Overland Flow Model Research Articles

A hybrid‐3D hillslope hydrological model for use in Earth system models

AbstractHillslope‐scale rainfall‐runoff processes leading to a fast catchment response are not explicitly included in land surface models (LSMs) for use in earth system models (ESMs) due to computational constraints. This study presents a hybrid‐3D hillslope hydrological model (h3D) that couples a 1‐D vertical soil column model with a lateral pseudo‐2D saturated zone and overland flow model for use in ESMs. By representing vertical and lateral responses separately at different spatial resolutions, h3D is computationally efficient. The h3D model was first tested for three different hillslope planforms (uniform, convergent and divergent). We then compared h3D (with single and multiple soil columns) with a complex physically based 3‐D model and a simple 1‐D soil moisture model coupled with an unconfined aquifer (as typically used in LSMs). It is found that simulations obtained by the simple 1‐D model vary considerably from the complex 3‐D model and are not able to represent hillslope‐scale variations in the lateral flow response. In contrast, the single soil column h3D model shows a much better performance and saves computational time by 2‐3 orders of magnitude compared with the complex 3‐D model. When multiple vertical soil columns are implemented, the resulting hydrological responses (soil moisture, water table depth, and base flow along the hillslope) from h3D are nearly identical to those predicted by the complex 3‐D model, but still saves computational time. As such, the computational efficiency of the h3D model provides a valuable and promising approach to incorporating hillslope‐scale hydrological processes into continental and global‐scale ESMs.

Chance-constrained overland flow modeling for improving conceptual distributed hydrologic simulations based on scaling representation of sub-daily rainfall variability

Lack of hydrologic process representation at the short time-scale would lead to inadequate simulations in distributed hydrological modeling. Especially for complex mountainous watersheds, surface runoff simulations are significantly affected by the overland flow generation, which is closely related to the rainfall characteristics at a sub-time step. In this paper, the sub-daily variability of rainfall intensity was considered using a probability distribution, and a chance-constrained overland flow modeling approach was proposed to capture the generation of overland flow within conceptual distributed hydrologic simulations. The integrated modeling procedures were further demonstrated through a watershed of China Three Gorges Reservoir area, leading to an improved SLURP-TGR hydrologic model based on SLURP. Combined with rainfall thresholds determined to distinguish various magnitudes of daily rainfall totals, three levels of significance were simultaneously employed to examine the hydrologic-response simulation. Results showed that SLURP-TGR could enhance the model performance, and the deviation of runoff simulations was effectively controlled. However, rainfall thresholds were so crucial for reflecting the scaling effect of rainfall intensity that optimal levels of significance and rainfall threshold were 0.05 and 10mm, respectively. As for the Xiangxi River watershed, the main runoff contribution came from interflow of the fast store. Although slight differences of overland flow simulations between SLURP and SLURP-TGR were derived, SLURP-TGR was found to help improve the simulation of peak flows, and would improve the overall modeling efficiency through adjusting runoff component simulations. Consequently, the developed modeling approach favors efficient representation of hydrological processes and would be expected to have a potential for wide applications.

Testing the impact of at‐source stormwater management on urban flooding through a coupling of network and overland flow models

In many urban catchments, stormwater flooding is a serious problem. Low‐impact development and the use of stormwater control measures (SCMs) have the potential to mitigate such flooding, but this potential is highly context specific and remains largely untested. In this study, we tested the potential of SCMs to impact stormwater flooding in a peri‐urban catchment by coupling a one‐dimensional 1D stormwater drainage model with a two‐dimensional (2D) overland flow model. We predicted consequent flood dynamics for a range of storm events and management scenarios. We found that realistically extensive application of rain tanks and infiltration trenches is most effective in mitigating stormwater flooding for common‐to‐rare storm events [5‐ to 20‐year annual recurrence interval (ARI)] with short durations, such that the events have rainfall depths that are similar to the retention (or storage) capacity of commonly used small‐scale SCMs. Retention of stormwater also has the ability to mitigate overland flow intensity, positively impacting pedestrian safety and reducing potential building damage for rare events with ARIs above 20 years. In addition, SCMs also provide benefits to alternative options such as pipe upgrades, through their protection of receiving waters and enhancement of urban landscape amenity. WIREs Water 2015, 2:291–300. doi: 10.1002/wat2.1078This article is categorized under: Engineering Water > Sustainable Engineering of Water Science of Water > Water Extremes

