Showing 6 results for Uncertainty Analysis
R Rostamian, S.f Mousavi, M Heidarpour, M Afyuni, K Abaspour,
Volume 12, Issue 46 (1-2009)
Abstract
Soil erosion is an important economical, social and environmental problem requiring intensive watershed management for its control. In recent years, modeling has become a useful approach for assessing the impact of various erosion-reduction approaches. ِDue to limited hydrologic data in mountainous watersheds, watershed modeling is, however, subject to large uncertainties. In this study, SWAT2000 was applied to simulate runoff and sediment discharge in Beheshtabad watershed, a sub-basin of Northern Karun catchment in central Iran, with an area of 3860 km2. Model calibration and uncertainty analysis were performed with SUFI-2. Four indices were used to assess the goodness of calibration, viz., P-factor, d-factor, R2 and Nash-Sutcliffe (NS). Runoff data (1996-2004) of six hydrometery stations were used for calibration and validation of this watershed. The results of monthly calibration p-factor, d-factor, R2 and NS values for runoff at the watershed outlet were 0.61, 0.48, 0.85 and 0.75, respectively, and for the validation, these statistics were 0.53, 0.38, 0.85 and 0.57, respectively. The values for calibration of sediment concentration at the watershed outlet were 0.55, 0.41, 0.55 and 0.52, respectively, and for the validation, these statistics were 0.69, 0.29, 0.60 and 0.27, respectively. In general, SWAT simulated runoff much better than sediment. Weak simulation of runoff at some months of the year might be due to under-prediction of snowmelt in this mountainous watershed, model’s assumptions in frozen and saturated soil layers, and lack of sufficient data. Improper simulation of sediment load could be attributed to weak simulation of runoff, insufficient data and periodicity of sediment data.
S Akhavan, J Abedi Koupaee, S.f Mousavi, K Abbaspour, M Afyuni, S.s Eslamian,
Volume 14, Issue 53 (10-2010)
Abstract
Temporal and spatial distribution of water components in watersheds, estimation of water quality, and uncertainties
associated with these estimations are important issues in freshwater studies. In this study, Soil and Water Assessment
Tool (SWAT) model was used to estimate components of freshwater availability: blue water (surface runoff plus deep
aquifer recharge), green water flow (actual evapotranspiration) and green water storage (soil water), in Hamadan-Bahar
watershed. Also, the Sequential Uncertainty Fitting program (SUFI2) was used to calibrate and validate the SWAT
model and do the uncertainty analysis. Degree of uncertainty is calculated by R-factor and P-factor parameters. In this
paper, results of calibration and validation are given for the river monthly discharge. In most stations, especially in
outlet of the watershed (Koshkabad station), simulation of river discharge was satisfactory. Values of R-factor in
calibration of monthly runoff were 0.4-0.8. These small values show good calibration of runoff in this watershed.
Values of P-factor were 20-60%. These small values show high uncertainty in estimations. For most stations of the
watershed, lack of data on river-water withdrawal caused poor simulation of base-flow and therefore the P-factor values
were low. Nash-Sutcliff (NS) coefficient was 0.3-0.8 after calibration, which shows good model calibration of outlet.
This study provided good information on the components of freshwater availability at spatial (sub-basin) and temporal
(monthly) scales with 95% prediction uncertainty ranges. The results of uncertainty analysis of components of
freshwater availability show that uncertainty ranges of average monthly blue water are larger than the other
components, because of its sensitivity to more parameters.
M. Karam, M. Afyuni, A. H. Khoshgoftarmanesh, M. A. Hajabbasi, H. Khademi, A. Abdi,
Volume 16, Issue 61 (10-2012)
Abstract
The task of modern agriculture is to safeguard the production of high quality food, in a sustainable natural environment under the precondition of pollution not exceeding accepted norms. The sustainability of current land use in agro-ecosystems can be assessed with respect to heavy metal accumulation in soils by balancing the input/ output fluxes. The objectives of this study were to model accumulation rate and the associated uncertainty of Zn in the agro-ecosystems of 3 arid and semi-arid provinces (Fars, Isfahan and Qom). Zinc accumulation rates in the agro-ecosystems were computed using a stochastic mass flux assessment (MFA) model with using Latin Hypercube sampling in combination with Monte-Carlo simulation procedures. Agricultural information including crop types, crop area and yield, kind and number of livestock, application rates of mineral fertilizers, compost and sewage sludge and also metal concentration in plants and soil amendments were used to quantify Zn fluxes and Zn accumulation rates. The results indicated that Zn accumulates considerably in agricultural lands of the studied townships especially in Najafabad (3009 g ha-1yr-1). The major Zn input routes to the agricultural soils (and due to agricultural activities) were manure and mineral fertilizers and the major part of the uncertainty in the Zn accumulation rate resulted from manure source.
