Examining Modeling Approaches for the Rainfall‐Runoff Process in Wildfire‐Affected Watersheds: Using San Dimas Experimental Forest

Wildfire can significantly change watershed hydrological processes resulting in increased risks for flooding, erosion, and debris flow. The goal of this study was to evaluate the predictive capability of hydrological models in estimating post‐fire runoff using data from the San Dimas Experimental Forest (SDEF), San Dimas, California. Four methods were chosen representing different types of post‐fire runoff prediction methods, including a Rule of Thumb, Modified Rational Method (MODRAT), HEC‐HMS Curve Number, and KINematic Runoff and EROSion Model 2 (KINEROS2). Results showed that simple, empirical peak flow models performed acceptably if calibrated correctly. However, these models do not reflect hydrological mechanisms and may not be applicable for predictions outside the area where they were calibrated. For pre‐fire conditions, the Curve Number approach implemented in HEC‐HMS provided more accurate results than KINEROS2, whereas for post‐fire conditions, the opposite was observed. Such a trend may imply fundamental changes from pre‐ to post‐fire hydrology. Analysis suggests that the runoff generation mechanism in the watershed may have temporarily changed due to fire effects from saturation‐excess runoff or subsurface storm dominated complex mechanisms to an infiltration‐excess dominated mechanism. Infiltration modeling using the Hydrus‐1D model supports this inference. Results of this study indicate that physically‐based approaches may better reflect this trend and have the potential to provide consistent and satisfactory prediction.     

A GIS‐ASSISTED DISTRIBUTED WATERSHED MODEL FOR SIMULATING FLOODING AND INUNDATION1

ABSTRACT: A distributed watershed model combining kinematic wave routing, 1‐D dynamic channel‐flow routing, and 2‐D diffusive overland‐flow routing has been developed to simulate flooding and inundation levels of large watersheds. The study watershed was linked to a GIS database and was divided into an upstream mountainous area and a downstream alluvial plain. A kinematic wave routing was adopted at the mountainous area to compute the discharge flowing into the alluvial plain. A 1‐D dynamic channel routing solving the St. Venant equations by the Preissmann method was performed for the main channel of the alluvial plain, whereas a 2‐D overland‐flow routing solving the diffusion wave equation with the Alternating Direction Explicit scheme was used for floodplains. The above two routings were connected by weir‐link discharge formula. The parameters in the model were calibrated and independently verified by single‐event storms. An example application of flooding/inundation analysis was conducted for the Taichung station and the Woozi depot (Taiwan High Speed Rail). Suggested inundation‐proofing measures ‐ including raising ground surface elevation of the station and depot and building a waterproofing exterior wall and their combination ‐ were investigated. It was concluded that building the waterproofing exterior wall had a strong tendency to decrease peak inundation depth.     

The Comparative Accuracy of Two Hydrologic Models in Simulating Warm‐Season Runoff for Two Small,Hillslope Catchments

This article assesses the performance of two hydrologic models in simulating warm‐season runoff for two upland, low‐yield micro‐catchments near Coshocton, Ohio. The two models, namely the Storm Water Management Model (SWMM) and the Gridded Surface‐Subsurface Hydrologic Analysis (GSSHA), were implemented with contrasting levels of complexity, with the former representing the catchments as lumped spatial units and computing evaporation only from standing water, and the latter incorporating fine‐scale variation in topography and soil properties and computing evapotranspiration from soil based on weather data. Our investigation began with uncalibrated model runs for 1990‐2003 except for 1994 using a priori parameter values. Then a set of calibration experiments were performed wherein the sensitivity of model performance to the length of calibration records was examined. Our results pointed to large errors associated with simulations from both models: even the calibrated models were unable to reproduce the seasonal and between‐catchment contrasts in runoff response. Using a priori parameter values, SWMM attained better results than GSSHA. However, with simple calibration, GSSHA outperformed SWMM in several respects. It was also found that extending the record of calibration rendered relatively minor changes to model performance. The practical and scientific implications of the findings are discussed.     