A novel approach to model dynamic flow interactions between storm sewer system and overland surface for different land covers in urban areas

In this study, we developed a novel approach to simulate dynamic flow interactions between storm sewers and overland surface for different land covers in urban areas. The proposed approach couples the one-dimensional (1D) sewer flow model (SFM) and the two-dimensional (2D) overland flow model (OFM) with different techniques depending on the land cover type of the study areas. For roads, pavements, plazas, and so forth where rainfall becomes surface runoff before entering the sewer system, the rainfall–runoff process is simulated directly in the 2D OFM, and the runoff is drained to the sewer network via inlets, which is regarded as the input to 1D SFM. For green areas on which rainfall falls into the permeable ground surface and the generated direct runoff traverses terrain, the deduction rate is applied to the rainfall for reflecting the soil infiltration in the 2D OFM. For flat building roofs with drainage facilities allowing rainfall to drain directly from the roof to sewer networks, the rainfall–runoff process is simulated using the hydrological module in the 1D SFM where no rainfall is applied to these areas in the 2D OFM. The 1D SFM is used for hydraulic simulations in the sewer network. Where the flow in the drainage network exceeds its capacity, a surcharge occurs and water may spill onto the ground surface if the pressure head in a manhole exceeds the ground elevation. The overflow discharge from the sewer system is calculated by the 1D SFM and considered a point source in the 2D OFM. The overland flow will return into the sewer network when it reaches an inlet that connects to an un-surcharged manhole. In this case, the inlet is considered as a point sink in the 2D OFM and an inflow to a manhole in the 1D SFM. The proposed approach was compared to other five urban flood modelling techniques with four rainfall events that had previously recorded inundation areas. The merits and drawbacks of each modelling technique were compared and discussed. Based on the simulated results, the proposed approach was found to simulate floodings closer to the survey records than other approaches because the physical rainfall–runoff phenomena in urban environment were better reflected.

Comparative analysis of two infiltration models for application in a physically based overland flow model

In the prediction of overland flow, infiltration is an essential component, which should be modeled accurately to achieve optimum runoff rates. Many mathematical models that simulate the details of runoff and erosion process in hillslopes, where the rill-interrill configuration significantly affects overland flow, employ Horton’s model for infiltration due to its simplicity. However, Horton’s model does not handle adequately antecedent moisture condition (AMC) of soil. In this study, the Green-Ampt infiltration model is incorporated into a physically based overland flow model, which was originally coupled with Horton’s equation in an effort to improve the overland flow model’s prediction ability. In so doing, the model used the Horton and Green-Ampt model as an infiltration component separately and simulated flow to directly compare which infiltration equation performs better with the overland flow model. Calibration using laboratory data produced good results for both Horton with NSE = 0.88 and r2 = 0.92 and Green-Ampt with NSE = 0.90 and r2 = 0.95 while in validation, the Horton-coupled model produced lower NSE = 0.64 and r2 = 0.84 than the Green-Ampt which produced NSE = 0.85 and r2 = 0.85. The results suggest that the Green-Ampt equation can improve the performance of the overland flow model with its ability to account more accurately the AMC and flow processes in the soil.

A new modeling approach for simulating microtopography-dominated, discontinuous overland flow on infiltrating surfaces

Realistic modeling of discontinuous overland flow on irregular topographic surfaces has been proven to be a challenge. This study is aimed to develop a new modeling framework to simulate the discontinuous puddle-to-puddle (P2P) overland flow dynamics for infiltrating surfaces with various microtopographic characteristics. In the P2P model, puddles were integrated in a well-delineated, cascaded drainage system to facilitate explicit simulation of their dynamic behaviors and interactions. Overland flow and infiltration were respectively simulated by using the diffusion wave model and a modified Green–Ampt model for the DEM-derived flow drainage network that consisted of a series of puddle-based units (PBUs). The P2P model was tested by using a series of data from laboratory overland flow experiments for various microtopography, soil, and rainfall conditions. The modeling results indicated that the hierarchical relationships and microtopographic properties of puddles significantly affected their connectivity, filling–spilling dynamics, and the associated threshold flow. Surface microtopography and rainfall characteristics also exhibited strong influences on the spatio-temporal distributions of infiltration rates, runoff fluxes, and unsaturated flow. The model tests demonstrated its applicability in simulating microtopography-dominated overland flow on infiltrating surfaces.