S. Akhavan, A. Jodi Hameze Abad,
Volume 19, Issue 72 (8-2015)
Abstract
Urmia Lake, located in north-west of Iran, has been exposed to various threats such as drought, construction of dams, land use changes and increased global temperature. Due to the importance of Urmia Lake, it is feasible to conduct different kinds of studies to identify the problems of its watershed. The main objective of this study was to evaluate SWAT program’s ability to simulate runoff in Urmia Lake watershed with an area of 52000 km2. The model was run for the 1980-1997 period. Calibration and validation periods were from 1980 to 1991 and from 1992 to 1997, respectively. The results of calibration for 10% and 85% of hydrometric stations were very good and suitable, respectively. Also, validation results for 25% and 45% of hydrometric stations were very good and suitable, respectively. These results show the high ability of SWAT model to simulate discharges in Urmia Lake watershed. Moreover, some factors influencing inflow to the lake in recent years were evaluated. The outcomes revealed that recent changes (dam cconstructions, climate change and land use change) in the watershed have caused inflow volume to the lake to decrease by 80%. So, if natural management conditions had prevailed in the watershed, the Lake’s conditions would have been much better.
F. Yosevfand, S. Shabanlou,
Volume 23, Issue 4 (2-2020)
Abstract
In this study, the groundwater level (GWL) of the Sarab Qanbar region located in the south of Kermanshah, Iran, was estimated using the Wavelet- Self- Adaptive Extreme Learning Machine (WA- SAELM) model. An artificial intelligence method called “Self- Adaptive Extreme Learning Machine” and the “Wavelet transform” method were implemented for developing the numerical model. First, by using the autocorrelation function (ACF), the partial autocorrelation function (PACF) and the effective lags in estimating GWL, eight distinctive SAELM and WA- SAELM models were developed. Later, the values of the observational well were normalized for estimating GWL. Next, the most optimized mother wavelet was chosen for the modeling. By evaluating the results of SAELM and WA- SAELM, it was concluded that the WA- SAELM models could estimate the values of the objective function with higher accuracy. Then, the superior model was introduced, showing that it could be very accurate in forecasting the GWL. In the test mode, for example, the values of R (correlation coefficient), Main absolute error (MAE) and the NSC- Sutcliffe efficiency coefficient (NSC) for the superior model were calculated to be 0.995, 0.988 and 0.990, respectively. Furthermore, an uncertainty analysis was conducted for the numerical models, proving that the superior model had an underestimated performance.
A.h. Azimi, S Shabanlou, F. Yosefvand, A. Rajabi, B. Yaghoubi,
Volume 25, Issue 4 (3-2022)
Abstract
In this research, the scour hole depth at the downstream of cross-vane structures with different shapes (i.e., J, I, U, and W) was simulated utilizing a modern artificial intelligence method entitled "Outlier Robust Extreme Learning Machine (ORELM)". The observational data were divided into two groups: training (70%) and test (30%). Then, using the input parameters including the ratio of the structure length to the channel width (b/B), the densimetric Froude number (Fd), the ratio of the difference between the downstream and upstream depths to the structure height (Δy/hst), and the structure shape factor (φ), eleven different ORELM models were developed for estimating the scour depth. Subsequently, the superior model and also the most effective input parameters were identified through the conduction of uncertainty analysis. The superior model simulated the scour values by the dimensionless parameters b/B, Fd, Δy/hst. For this model, the values of the correlation coefficient (R), the variance accounted for (VAF), and the Nash-Sutcliffe efficiency (NSC) for the superior model in the test mode were obtained 0.956, 91.378, and 0.908, respectively. Also, the dimensionless parameters b/B and Δy/hst were detected as the most effective input parameters. Furthermore, the results of the superior model were compared with the extreme learning machine model and it was concluded that the ORELM model was more accurate. Moreover, an uncertainty analysis exhibited that the ORELM model had an overestimated performance. Besides, a partial derivative sensitivity analysis (PDSA) model was performed for the superior model.