GIS‐Based Predictive Models of Hillslope Runoff Generation Processes1

Abstract: Successful nonpoint source pollution control using best management practice placement is a complex process that requires in‐depth knowledge of the locations of runoff source areas in a watershed. Currently, very few simulation tools are capable of identifying critical runoff source areas on hillslopes and those available are not directly applicable under all runoff conditions. In this paper, a comparison of two geographic information system (GIS)‐based approaches: a topographic index model and a likelihood indicator model is presented, in predicting likely locations of saturation excess and infiltration excess runoff source areas in a hillslope of the Savoy Experimental Watershed located in northwest Arkansas. Based on intensive data collected from a two‐year field study, the spatial distributions of hydrologic variables were processed using GIS software to develop the models. The likelihood indicator model was used to produce probability surfaces that indicated the likelihood of location of both saturation and infiltration excess runoff mechanisms on the hillslope. Overall accuracies of the likelihood indicator model predictions varied between 81 and 87% for the infiltration excess and saturation excess runoff locations respectively. On the basis of accuracy of prediction, the likelihood indicator models were found to be superior (accuracy 81‐87%) to the predications made by the topographic index model (accuracy 69.5%). By combining statistics with GIS, runoff source areas on a hillslope can be identified by incorporating easily determined hydrologic measurements (such as bulk density, porosity, slope, depth to bed rock, depth to water table) and could serve as a watershed management tool for identifying critical runoff source areas in locations where the topographic index or other similar methods do not provide reliable results.     

Rainfall Runoff Relationships for a Cloud Forest Watershed in Central America: Implications for Water Resource Engineering1

Caballero, Luis A., Alon Rimmer, Zachary M. Easton, and Tammo S. Steenhuis, 2012. Rainfall Runoff Relationships for a Cloud Forest Watershed in Central America: Implications for Water Resource Engineering. Journal of the American Water Resources Association (JAWRA) 48(5): 1022‐1031. DOI: 10.1111/j.1752‐1688.2012.00668.x Abstract: Understanding the basic relationships between rainfall and runoff is vital for effective management and utilization of scarce water resources. Especially, this is important in Central America with widespread potable water shortage during the dry months of the monsoon. Potential good water sources are cloud forests, but little information concerning its potential is available to water supply engineers. Our objective is to define rainfall‐runoff‐base flow relationships for a cloud forest catchment. Flumes were installed for measuring river flow in four subwatersheds in La Tigra National Park, Honduras. One of the four watersheds was a 636‐ha subwatershed (WS1) with 60% cloud forest coverage. Precipitation averaged 1,130 mm/yr over the entire basin. About half of the total rainfall became runoff for the cloud forest watershed whereas, for the adjacent undisturbed forested watershed, the total discharge was <20% of the amount of precipitation. Infiltration rates were generally greater than rainfall rates. Therefore, most rainfall infiltrated into the soil, especially in the upper, steep, and well‐drained portions of the watershed. Direct runoff was generated from saturated areas near the river and exposed bedrock. This research provides compelling evidence that base flow is the primary contributor to streamflow during both wet and dry seasons in cloud forest catchments. Protecting these flow processes over time is critical for the sustained provision of potable water.     

Incorporating Variable Source Area Hydrology into a Spatially Distributed Direct Runoff Model1

Buchanan, Brian, Zachary M. Easton, Rebecca Schneider, and M. Todd Walter, 2011. Incorporating Variable Source Area Hydrology Into a Spatially Distributed Direct Runoff Model. Journal of the American Water Resources Association (JAWRA) 48(1): 43‐60. DOI: 10.1111/j.1752‐1688.2011.00594.x Abstract: Few hydrologic models simulate both variable source area (VSA) hydrology, and runoff‐routing at high enough spatial resolutions to capture fine‐scale hydrologic pathways connecting VSA to the stream network. This paper describes a geographic information system‐based operational model that simulates the spatio‐temporal dynamics of VSA runoff generation and distributed runoff‐routing, including through complex artificial drainage networks. The model combines the Natural Resource Conservation Service’s Curve Number (CN) equation for estimating storm runoff with the topographic index concept for predicting the locations of VSA and a runoff‐routing algorithm into a new spatially distributed direct hydrograph (SDDH) model (SDDH‐VSA). Using a small agricultural watershed in central New York, SDDH‐VSA results were compared to those from a SDDH model using the traditional land use assumptions for the CN (SDDH‐CN). The SDDH‐VSA model generally agreed better with observed discharge than the SDDH‐CN model (average, Nash‐Sutcliffe efficiency of 0.69 vs. 0.58, respectively) and resulted in more realistic spatial patterns of runoff‐generating areas. The SDDH approach did not correctly capture the timing of runoff from small storms in dry periods. Despite this type of limitation, SDDH‐VSA extends the applicability of the SDDH technique to VSA conditions, providing a basis for new tools to help identify critical management areas and assess water quality risks due to landscape alterations.     