Overland Flow Modeling with the Shallow Water Equations Using a Well-Balanced Numerical Scheme: Better Predictions or Just More Complexity

In the last decades, several physically based hydrological modeling approaches of various complexities have been developed that solve shallow water equations or their approximations using various numerical methods. Users of the model may not necessarily know the different hypotheses underlying these development and simplifications, and it might therefore be difficult to judge if a code is well adapted to their objectives and test case configurations. This paper aims to compare the predictive abilities of different models and evaluate potential gain by using an advanced numerical scheme for modeling runoff. Four different codes are presented, each based on either shallow water or kinematic wave equations, and using either the finite volume or finite difference method. These four numerical codes are compared with different test cases, allowing to emphasize their main strengths and weaknesses. Results show that, for relatively simple configurations, kinematic wave equations solved with the finite volume method represent an interesting option. Nevertheless, as it appears to be limited in case of discontinuous topography or strong spatial heterogeneities, for these cases they advise the use of shallow water equations solved with the finite volume method.

Numerical Modeling of Rainfall-Generated Overland Flow Using Nonlinear Shallow-Water Equations

AbstractThis paper presents a physics-based two-dimensional (2D) numerical model to simulate overland flows typically generated by rainfall of a storm event or multiple events in natural terrain with complex topography, landform, soil characteristics, and land use. The model is based on the 2D fully nonlinear shallow-water equations (SWEs) in which tempospatial variations of rainfall intensity and infiltration are taken into account as source and sink terms, respectively. The Green-Ampt equation is used to simulate infiltration. Due to strong nonlinearity of the coupled dynamic processes in overland flow and stormwater runoff, special efforts were made to solve this coupled-flow system under the conditions of unsteady rainfall intensity and natural terrain. To attain the modeling capabilities for multiple flow regimes including subcritical, supercritical, and the transitions, a second-order central-upwind shock-capturing scheme, which is well balanced and depth-positivity preserving, is used to solve the ...

Stability of an overland flow scheme in the framework of a fully coupled eco-hydrological model based on the Macroscopic Cellular Automata approach

Cellular Automata are often used for modeling the evolution in time of environmental systems mainly because they are directly compatible with parallel programming. Nevertheless, defining the optimal time step criterion for integrating forward in time numerical processes can further enhance model computational efficiency. To this aim, a numerical stability analysis of an original overland flow model, within the framework of a fully coupled eco-hydrological system based on the Macroscopic Cellular Automata paradigm, is performed. According to the other modules of the system describing soil water flow, soil-surface-atmosphere fluxes and vegetation dynamics, overland flow model equations were derived through a direct discrete formulation (i.e. no differential equations were discretized), adopting the diffusion wave model as an approximation of the full De Saint Venant equations and including the capability of accounting for specific processes, such as the increasing roughness effects due to vegetation growth or surface-soil water exchanges. Suitable formulations of robust tools usually applied in the stability analyses, such as Courant–Friedrichs–Lewy and von Neumann conditions, were initially derived for the CA-based overland flow model. Afterwards, the theoretical stability conditions were compared to experimental time step constraints through several numerical simulations of a 5-h rain event. Specifically, adopting a constant (i.e. not adaptive) time step for simulations, and discretizing head losses in a way that increases model stability, experimental upper limits preventing numerical instability were found for 13 test cases with different slopes, precipitation intensities, vegetation densities and depths of surface depressions. Even though von Neumann condition and experimental values were well positively correlated, the latter were almost always sensibly lower, excluding cases when free surface gradients tended to zero. Therefore, based on the original method, two alternative criteria were developed. Numerical tests showed that the joint use of these criteria greatly helps in finding the optimal time steps for convergent and stable simulations of the overland flow model.

