A time-varying risk analysis is proposed for an adaptive design framework in nonstationary condit... more A time-varying risk analysis is proposed for an adaptive design framework in nonstationary conditions arising from climate change. A Bayesian, dynamic conditional copula is developed for modeling the time-varying dependence structure between mixed continuous and discrete multiattributes of multidi-mensional hydrometeorological phenomena. Joint Bayesian inference is carried out to fit the marginals and copula in an illustrative example using an adaptive, Gibbs Markov Chain Monte Carlo (MCMC) sampler. Posterior mean estimates and credible intervals are provided for the model parameters and the Deviance Information Criterion (DIC) is used to select the model that best captures different forms of nonstationarity over time. This study also introduces a fully Bayesian, time-varying joint return period for multivariate time-dependent risk analysis in nonstationary environments. The results demonstrate that the nature and the risk of extreme-climate multidimensional processes are changed over time under the impact of climate change, and accordingly the long-term decision making strategies should be updated based on the anomalies of the nonstationary environment.
s u m m a r y Accurate projection of global warming on the probabilistic behavior of hydro-climat... more s u m m a r y Accurate projection of global warming on the probabilistic behavior of hydro-climate variables is one of the main challenges in climate change impact assessment studies. Due to the complexity of climate-associated processes, different sources of uncertainty influence the projected behavior of hydro-climate variables in regression-based statistical downscaling procedures. The current study presents a comprehensive methodology to improve the predictive power of the procedure to provide improved projections. It does this by minimizing the uncertainty sources arising from the high-dimensionality of atmospheric predictors, the complex and nonlinear relationships between hydro-climate predictands and atmospheric predictors, as well as the biases that exist in climate model simulations. To address the impact of the high dimensional feature spaces, a supervised nonlinear dimensionality reduction algorithm is presented that is able to capture the nonlinear variability among projectors through extracting a sequence of principal components that have maximal dependency with the target hydro-climate variables. Two soft-computing nonlinear machine-learning methods, Support Vector Regression (SVR) and Relevance Vector Machine (RVM), are engaged to capture the nonlinear relationships between predictand and atmospheric predictors. To correct the spatial and temporal biases over multiple time scales in the GCM predictands, the Multivariate Recursive Nesting Bias Correction (MRNBC) approach is used. The results demonstrate that this combined approach significantly improves the downscaling procedure in terms of precipitation projection.
This study presents a comparison of two ecological and hydrological approaches for determining en... more This study presents a comparison of two ecological and hydrological approaches for determining environmental water requirements of the Gavkhuni wetland in Iran. Considering the ecological approach, water requirements of the wetland are determined based on a relationship between the biological requirements of a preferred biota (Artemia) and the inflow data on the wetland. The hydrological approach uses a flow duration curve (FDC) to gain a dependable base flow, whose water quality parameters remain at acceptable levels. The results reveal that there is no significant difference between the water requirement calculated by the two approaches, however the hydrological approach, which needs little information comparatively, may be used for those wetlands devoid of enough ecological information, especially in arid and semi-arid regions.
The present study applies different time-series models for forecasting daily and monthly snow wat... more The present study applies different time-series models for forecasting daily and monthly snow water equivalent (SWE) data in Ontario, Canada, during 1987-2011. For daily time series, which showed a significant negative trend, four categories of the autoregressive moving-average (ARMA) and ARMA model with exogenous variables (ARMAX) were applied. The North Atlantic Oscillation, Southern Oscillation Index and Pacific/North American Pattern, as large-scale atmospheric anomalies, as well as temperature time series are considered as exogenous variables for ARMAX models. According to the multicriteria performance evaluation, a time-trend ARMAX model demonstrated the best performance for modelling and forecasting daily SWE. Two models, seasonal autoregressive integrated moving average (SARIMA) and SARIMA with exogenous variables (SARIMAX), were also fitted to the monthly SWE time series. The results revealed that the SARIMAX model showed a better performance than the SARIMA model according to multicriteria evaluation. The three nonparametric tests, Wilcoxon, Levene and Kolmogorov-Smirnov for forecasting evaluation demonstrated that the selected time-series models had enough reliability for short-term SWE forecasting in Ontario. The results of this study also demonstrate the importance of incorporating both trend and appropriate exogenous variables for SWE time-series modelling and forecasting.
