Evaluation of Global and Regional Soil Maps in Flow Forecasting using SWAT Model (Talar Watershed, of Mazandaran Province)

Document Type : Research Paper


1 Department of Engineering and Agricultural Technology, Faculty of Engineering and Technical, Payam-e- Noor University (PNU), Tehran, Iran

2 Department of Water Resources Engineering, Faculty Civil, Water and Environmental Engineering, Shahid Beheshti University, Tehran, Iran


In most developing countries soil georeferenced data is not yet available. The FAO prepared the global map through the Coordinated Global Soil Database. The aim of this study was to evaluate the required details of the regional map in comparison with the soil global map of Talar watershed using the SWAT model. In this model, climatic data of 2004-2017 were used to simulate runoff. The first two years were used as model warm-up period, 2005-2015 for calibration and 2016-2017 for model validation. Sensitivity and uncertainty analysis, was done in SWAT-CUP software with the help of SUFI-2 algorithm. The evaluation of the model was done using the coefficients of explanation (R2) and Nash-Sutcliffe (NSE) statistics. The available soil water capacity (SOL_AWC) parameter was determined as a sensitive parameter. Before calibration, 32.7% (global) and 29.3% (regional) precipitation was simulated as daily flow. After recalibration based on the regional soil map (NSE=0.56 and R2=0.74) and global (NSE=0.56 and R2=0.64) and in the verification stage according to the regional soil map (NSE=0.59 and R2=0.76) and global (NSE=0.62) and R2=0.75) was estimated. The results showed the acceptable performance of the model in flow simulation. Sub-basin 9 (deciduous forest) and 14 (summer pasture) had the lowest and highest share in runoff production, respectively. The regional map provided more reliable results. Although the additional information provided by the regional map, while changing the optimized parameters in the model, has no effect on the outflow of the basin, the results indicated that the regional soil map has no significant effect on the flow prediction, and if it is not available, the global soil map information can be used.
Extended Abstract
Due to the limitation of measurement methods in flood control projects and the need to have a method to generalize existing statistics to basins without statistics or places where measurement is not possible, simulating changes in Future hydrology is one of the main reasons for hydrological simulation. Hydrologists and water resource managers have widely used hydrological models to simulate such changes in the past decades. One of the hydrological models is the Soil and Water Assessment Tools or SWAT model, which was presented in the United States to evaluate the effects of conservation agriculture on hydrological processes and water quality at the watershed scale in 1998. In the SWAT model, the watershed is divided into several sub-basins. Using sub-basins in simulation is beneficial, especially for areas with complex soil characteristics and land use. Then these sub-basins are divided into hydrological response units (HRUs) with the same soil characteristics, land use and management. One important and effective data in this model is soil data and map. Soil data and its related characteristics, including permeability and water retention capacity, determine the amount of runoff production. In most research studies and practical plans, the lack of appropriate soil information prevents researchers and decision-makers from using hydrological models to determine sub-basins sensitive to runoff production. While the existence of soil information (global maps) can make it possible to use hydrological models in basins without measured soil data. This information is such that due to the need for exorbitant cost in stratification and creation of soil profile, usually most areas of the country lack regional soil data, and the need to use the global soil map becomes necessary in such areas. Since this research compares two maps in two different scales for the first time, the main goal of this research is to prepare global and regional soil maps for the Talar watershed of Mazandaran province as one of the most effective SWAT model inputs, the accuracy of this model with two different inputs from soil maps will also be checked in predicting the runoff flow.
