Search

Search Results

• 1. [Article] Identification of Appropriate Lags and Temporal Resolutions for Low Flow Indicators in the River Rhine to Forecast Low Flows with Different Lead Times

  The aim of this paper is to assess the relative importance of low flow indicators for the River Rhine and to identify their appropriate temporal lag and resolution. This is done in the context of low flow ...

  Citation × Citation Citation Abstract Copy Title: Identification of Appropriate Lags and Temporal Resolutions for Low Flow Indicators in the River Rhine to Forecast Low Flows with Different Lead Times Author: Demirel, Mehmet, Booij, Martijn, Hoekstra, Arjen Year: 2012 The aim of this paper is to assess the relative importance of low flow indicators for the River Rhine and to identify their appropriate temporal lag and resolution. This is done in the context of low flow forecasting with lead times of 14 and 90 days. First, the Rhine basin is subdivided into seven sub-basins. By considering the dominant processes in the sub-basins, five low flow indicators were selected: precipitation, potential evapotranspiration, groundwater storage, snow storage and lake storage. Correlation analysis was then carried out to determine the relationship between observed low flows and preselected indicators with varying lags (days) and temporal resolutions (from 1 day to 7 months). The results show that the most important low flow indicators in the Alpine sub-basins for forecasts with a lead time of 14 days are potential evapotranspiration with a large lag and temporal resolution, and lake levels with a small lag and temporal resolution. In the other sub-basins groundwater levels with a small lag and temporal resolution are important in addition to potential evapotranspiration with a large lag and temporal resolution. The picture is slightly different for forecasts with a lead time of 90 days. The snow storage in the Alpine sub-basins and the precipitation in the other sub-basins also become relevant for low flows. Consequently, the most important low flow indicators in the Alpine sub-basins for forecasts with a lead time of 90 days are potential evapotranspiration with a large lag and temporal resolution, lake levels with a small lag and temporal resolution and snow storage with a small lag and large temporal resolution. The resultant correlation maps provide appropriate lags and temporal resolutions for indicators to forecast low flows in the River Rhine with different lead times. Title: Identification of Appropriate Lags and Temporal Resolutions for Low Flow Indicators in the River Rhine to Forecast Low Flows with Different Lead Times Author: Demirel, Mehmet, Booij, Martijn, Hoekstra, Arjen Year: 2012 The aim of this paper is to assess the relative importance of low flow indicators for the River Rhine and to identify their appropriate temporal lag and resolution. This is done in the context of low flow forecasting with lead times of 14 and 90 days. First, the Rhine basin is subdivided into seven sub-basins. By considering the dominant processes in the sub-basins, five low flow indicators were selected: precipitation, potential evapotranspiration, groundwater storage, snow storage and lake storage. Correlation analysis was then carried out to determine the relationship between observed low flows and preselected indicators with varying lags (days) and temporal resolutions (from 1 day to 7 months). The results show that the most important low flow indicators in the Alpine sub-basins for forecasts with a lead time of 14 days are potential evapotranspiration with a large lag and temporal resolution, and lake levels with a small lag and temporal resolution. In the other sub-basins groundwater levels with a small lag and temporal resolution are important in addition to potential evapotranspiration with a large lag and temporal resolution. The picture is slightly different for forecasts with a lead time of 90 days. The snow storage in the Alpine sub-basins and the precipitation in the other sub-basins also become relevant for low flows. Consequently, the most important low flow indicators in the Alpine sub-basins for forecasts with a lead time of 90 days are potential evapotranspiration with a large lag and temporal resolution, lake levels with a small lag and temporal resolution and snow storage with a small lag and large temporal resolution. The resultant correlation maps provide appropriate lags and temporal resolutions for indicators to forecast low flows in the River Rhine with different lead times. Close

• 2. [Article] Validation of an ANN Flow Prediction Model Using a Multt-Station Cluster Analysis

  The objective of this study is to validate a flow prediction model for a hydrometric station using a multistation criterion in addition to standard single-station performance criteria. In this contribution ...

