Application of semi‐distributed hydrological model for basin level water balance of the Ken basin of Central India

In the present study, a semi‐distributed hydrological model soil and water assessment tool (SWAT) has been employed for the Ken basin of Central India to predict the water balance. The entire basin was divided into ten sub basins comprising 107 hydrological response units on the basis of unique slope, soil and land cover classes using SWAT model. Sensitivity analysis of SWAT model was performed to examine the critical input variables of the study area. For Ken basin, curve number, available water capacity, soil depth, soil evaporation compensation factor and threshold depth of water in the shallow aquifer (GWQ_MN) were found to be the most sensitive parameters. Yearly and monthly calibration (1985–1996) and validation (1997–2009) were performed using the observed discharge data of the Banda site in the Ken basin. Performance evaluation of the model was carried out using coefficient of determination, Nash–Sutcliffe efficiency, root mean square error‐observations standard deviation ratio, percent bias and index of agreement criterion. It was found that SWAT model can be successfully applied for hydrological evaluation of the Ken basin, India. The water balance analysis was carried out to evaluate water balance of the Ken basin for 25 years (1985–2009). The water balance exhibited that the average annual rainfall in the Ken basin is about 1132 mm. In this, about 23% flows out as surface run‐off, 4% as groundwater flow and about 73% as evapotranspiration. Copyright © 2013 John Wiley & Sons, Ltd.     

SWAT model application and prediction uncertainty analysis in the Lake Tana Basin,Ethiopia

Lake Tana Basin is of significant importance to Ethiopia concerning water resources aspects and the ecological balance of the area. Many years of mismanagement, wetland losses due to urban encroachment and population growth, and droughts are causing its rapid deterioration. The main objective of this study was to assess the performance and applicability of the soil water assessment tool (SWAT) model for prediction of streamflow in the Lake Tana Basin, so that the influence of topography, land use, soil and climatic condition on the hydrology of Lake Tana Basin can be well examined. The physically based SWAT model was calibrated and validated for four tributaries of Lake Tana. Sequential uncertainty fitting (SUFI‐2), parameter solution (ParaSol) and generalized likelihood uncertainty estimation (GLUE) calibration and uncertainty analysis methods were compared and used for the set‐up of the SWAT model. The model evaluation statistics for streamflows prediction shows that there is a good agreement between the measured and simulated flows that was verified by coefficients of determination and Nash Sutcliffe efficiency greater than 0·5. The hydrological water balance analysis of the basin indicated that baseflow is an important component of the total discharge within the study area that contributes more than the surface runoff. More than 60% of losses in the watershed are through evapotranspiration. Copyright © 2009 John Wiley & Sons, Ltd.     

Simulation of stream flow components in a mountainous catchment in northern Thailand with SWAT,using the ANSELM calibration approach

Highland agriculture is intensifying rapidly in South‐East Asia, leading to alarmingly high applications of agrochemicals. Understanding the fate of these contaminants requires carefully planned monitoring programmes and, in most cases, accurate simulation of hydrological pathways into and through water bodies. We simulate run‐off in a steep mountainous catchment in tropical South‐East Asia. To overcome calibration difficulties related to the mountainous topography, we introduce a new calibration method, named A Nash–Sutcliffe Efficiency Likelihood Match (ANSELM), that allows the assignment of optimal parameters to different hydrological response units in simulations of stream discharge with the Soil and Water Assessment Tool (SWAT) hydrological model. ANSELM performed better than the Parasol calibration tool built into SWAT in terms of model efficiency and computation time. In our simulation, the most sensitive model parameters were those related to base flow generation, surface run‐off generation, flow routing and soil moisture change. The coupling of SWAT with ANSELM yielded reasonable simulations of both wet‐season and dry‐season storm hydrographs. Nash–Sutcliffe model efficiencies for daily stream flow during two validation years were 0.77 and 0.87. These values are in the upper range or even higher than those reported for other SWAT model applications in temperate or tropical regions. The different flow components were realistically simulated by SWAT, and showed a similar behaviour in all the study years, despite inter‐annual climatic differences. The realistic partitioning of total stream flow into its contributing components will be an important factor for using this hydrological model to simulate solute transport in the future. Copyright © 2014 John Wiley & Sons, Ltd.     