The first step to implement the SWAT model is to create hydrological response units (HRUs), which form the basis of the work in this study and the Talar watershed as one of the essential sub-basins of the Mazandaran Sea. HRUs are produced by integrating the map of sub-areas created in the study area (through DEM map), soil map, land use and slope. Finally, based on the input maps to the model, 14 sub-basins and 45 hydrological response units (HRUs) were produced in the study area. According to the input data type (minimum and maximum temperature information), the Hargreaves method was used to estimate daily reference evaporation-transpiration when the data of sunshine hours, relative humidity and wind speed were unavailable. In the simulation, the observational data of 2005-2015 were used for calibration and 2016-2017 for model validation. In order to adapt the model to the existing environmental conditions in the region, the first two years of simulation (2004-2005) were considered to warm up the model. After the calibration, the model's accuracy was measured using the obtained parameters and the observed values that were not used in the calibration stage. In case of acceptable simulation, the model will be ready for use. In order to evaluate the efficiency and accuracy of the SWAT model, two coefficients of determination (R2) and Nash-Sutcliffe coefficient of efficiency (NSE) were used. According to the research based on soil maps, firstly, a global soil map was prepared according to the global raster map, and after analysis in GIS, the soil type of the area was determined and entered into the SWAT database. On the other hand, having the data and soil layers of the region, a regional soil map was prepared, and this map was prepared for entering the SWAT model. After preparing the soil maps (global and regional), the model was implemented twice (the first time by entering the global soil map and the second time by entering the regional soil map). Then, after implementation with two different projects, the outputs of the model were extracted and compared with each other. Finally, to check the model's accuracy using the above maps' inflow forecasting, it was calibrated and verified through SWAT-CUP.
Results and Discussion
At first, by running the model in two simulation modes (based on the regional and global soil map) in the GIS environment, the Talar watershed was divided into 14 sub-basins and 45 hydrological response units. The results showed that approximately 29.3% of the basin's precipitation and 32.7% of it were lost as base flow or surface flow, respectively, in the global and regional soil maps. Therefore, in these two cases, the simulation resulted in the lowest amount of surface runoff (in the soil map of the region) and the highest amount of evaporation-transpiration (in the global soil map). By analyzing the general sensitivity of the desired parameters, the available soil water capacity parameter (SOL_AWC) was determined as the most sensitive parameter. After the recalibration of the model based on the regional soil map (NSE=0.56 and R2=0.74) and global (NSE=0.56 and R2=0.64) and in the verification stage also according to the regional soil map (NSE=0.59 and R2=0.76) and global (NSE= 0.62 and R2 = 0.75) were estimated. Also, the simulated flow (surface and base flow) decreased after recalibrating the model. Evaporation-transpiration and feeding to the deep aquifer increased due to the reduction of capillary rise from the shallow aquifer and the reduction of the minimum amount of water storage required in the aquifer for the base flow event. Despite improving the model's performance in Talar watershed during calibration and validation using the regional soil map, the simulated discharge at the outlet was higher than the observed discharge. So, in the calibration period, compared to the validation period, the model predicted the peak discharge much better. Using the global soil map, the model overestimated the discharge compared to the regional soil map. Therefore, the global soil map showed less accuracy than the regional soil map in the SWAT model, and the results of this research also showed that using soil information with good resolution can improve flow prediction.
In this study, we used soil information from regional and global soil maps to evaluate the effect of soil data representation and spatial variability. Flow forecasting was done in basin hydrology simulation using the SWAT hydrological model. Before calibrating the model, other water balance components decreased compared to calibrating period except for evapotranspiration. There is no significant difference between the coefficients obtained in the calibration and validation periods of the model. Therefore, the obtained coefficients show the model's efficiency in the direction of runoff simulation in Talar watershed. The results indicate that the use of regional soil information with more spatial details does not significantly affect flow prediction, and global soil information in the form of a global map can be used if this information is unavailable. The additional information provided in the regional soil map, although it can change the optimized parameters in the model, does not affect the watershed discharge. Our simulation results showed the effect of the global soil map used on flow forecasting for Talar watershed, and good resolution soil information (such as regional soil map) can also improve flow forecasting.
There is no funding support.
 Authors’ Contribution
All of the authors approved thecontent of the manuscript and agreed on all aspects of the work.
Conflict of Interest
Authors declared no conflict of interest.
We are grateful to all the scientific consultants of this paper


Abbaspour, K.C. (2011). SWAT-CUP: SWAT Calibration and Uncertainty Programs- A User Manual. Eawag: Swiss Federal Institute of Aquatic Science and Technology, 100 pp.
