FLOOD MODELLING WITH GIS-DERIVED DISTRIBUTED UNIT HYDROGRAPHS

The concept of a spatially distributed unit hydrograph is based on the fact that the unit hydrograph can be derived from the time–area curve of a watershed by the S-curve method. The time–area diagram is a graph of cumulative drainage area contributing to discharge at the watershed outlet within a specified time of travel. Accurate determination of the time–area diagram is made possible by using a GIS. The GIS is used to describe the connectivity of the links in the watershed flow network and to calculate distances and travel times to the watershed outlet for various points within the watershed. Overland flow travel times are calculated by the kinematic wave equation for time to equilibrium; channel flow times are based on the Manning and continuity equations. To account for channel storage, travel times for channel reaches are increased by a percentage depending on the channel reach length and geometry. With GIS capability for rainfall mapping, the assumption of a uniform spatial rainfall distribution is no longer necessary; hence the term, spatially distributed unit hydrograph. An example of the application for the Waiparous Creek in the Alberta Foothills is given. IDRISI is used to develop a simple digital elevation model of the 229 km² watershed, using 1 km × 1 km grid cells. A grid of flow directions is developed and used to create an equivalent channel network. Excess rainfall for each 1 km × 1 km cell is individually computed by the Soil Conservation Service (SCS) runoff curve method and routed through the equivalent channel network to obtain the time–area curve. The derived unit hydrograph gave excellent results in simulating an observed flood hydrograph. The distributed unit hydrograph is no longer a lumped model, since it accounts for internal distribution of rainfall and runoff. It is derived for a watershed without the need for observed rainfall and discharge data, because it is essentially a geomorphoclimatic approach. As such, it allows the derivation of watershed responses (hydrographs) to inputs of various magnitudes, thus eliminating the assumption of proportionality of input and output if needed. The superposition of outputs is retained in simulating flood hydrographs by convolution, since it has been shown that some non-linear systems satisfy the principle of superposition. The distributed unit hydrograph appears to be a very promising rainfall runoff model based on GIS technology.     

UPLAND EROSION MODELING WITH CASC2D-SED

1 INTRODUCTION Erosion and sedimentation embody the processes of detachment, transport, and deposition of soil particles. Erosion and subsequent deposition can cause major problems. Erosion reduces productivity of cropland, sediment degrades water quality and may carry soil adsorbed polluting chemicals. Deposition in irrigation canals, stream channels and reservoirs reduces structural capacity and requires costly removal. Ideally, an erosion model should describe the physical processes…     

A spatially distributed Clark’s unit hydrograph based hybrid hydrologic model (Distributed-Clark)

ABSTRACT

A hybrid hydrologic model (Distributed-Clark), which is a lumped conceptual and distributed feature model, was developed based on the combined concept of Clark’s unit hydrograph and its spatial decomposition methods, incorporating refined spatially variable flow dynamics to implement hydrological simulation for spatially distributed rainfall–runoff flow. In Distributed-Clark, the Soil Conservation Service (SCS) curve number method is utilized to estimate spatially distributed runoff depth and a set of separated unit hydrographs is used for runoff routing to obtain a direct runoff flow hydrograph. Case studies (four watersheds in the central part of the USA) using spatially distributed (Thiessen polygon-based) rainfall data of storm events were used to evaluate the model performance. Results demonstrate relatively good fit to observed streamflow, with a Nash-Sutcliffe efficiency (E_NS) of 0.84 and coefficient of determination (R²) of 0.86, as well as a better fit in comparison with outputs of spatially averaged rainfall data simulations for two models including HEC-HMS.     

Assessing impervious area ratios of grid-based land-use classifications on the example of an urban watershed

ABSTRACT

When applying a distributed hydrological model in urban watersheds, grid-based land-use classification data with 10 m resolution are typically used in Japan. For urban hydrological models, the estimation of the impervious area ratio (IAR) of each land-use classification is a crucial factor for accurate runoff analysis. In order to assess the IAR accurately, we created a set of vector-based “urban landscape GIS delineation” data for a typical urban watershed in Tokyo. By superimposing the vector-based delineation map on the grid-based map, the IAR of each grid-based land-use classification was estimated, after calculating the IARs of all grid cells in the entire urban watershed. As a result, we were able to calculate the frequency distribution of IAR for each land-use classification, as well as the spatial distribution of IARs for the urban watershed. It is evident from the results that the reference values of IAR for the land-use classifications were estimated very roughly and inherited errors of between about 7% and 70%, which corresponds to more than 100 mm increase of direct runoff for the 1500 mm annual average precipitation.