1] The present study performs the spatial and temporal trend analysis of the annual and 24-hr max... more 1] The present study performs the spatial and temporal trend analysis of the annual and 24-hr maximum rainfall of a set of 145 precipitation gauging stations of Iran. The study shows that the annual rainfall is decreasing at 67% of the stations while the 24-hr maximum rainfall is increasing at 50% of the stations. The negative trends of annual rainfall are mostly observed in northern and northwestern regions, whereas positive trends of 24-hr maximum rainfall are mostly located in arid and semiarid regions of Iran. However, the Kolmogorov-Smirnov test for Mann-Kendall (MK) statistics show that the regional trend of annual rainfall is significant, but it is not significant for 24-hr maximum rainfall. On the other hand, the sequential MK test reveals that the trends of annual rainfall and 24-hr maximum rainfall began since 1970s for most of the stations. The negative trend of the rainfall for most of the country may show the initial stages of climate change in Iran, but further information and analysis is required for future studies.
Improving techniques of flood frequency estimation at ungauged catchments is one of the major cha... more Improving techniques of flood frequency estimation at ungauged catchments is one of the major challenges for hydrologists, especially in arid and semi-arid regions with insufficient information. Recently, popularity of flood seasonality-based descriptors has increased among hydrologists for delineating of hydrologically homogenous regions.
Iran is a large country with diverse geophysical and climatic conditions which are influenced by ... more Iran is a large country with diverse geophysical and climatic conditions which are influenced by both large atmospheric circulation patterns and local effects. The density of rainfall station network of Iran is not enough for rainfall estimation at ungauged regions. Therefore, rainfall regionalization should be used to extend rainfall data to regions where rainfall data are not available. The aim of this study is to use cluster analysis and L-moment methods together to quantify regional rainfall patterns of Iran using annual rainfall of 137 stations for the period of 1952-2003. The cluster analysis follows "Ward's method" and shows eight regions of rainfall in Iran. The homogeneity test of L-moments shows that most of these regions are homogeneous. Using the goodness-of-fit test, Z Dist , the regional frequency distribution functions for each group are then selected. The 3-parameter Log Normal (LN3), Generalized Extreme Value (GEV) and generalize logistic (GLOG) distributions are selected for the first, second and the remaining 6 regions of rainfall over Iran, respectively. However, because of different rainfall generating mechanisms in Iran such as elevation, sea neighborhood and large atmospheric circulation systems, no parent distribution could be found for the entire country.
Flood hazard mapping Ungauged reaches High resolution Cartosat-1 based DTM HEC RAS Flood frequenc... more Flood hazard mapping Ungauged reaches High resolution Cartosat-1 based DTM HEC RAS Flood frequency analysis Arid regions s u m m a r y This study presents an exhaustive methodology of floodplain mapping at ungauged rivers. To present our methodology, we selected the Halilrud basin and Jiroft city in southeastern Iran as an example of hazardous regions. To estimate flood quantiles in different return periods at ungauged reaches, we used regional flood frequency analysis. By using the well-known L-moments approach and related criteria, a homogeneous region was formed and the 3-parameter Log normal distribution was identified as the robust regional frequency distribution. The hydro-geomorphic characteristics and the land use properties of the catchments were then extracted using RS&GIS techniques to establish multivariate regional regression models between hydro-geomorphic characteristics and flood quantiles. After delineation of the catchments for the ungauged reaches, flood quantiles as an important factor in flood inundation at outlets of these reaches with different probabilities were estimated using the regional regression models. To delineate flood hazard maps and to enhance the accuracy of the hydraulic modeling, we applied satellite stereoscope images of Cartosat-1 along with the Rational Polynomial Coefficients to extract a high resolution DTM and detailed parameterization of the channel required by 1D hydraulic model HEC-RAS. The GIS-based HEC-Geo RAS pre-and post-processor were also used for careful optimization of the geometry features for real visualization of the flood prone areas. Information of some historical flood events was also used to evaluate the hydraulic model performance in predicting flood inundations. Finally, vulnerable areas were crossed with extracted land use mapping from IRS-P6 satellite images to differentiate the critical infrastructures and the valuable land use classes affected by floods in different return periods.