A computationally efficient method for uncertainty analysis of SWAT model simulations

The physically based distributed hydrological models are ideal for hydrological simulations; however most of such models do not use the basic equations pertaining to mass, energy and momentum conservation, to represent the physics of the process. This is plausibly due to the lack of complete understanding of the hydrological process. The soil and water assessment tool (SWAT) is one such widely accepted semi-distributed, conceptual hydrological model used for water resources planning. However, the over-parameterization, difficulty in its calibration process and the uncertainty associated with predictions make its applications skeptical. This study considers assessing the predictive uncertainty associated with distributed hydrological models. The existing methods for uncertainty estimation demand high computational time and therefore make them challenging to apply on complex hydrological models. The proposed approach employs the concepts of generalized likelihood uncertainty estimation (GLUE) in an iterative procedure by starting with an assumed prior probability distribution of parameters, and by using mutual information (MI) index for sampling the behavioral parameter set. The distributions are conditioned on the observed information through successive cycles of simulations. During each cycle of simulation, MI is used in conjunction with Markov Chain Monte Carlo procedure to sample the parameter sets so as to increase the number of behavioral sets, which in turn helps reduce the number of cycles/simulations for the analysis. The method is demonstrated through a case study of SWAT model in Illinois River basin in the USA. A comparison of the proposed method with GLUE indicates that the computational requirement of uncertainty analysis is considerably reduced in the proposed approach. It is also noted that the model prediction band, derived using the proposed method, is more effective compared to that derived using the other methods considered in this study.     

Evaluating the SWAT's snow hydrology over a Northern Quebec watershed

The snow treatment becomes an important component of Soil and Water Assessment Tool (SWAT)’s hydrology when spring flows are dominated by snow melting. However, little is known about SWAT's snow hydrology performance because most studies using SWAT were conducted in rainfall‐driven catchments. To fill this gap, the present study aims to evaluate the ability of SWAT in simulating snow‐melting‐dominated streamflow in the Outardes Basin in Northern Quebec. SWAT performance in simulating snowmelt is evaluated against observed streamflow data and compared to simulations from the operationally used Streamflow Synthesis and Reservoir Regulation (SSARR) model over that catchment. The SWAT 5‐year calibration showed a satisfactory performance at the daily and seasonal time scales with low volume biases. The SWAT validation was conducted over two (17‐year and 15‐year) periods. Performances were similar to the calibration period in simulating the daily and seasonal streamflows again with low model biases. The spring‐snowmelt‐generated peak flow was accurately simulated by SWAT both in magnitude and timing. When SWAT's results are compared to SSARR, similar performances in simulating the daily discharges were observed. SSARR simulates more accurately streamflow generated at the snowmelt onset whereas SWAT better predicts streamflow in summer, fall and winter. SWAT provided reasonable streamflow simulations for our snow‐covered catchment, but refinement of the process‐driven baseflow during the snowmelt onset could improve spring performances. Therefore, SWAT becomes an attractive tool for evaluating water resources management in Nordic environments when a distributed model is preferred or when water quality information (e.g. temperature) is required. Copyright © 2013 John Wiley & Sons, Ltd.     

Assessment of the effectiveness of a multi-site stochastic weather generator on hydrological modelling in the Red Deer River watershed,Canada

ABSTRACT

To improve the convergence of multiple-site weather generators (SWGs) based on the brute force algorithm (MBFA), a genetic algorithm (GA) is proposed to search the overall optimal correlation matrix. Precipitation series from weather generators are used as input to the hydrological model, the soil and water assessment tool (SWAT), to generate runoff over the Red Deer watershed, Canada for further runoff analysis. The results indicate that the SWAT model using SWG-generated data accurately represents the mean monthly streamflow for most of the months. The multi-site generators were capable of better representing the monthly streamflow variability, which was notably underestimated by the single-site version. In terms of extreme flows, the proposed method reproduced the observed extreme flow with smaller bias than MBFA, while the single-site generator significantly underestimated the annual maximum flows due to its poor capability in addressing partial precipitation correlations.     

