjusted during the calibration process and are not necessarily equal to the measured variables. But to maintain the physical relationship these parameters should be similar in magnitude and behavior to the measured values. The use of fitted versus physically determined parameters is a major issue in the application of rainfall-runoff models. Fitted parameters are less likely to be consistent from one data set to another, and models that use these parameters are less appropriate for extrapolation. In general, lumped models and most conceptual models use fitted parameters. However, it has been indicated that fitted parameters cannot reliably be transferred for use on ungaged watersheds. Thus, empirically derived parameter methods are often used with the lumped conceptual models for ungaged sites. Distributed and quasi-distributed conceptual models can use a combination of fitted, physically determined, and empirical parameters. Distributed hydrodynamic models primarily use measured or physically determined parameters, with some empirically derived parameters. Empirically derived parameters are developed by the regression analysis of either fitted or physically determined parameters. Empirically derived parameters may vary in the amount of physical interpretation that can be associated with their values. This category of parameters includes the Soil Conservation Service (SCS) runoff curve numbers that were developed for estimating rainfall losses on ungaged watersheds. Many of these empirically fixed relationships are required for parameterization of selected components in all models, including the models that are more physically based. In physically-based distributed models processes are represented by one or more partial differential equations and equations and parameters are distributed in space. The
18
principal mode of operation of a physically-based distributed model is illustrated in the following figure. Contrary to the lumped conceptual models a physically-based model does not consider the water flows in an area to take place a few storage units. Instead, the flows of water and energy are directly calculated from the governing continuum (partial differential) equations, such as the Saint Venant equations for overland and channel flow, Richards equation for unsaturated zone flow and Boussinesqs equation for groundwater flow. Distributed models are applied to catchments with complex channel network, varying spatial distribution of land use, soil type and vegetation cover, with complex aquifer system below the soil surface, etc.
(5) Channel flow routing models Hydrological and hydrodynamic approaches to channel flow routing can usually be shown to have a common basis in the St. Venant equations, and though them to the physical properties of the river channel and its floodplain. As a consequence, application to ungauged river channels has a natural physical basis. However, even for the most refined hydrodynamic river model, channel geometry simplification and the inherently empirical nature of roughness normally means there is benefit in model calibration for gauged sites and transfer of the experience gained for application to ungauged reaches. Hydrological approaches combine simple mass balance water storage accounting with a simplified momentum equation linking channel storage to water level or flow. The simplifications involved can make the links to channel properties less direct in physical terms, but can ease practical application and the building up of experience for use in modelling ungauged reaches. Simpler hydrological approaches are normally preferred where backwater influences from tides, river controls and confluences are not dominant. The hydrodynamic approach is sometimes distinguished by models providing estimates of both river flow and level for situations where there is no unique relation between these two quantities. However, the distinction between hydrological and hydrodynamic (hydraulic) approaches is largely artificial with a spectrum of levels of simplification. A popular method of hydrological routing is the one in which reach storage is a linear function of a weighted combination of the reach inflow and outflow. It is possible to relate this back to the underpinning St. Venant equation and in this way establish relations with channel properties applicable to ungauged reaches. There are different ways of doing this leading to different variants.
19