Reliable estimation of low flows at ungauged catchments is one of the major challenges in water-r... more Reliable estimation of low flows at ungauged catchments is one of the major challenges in water-resources planning and management. This study aims at providing at-site and ungauged sites low-flow frequency analysis using regionalization approach. A two-stage delineating homogeneous region is proposed in this study. Clustering sites with similar low-flow L-moment ratios is initially conducted, and L-moment-based discordancy and heterogeneity measures are then used to detect unusual sites. Based on the goodness-of-fit test statistic, the best-fit regional model is identified in each hydrologically homogeneous region. The relationship between mean annual 7-day minimum flow and hydro-geomorphic characteristics is also constructed in each homogeneous region associated with the derived regional model for estimating various low-flow quantiles at ungauged sites. Uncertainty analysis of model parameters and low-flow estimations is carried out using the Bayesian inference. Applied in Sefidroud basin located in northwestern Iran, two hydrologically homogeneous regions are identified, i.e. the east and west regions. The best-fit regional model for the east and west regions are generalized logistic and Pearson type III distributions, respectively. The results show that the proposed approach provides reasonably good accuracy for at-site as well as ungauged-site frequency analysis. Besides, interval estimations for model parameters and low flows provide uncertainty information, and the results indicate that Bayesian confidence intervals are significantly reduced when comparing with the outcomes of conventional t-distribution method.
This paper presents a methodology for regional frequency analysis and spatial pattern features of... more This paper presents a methodology for regional frequency analysis and spatial pattern features of the Annual Maximum Dry Spell Length (AMDSL) as an indicator of drought conditions using the well-known L-moments approach and statistical-based methods. Applying Ward's cluster-analysis method identifies eight regions with distinctive AMDSL behaviours for Iran. Homogeneity testing indicates that most of these regions are homogenous. The goodness-of-fit test Z Dist shows that Generalized Logistic; Generalized Extreme Value and Pearson type III, distributions fit best for most regions. The spatial pattern of L-Moment statistics demonstrates that although the northwestern and northern parts of the country experience short dry spells, these periods are inconstant, and extreme dry spell events may happen in these areas. Almost all spatial mapping of AMDSLs at different probabilistic levels demonstrates that dry spells increase gradually from west to east and from north to south, and the southern parts (especially along the Persian Gulf and Oman Sea) and central areas, including most agricultural lands, stand out as the most sensitive to soil moisture deficits, because of longer lasting droughts.
This study compared three hydrological methods e low flow frequency analysis, flow duration curve... more This study compared three hydrological methods e low flow frequency analysis, flow duration curve (FDC), and concentration-discharge (C-D) modeling e to determine the water requirement of the Gavkhuni wetland located in an arid region of Iran. Due to the effects of water impounding upstream on the hydrological regime of the Gavkhuni's inflow after construction of the Zayandehrud dam, none of the methods produced satisfactory results.
Hydrologic drought is a type of drought which directly affects the water supply of a region. Long... more Hydrologic drought is a type of drought which directly affects the water supply of a region. Long streamflow dry spells or streamflow under a specific threshold are usually considered as hydrologic drought. The annual extreme hydrologic dry spell length ͑AEHDSL͒ data of the Halilrud basin in the southeastern semiarid region of Iran were considered to estimate the return period of hydrologic drought and the associated risk in this region. The method of L-moments was applied to check discordant stations and test the homogeneity of the region which consists of 15 gauging watersheds. One discordant station was found and the region was homogeneous according to the homogeneity measure after removing the discordant station. The three-parameter lognormal distribution was found to be representative of the regional distribution for the entire region based on the goodness-of-fit test. For prediction in ungauged basins, the AEHDSL regional regression was developed for the region. The regression model indicates that the vegetation cover and relief of watersheds play important roles in the hydrologic drought length of the Halilrud basin. These two variables control the infiltration and hydraulic slopes of a watershed, respectively.