Editor D. Koutsoyiannis; Guest editor E. Volpi     

Integration of gauge and radar rainfall to enable best simulation of hydrological parameters

This study is about use of spatially distributed rain in physically based hydrological models. In recent years, spatially distributed radar rainfall data have become available. The distributed radar rain is used to precisely model hydrologic processes and it is more realistic than the past practice of distribution methods like Thiessen polygons. Radar provides a highly accurate spatial distribution of rainfall and greatly improves the basin average rainfall estimates. However, quantification of the exact amount of rainfall from radar observation is relatively difficult. Thus, the fundamental idea of this study is to apply hourly gauge and radar rainfall data in a distributed hydrological model to simulate hydrological parameters. Hence the comparison is made between the outcomes of the WetSpa model from radar rainfall distribution and gauge rainfall distributed by the Thiessen polygon technique. The comparative plots of the hydrograph and the results of hydrological components such as evapotranspiration, surface runoff, soil moisture, recharge and interflow, reflect the spatially distributed radar input performing well for model outflow simulation.

EDITOR D. Koutsoyiannis

ASSOCIATE EDITOR F. Pappenberger     

Physically-based distributed soil erosion and sediment yield model (DREAM) for simulating individual storm events

Abstract

A relatively simple process-oriented, physically-based distributed (PBD) hydrological model, the distributed runoff and erosion assessment model (DREAM), is described, and a validation study conducted in the semi-forested watershed of Pathri Rao, in the Garhwal Himalayas, India, is reported. DREAM takes account of watershed heterogeneity as reflected by land use, soil type, topography and rainfall, measured in the field or estimated through remote sensing, and generates estimates of runoff and sediment yield in spatial and temporal domains. The model is based on simultaneous solution of flow dynamics, based on kinematic wave theory, followed by solution of soil erosion dynamics. As the storm rainfall proceeds, the process of overland flow generation is dependent on the interception storage and infiltration rates. The components of the soil erosion model have been modified to provide better prediction of sediment flow rates and sediment yields. The validation study conducted to test the performance of the model in simulating soil erosion and sediment yield during different storm events monitored in the study watershed showed that the model outputs are satisfactory. Details of a sensitivity analysis, model calibration and the statistical evaluation of the results obtained are also presented and discussed. It is noteworthy that the distributed nature of the model combined with the use of geographical information system (GIS) techniques permits the computation and representation of the spatial distribution of sediment yield for simulated storm events, and a map of the spatial distribution of sediment yield for a simulated storm event is presented to highlight this capability.

Citation Ramsankaran, R., Kothyari, U.C., Ghosh, S.K., Malcherek, A., and Murugesan, K., 2013. Physically-based distributed soil erosion and sediment yield model (DREAM) for simulating individual storm events. Hydrological Sciences Journal, 58 (4), 872–891.     

Field investigation and modelling of coupled stream discharge and shallow water‐table dynamics in a small Mediterranean catchment (Sardinia)

In the semi‐arid Mediterranean environment, the rainfall–runoff relationships are complex because of the markedly irregular patterns in rainfall, the seasonal mismatch between evaporation and rainfall, and the spatial heterogeneity in landscape properties. Watersheds often display considerable non‐linear threshold behavior, which still make runoff generation an open research question. Our objectives in this context were: to identify the primary processes of runoff generation in a small natural catchment; to test whether a physically based model, which takes into consideration only the primary processes, is able to predict spatially distributed water‐table and stream discharge dynamics; and to use the hydrological model to increase our understanding of runoff generation mechanisms. The observed seasonal dynamics of soil moisture, water‐table depth, and stream discharge indicated that Hortonian overland‐flow was negligible and the main mechanism of runoff generation was saturated subsurface‐flow. This gives rise to base‐flow, controls the formation of the saturated areas, and contributes to storm‐flow together with saturation overland‐flow. The distributed model, with a 1D scheme for the kinematic surface‐flow, a 2D sub‐horizontal scheme for the saturated subsurface‐flow, and ignoring the unsaturated flow, performed efficiently in years when runoff volume was high and medium, although there was a smoothing effect on the observed water‐table. In dry years, small errors greatly reduced the efficiency of the model. The hydrological model has allowed to relate the runoff generation mechanisms with the land‐use. The forested hillslopes, where the calibrated soil conductivity was high, were never saturated, except at the foot of the slopes, where exfiltration of saturated subsurface‐flow contributed to storm‐flow. Saturation overland‐flow was only found near the streams, except when there were storm‐flow peaks, when it also occurred on hillslopes used for pasture, where soil conductivity was low. The bedrock–soil percolation, simulated by a threshold mechanism, further increased the non‐linearity of the rainfall–runoff processes. Copyright © 2013 John Wiley & Sons, Ltd.     