Combining the SWAT model with sequential uncertainty fitting algorithm for streamflow prediction and uncertainty analysis for the Lake Dianchi Basin,China

Streams play an important role in linking the land with lakes. Nutrients released from agricultural or urban sources flow via streams to lakes, causing water quality deterioration and eutrophication. Therefore, accurate simulation of streamflow is helpful for water quality improvement in lake basins. Lake Dianchi has been listed in the ‘Three Important Lakes Restoration Act’ in China, and the degradation of its water quality has been of great concern since the 1980s. To assist environmental decision making, it is important to assess and predict hydrological processes at the basin scale. This study evaluated the performance of the soil and water assessment tool (SWAT) and the feasibility of using this model as a decision support tool for predicting streamflow in the Lake Dianchi Basin. The model was calibrated and validated using monthly observed streamflow values at three flow stations within the Lake Dianchi Basin through application of the sequential uncertainty fitting algorithm (SUFI‐2). The results of the autocalibration method for calibrating and the prediction uncertainty from different sources were also examined. Together, the p‐factor (the percentage of measured data bracketed by 95% prediction of uncertainty, or 95PPU) and the r‐factor (the average thickness of the 95PPU band divided by the standard deviation of the measured data) indicated the strength of the calibration and uncertainty analysis. The results showed that the SUFI‐2 algorithm performed better than the autocalibration method. Comparison of the SUFI‐2 algorithm and autocalibration results showed that some snowmelt factors were sensitive to model output upstream at the Panlongjiang flow station. The 95PPU captured more than 70% of the observed streamflow at the three flow stations. The corresponding p‐factors and r‐factors suggested that some flow stations had relatively large uncertainty, especially in the prediction of some peak flows. Although uncertainty existed, statistical criteria including R² and Nash–Sutcliffe efficiency were reasonably determined. The model produced a useful result and can be used for further applications. Copyright © 2012 John Wiley & Sons, Ltd.     

Parameter estimation and uncertainty analysis of SWAT model in upper reaches of the Heihe river basin

Heihe river basin, the second largest inland river basin in China, has attracted more attention in China due to the ever increasing water resources and eco‐environmental problems. In this article, SWAT (Soil and Water Assessment Tool; http://www.brc.tamus.edu/swat/ ) model was applied to upper reaches of the basin for better understanding of the hydrological process over the watershed. Parameter uncertainty and its contribution on model simulation are the main foci. In model calibration, the aggregate parameters instead of the original parameters in SWAT model were used to reduce the computing effort. The Bayesian approach was employed for parameter estimation and uncertainty analysis because its posterior distribution provides not only parameter estimation but also uncertainty analysis without normality assumption. The results indicated that: (1) SWAT model performs satisfactorily in this watershed as a whole, although some low and high flows were under‐ or overestimated, particularly in dry (e.g. 1991) and wet (e.g. 1996) years; (2) all calibrated parameters were not normally distributed (essentially positively or negatively skewed) and the parameter uncertainties were relatively small; and (3) the contributions of parameter uncertainty on model simulation uncertainty were relatively small. Copyright © 2009 John Wiley & Sons, Ltd.     

Comparison of daily and sub-daily SWAT models for daily streamflow simulation in the Upper Huai River Basin of China

Despite the significant role of precipitation in the hydrological cycle, few studies have been conducted to evaluate the impacts of the temporal resolution of rainfall inputs on the performance of SWAT (soil and water assessment tool) models in large-sized river basins. In this study, both daily and hourly rainfall observations at 28 rainfall stations were used as inputs to SWAT for daily streamflow simulation in the Upper Huai River Basin. Study results have demonstrated that the SWAT model with hourly rainfall inputs performed better than the model with daily rainfall inputs in daily streamflow simulation, primarily due to its better capability of simulating peak flows during the flood season. The sub-daily SWAT model estimated that 58 % of streamflow was contributed by baseflow compared to 34 % estimated by the daily model. Using the future daily and 3-h precipitation projections under the RCP (Representative Concentration Pathways) 4.5 scenario as inputs, the sub-daily SWAT model predicted a larger amount of monthly maximum daily flow during the wet years than the daily model. The differences between the daily and sub-daily SWAT model simulation results indicated that temporal rainfall resolution could have much impact on the simulation of hydrological process, streamflow, and consequently pollutant transport by SWAT models. There is an imperative need for more studies to examine the effects of temporal rainfall resolution on the simulation of hydrological and water pollutant transport processes by SWAT in river basins of different environmental conditions.     