An object-oriented overland flow solver for watershed flood inundation predictions: case study of Ulus basin, Turkey

Abstract This paper presents an object-oriented two-dimensional (2-D) overland flow model and its application in simulating flood flows over Ulus basin, located in the north of Turkey adjacent to the Black Sea. A new coding implementation according to the class environment created in object oriented C++ programming language is carried out in structuring and building the solver. The model is based on the Godunov type finite volume scheme on unstructured triangular meshes. A mass balance preserving wet/dry boundary solution algorithm is integrated in the numerical scheme to satisfy the positive-depth condition and minimize the numerical instability when treating the propagation of wave front in regions of dry bed. The balance between bed slope and flux terms is also preserved for still water conditions on irregular topography. The 2-D solver is verified by simulating selected dam break cases, where good agreement with measured data is achieved. For the simulation of flood flows in the Ulus basin, in general, the simulated outflow hydrograph is found to compare well with the recorded data. A selected inundation map that is extracted from the model results is also presented to show the water surface level in the Floodplain.

Experimental validation of a 2D overland flow model using high resolution water depth and velocity data

• A 2D overland flow model is validated using high resolution depth and velocity data. • Manning and Keulegan formulations produce similar and satisfactory results. • More complex formulations do not improve model performance. • The friction coefficient can be related to the micro and macro-rugosity of the bed. • The model might be used as the hydraulic base of a distributed soil erosion model. This paper presents a validation of a two-dimensional overland flow model using empirical laboratory data. Unlike previous publications in which model performance is evaluated as the ability to predict an outlet hydrograph, we use high resolution 2D water depth and velocity data to analyze to what degree the model is able to reproduce the spatial distribution of these variables. Several overland flow conditions over two impervious surfaces of the order of one square meter with different micro and macro-roughness characteristics are studied. The first surface is a simplified representation of a sinusoidal terrain with three crests and furrows, while the second one is a mould of a real agricultural seedbed terrain. We analyze four different bed friction parameterizations and we show that the performance of formulations which consider the transition between laminar, smooth turbulent and rough turbulent flow do not improve the results obtained with Manning or Keulegan formulas for rough turbulent flow. The simulations performed show that using Keulegan formula with a physically-based definition of the bed roughness coefficient, a two-dimensional shallow water model is able to reproduce satisfactorily the flow hydrodynamics. It is shown that, even if the resolution of the topography data and numerical mesh are high enough to include all the small scale features of the bed surface, the roughness coefficient must account for the macro-roughness characteristics of the terrain in order to correctly reproduce the flow hydrodynamics.

DEVELOPMENT OF FLOOD ROUTING MODELS FOR WANG RIVER BASIN

Nowadays, severe flooding frequently occurs in various parts of Thailand resulted from changes in climatic condition and land use patterns. The flooding has caused great damages to properties and lives and affects country economy. Experience from the most severe flooding in the northern and central regions of Thailand in the year 2011 reveals that reliable flood warning system is still lagging. For flood warning purpose, it is necessary to have an accurate flood routing system. This study is aimed at developing mathematical models for flood routing so as to provide data for flood warning. Two different models are developed, i.e., kinematic overland flow model and kinematic stream flow model. The finite element method with Galerkin’s weighted residual technique is used in model development. The second-order Runge-Kutta method is applied to solve the set of differential equations obtained from finite element formulation. The developed models are applied to simulate flows in the Wang river basin in the northern region of Thailand during July 1 – October 31, 2011 when severe flooding occurred in this region. Model calibration is made by adjusting some parameters in the models and comparing the obtained results with measured data recorded by RID at 5 stream flow gauge stations along the Wang river. For correlation analysis, three statistical indices are determined, these include coefficient of determination, R2 , Nash–Sutcliffe model efficiency coefficient, NSE, and coefficient of variation of the root mean square error, CV(RMSE). It is found that the model results at the upstream portion of the river satisfactorily agree with the observed data, with the values of R2 greater than 0.55 and CV(RMSE) less than 0.57. For the downstream portion of the river, there are remarkable differences between the model results and the observed data. The values of R2 are less than 0.35, CV(RMSE) greater than 0.76, and the NSE values are less than 0.16. This might be due to some errors in the input data, including rainfall pattern, topography, land use, river cross-sectional area, and water seepage along the river. More detailed field investigation and model calibration are still needed.