3.4 MODEL CALIBRATION, VALIDATION, AND VERIFICATION: Calibration, validation, and verification are the three crucial steps for the proper application of a model. Calibration is the process of modifying model parameters to reduce the error between the simulated streamflow and some portion of the observed flow record. Model validation tests the ability of the model to estimate runoff for periods outside that used to calibrate the model. Model verification investigates the range of conditions over which the model will produce acceptable results. In normal application of a model to a gaged watershed, calibration is often the only procedure of the three that is followed. Model validation and verification are often not considered practical. If essential information about these two procedures is to be obtained, then it is normally up to the model developers and researchers. Some explanation of model verification is especially important for applications to ungaged watersheds when calibration and validation cannot be achieved. Associated with the procedures of calibration, validation, and verification are three separate issues involving model application: flexibility, divergence, and extrapolation. Model flexibility describes the capability of a model to calibrate for a variety of different watersheds and flow conditions. Model divergence defines the relative accuracy of the model between the calibrated period and the validated period. Model extrapolation is the use of a model to describe hydrologic conditions outside of the range used for calibration and validation. These three issues are discussed below in context with other concerns related to calibration, validation, and verification.
CHAPTER 4 APPLICATION OF RAINFALL-RUNOFF MODELS The tasks for which rainfall-runoff models are used are diverse, and the scale of applications ranges from small catchments, of the order of a few hectares, to that of global models. Typical tasks for hydrological simulation models include: modelling of gauged catchments (e.g. modelling of river behaviour, real-time flood forecasting, adjusting and evaluation of water resource management); runoff estimation of ungauged catchments; effects of rivers activity (erosion, sedimentation); prediction of catchment response to changed conditions (e.g. land use change, climate change) and water quality investigations
20
(e.g. nutrients, migration of microbes, salinity and alkalinity of soils, acid precipitation, nonpoint source pollution). In contemporary practise, rainfall-runoff models are standard tools routinely used for hydrological investigations in engineering and environmental science. Also the topic of watershed management gains an increased attention. Some of the models are also employed in military operations
4.1 MODELING APPROACHES: The approaches used for rainfall runoff modeling, over a wide range, 4 methods from black-box models to very detailed deterministic/conceptual models. The determonistic conceptual models need a thorough understanding of the physics involved a large amount of data for calibration and validation purposes and are computationally demanding. There is m abundance of literature available in the area of rainfall- runoff modeling using deterministic conceptual methods. Recently, artificial neural networks (ANNs) have been proposed as efficient tools for modeling complex physical systems. The application of ANNs to the field of rainfall runoff modeling, which is popularly known as hydro informatics, started in the 1990s.
The choice of a particular rainfall-runoff model, the types of input, and modeling approach are functions of both the desired products of the modeling effort and the complexity of the watershed. Five approaches are identified for quantifying infrequent flood events and their frequencies. These approaches are based on a combination of precipitation inputs into the hydrologic model and frequency analysis. The first three approaches (frequency analysis, continuous modeling with historical precipitation, and design storm modeling) are commonly used.
4.1.1 Channel flow routing models Channel flow routing models are used to translate a flow hydrograph from an upstream site to one downstream. Where the downstream flow influences this translation via backwater control, this situation is treated separately here under hydrodynamic river models. A modelled river reach is normally sub-divided into sub-reaches with nodes at their boundaries. Assigning a boundary node to a target ungauged location provides a simple example of the use of a channel flow routing model as an indirect modelling approach for ungauged forecasting. Ungauged lateral inflows commonly bring further complexity and lessen forecast accuracy. Simple scaling methods or rainfall-runoff models may be used to represent such ungauged lateral inflows. A lesser form of ungauged problem is where only river level measurements are available and a stage-discharge relation cannot readily be established via a current metering field programme. The stage-discharge relation may be embedded within the channel flow routing model and its form and parameters calibrated along with those of the routing model. Some channel flow routing models can be linked directly to the St. Venant equations of
21