Reservoir rainfall–runoff geomorphological model. I: application and parameter analysis

This article describes and formulates a model designed to simulate runoff in wet weather events, called reservoir rainfall–runoff geomorphological model (R³GeM). In these wetlands, soil saturation is the main mechanism for the generation of surface runoff. To determine the saturated areas, the model applies a relationship based on the topographic index, between watershed storage and saturated surface. Precipitation is separated into surface runoff by saturation, subsurface runoff and losses; then, the flow of surface and subsurface runoff is performed. This hydrological model has five parameters and has been implemented in 37 events in Aixola watershed and 15 in Oiartzun watershed, both located on the Cantabrian coast of Spain. We analysed the influence of these five parameters in their behaviour, and we have proven, noting the efficiency gains, that the proposed model is valid to simulate the rainfall–runoff process. Copyright © 2012 John Wiley & Sons, Ltd.     

大尺度水文模型TOPX构建及其与区域环境系统集成模式RIEMS的耦合

基于改进型SIMTOP参数化径流方案和新安江模型的三层土壤水量平衡计算方法,本文构建了一个输入数据和率定参数较少、同时具有地形指数尺度转换机制、较好描述二维水文过程的简单高效的大尺度水文模型TOPX,并将其与区域环境系统集成模式RIEMS紧密耦合,以增强区域气候模式对大尺度流域径流量的定量数值模拟能力．TOPX模型在酉水河流域和泾河流域的离线测试表明：该模型对小尺度流域的径流量模拟精度较高,能够较好地描述流域水文变化过程;同时,该模型在大尺度上具有较强的分布式模拟能力,能够捕捉陆面水文过程的主要特征和时空演变特点．TOPX与RIEMS的耦合模式在泾河流域进行了在线测试,借助TOPX模型中的地形指数降尺度转换和水文过程产汇流机制,耦合模式实现了利用区域气候模式模拟的气象资料来驱动水文模型进行大尺度流域日径流量的模拟．进一步分析还表明：区域气候模式RIEMS模拟的降水时空分布数据的精度是影响耦合模式对径流量模拟效果的关键因素．     

Multi‐site calibration using a grid‐based event rainfall–runoff model: a case study of the upstream areas of the Nakdong River basin in Korea

Most runoff analyses using a grid‐based distributed model use one parameter group calibrated at the outlet of a watershed, instead of dividing the watershed into subwatersheds. Significant differences between the observed value and the simulation result of the subwatersheds can occur if just one parameter group is used in all subwatersheds that have different hydrological characteristics from each other. Therefore, to improve the simulation results of the subwatersheds within a watershed, a model calibrated at every subwatershed needs to be used to reflect the characteristics of each subwatershed. In this study, different parameter groups were set up for one or two sites using a distributed model, the GRM (Grid based Rainfall‐runoff Model), and the evaluations were based on the results of rainfall–runoff analysis, which uses a multi‐site calibration (MSC) technique to calibrate the model at the outlet of each site. The Hyangseok watershed in Naeseong River, which is a tributary of Nakdong River in Korea, was chosen as the study area. The watershed was divided into five subwatersheds each with a subwatershed outlet that was applied to the calibration sites . The MSC was applied for five cases. When a site was added for calibration in a watershed, the runoff simulation showed better results than the calibration of only one site at the most downstream area of the watershed. The MSC approach could improve the simulation results on the calibrated sites and even on the non‐calibrated sites, and the effect of MSC was improved when the calibrated site was closer to the runoff site. Copyright © 2014 John Wiley & Sons, Ltd.     