OPTIMUM USE OF PADDY FIELDS FOR FLOOD MITIGATION1

In Taiwan, the continuously increasing levels of rice imports are likely to result in surplus paddy fields. Hence, the surplus paddy fields may be developed into wetlands to increase ground water recharge, provide appropriate environments for wildlife, and most importantly, store flood water. This study developed a hydrological model incorporating the distributed rainfall‐runoff model based on the kinematics wave approach and the distributed tank model for simulation, respectively, in mountainous and flat areas. The hydrological model was found to simulate the rainfall‐runoff behavior well in the study area. Furthermore, a decision method based on the genetic algorithm concepts was proposed to give policy makers the optimal location and area size of paddy fields to construct wetlands for flood mitigation.     

Simulation of Surface Water for Un‐Gauged Areas with Storage‐Attenuation Wetlands1

Abstract: A nontraditional application of the Hydrological Simulation Program – FORTRAN (HSPF) model to simulate freshwater discharge to upper Charlotte Harbor along Florida’s west coast was performed. This application was different from traditional HSPF applications in three ways. First, the domain of the model was defined based on the hydraulic characteristics of the landforms using small distributed parameter discretization. Second, broad wetland land forms, representing more than 20% of this area, were simulated as reaches with storage‐attenuation characteristics and not as pervious land segments (PERLNDs). Finally, the reach flow‐tables (F‐Tables) were configured in a unique way to be calibrated representing the uncertainty of the storage‐attenuation process. Characterizing wetlands as hydrography elements allows flow from the wetlands to be treated as a stage‐dependent flux. The study was conducted for the un‐gauged portion of the Peace and Myakka rivers in west‐central Florida. Due to low gradient tidal influences, a large portion of the basin is un‐gauged. The objective of this study was to simulate stream flow discharges and to estimate freshwater inflow from these un‐gauged areas to upper Charlotte Harbor. Two local gauging stations were located within the model domain and were used for calibration. Another gauge with a shorter period of record was used for verification. A set of global hydrologic parameters were selected and tested using the parameter optimization software (PEST) during the calibration. Model results were evaluated using PEST and well‐known statistical indices. The correlation coefficients were very high (0.899 and 0.825) for the two calibration stations. Further testing of this approach appears warranted for watersheds with significant wetlands coverage.     

Generalized Nondimensional Depth‐Discharge Rating Curves Tested on Florida Streamflow1

Abstract: A discharge rating is a relationship between stage and discharge at a specific point in a river stream or lake outlet structure. Rating curves are useful for interpolating and perhaps extrapolating flow measurements and for use directly in storage routing models. However, rating data and stations are limited. A generalized nondimensional mathematical expression that describes the rating relation of depth and discharge has been developed and tested against observations from 46 stations in West‐Central Florida. Three approaches were tested in sequence to select the best fit. The proposed model is a log‐linear equation with zero intercept and a slope that fits more than 50% of the stations were analyzed. The model is normalized by the depth and discharge values at 10% exceedance from data published by the U.S. Geological Survey. For ungauged applications, Q₁₀ and d₁₀ were derived from a relationship shown to be reasonably well correlated to the watershed drainage area. The average relative error for this parameter set shows that for the flow range up to the Q₁₀ discharge, better than 30% agreement with the USGS rating data can be expected for about 50% of the stations. Further analysis is required to determine why so many stations exhibit such similar behavior and to identify the criteria or parameters governing the differences.     