A time-varying risk analysis is proposed for an adaptive design framework in nonstationary condit... more A time-varying risk analysis is proposed for an adaptive design framework in nonstationary conditions arising from climate change. A Bayesian, dynamic conditional copula is developed for modeling the time-varying dependence structure between mixed continuous and discrete multiattributes of multidi-mensional hydrometeorological phenomena. Joint Bayesian inference is carried out to fit the marginals and copula in an illustrative example using an adaptive, Gibbs Markov Chain Monte Carlo (MCMC) sampler. Posterior mean estimates and credible intervals are provided for the model parameters and the Deviance Information Criterion (DIC) is used to select the model that best captures different forms of nonstationarity over time. This study also introduces a fully Bayesian, time-varying joint return period for multivariate time-dependent risk analysis in nonstationary environments. The results demonstrate that the nature and the risk of extreme-climate multidimensional processes are changed over time under the impact of climate change, and accordingly the long-term decision making strategies should be updated based on the anomalies of the nonstationary environment.
s u m m a r y Accurate projection of global warming on the probabilistic behavior of hydro-climat... more s u m m a r y Accurate projection of global warming on the probabilistic behavior of hydro-climate variables is one of the main challenges in climate change impact assessment studies. Due to the complexity of climate-associated processes, different sources of uncertainty influence the projected behavior of hydro-climate variables in regression-based statistical downscaling procedures. The current study presents a comprehensive methodology to improve the predictive power of the procedure to provide improved projections. It does this by minimizing the uncertainty sources arising from the high-dimensionality of atmospheric predictors, the complex and nonlinear relationships between hydro-climate predictands and atmospheric predictors, as well as the biases that exist in climate model simulations. To address the impact of the high dimensional feature spaces, a supervised nonlinear dimensionality reduction algorithm is presented that is able to capture the nonlinear variability among projectors through extracting a sequence of principal components that have maximal dependency with the target hydro-climate variables. Two soft-computing nonlinear machine-learning methods, Support Vector Regression (SVR) and Relevance Vector Machine (RVM), are engaged to capture the nonlinear relationships between predictand and atmospheric predictors. To correct the spatial and temporal biases over multiple time scales in the GCM predictands, the Multivariate Recursive Nesting Bias Correction (MRNBC) approach is used. The results demonstrate that this combined approach significantly improves the downscaling procedure in terms of precipitation projection.
This study presents a comparison of two ecological and hydrological approaches for determining en... more This study presents a comparison of two ecological and hydrological approaches for determining environmental water requirements of the Gavkhuni wetland in Iran. Considering the ecological approach, water requirements of the wetland are determined based on a relationship between the biological requirements of a preferred biota (Artemia) and the inflow data on the wetland. The hydrological approach uses a flow duration curve (FDC) to gain a dependable base flow, whose water quality parameters remain at acceptable levels. The results reveal that there is no significant difference between the water requirement calculated by the two approaches, however the hydrological approach, which needs little information comparatively, may be used for those wetlands devoid of enough ecological information, especially in arid and semi-arid regions.
The present study applies different time-series models for forecasting daily and monthly snow wat... more The present study applies different time-series models for forecasting daily and monthly snow water equivalent (SWE) data in Ontario, Canada, during 1987-2011. For daily time series, which showed a significant negative trend, four categories of the autoregressive moving-average (ARMA) and ARMA model with exogenous variables (ARMAX) were applied. The North Atlantic Oscillation, Southern Oscillation Index and Pacific/North American Pattern, as large-scale atmospheric anomalies, as well as temperature time series are considered as exogenous variables for ARMAX models. According to the multicriteria performance evaluation, a time-trend ARMAX model demonstrated the best performance for modelling and forecasting daily SWE. Two models, seasonal autoregressive integrated moving average (SARIMA) and SARIMA with exogenous variables (SARIMAX), were also fitted to the monthly SWE time series. The results revealed that the SARIMAX model showed a better performance than the SARIMA model according to multicriteria evaluation. The three nonparametric tests, Wilcoxon, Levene and Kolmogorov-Smirnov for forecasting evaluation demonstrated that the selected time-series models had enough reliability for short-term SWE forecasting in Ontario. The results of this study also demonstrate the importance of incorporating both trend and appropriate exogenous variables for SWE time-series modelling and forecasting.