An improved process-based representation of stream solute transport in the soil and water assessment tools

Hydrological models have long been used to study the interactions between land, surface and groundwater systems, and to predict and manage water quantity and quality. The soil and water assessment tool (SWAT), a widely used hydrological model, can simulate various ecohydrological processes on land and subsequently route the water quality constituents through surface and subsurface waters. So far, in-stream solute transport algorithms of the SWAT model have only been minimally revised, even though it has been acknowledged that an improvement of in-stream process representation can contribute to better model performance with respect to water quality. In this study, we aim to incorporate a new and improved solute transport model into the SWAT model framework. The new process-based model was developed using in-stream process equations from two well established models—the One-dimensional Transport with Inflow and Storage model and the Enhanced Stream Water Quality Model. The modified SWAT model (Mir-SWAT) was tested for water quality predictions in a study watershed in Germany. Compared to the standard SWAT model, Mir-SWAT improved dissolved oxygen (DO) predictions by removing extreme low values of DO (<6 mg/L) simulated by SWAT. Phosphate concentration peaks were reduced during high flows and a better match of daily predicted and measured values was attained using the Mir-SWAT model (R² = 0.17, NSE = −0.65, RSR = 1.29 with SWAT; R² = 0.28, NSE = −0.04, RSR = 1.02 with Mir-SWAT). In addition, Mir-SWAT performed better than the SWAT model in terms of Chlorophyll-a content particularly during winter months, improving the NSE and RSR for monthly average Chl-a by 74 and 42%, respectively. With the new model improvements, we aim to increase confidence in the stream solute transport component of the model, improve the understanding of nutrient dynamics in the stream, and to extend the applicability of SWAT for reach-scale analysis and management.     

Improved simulation of river water and groundwater exchange in an alluvial plain using the SWAT model

Hydrological interaction between surface and subsurface water systems has a significant impact on water quality, ecosystems and biogeochemistry cycling of both systems. Distributed models have been developed to simulate this function, but they require detailed spatial inputs and extensive computation time. The soil and water assessment tool (SWAT) model is a semi‐distributed model that has been successfully applied around the world. However, it has not been able to simulate the two‐way exchanges between surface water and groundwater. In this study, the SWAT‐landscape unit (LU) model – based on a catena method that routes flow across three LUs (the divide, the hillslope and the valley) – was modified and applied in the floodplain of the Garonne River. The modified model was called SWAT‐LUD. Darcy's equation was applied to simulate groundwater flow. The algorithm for surface water‐level simulation during flooding periods was modified, and the influence of flooding on groundwater levels was added to the model. Chloride was chosen as a conservative tracer to test simulated water exchanges. The simulated water exchange quantity from SWAT‐LUD was compared with the output of a two‐dimensional distributed model, surface–subsurface water exchange model. The results showed that simulated groundwater levels in the LU adjoining the river matched the observed data very well. Additionally, SWAT‐LUD model was able to reflect the actual water exchange between the river and the aquifer. It showed that river water discharge has a significant influence on the surface–groundwater exchanges. The main water flow direction in the river/groundwater interface was from groundwater to river; water that flowed in this direction accounted for 65% of the total exchanged water volume. The water mixing occurs mainly during high hydraulic periods. Flooded water was important for the surface–subsurface water exchange process; it accounted for 69% of total water that flowed from the river to the aquifer. The new module also provides the option of simulating pollution transfer occurring at the river/groundwater interface at the catchment scale. Copyright © 2015 John Wiley & Sons, Ltd.     

Modification of the SWAT model to simulate regional groundwater flow using a multicell aquifer

The soil and water assessment tool (SWAT) has been widely used and thoroughly tested in many places in the world. The application of the SWAT model has pointed out that 2 of the major weaknesses of SWAT are related to the nonspatial reference of the hydrologic response unit concept and to the simplified groundwater concept, which contribute to its low performance in baseflow simulation and its inability to simulate regional groundwater flow. This study modified the groundwater module of SWAT to overcome the above limitations. The modified groundwater module has 2 aquifers. The local aquifer, which is the shallow aquifer in the original SWAT, represents a local groundwater flow system. The regional aquifer, which replaces the deep aquifer of the original SWAT, represents intermediate and regional groundwater flow systems. Groundwater recharge is partitioned into local and regional aquifer recharges. The regional aquifer is represented by a multicell aquifer (MCA) model. The regional aquifer is discretized into cells using the Thiessen polygon method, where centres of the cells are locations of groundwater observation wells. Groundwater flow between cells is modelled using Darcy's law. Return flow from cell to stream is conceptualized using a non‐linear storage–discharge relationship. The SWAT model with the modified aquifer module, the so‐called SWAT‐MCA, was tested in 2 basins (Wipperau and Neetze) with porous aquifers in a lowland area in Lower Saxony, Germany. Results from the Wipperau basin show that the SWAT‐MCA model is able (a) to simulate baseflow in a lowland area (where baseflow is a dominant source of streamflow) better than the original model and (b) to simulate regional groundwater flow, shown by the simulated groundwater levels in cells, quite well.     