Effect of moving storms on attainment of equilibrium discharge

Although rainfall is assumed spatially uniform in conventional hydrological modelling for rainfall–runoff simulations, moving storms have been shown to have substantial influence on flow hydrographs. In this study, criteria for attainment of the equilibrium discharge from watersheds subjected to moving storms were examined. Non-linear numerical kinematic-wave models were developed to simulate runoff from an overland plane and from a V-shaped catchment. Dimensional analysis was applied to obtain the independent variables to be used as control factors in performing a series of numerical tests. The results indicate that, for storms moving downstream, runoff can attain equilibrium discharge even though the storm length is shorter than the watershed length and the rainfall duration is less than the time to equilibrium of the watershed for stationary uniform storms. The phenomenon of attainment of equilibrium discharge from watersheds subjected to moving storms is contradictory to conventional hydrologic design, which assumes the storm duration must equal the time to equilibrium to attain the maximum discharge. Copyright © 2007 John Wiley & Sons, Ltd.     

An analysis of the impact of spatial variability in rainfall on runoff and sediment predictions from a distributed model

This paper investigates the effect of introducing spatially varying rainfall fields to a hydrological model simulating runoff and erosion. Pairs of model simulations were run using either spatially uniform (i.e. spatially averaged) or spatially varying rainfall fields on a 500‐m grid. The hydrological model used was a simplified version of Thales which enabled runoff generation processes to be isolated from hillslope averaging processes. Both saturation excess and infiltration excess generation mechanisms were considered, as simplifications of actual hillslope processes. A 5‐year average recurrence interval synthetic rainfall event typical of temperate climates (Melbourne, Australia) was used. The erosion model was based on the WEPP interrill equation, modified to allow nonlinear terms relating the erosion rate to rainfall or runoff‐squared. The model results were extracted at different scales to investigate whether the effects of spatially varying rainfall were scale dependent. A series of statistical metrics were developed to assess the variability due to introducing the spatially varying rainfall field. At the catchment (approximately 150 km²) scale, it was found that particularly for saturation excess runoff, model predictions of runoff were insensitive to the spatial resolution of the rainfall data. Generally, erosion processes at smaller sub‐catchment scales, particularly when the sediment generation equation had non linearity, were more sensitive to spatial rainfall variability. Introducing runon infiltration reduced the total runoff and sediment yield at all scales, and this process was also most sensitive to the rainfall resolution. Copyright © 2011 John Wiley & Sons, Ltd.     

Scenario-based assessment and management of water resources in mid-Himalayan micro-watersheds with limited data availability

Abstract

A conceptual hydrological water balance model has been developed and employed to assess the water availability in a mid-Himalayan watershed and analyse the dynamics of water supply and its utilization under different scenarios arising out of present and future developments. The model was calibrated and validated using daily runoff and rainfall data at different locations in the watershed. The water availability in streams and springs to meet the human, animal and crop requirements was assessed, and was found to vary in different quarters of the year with inequalities existing in different parts of the watershed. The model was successfully applied to analyse the impact of land-use changes and weather aberrations on water availability in the present and future scenarios. The drought scenarios are more critical in causing water scarcity in a given location, compared to the impact of land-use changes. The findings can be applied for assessing, planning and allocation of water resources among different sectors of water use in hilly areas and to make informed decisions during critical periods of water scarcity.     

Hydrological modelling of a small watershed using generated rainfall in the soil and water assessment tool model

An adequately tested soil and water assessment tool (SWAT) model was applied to the runoff and sediment yield of a small agricultural watershed in eastern India using generated rainfall. The capability of the model for generating rainfall was evaluated for a period of 18 years (1981–1998). The watershed and subwatershed boundaries, drainage networks, slope, soil series and texture maps were generated using a geographical information system (GIS). A supervised classification method was used for land‐use/cover classification from satellite imageries. Model simulated monthly rainfall for the period of 18 years was compared with observations. Simulated monthly rainfall, runoff and sediment yield values for the monsoon season of 8 years (1991–1998) were also compared with their observed values. In general monthly average rainfall predicted by the model was in close agreement with the observed monthly average values. Also, simulated monthly average values of surface runoff and sediment yield using generated rainfall compared well with observed values during the monsoon season of the years 1991–1998. Results of this study revealed that the SWAT model can generate monthly average rainfall satisfactorily and thereby can produce monthly average values of surface runoff and sediment yield close to the observed values. Therefore, it can be concluded that the SWAT model could be used for developing a multiple year management plan for the critical erosion prone areas of a small watershed. Copyright © 2004 John Wiley & Sons, Ltd.     