成都市某区暴雨径流过程模拟分析

为提高城市雨洪管理的效率,最大限度地减少暴雨洪水带来的危害,针对城市防洪排涝的需要,在分析成都市某区降雨径流规律后,建立了该区暴雨径流模拟的数学模型,有助于采取相应措施充分利用雨洪资源。通过对雨洪过程模拟验证表明,模型适合该区域的实际情况,具有一定的合理性和可靠性。     

Calibration and Validation of ADAPT and SWAT for Field‐Scale Runoff Prediction1

Abstract: The pollutant reduction possible with a given agricultural best‐management practice (BMP) is complex and site‐specific. Water‐quality models can evaluate BMPs, but model results are often limited by the lack of calibrated parameters for a given BMP. This study calibrated runoff prediction of two models (ADAPT and SWAT) for individual field plots having one till and two no‐till management practices. The factors used for runoff calibration were curve number II (CN_II) and saturated hydraulic conductivity (Ksat) for ADAPT, and CN_II, Ksat, and available water capacity for SWAT. Results were evaluated using coefficient of determination (R²), Nash‐Sutcliffe efficiency (E_f), root‐mean square error, median‐based E_f, and sign tests. Results indicated that for ADAPT, the best‐fit CN_II was 66 for the NT/SB (no‐till plot with surface‐broadcast fertilizer) treatment, 68 for the NT/DB (no‐till with deep‐banded fertilizer) treatment, and 70 for the tilled plot, whereas for SWAT the best‐fit CN_II was much higher, 86, for all treatments. Neither agreed with the textbook CN_II, 78, for sorghum in silty clay loam soil. The best‐fit model parameters for both runoff calibration phases had excellent correlation to monthly totals and moderate correlation to individual events.     

Ensemble Streamflow Forecast: A GLUE‐Based Neural Network Approach1

Abstract: While training a Neural Network to model a rainfall‐runoff process, generally two aspects are considered: its capability to be able to describe the complex nature of the processes being modeled and the ability to generalize so that novel samples could be mapped correctly. The general conclusion is that, the smallest size network capable of representing the sample distribution is the best choice, as far as generalization is concerned. Oftentimes input variables are selected a priori in what is called an explanatory data analysis stage and are not part of the actual network training and testing procedures. When they are, the final model will have only a “fixed” type of inputs, lag‐space, and/or network structure. If one of these constituents was to change, one would obtain another equally “optimal” Neural Network. Following Beven and others' generalized likelihood uncertainty estimate approach, a methodology is introduced here that accounts for uncertainties in network structures, types of inputs, and their lag‐space relationships by looking at a population of Neural Networks rather than target in getting a single “optimal” network. It is shown that there is a wide array of networks that provide “similar” results, as seen by a likelihood measure, for different types of inputs, lag‐space, and network size combinations. These equally optimal networks expose the range of uncertainty in streamflow predictions and their expected value results in a better performance than any of the single network predictions.     

GENETIC PROGRAMMING AND ITS APPLICATION IN REAL‐TIME RUNOFF FORECASTING1

ABSTRACT: Genetic programming (GP), a relatively new evolutionary technique, is demonstrated in this study to evolve codes for the solution of problems. First, a simple example in the area of symbolic regression is considered. GP is then applied to real‐time runoff forecasting for the Orgeval catchment in France. In this study, GP functions as an error updating scheme to complement a rainfall‐runoff model, MIKE11/NAM. Hourly runoff forecasts of different updating intervals are performed for forecast horizons of up to nine hours. The results show that the proposed updating scheme is able to predict the runoff quite accurately for all updating intervals considered and particularly for updating intervals not exceeding the time of concentration of the catchment. The results are also compared with those of an earlier study, by the World Meteorological Organization, in which autoregression and Kalman filter were used as the updating methods. Comparisons show that GP is a better updating tool for real‐time flow forecasting. Another important finding from this study is that nondimensionalizing the variables enhances the symbolic regression process significantly.     

成都市径流污染的概念性模型

概念性模型是根据城市径流形成过程及对水体影响变化的物理机理建立的一套数学模型，模型及参数均有明确的物理意义。本文用概念性模型求出了成都市街道地表物的累积量，降雨径流污染负荷量，并模拟了污染物对受纳水体的影响。     

A QUASI-LINEAR SPATIALLY DISTRIBUTED CELL MODEL FOR THE SURFACE RUNOFF SYSTEM1

ABSTRACT: The paper presents a spatially distributed model consisting of cells that are interconnected in a pattern similar to the major drainage network of the watershed. Each cell receives as input the rainfall excess for the area represented by the cell as well as inflows from cells located upstream. Outflow from the cell is derived by routing the total input through the cell assuming it to be a linear reservoir during the storm. The time constant of the cells is however allowed to vary from storm to storm so that the model may be described as a quasi-linear model. The model was tested with rainfall excess and direct surface runoff data available for a medium size watershed with satisfactory results. The time constant was found to be related to the rainfall excess of the storms studied, its value decreasing with the increase in the total rainfall excess.     