On the use of multi‐algorithm,genetically adaptive multi‐objective method for multi‐site calibration of the SWAT model

With the availability of spatially distributed data, distributed hydrologic models are increasingly used for simulation of spatially varied hydrologic processes to understand and manage natural and human activities that affect watershed systems. Multi‐objective optimization methods have been applied to calibrate distributed hydrologic models using observed data from multiple sites. As the time consumed by running these complex models is increasing substantially, selecting efficient and effective multi‐objective optimization algorithms is becoming a nontrivial issue. In this study, we evaluated a multi‐algorithm, genetically adaptive multi‐objective method (AMALGAM) for multi‐site calibration of a distributed hydrologic model—Soil and Water Assessment Tool (SWAT), and compared its performance with two widely used evolutionary multi‐objective optimization (EMO) algorithms (i.e. Strength Pareto Evolutionary Algorithm 2 (SPEA2) and Non‐dominated Sorted Genetic Algorithm II (NSGA‐II)). In order to provide insights into each method's overall performance, these three methods were tested in four watersheds with various characteristics. The test results indicate that the AMALGAM can consistently provide competitive or superior results compared with the other two methods. The multi‐method search framework of AMALGAM, which can flexibly and adaptively utilize multiple optimization algorithms, makes it a promising tool for multi‐site calibration of the distributed SWAT. For practical use of AMALGAM, it is suggested to implement this method in multiple trials with relatively small number of model runs rather than run it once with long iterations. In addition, incorporating different multi‐objective optimization algorithms and multi‐mode search operators into AMALGAM deserves further research. Copyright © 2009 John Wiley & Sons, Ltd.     

Simulating daily soil water under foothills fescue grazing with the soil and water assessment tool model (Alberta,Canada)

Grazing is common in the foothills fescue grasslands and may influence the seasonal soil‐water patterns, which in turn determine range productivity. Hydrological modelling using the soil and water assessment tool (SWAT) is becoming widely adopted throughout North America especially for simulation of stream flow and runoff in small and large basins. Although applications of the SWAT model have been wide, little attention has been paid to the model's ability to simulate soil‐water patterns in small watersheds. Thus a daily profile of soil water was simulated with SWAT using data collected from the Stavely Range Sub‐station in the foothills of south‐western Alberta, Canada. Three small watersheds were established using a combination of natural and artificial barriers in 1996–97. The watersheds were subjected to no grazing (control), heavy grazing (2·4 animal unit months (AUM) per hectare) or very heavy grazing (4·8 AUM ha^?1). Soil‐water measurements were conducted at four slope positions within each watershed (upper, middle, lower and 5 m close to the collector drain), every 2 weeks annually from 1998 to 2000 using a downhole CPN 503 neutron moisture meter. Calibration of the model was conducted using 1998 soil‐water data and resulted in Nash–Sutcliffe coefficient (EF or R²) and regression coefficient of determination (r²) values of 0·77 and 0·85, respectively. Model graphical and statistical evaluation was conducted using the soil‐water data collected in 1999 and 2000. During the evaluation period, soil water was simulated reasonably with an overall EF of 0·70, r² of 0·72 and a root mean square error (RMSE) of 18·01. The model had a general tendency to overpredict soil water under relatively dry soil conditions, but to underpredict soil water under wet conditions. Sensitivity analysis indicated that absolute relative sensitivity indices of input parameters in soil‐water simulation were in the following order; available water capacity > bulk density > runoff curve number > fraction of field capacity (FFCB) > saturated hydraulic conductivity. Thus these data were critical inputs to ensure reasonable simulation of soil‐water patterns. Overall, the model performed satisfactorily in simulating soil‐water patterns in all three watersheds with a daily time‐step and indicates a great potential for monitoring soil‐water resources in small watersheds. Copyright © 2004 John Wiley & Sons, Ltd.     