Geospatially based distributed rainfall‐runoff modelling for simulation of internal and outlet responses in a semi‐forested lower Himalayan watershed

This study presents a Geographic Information System (GIS)‐based distributed rainfall‐runoff model for simulating surface flows in small to large watersheds during isolated storm events. The model takes into account the amount of interception storage to be filled using a modified Merriam ( 1960 ) approach before estimating infiltration by the Smith and Parlange ( 1978 ) method. The mechanics of overland and channel flow are modelled by the kinematic wave approximation of the Saint Venant equations which are then numerically solved by the weighted four‐point implicit finite difference method. In this modelling the watershed was discretized into overland planes and channels using the algorithms proposed by Garbrecht and Martz ( 1999 ). The model code was first validated by comparing the model output with an analytical solution for a hypothetical plane. Then the model was tested in a medium‐sized semi‐forested watershed of Pathri Rao located in the Shivalik ranges of the Garhwal Himalayas, India. Initially, a local sensitivity analysis was performed to identify the parameters to which the model outputs like runoff volume, peak flow and time to peak flow are sensitive. Before going for model validation, calibration was performed using the Ordered‐Physics‐based Parameter Adjustment (OPPA) method. The proposed Physically Based Distributed (PBD) model was then evaluated both at the watershed outlet as well as at the internal gauging station, making this study a first of its kind in Indian watersheds. The results of performance evaluation indicate that the model has simulated the runoff hydrographs reasonably well within the watershed as well as at the watershed outlet with the same set of calibrated parameters. The model also simulates, realistically, the temporal variation of the spatial distribution of runoff over the watershed and the same has been illustrated graphically. Copyright © 2011 John Wiley & Sons, Ltd.     

Estimating effective impervious area in urban watersheds using land cover,soil character and asymptotic curve number

Knowledge of the effective impervious area (EIA) or the degree to which impervious surfaces are hydraulically connected to the drainage system is useful for improving hydrological and environmental models and assessing the effectiveness of green stormwater infrastructure in urban watersheds. The goal of this research is to develop a method to estimate EIA fraction in urban watersheds using readily available data. Since EIA is dependent on rainfall–runoff response and cannot be solely determined based on the physical characteristics of a watershed, the EIA is linked with the asymptotic curve number (CN), a watershed index that represents runoff characteristics. In order for the method to be applicable to ungauged watersheds, the asymptotic CN is predicted using land cover and soil data from 35 urban catchments in Minnesota and Texas, USA. Similar data from 11 other urban catchments in Wisconsin and Texas, USA, are used to validate the results. A set of runoff depth versus EIA fraction curves is also developed to assess the impact of EIA reduction on discharge from an urban watershed in land-use planning studies.     

SWAT-based hydrological modelling of tropical land-use scenarios

Abstract

The Hulu Langat basin, a strategic watershed in Malaysia, has in recent decades been exposed to extensive changes in land-use and consequently hydrological conditions. In this work, the impact of Land Use and Cover Change (LUCC) on hydrological conditions (water discharge and sediment load) of the basin were investigated using the Soil and Water Assessment Tool (SWAT). Four land-use scenarios were defined for land-use change impact analysis, i.e. past, present (baseline), future and water conservation planning. The land-use maps, dated 1984, 1990, 1997 and 2002, were defined as the past scenarios for LUCC impact analysis. The present scenario was defined based on the 2006 land-use map. The 2020 land-use map was simulated using a cellular automata-Markov model and defined as the future scenario. Water conservation scenarios were produced based on guidelines published by Malaysia’s Department of Town and Country Planning and Department of Environment. Model calibration and uncertainty analysis was performed using the Sequential Uncertainty Fitting (SUFI-2) algorithm. The model robustness for water discharge simulation for the period 1997–2008 was good. However, due to uncertainties, mainly resulting from intense urban development in the basin, its robustness for sediment load simulation was only acceptable for the calibration period 1997–2004. The optimized model was run using different land-use maps over the periods 1997–2008 and 1997–2004 for water discharge and sediment load estimation, respectively. In comparison to the baseline scenario, SWAT simulation using the past and conservative scenarios showed significant reduction in monthly direct runoff and monthly sediment load, while SWAT simulation based on the future scenario showed significant increase in monthly direct runoff, monthly sediment load and groundwater recharge.