1] The present study performs the spatial and temporal trend analysis of the annual and 24-hr max... more 1] The present study performs the spatial and temporal trend analysis of the annual and 24-hr maximum rainfall of a set of 145 precipitation gauging stations of Iran. The study shows that the annual rainfall is decreasing at 67% of the stations while the 24-hr maximum rainfall is increasing at 50% of the stations. The negative trends of annual rainfall are mostly observed in northern and northwestern regions, whereas positive trends of 24-hr maximum rainfall are mostly located in arid and semiarid regions of Iran. However, the Kolmogorov-Smirnov test for Mann-Kendall (MK) statistics show that the regional trend of annual rainfall is significant, but it is not significant for 24-hr maximum rainfall. On the other hand, the sequential MK test reveals that the trends of annual rainfall and 24-hr maximum rainfall began since 1970s for most of the stations. The negative trend of the rainfall for most of the country may show the initial stages of climate change in Iran, but further information and analysis is required for future studies.
Improving techniques of flood frequency estimation at ungauged catchments is one of the major cha... more Improving techniques of flood frequency estimation at ungauged catchments is one of the major challenges for hydrologists, especially in arid and semi-arid regions with insufficient information. Recently, popularity of flood seasonality-based descriptors has increased among hydrologists for delineating of hydrologically homogenous regions.
Iran is a large country with diverse geophysical and climatic conditions which are influenced by ... more Iran is a large country with diverse geophysical and climatic conditions which are influenced by both large atmospheric circulation patterns and local effects. The density of rainfall station network of Iran is not enough for rainfall estimation at ungauged regions. Therefore, rainfall regionalization should be used to extend rainfall data to regions where rainfall data are not available. The aim of this study is to use cluster analysis and L-moment methods together to quantify regional rainfall patterns of Iran using annual rainfall of 137 stations for the period of 1952-2003. The cluster analysis follows "Ward's method" and shows eight regions of rainfall in Iran. The homogeneity test of L-moments shows that most of these regions are homogeneous. Using the goodness-of-fit test, Z Dist , the regional frequency distribution functions for each group are then selected. The 3-parameter Log Normal (LN3), Generalized Extreme Value (GEV) and generalize logistic (GLOG) distributions are selected for the first, second and the remaining 6 regions of rainfall over Iran, respectively. However, because of different rainfall generating mechanisms in Iran such as elevation, sea neighborhood and large atmospheric circulation systems, no parent distribution could be found for the entire country.
Flood hazard mapping Ungauged reaches High resolution Cartosat-1 based DTM HEC RAS Flood frequenc... more Flood hazard mapping Ungauged reaches High resolution Cartosat-1 based DTM HEC RAS Flood frequency analysis Arid regions s u m m a r y This study presents an exhaustive methodology of floodplain mapping at ungauged rivers. To present our methodology, we selected the Halilrud basin and Jiroft city in southeastern Iran as an example of hazardous regions. To estimate flood quantiles in different return periods at ungauged reaches, we used regional flood frequency analysis. By using the well-known L-moments approach and related criteria, a homogeneous region was formed and the 3-parameter Log normal distribution was identified as the robust regional frequency distribution. The hydro-geomorphic characteristics and the land use properties of the catchments were then extracted using RS&GIS techniques to establish multivariate regional regression models between hydro-geomorphic characteristics and flood quantiles. After delineation of the catchments for the ungauged reaches, flood quantiles as an important factor in flood inundation at outlets of these reaches with different probabilities were estimated using the regional regression models. To delineate flood hazard maps and to enhance the accuracy of the hydraulic modeling, we applied satellite stereoscope images of Cartosat-1 along with the Rational Polynomial Coefficients to extract a high resolution DTM and detailed parameterization of the channel required by 1D hydraulic model HEC-RAS. The GIS-based HEC-Geo RAS pre-and post-processor were also used for careful optimization of the geometry features for real visualization of the flood prone areas. Information of some historical flood events was also used to evaluate the hydraulic model performance in predicting flood inundations. Finally, vulnerable areas were crossed with extracted land use mapping from IRS-P6 satellite images to differentiate the critical infrastructures and the valuable land use classes affected by floods in different return periods.