Abbaspour, K.C., Rouholahnejad, E., Vaghefi, S., Srinivasan, R., Yang, H. & Klove, B. (2015). A Continental-scale hydrology and water quality model for Europe: Calibration and uncertainty of a high-resolution large-scale SWAT model. Journal of Hydrology, 524, 733–752. doi: 10.1016/j.jhydrol.2015.03.027.
Adem, A.A., Dile, W.T., Worqlul, A.W., Ayana, E.K., Tilahun, S.A., Steenhuis, T.S. (2020). Assessing Digital Soil Inventories for Predicting Streamflow in the Headwaters of the Blue Nile. Hydrology, 7 (8), 1-19. doi: 10.3390/hydrology7010008.
Arnold, J.G., Moriasi, D.N., Gassman, P.W., Abbaspour, K.C., White, M., Srinivasan, J., Santhi, R. C., Harmel, R. D., van Griensven, A., Van Liew, M., Kannan, W.N., & Jha, M.K. (2012). SWAT: model use, calibration and validation. American Society of Agricultural and Biological Engineers, 55(4), 1491-1508. doi: 10.13031/2013.42256.
Arnold, J.G., Srinivasan, R., Muttiah, R.S., & Williams, J.R. (1998). Large area hydrologic modeling and assessment part I: model development. Journal of the American Water Resources Association, 34(1), 73-89. doi: 10.1111/j.1752-1688.1998.tb05961.x.
Boluwade, A., & Madramootoo, C. (2013). Modeling the impacts of spatial heterogeneity in the castor watershed on runoff, sediment, and phosphorus loss using SWAT: I. Impacts of spatial variability of soil properties. Water, Air, Soil Pollution, 224, 1692. doi: 10.1007/s11270-013-1692-0.
Bossa, A., Diekkrüger, B., Igué, A., & Gaiser, T. (2012). Analyzing the effects of different soil databases on modeling of hydrological processes and sediment yield in Benin (West Africa). Geoderma, 173, 61–74. doi: 10.1016/j.geoderma.2012.01.012.
Chaplot, V. (2005). Impact of DEM mesh size and soil map scale on SWAT runoff, sediment, and NO3–N loads predictions. Journal of Hydrology, 312, 207–222. doi: 10.1016/j.jhydrol.2005.02.017.
Farrokhzadeh, B., Ildoromi, A., Ataeian, B., & Nourouzi. (2016). Evaluation of the Effects of Land Use Changes on Suspended Load Using SWAT Model (Case Study: Yalfan Watershed, Hamedan Province). Environental Erosion Research, 5: 3 (19), 28-46. [In Persian]
Gholami, A., Shahedi, K., Habibnejad Rooshan, M., & Vafakhah, M., & Soleimani, K. (2017). Assesment About Efficiency of SWAT Semi-Distribution Model for Simulation of Streamflow (Case Study in Talar Watershed, Mazandaran Province). Iranian Journal of Soil and Water Research, 48 (3), 463-476. doi: 10.22059/IJSWR.2017.63414. [In Persian].
Golshan, G., Esmali Ouri, A., & Khosravi, Kh. (2018). Flood Susceptibility assessments Using Frequency Ratio model in Talar Watershed. Journal of Natural Environmental Hazards, 7 (15), 1-16. doi: 10.22111/JNEH.2017.3120. [In Persian].
Hargreaves, G., & Samani, Z.A. (1985). Reference crop evapotranspiration from temperature. Applid Engineering Agriculture, 1(2), 96–99. doi: 10.13031/2013.26773.
Kavian, A., & Mohammadi, M. (2019). Effects of Digital Elevation Models (DEM) Spatial Resolution on Hydrological Simulation. Journal of Watershed Management Research, 10 (19), 36-45. doi: 10.29252/jwmr.10.19.36. [In Persian].
Khosravi, M., Zolfaghari, A.A., Kaboli, S.H., & Ghafari, H. (2022). Investigation of SWAT model efficiency in spatial analysis of runoff in watersheds without local soil data (Case Study: Damghanrood Watershed). Applied Soil Research, 10 (1), 129-143. [In Persian]
Kuo, W.L., Steenhuis, T.S., McCulloch, C.E., Mohler, C.L.,Weinstein, D.A., DeGloria, S.D., & Swaney, D.P. (1999). Effect of grid size on runoff and soil moisture for a variable-source-area hydrology model. Water Resources Research, 35, 3419–3428. doi: 10.1029/1999WR900183.