open channel flow and through them to the properties of the river channel and its floodplain. This can provide a direct basis for application to ungauged sites but, on account of the simplifications involved, is likely to benefit greatly from experience gained in modelling similar river reaches that are gauged. Channel flow routing models have a common basis in the St. Venant equations and their simplification. This provides a formal link to channel properties, concerning geometry and resistance (roughness), and a sound basis for application to ungauged channel reaches. Simplifications of representation and of channel geometry, together with the essentially empirical nature of roughness, means that there will normally be benefit in model calibration at gauged sites and transfer of this experience to ungauged sites. This applies even for the most refined hydraulic models. 4.1.2 Flood mapping tools Flood mapping tools facilitate the mapping of water levels continuously over an area so the ungauged location is most typical. The tool may serve wholly as a visual display facility with the information mapped deriving from observed (remotely-sensed imagery) and/or modelled sources. The mapping tool may be provided as an intrinsic component of a 1-D or 2-D hydrodynamic river modelling system. There is a developing opportunity for area-wide hydrological models to map inundation extent and depth at an indicative level and with UK coverage. The river flow volume along the entire river network can also be mapped in intensity-coded line form. Simple geomorphological relations on channel geometry linked to grid-to-grid flow routing models and DTMs provide the modelling support to such products. (1) Animated spatial displays of observed and modelled water levels are useful to depict the spatial extent and severity of flood inundation. It is common for some form of GIS (Geographical Information System) to be used to provide this functionality. The degree to which the GIS itself is used for inference of mapped information or an external model or observations will depend on the detail of the application. (2) While flood mapping tools are commonly used with 1-D, 2-D and 3-D hydrodynamic model outputs, there is also great scope to use distributed hydrological forecasting model outputs to produce spatial maps of river flow, flood inundation and related quantities over time. Some early prototyping of these opportunities has been done using the Grid-to-Grid hydrological model. Model outputs in gridded form are exported to HYRAD and displayed as
22
animated images of river flows propagating down the modelled river network along with fields of soil moisture deficit and local runoff. Also, time-series hydrographs can be extracted and viewed for any location (gauged or ungauged) down the river network. Further work leading to operational implementation is recommended here.
CHAPTER 5 THE ROLE OF REMOTE SENSING (RS), GEOGRAPHIC INFORMATION SYSTEM (GIS) AND DIGITAL ELEVATION MODEL (DEM) RAINFALL RUNOFF MODELLING The very quick developments in RS and GIS technology have played a critical role of application of RS and GIS in watershed modelling in general and rainfall runoff modelling in particular. The reason is that RS and GIS have contributed critical information as input of the models. Actually nowadays, we hardly find any rainfall runoff models that do not utilise RS and GIS data. Several scientists have introduced RS and GIS as powerful tools in rainfall runoff modelling.
23
Weather radar is a ground-based form of remote sensing configured for rainfall measurement. There are other important forms of monitoring by remote-sensing that are satellite-based. Some have already been commented on, especially as a source of elevation and land cover data. Whilst these datasets are often considered static, there is now increasing availability of time-history spatial datasets of leaf area index, snow cover, area of flood inundation and surface soil moisture. These have relevance both to the monitoring and modelling/forecasting of ungauged areas. An exciting prospect is the ability to remotely sense river level (and width) from which to develop flow discharge estimates. A combination of GPS (global positioning system) technology and a tethered floating buoy has been investigated in field trials and through computer simulation of anticipated satellite position systems. A system, usually computer based, for the input, storage, retrieval, analysis and display of interpreted geographic data. The database is typically composed of map-like spatial representations, often called coverages or layers. These layers may involve a three dimensional matrix of time, location, and attribute or activity. A GIS may include digital line graph (DLG) data, Digital Elevation Models (DEM), geographic names, land-use characterizations, land ownership, land cover, registered satellite and/or areal photography along with any other associated or derived geographic data. GIS processing becomes a critical step in hydrologic modelling since it contributes to generating model parameter distribution in spatial manner. In these applications, the GIS processing steps such as data storing, map overlaying, map analysis etc. have helped to derive hydrologic parameters from soil, land cover, rainfall maps etc. With respect to GIS processing products, Digital Elevation Models (DEM) are more important in rainfall runoff modelling. The development of DEM processing algorithms as well as relevant softwares to extract hydrologic information from DEM is increasing and makes it widely applied.
24
25
CHAPTER 5 CASE STUDY