Comparison of radar and gauge precipitation data in watershed models across varying spatial and temporal scales

Precipitation is a key control on watershed hydrologic modelling output, with errors in rainfall propagating through subsequent stages of water quantity and quality analysis. Most watershed models incorporate precipitation data from rain gauges; higher‐resolution data sources are available, but they are associated with greater computational requirements and expertise. Here, we investigate whether the Multisensor Precipitation Estimator (MPE or Stage IV Next‐Generation Radar) data improve the accuracy of streamflow simulations using the Soil and Water Assessment Tool (SWAT), compared with rain gauge data. Simulated flows from 2002 to 2010 at five timesteps were compared with observed flows for four nested subwatersheds of the Neuse River basin in North Carolina (21‐, 203‐, 2979‐, and 10 100‐km² watershed area), using a multi‐objective function, informal likelihood‐weighted calibration approach. Across watersheds and timesteps, total gauge precipitation was greater than radar precipitation, but radar data showed a conditional bias of higher rainfall estimates during large events (>25–50 mm/day). Model parameterization differed between calibrations with the two datasets, despite the fact that all watershed characteristics were the same across simulation scenarios. This underscores the importance of linking calibration parameters to realistic processes. SWAT simulations with both datasets underestimated median and low flows, whereas radar‐based simulations were more accurate than gauge‐based simulations for high flows. At coarser timesteps, differences were less pronounced. Our results suggest that modelling efforts in watersheds with poor rain gauge coverage can be improved with MPE radar data, especially at short timesteps. Published 2013. This article is a U.S. Government work and is in the public domain in the USA.     

Model calibration using hydropedological insights to improve the simulation of internal hydrological processes using SWAT+

Soils affect the distribution of hydrological processes by partitioning precipitation into different components of the water balance. Therefore, understanding soil-water dynamics at a catchment scale remains imperative to future water resource management. In this study, the value of hydropedological insights was examined to calibrate a processes-based model. Soil morphology was used as soft data to assist in the calibration of the Soil Water Assessment Tool (SWAT+) model at five different catchment scales (48, 56, 174, 674, and 2421 km²) in the Sabie River catchment, South Africa. The aim of this study was to calibrate the SWAT+ model to accurately simulate long-term monthly streamflow predictions as well as to reflect internal soil hydrological processes using a procedure focusing on hydropedology as a calibration tool in a multigauge system. Results indicated that calibration improved streamflow predictions where R² improved by 2%–8%. Nash-Sutcliffe Efficiency (NSE) improved from negative correlations to values exceeding 0.5 at four of the five catchment scales compared to the uncalibrated model. Results confirm that soil mapping units can be calibrated individually within SWAT+ to improve the representation of hydrological processes. Particularly, the spatial linkage between hydropedology and hydrological processes, which is captured within the soil map of the catchment, can be adequately reflected within the model simulations after calibration. This research will lead to an improved understanding of hydropedology as soft data to improve hydrological modelling accuracy.     

Predicting phosphorus dynamics in complex terrains using a variable source area hydrology model

Phosphorus (P) loss from agricultural watersheds has long been a critical water quality problem, the control of which has been the focus of considerable research and investment. Preventing P loss depends on accurately representing the hydrological and chemical processes governing P mobilization and transport. The Soil and Water Assessment Tool (SWAT) is a watershed model commonly used to predict run‐off and non‐point source pollution transport. SWAT simulates run‐off employing either the curve number (CN) or the Green and Ampt methods, both assume infiltration‐excess run‐off, although shallow soils underlain by a restricting layer commonly generate saturation‐excess run‐off from variable source areas (VSA). In this study, we compared traditional SWAT with a re‐conceptualized version, SWAT‐VSA, that represents VSA hydrology, in a complex agricultural watershed in east central Pennsylvania. The objectives of this research were to provide further evidence of SWAT‐VSA's integrated and distributed predictive capabilities against measured surface run‐off and stream P loads and to highlight the model's ability to drive sub‐field management of P. Thus, we relied on a detailed field management database to parameterize the models. SWAT and SWAT‐VSA predicted discharge similarly well (daily Nash–Sutcliffe efficiencies of 0.61 and 0.66, respectively), but SWAT‐VSA outperformed SWAT in predicting P export from the watershed. SWAT estimated lower P loss (0.0–0.25 kg ha^?1) from agricultural fields than SWAT‐VSA (0.0–1.0+ kg ha^?1), which also identified critical source areas – those areas generating large run‐off and P losses at the sub‐field level. These results support the use of SWAT‐VSA in predicting watershed‐scale P losses and identifying critical source areas of P loss in landscapes with VSA hydrology. Copyright © 2014 John Wiley & Sons, Ltd.