An All-Season Flash Flood Forecasting System for Real-Time Operations

Flash floods can cause extensive damage to both life and property, especially because they are difficult to predict. Flash flood prediction requires high-resolution meteorological observations and predictions, as well as calibrated hydrological models, which should effectively simulate how a catchment filters rainfall inputs into streamflow. Furthermore, because of the requirement of both hydrological and meteorological components in flash flood forecasting systems, there must be extensive data handling capabilities built in to force the hydrological model with a variety of available hydrometeorological data and predictions, as well as to test the model with hydrological observations. The authors have developed a working prototype of such a system, called KINEROS/hsB-SM, after the hydrological models that are used: the Kinematic Erosion and Runoff (KINEROS) and hillslope-storage Boussinesq Soil Moisture (hsB-SM) models. KINEROS is an event-based overland flow and channel routing model that is designed to simulate flash floods in semiarid regions where infiltration excess overland flow dominates, while hsB-SM is a continuous subsurface flow model, whose model physics are applicable in humid regions where saturation excess overland flow is most important. In addition, KINEROS/hsB-SM includes an energy balance snowmelt model, which gives it the ability to simulate flash floods that involve rain on snow. There are also extensive algorithms to incorporate high-resolution hydrometeorological data, including stage III radar data (5 min, 1° by 1 km), to assist in the calibration of the models, and to run the model in real time. The model is currently being used in an experimental fashion at the National Weather Service Binghamton, New York, Weather Forecast Office.

A web GIS based integrated flood assessment modeling tool for coastal urban watersheds

Urban flooding has become an increasingly important issue in many parts of the world. In this study, an integrated flood assessment model (IFAM) is presented for the coastal urban flood simulation. A web based GIS framework has been adopted to organize the spatial datasets for the study area considered and to run the model within this framework. The integrated flood model consists of a mass balance based 1-D overland flow model, 1-D finite element based channel flow model based on diffusion wave approximation and a quasi 2-D raster flood inundation model based on the continuity equation. The model code is written in MATLAB and the application is integrated within a web GIS server product viz: Web Gram Server™ (WGS), developed at IIT Bombay, using Java, JSP and JQuery technologies. Its user interface is developed using open layers and the attribute data are stored in MySQL open source DBMS. The model is integrated within WGS and is called via Java script. The application has been demonstrated for two coastal urban watersheds of Navi Mumbai, India. Simulated flood extents for extreme rainfall event of 26 July, 2005 in the two urban watersheds of Navi Mumbai city are presented and discussed. The study demonstrates the effectiveness of the flood simulation tool in a web GIS environment to facilitate data access and visualization of GIS datasets and simulation results.

Experimental study of the water depth and rainfall intensity effects on the bed roughness coefficient used in distributed urban drainage models

Summary The work presented in this paper analyses the effect of water depth and rainfall intensity on the surface roughness coefficients used in overland flow models based on the shallow water equations. The relation between the Manning coefficient and the water depth and rainfall intensity has been quantified using different methodologies based on the analysis of two sets of experimental data. In the first set uniform overland flow conditions were generated, and the bed roughness coefficient was computed from direct measurements of the water depth and discharge. In the second set of experiments, unsteady rainfall–runoff transformations with different rainfall intensities were generated in a flume and computed with a shallow water model in which different bed friction formulations were implemented and calibrated. Results show that for very low water depth values there is a significant increase in the surface resistance, which is not captured by any standard bed friction formulation. Rainfall intensity also increases surface resistance especially as the water depth diminishes below a critical threshold. Using a Reynolds dependent formulation for the Manning coefficient improves model predictions.

Evaluation of overland flow model for a hillslope using laboratory flume data

Comprehensive modelling of overland flow requires models for both rill and interrill area overland flow. Evaluation of a physically based mathematical model for simulating overland flow generated on rill and interrill areas of hillslope was done using a data set gathered from a laboratory experimental setup. A rainfall simulator has been constructed together with a 6.50 m × 1.36 m erosion flume that can be given adjustable slopes in both longitudinal and lateral directions. The model was calibrated and validated using the experimental results from the setup of the flume having 5% lateral and 10% longitudinal slopes where rainfall intensities of 105 and 45 mm/hr were induced with the use of nozzles. Results show that for the given slope combination, the model was capable of simulating the flow coming from the rill and interrill areas for the two different rainfall intensities. It was found that significantly more of the flow occurred in the form of the rill flow. The model studied here can be used for the better prediction of overland flow and can also be used as a building block for an associated erosion and sediment transport model.