Reliable estimation of low flows at ungauged catchments is one of the major challenges in water-r... more Reliable estimation of low flows at ungauged catchments is one of the major challenges in water-resources planning and management. This study aims at providing at-site and ungauged sites low-flow frequency analysis using regionalization approach. A two-stage delineating homogeneous region is proposed in this study. Clustering sites with similar low-flow L-moment ratios is initially conducted, and L-moment-based discordancy and heterogeneity measures are then used to detect unusual sites. Based on the goodness-of-fit test statistic, the best-fit regional model is identified in each hydrologically homogeneous region. The relationship between mean annual 7-day minimum flow and hydro-geomorphic characteristics is also constructed in each homogeneous region associated with the derived regional model for estimating various low-flow quantiles at ungauged sites. Uncertainty analysis of model parameters and low-flow estimations is carried out using the Bayesian inference. Applied in Sefidroud basin located in northwestern Iran, two hydrologically homogeneous regions are identified, i.e. the east and west regions. The best-fit regional model for the east and west regions are generalized logistic and Pearson type III distributions, respectively. The results show that the proposed approach provides reasonably good accuracy for at-site as well as ungauged-site frequency analysis. Besides, interval estimations for model parameters and low flows provide uncertainty information, and the results indicate that Bayesian confidence intervals are significantly reduced when comparing with the outcomes of conventional t-distribution method.
This paper presents a methodology for regional frequency analysis and spatial pattern features of... more This paper presents a methodology for regional frequency analysis and spatial pattern features of the Annual Maximum Dry Spell Length (AMDSL) as an indicator of drought conditions using the well-known L-moments approach and statistical-based methods. Applying Ward's cluster-analysis method identifies eight regions with distinctive AMDSL behaviours for Iran. Homogeneity testing indicates that most of these regions are homogenous. The goodness-of-fit test Z Dist shows that Generalized Logistic; Generalized Extreme Value and Pearson type III, distributions fit best for most regions. The spatial pattern of L-Moment statistics demonstrates that although the northwestern and northern parts of the country experience short dry spells, these periods are inconstant, and extreme dry spell events may happen in these areas. Almost all spatial mapping of AMDSLs at different probabilistic levels demonstrates that dry spells increase gradually from west to east and from north to south, and the southern parts (especially along the Persian Gulf and Oman Sea) and central areas, including most agricultural lands, stand out as the most sensitive to soil moisture deficits, because of longer lasting droughts.
This study compared three hydrological methods e low flow frequency analysis, flow duration curve... more This study compared three hydrological methods e low flow frequency analysis, flow duration curve (FDC), and concentration-discharge (C-D) modeling e to determine the water requirement of the Gavkhuni wetland located in an arid region of Iran. Due to the effects of water impounding upstream on the hydrological regime of the Gavkhuni's inflow after construction of the Zayandehrud dam, none of the methods produced satisfactory results.
Hydrologic drought is a type of drought which directly affects the water supply of a region. Long... more Hydrologic drought is a type of drought which directly affects the water supply of a region. Long streamflow dry spells or streamflow under a specific threshold are usually considered as hydrologic drought. The annual extreme hydrologic dry spell length ͑AEHDSL͒ data of the Halilrud basin in the southeastern semiarid region of Iran were considered to estimate the return period of hydrologic drought and the associated risk in this region. The method of L-moments was applied to check discordant stations and test the homogeneity of the region which consists of 15 gauging watersheds. One discordant station was found and the region was homogeneous according to the homogeneity measure after removing the discordant station. The three-parameter lognormal distribution was found to be representative of the regional distribution for the entire region based on the goodness-of-fit test. For prediction in ungauged basins, the AEHDSL regional regression was developed for the region. The regression model indicates that the vegetation cover and relief of watersheds play important roles in the hydrologic drought length of the Halilrud basin. These two variables control the infiltration and hydraulic slopes of a watershed, respectively.
Uploads
Papers by Ali Sarhadi