HYDROLOGIC PARAMETERIZATION OF WATERSHEDS FOR RUNOFF PREDICTION USING SWAT1

ABSTRACT: The use of continuous time, distributed parameter hydrologic models like SWAT (Soil and Water Assessment Tool) has opened several opportunities to improve watershed modeling accuracy. However, it has also placed a heavy burden on users with respect to the amount of work involved in parameterizing the watershed in general and in adequately representing the spatial variability of the watershed in particular. Recent developments in Geographical Information Systems (GIS) have alleviated some of the difficulties associated with managing spatial data. However, the user must still choose among various parameterization approaches that are available within the model. This paper describes the important parameterization issues involved when modeling watershed hydrology for runoff prediction using SWAT with emphasis on how to improve model performance without resorting to tedious and arbitrary parameter by parameter calibration. Synthetic and actual watersheds in Indiana and Mississippi were used to illustrate the sensitivity of runoff prediction to spatial variability, watershed decomposition, and spatial and temporal adjustment of curve numbers and return flow contribution. SWAT was also used to predict stream runoff from actual watersheds in Indiana that have extensive subsurface drainage. The results of this study provide useful information for improving SWAT performance in terms of stream runoff prediction in a manner that is particularly useful for modeling ungaged watersheds wherein observed data for calibration is not available.     

Stream and Retention Pond Thermal Response to Heated Summer Runoff From Urban Impervious Surfaces1

Abstract: Runoff from parking lots during summer storms injects surges of hot water into receiving water bodies. We present temperature data collected near urban storm sewer outfalls in Blacksburg, Virginia, using arrays of sensors in a stream and a stormwater pond. Surges occurred roughly a dozen times per month, ranging up to 8.1°C with average duration 2 h in the stream and up to 11.2°C with average duration 7 h in the pond. Surges were larger in the pond due to a larger contributing watershed, no dilution by upstream water, and cool background temperatures near the outfall. Surges began abruptly, warming at rates averaging 0.2°C/min for periods of 5‐20 min. Surges dissipated as they propagated into the water bodies, travelling further in the stream (>19 m) than the pond (～10 m) consistent with greater advection in the stream. Surges were largest and most frequent in the afternoon but occurred at all times of day and night. Stream surges exhibited two phases: an early high‐temperature low‐volume input from the storm sewer and a later low‐temperature high‐volume input from upstream. Surges at the pond did not exhibit two phases, consistent with inputs only from storm sewers. Surges are likely common in urban areas, and may cumulatively have consequences for aquatic organisms, biogeochemical process rates, and even human health. Such effects may be compounded by urban heat islands and climate change, so prevention or mitigation should be considered.     

Grass‐Shrub Riparian Buffer Removal of Sediment,Phosphorus, and Nitrogen From Simulated Runoff1

Abstract: Riparian buffer forests and vegetative filter strips are widely recommended for improving surface water quality, but grass‐shrub riparian buffer system (RBSs) are less well studied. The objective of this study was to assess the influence of buffer width and vegetation type on the key processes and overall reductions of total suspended solids (TSS), phosphorus (P), and nitrogen (N) from simulated runoff passed through established (7‐year old) RBSs. Nine 1‐m RBS plots, with three replicates of three vegetation types (all natural selection grasses, two‐segment buffer with native grasses and plum shrub, and two‐segment buffer with natural selection grasses and plum shrub) and widths ranging from 8.3 to 16.1 m, received simulated runoff having 4,433 mg/l TSS from on‐site soil, 1.6 mg/l total P, and 20 mg/l total N. Flow‐weighted samples were collected by using Runoff Sampling System (ROSS) units. The buffers were very efficient in removal of sediments, N, and P, with removal efficiencies strongly linked to infiltration. Mass and concentration reductions averaged 99.7% and 97.9% for TSS, 91.8% and 42.9% for total P, and 92.1% and 44.4% for total N. Infiltration alone could account for >75% of TSS removal, >90% of total P removal, and >90% of total N removal. Vegetation type induced significant differences in removal of TSS, total P, and total N. These results demonstrate that adequately designed and implemented grass‐shrub buffers with widths of only 8 m provide for water quality improvement, particularly if adequate infiltration is achieved.