A model for distributed GIUH-based flow routing on natural and anthropogenic hillslopes

Attempts to reduce the number of parameters in distributed rainfall–runoff models have not yet resulted in a model that is accurate for both natural and anthropogenic hillslopes. We take on the challenge by proposing a distributed model for overland flow and channel flow based on a combination of a linear response time distribution and the hillslope geomorphologic instantaneous unit hydrograph (GIUH), which can be calculated with only a digital elevation model and a map with field boundaries and channel network as input. The spatial domain is subdivided into representative elementary hillslopes (REHs) for each of which we define geometric and flow velocity parameters and compute the GIUH. The catchment GIUH is given by the sum of all REH responses. While most distributed models only perform well on natural hillslopes, the advantage of our approach is that it can also be applied to modified hillslopes with for example a rectangular drainage network and terrace cultivation. Tests show that the REH-GIUH approach performs better than classical routing functions (exponential and gamma). Simulations of four virtual hillslopes suggest that peak flow at the catchment outlet is directly related to drainage density. By combining the distributed flow routing model with a lumped-parameter infiltration model, we were also able to demonstrate that terrace cultivation delays the response time and reduces peak flow in comparison to the same hillslope, but with a natural stream network. The REH-GIUH approach is a first step in the process of coupling distributed hydrological models to erosion and water quality models at the REH (associated with agricultural management) and at the catchment scale (associated with the evaluation of the environmental impact of human activities). It furthermore provides a basis for the development of models for large catchments and urban or peri-urban catchments. Copyright © 2013 John Wiley & Sons, Ltd.

Self-Adaptive Kinematic-Dynamic Model for Overland Flow

AbstractThe authors developed an innovative two-dimensional finite-volume model for overland flow, which is capable of simultaneously applying the kinematic wave approximation and the full Saint-Venant (dynamic) equations to different regions of the modeled domain. A locally based criterion is employed to determine the applicability limits for the kinematic approximation. The model assesses the flow and topography on a cell-by-cell basis at each time step over an unstructured, triangular grid and determines the appropriate solution method for each individual cell. Several case studies are shown that illustrate the adaptive kinematic-dynamic model’s flexibility and robustness. The effects of the user-defined limit for the kinematic wave approximation and the abrupt switch from one solver to another are explored for the adaptive kinematic-dynamic model.

Evaluating manure release parameters for nonpoint contaminant transport model KINEROS2/STWIR

Release of manure components is an important element of modeling applications in environmental water quality. The scarcity of experimental data and the multiplicity of the approaches for modeling release kinetics of the manure components introduce uncertainty and reduce reliability of overland flow and contaminant transport models. The goal of this study was to estimate release parameters for different manure components and provide input for KINEROS2/STWIR model developed for pathogen risk assessment associated with livestock operations. The objectives of this work were to evaluate reliability and robustness of the manure release parameters estimated based on individual and grouped release kinetics of soluble, particulate and combination of particulate and soluble materials from surface applied manure. The parameters of Bradford–Schijven model were evaluated from the experimental data on release of chloride, water-extractable phosphate-P, total bioactive P, organic carbon, enterococci and E. coli from surface applied manure measured in the runoff-box and runoff-plot experiments. The results showed that release of different manure components from surface applied manure can be reliably predicted with just a single set of parameters characterizing the kinetics of manure mass release. We demonstrated that the manure release parameters could be estimated more reliably when the model fit was performed using data for different manure components pooled together, while the model fit to a single release curve produced correlated parameters. The model parameters appeared to be robust and transferable from the calibration to validation datasets without any or with only minor losses of the model accuracy.

Real-time rainfall monitoring and flood inundation forecasting for the city of Kolkata

Kolkata, the capital city of the Indian state of West Bengal, is prone to occasional flooding at times of moderate to heavy rainfall owing to relatively flat terrain and tide-dominated drainage outfalls. This article describes an instrumentation network that attempts to capture the distributed rainfall information in real time over the city with the help of an array of automatic rain gauges. Digital water level sensors, which monitor the elevations of the drainage channels, provide information about the water level in real time. Both rainfall and water level data, transmitted through GSM technology as SMS to a server at the Indian Institute of Technology (IIT) Kharagpur, get uploaded to a website continuously. Future plans include linking the rainfall and channel water level data with an overland flow and inundation numerical simulation model for predicting possible inundation scenarios in the city. Plans also propose to utilise the real-time data of the Kolkata Doppler Weather Radar for supplementing the ground rainfall data for producing longer lead-time forecasts.