DISTRIBUTED RAINFALL RUNOFF MODELING Published By: Dilip Kumar, Rajib Kumar Bhattacharjya Published in: International Journal of Earth Sciences and Engineering ISSN 0974-5904, Volume 04, No 06 SPL, October 2011, pp. 270-275 The present study develops a distributed approach to simulate the rainfall runoff process of a catchment. The catchment area has been divided in to the numbers of divisions equal to the numbers of rain gauge station. The rainfall in a particular rain gauge is considered as uniformly distributed over the entire sub catchments. Spatially distributed catchment characteristics have been obtained from the 90 m resolution SRTM digital elevation data. A lump model is also developed using average rainfall of the catchment. In case of lump model, average rainfall is calculated using thessian polygon method. In order to estimate runoff from rainfall events, loss rate or infiltration parameters for the basin have to be calculated, which is a basic input for further rainfall runoff modelling. The infiltration capacity of the basin depends on the land use and soil property. Hortons and Green-Ampt equations are most commonly used equations for estimation infiltration of a basin. Curve Number (CN) method is also a widely used method for estimating infiltration characteristics of the watershed, based on the land use property and soil property. Therefore the estimation of infiltration parameters or curve number of the basin is made initially. An inverse model is formulated and solved for estimating the curve numbers for the lump and distributed models. METHODOLOGY: HEC-HMS Model: HEC-HEC-HMS is hydrologic modelling software developed by the US Army Corps of Engineers Hydrologic Engineering Centre (HEC). It is designed to simulate the precipitation runoff processes of watershed systems in a wide range of geographic areas such as large river basins and small urban or natural watersheds. In HEC-HMS, the base flow
26
model is applied both at the start of simulation of a storm event, and later in the event as the delayed subsurface flow reaches the watershed channels. Three alternative models of base flow such as constant monthly varying value, exponential recession model, and linear reservoir volume accounting model are included. STUDY AREA AND DATA USED: Overview of study area: Considering the land and water problems and the availability of hydrological, meteorological, soil, and other collateral data, the Ranganadi watershed was selected as the study area for the present study, as shown in figure 2. The study area is located between 9402'34" E longitude and 2714'01" N latitude in the Brahmaputra River basin of India. It has an area of 1,920.68 km2 encircling five sub watershed, namely Yazali, Pingrove, Did, Mangio, Peprong. All these five are raingauge stations, which are considered as outlet location of sub watershed in the study. Again for this study, Ranganadi dam site was taken as the main outlet of the watershed which is located at 9344'28"E longitude and 2724'32"N latitude. Data acquisition: The data used in this study were (a) daily rainfall data of the five raingage stations (Yazali, Pingrove, Did, Mangio, and Peprong.) for the 3-year period (20062008) (b) daily discharge data of the watershed at main outlet for the 3- year period (2006 2008) (c) digital Elevation Model (DEM) of the Ranganadi River basin was acquired from the SRTM Site. Preparation of model inputs:
27
The rainfall records for five raingauge stations are available. These raingauge stations are Yazali, Pingrove, Did, Mangio, and Peprong. For the distributed model the rainfall records observed at a particular raingauge station is consider as uniformly distributed over the entire sub catchment. This distributed rainfall records are directly used in the model developed using distributing approach. Thesian polygon method is used for this purpose. Figure 3 shows the thesian polygons. For the lumped model average rainfall was calculated
Figure 2 Calculation of distributed rainfall pattern by thesian
using WMS (web map sevice). The SRTM digital elevation data is used to delineate the catchment watershed and generation of stream network. Figure 5 shows the DEM of study
28
area and water flow direction, which is calculated using TOPAZ. The watershed area has been further sub divided into the number of rain gauge station available in the watershed. There are five raingauge stations available in the Ranganadi catchment. The sub watersheds are shown in the figure 4. Basin processing module of WMS was used for the generation of background map file of the study area which in turn was used as an input to the HEC-HMS model (Fig.5). The other model input like CN of watershed is assumed for calibration purpose, as shown in table 1. Table 2 shows the basic model input which is described earlier.
29
Figure 3 Stream network of the study area (Distributed approach)
Calibration and Validation of the Models The successful application of a hydrologic watershed model depends on how well the model is calibrated, which in turn depends on the technical capability of hydrological model as well as the quality of input data. HEC-HMS watershed model were calibrated using daily rainfall data (Jan. to December) and stream flow data of 1.5 years (Jan.2006May2007). The objective of the model calibration was to match simulated volumes, peaks, and timing of hydrographs with the observed ones. For simulating stream flow by the HEC-HMS model, the SCS unit hydrograph transform method was used to compute direct surface runoff hydrographs, the SCS curve number loss method to compute runoff volumes, and the constant monthly method was used for base flow separation. Initial abstraction (Ia), SCS lag time, and Muskingum constant (K&X) were considered as HEC-HMS calibration parameter. These model parameters were estimated using the optimization algorithm available in HECHMS. After each parameter adjustment and corresponding simulation run, the simulated and observed stream flow hydrographs were visually compared.