Lal, R. (2005). Soil erosion and carbon dynamics. Soil and Tillage Research, 81(2), 137-142. doi: 10.1016/j.still.2004.09.002.
Mohseni, B., Shahedi, K., Habibnejad-Roshan, M., & Darzi-Naftchali, A. (2021). Quantitative simulation of surface water and groundwater resources in Behshahr - Bandar-e-Gaz Plain using SWAT model. Journal of Water and Soil Resources Conservation (WSRCJ), 10 (4), 109-126. doi: 10.30495/WSRCJ.2021.18084. [In Persian].
Mohseni, B., Shahedi, K., Habibnejad-Roshan, M., & Darzi-Naftchali, A. (2022). Improving groundwater sustainability through conservation strategies in a critical-prohibited coastal plain. Physics and Chemistry of the Earth, 127, 103176. doi: 10.1016/j.pce.2022.103176.
Moradi, A., Najafinejad, A., Ownegh, M., & Komaki, Ch. B. (2019). Assessment of the impacts of different Digital Elevation Models on runoff and suspended sediment estimations using SWAT model, case study: Galikesh Watershed, Golestan Province. Watershed Engineering and Management, 11 (1), 62-75. doi:10.22092/ijwmse.2019.118433. [In Persian].
Moriasi, D.N., Arnold, J.G., Van Liew, M.W., Bingner, R.L., Harmel, R.D. & Veith, T.L. (2007). Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Transactions of the ASABE, 50 (3), 885-900. doi: 10.13031/2013.23153.
Nash, J.E., & Sutcliffe, J.V. (1970). River flow forecasting through conceptual, models part I- A discussion of principles. Journal of Hydrology, 10, 282–290. doi: 10.1016/0022-1694 (70) 90255-6.
Oeurng, C., Sauvage, S., & Sanchez-Perez, J. (2011). Assessment of hydrology, sediment and particulate organic carbon yield in a large agricultural catchment using the SWAT model. Journal of Hydrology, 401, 145–153. doi: 10.1016/j.jhydrol.2011.02.017.
Ogden, F., Pradhan, N., Downer, C., & Zahner, J.A. (2011). “Relative importance of impervious area, drainage density, width function, and subsurface storm drainage on flood runoff from an urbanized catchment.” Water Resources Research, 47 (12). doi:10.1029/ 2011WR010550.
Oruç, H.N., Çelen, M., Gülgen, F., Öncel, M.S., Vural, S., & Kılıç, B. (2022). Evaluating the effects of soil data quality on the SWAT runoff prediction Performance; A case study of Saz-Cayirova catchment, Turkey. Urban Water Journal, 19, 1-16. doi: 10.1080/1573062X.2022.2056060.
Rezaei Moghaddam, M.H., Mokhtari, D., & Shafieimehr., M. (2021). Calibration and validation the SWAT model in the simulation of runoff and sediment in Shahr Chai of Miyaneh. Quarterly Journal of Geography and Planning, 25 (76), 129-139. doi: 10.22034/GP.2020.40775.2656. [In Persian].
Rezazdeh, M.S., Bakhtiari, B., Abbaspour, K & Ahmadi, M.M. (2018). Simulation of Runoff, Sediment and Evapotranspiration through Management Scenarios to Reduce Sediment Load Using SWAT Model. Iran-Watershed Management Science & Engineering, 12 (40), 41-51. [In Persian]
Saraie, B., Talebi, A., Mazidi, A., & Parvizi, S. (2020). Prioritization of Sardab-Rood watershed from flooding viewpoint using the SWAT model. Journal of Natural Environmental Hazards, 23 (9) و 85-98. doi: 10.22111/JNEH.2019.29033.1500. [In Persian].
Sohrabizadeh, Z., Sharifi Moghadam, E., & Hakimzadeh, M.A. (2018). Trend Changes Analysis of the Water Quality in the Talar River Watershed Using the Man-Kendall (MK) Test. Journal of Natural Ecosystems of Iran, 9 (3), 1-20. [In Persian]