Editor D. Koutsoyiannis; Associate editor C. Perrin     

Identification of physical processes inherent in artificial neural network rainfall runoff models

The emergence of artificial neural network (ANN) technology has provided many promising results in the field of hydrology and water resources simulation. However, one of the major criticisms of ANN hydrologic models is that they do not consider/explain the underlying physical processes in a watershed, resulting in them being labelled as black‐box models. This paper discusses a research study conducted in order to examine whether or not the physical processes in a watershed are inherent in a trained ANN rainfall‐runoff model. The investigation is based on analysing definite statistical measures of strength of relationship between the disintegrated hidden neuron responses of an ANN model and its input variables, as well as various deterministic components of a conceptual rainfall‐runoff model. The approach is illustrated by presenting a case study for the Kentucky River watershed. The results suggest that the distributed structure of the ANN is able to capture certain physical behaviour of the rainfall‐runoff process. The results demonstrate that the hidden neurons in the ANN rainfall‐runoff model approximate various components of the hydrologic system, such as infiltration, base flow, and delayed and quick surface flow, etc., and represent the rising limb and different portions of the falling limb of a flow hydrograph. Copyright © 2004 John Wiley & Sons, Ltd.     

Spatially distributed tracer‐aided modelling to explore water and isotope transport,storage and mixing in a pristine,humid tropical catchment

Rapidly transforming headwater catchments in the humid tropics provide important resources for drinking water, irrigation, hydropower, and ecosystem connectivity. However, such resources for downstream use remain unstudied. To improve understanding of the behaviour and influence of pristine rainforests on water and tracer fluxes, we adapted the relatively parsimonious, spatially distributed tracer‐aided rainfall–runoff (STARR) model using event‐based stable isotope data for the 3.2‐km² San Lorencito catchment in Costa Rica. STARR was used to simulate rainforest interception of water and stable isotopes, which showed a significant isotopic enrichment in throughfall compared with gross rainfall. Acceptable concurrent simulations of discharge (Kling–Gupta efficiency [KGE] ~0.8) and stable isotopes in stream water (KGE ~0.6) at high spatial (10 m) and temporal (hourly) resolution indicated a rapidly responding system. Around 90% of average annual streamflow (2,099 mm) was composed of quick, near‐surface runoff components, whereas only ~10% originated from groundwater in deeper layers. Simulated actual evapotranspiration (ET) from interception and soil storage were low (~420 mm/year) due to high relative humidity (average 96%) and cloud cover limiting radiation inputs. Modelling suggested a highly variable groundwater storage (~10 to 500 mm) in this steep, fractured volcanic catchment that sustains dry season baseflows. This groundwater is concentrated in riparian areas as an alluvial–colluvial aquifer connected to the stream. This was supported by rainfall–runoff isotope simulations, showing a “flashy” stream response to rainfall with only a moderate damping effect and a constant isotope signature from deeper groundwater (~400‐mm additional mixing volume) during baseflow. The work serves as a first attempt to apply a spatially distributed tracer‐aided model to a tropical rainforest environment exploring the hydrological functioning of a steep, fractured‐volcanic catchment. We also highlight limitations and propose a roadmap for future data collection and spatially distributed tracer‐aided model development in tropical headwater catchments.     

Influences of spatially heterogeneous roughness on flow hydrographs

Spatially uniform roughness, which although may not be adequate in all cases, is often conveniently adopted for rainfall-runoff simulations in conventional watershed hydrology. In this study, effects of spatially heterogeneous roughness on hydrological response were investigated systematically. Three types of surface roughness scenarios, including: roughness decreasing in a downstream direction (Scenario 1), roughness increasing in a downstream direction (Scenario 2) and roughness distributed at random (Scenario 3), were assigned to an overland plane. A non-inertia wave model was developed to generate hydrographs at the end of the overland plane for certain rainstorms. The hydrographs in the three scenarios showed that when rainfall duration was less than the time of concentration of the overland plane, peak discharges generated from the spatially heterogeneous roughness scenarios significantly exceeded those from the spatially uniform roughness condition. The temporal equilibrium discharge which was usually observed on an overland plane under the spatially uniform roughness condition was not found in the present study. The flow peak in Scenario 1 occurred before that in Scenario 2. Nevertheless, the runoff peak could occur at the preceding, the middle or the later part of the hydrograph in Scenario 3, and this depended on the spatial distribution of the roughness. Reasons for the variation of peak discharge and the time to peak discharge were analyzed in detail based on water depth and flow velocity. Spatially heterogeneous roughness had an obviously influence on runoff generation, which ought to be handled with care in hydrological simulations.