30
RESULTS AND DISCUSSION: Calibration results of HEC-HMS: The rainfall runoff data recorded in the Ranganadi catchment have been used to calibrate and validate the developed model. The geomorphologic information of the catchment has been extracted from the SRTM digital elevation data. Table 3 shows the initial and optimized parameters of the lump approach. Similarly table 4 shows the parameters of the distributed approach. It is clear from these tables that the values of the calibrated parameters for the model vary from sub watershed to sub watershed. These parameters have been optimized using the optimization tools available in HEC-HMS, as discussed earlier. The variation in Ia values is attributed to the variation in antecedent moisture condition (AMC) over the years and the variation in SCS lag time is attributed to the varying observed stream flow over the years.
Performance evaluation using graphical indicators Visual checking of observed and simulated stream flow hydrographs, a comparison of the observed stream flow hydrograph with the simulated one by HEC-HMS as Lumped modelling approach as well as by Distributed modelling approach as shown in Fig.4-7. It is apparent from these figures that although there is a similar trend between the observed and simulated stream flow hydrographs, the peaks of the two hydrographs do not match reasonably at lean period of rainfall. As discussed earlier that an objective functions is a mathematical tool to measure the goodness of fit between the observed and generated hydrographs. To find the lowest objective function value and optimum parameter values are the main objectives behind our optimization trial. The univariate gradient method computes and adjusts one parameter at a time while locking the other parameters. Alternatively, the Nelder and Mead method evaluates all parameters simultaneously and determines which parameter to adjust. The
31
search algorithms are also known as optimization methods. The optimal objective function value is closed to zero.
Figure 4 Simulated Vs Observed Stream flow Hydrograph
Figure 5 Simulated Vs Observed Stream flow Hydrograph for lumped modelling approach during validation
32
Figure 6 Simulated Vs Observed Streamflow Hydrograph
Figure 7 Variation of objective function during distributed
Conclusions: Based on the analysis of the results obtained in this study, the following conclusions could be drawn:
33
1. Based on the statistical and graphical indicators used in this study, it was found that the HEC-HMS Distributed approach simulated daily stream flow is better than the Lumped simulated stream flow. 2. Although there is a reasonably good matching between observed and simulated stream flow hydrographs for both HEC-HMS Distributed and HEC-HMS Lumped modelling approach, the hydrographs do not match well for lean period of rainfall season. Overall, it is concluded that the HEC-HMS model is reliable for estimating infiltration parameters and for simulating daily stream flow in the Ranganadi River basin of North- Eastern India. Therefore, the use of HEC-HMS model may be used for future studies on hydrological modelling in this basin. It may also be noted that only three years of rainfall runoff data are used in the study. For modelling purpose these small duration data may not be suitable.
34
CHAPTER 6 REFERENCES
[1] Chow VT, Maidment DR, Mays LW (1988) Applied Hydrology. McGraw Hill, New York, USA [2] Subramanya, Engineering Hydrology [3] Keith J. Beven, Rainfall-Runoff Modelling: The Primer [4] David S. Bowles, P. Enda O'Connell, Recent Advances in the Modeling of Hydrologic Systems [5] Thorsten Wagener, Howard Wheater, Hoshin Vijai Gupta, Rainfall-Runoff Modelling In Gauged And Ungauged Catchments [6] Dilip Kumar and Rajib Kumar Bhattacharjya, Distributed Rainfall Runoff Modeling, October 2011, International Journal of Earth Sciences and Engineering, ISSN 0974-5904, Volume 04, No 06 SPL, pp. 270-275 [7] M. Ruslin Anwar, The rainfall-runoff model using of the watershed physical characteristic approach, December 2011, International Journal of Civil & Environmental Engineering IJCEE-IJENS Vol: 11 No: 06 [8] Jai Vaze, Phillip Jordan, Richard Beecham, Andrew Frost, Gregory Summerell, Guidelines for Rainfall-Runoff Modelling: Towards Best Practice Model Application, December 2011, eWater Cooperative Research Centre [9] Katarna Dubkov, Rainfall-runoff modelling: Its development, classification And possible applications, ACTA GEOGRAPHICA UNIVERSITATIS COMENIANAE, Vol. 54, 2010, No. 2, pp. 173-181 [10] Nguyen Hong Quan, Rainfall runoff modelling in the can le catchment, Saigaon river basin, March 2006, International institute for geoinformation science and earth observation, Enschede., Netherland.
35

Rainfall Runoff Modelling

Uploaded by

Copyright:

Available Formats

Rainfall Runoff Modelling

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Rainfall Runoff Modelling

Uploaded by

Copyright:

Available Formats

RAINFALL RUNOFF MODELLING