  Citation × Citation Citation Abstract Copy Title: Validation of an ANN Flow Prediction Model Using a Multt-Station Cluster Analysis Author: Demirel, Mehmet, Booij, Martijn, Kahya, Ercan Year: 2012 The objective of this study is to validate a flow prediction model for a hydrometric station using a multistation criterion in addition to standard single-station performance criteria. In this contribution we used cluster analysis to identify the regional flow height, i.e., water-level patterns and validate the output of an artificial neural network (ANN) model of the Alportel River in Portugal. Measurements of precipitation, temperature, and flow height were used as input variables to the ANN model with a lead time of 12 h. The lead time of 12 h is assumed to be appropriate for a short-term hydrological prediction since it is meaningful for physical processes. The ANN model with three inputs, four hidden neurons, and ten epochs was tested using the new model-validation criterion. The high performance of the model (i.e., Nash-Sutcliffe coefficient is equal to 0.922) was confirmed by the cluster-analysis criterion. It can be concluded that a multistation-based approach can be used as an additional validation criterion and might result in a rejection of a model which initially passed a single-station validation criterion. Title: Validation of an ANN Flow Prediction Model Using a Multt-Station Cluster Analysis Author: Demirel, Mehmet, Booij, Martijn, Kahya, Ercan Year: 2012 The objective of this study is to validate a flow prediction model for a hydrometric station using a multistation criterion in addition to standard single-station performance criteria. In this contribution we used cluster analysis to identify the regional flow height, i.e., water-level patterns and validate the output of an artificial neural network (ANN) model of the Alportel River in Portugal. Measurements of precipitation, temperature, and flow height were used as input variables to the ANN model with a lead time of 12 h. The lead time of 12 h is assumed to be appropriate for a short-term hydrological prediction since it is meaningful for physical processes. The ANN model with three inputs, four hidden neurons, and ten epochs was tested using the new model-validation criterion. The high performance of the model (i.e., Nash-Sutcliffe coefficient is equal to 0.922) was confirmed by the cluster-analysis criterion. It can be concluded that a multistation-based approach can be used as an additional validation criterion and might result in a rejection of a model which initially passed a single-station validation criterion. Close

• 3. [Article] The Skill of Seasonal Ensemble Low-Flow Forecasts in the Moselle River for Three Different Hydrological Models

  This paper investigates the skill of 90-day low-flow forecasts using two conceptual hydrological models and one data-driven model based on Artificial Neural Networks (ANNs) for the Moselle River. The three ...

  Citation × Citation Citation Abstract Copy Title: The Skill of Seasonal Ensemble Low-Flow Forecasts in the Moselle River for Three Different Hydrological Models Author: Demirel, Mehmet C., Booij, Martijn, Hoekstra, Arjen Year: 2015 This paper investigates the skill of 90-day low-flow forecasts using two conceptual hydrological models and one data-driven model based on Artificial Neural Networks (ANNs) for the Moselle River. The three models, i.e. HBV, GR4J and ANN-Ensemble (ANN-E), all use forecasted meteorological inputs (precipitation P and potential evapotranspiration PET), whereby we employ ensemble seasonal meteorological forecasts. We compared low-flow forecasts for five different cases of seasonal meteorological forcing: (1) ensemble P and PET forecasts; (2) ensemble P forecasts and observed climate mean PET; (3) observed climate mean P and ensemble PET forecasts; (4) observed climate mean P and PET and (5) zero P and ensemble PET forecasts as input for the models. The ensemble P and PET forecasts, each consisting of 40 members, reveal the forecast ranges due to the model inputs. The five cases are compared for a lead time of 90 days based on model output ranges, whereas the models are compared based on their skill of low-flow forecasts for varying lead times up to 90 days. Before forecasting, the hydrological models are calibrated and validated for a period of 30 and 20 years respectively. The smallest difference between calibration and validation performance is found for HBV, whereas the largest difference is found for ANN-E. From the results, it appears that all models are prone to over-predict runoff during low-flow periods using ensemble seasonal meteorological forcing. The largest range for 90-day low-flow forecasts is found for the GR4J model when using ensemble seasonal meteorological forecasts as input. GR4J, HBV and ANN-E under-predicted 90-day-ahead low flows in the very dry year 2003 without precipitation data. The results of the comparison of forecast skills with varying lead times show that GR4J is less skilful than ANN-E and HBV. Overall, the uncertainty from ensemble Pforecasts has a larger effect on seasonal low-flow forecasts than the uncertainty from ensemble PET forecasts and initial model conditions. Title: The Skill of Seasonal Ensemble Low-Flow Forecasts in the Moselle River for Three Different Hydrological Models Author: Demirel, Mehmet C., Booij, Martijn, Hoekstra, Arjen Year: 2015 This paper investigates the skill of 90-day low-flow forecasts using two conceptual hydrological models and one data-driven model based on Artificial Neural Networks (ANNs) for the Moselle River. The three models, i.e. HBV, GR4J and ANN-Ensemble (ANN-E), all use forecasted meteorological inputs (precipitation P and potential evapotranspiration PET), whereby we employ ensemble seasonal meteorological forecasts. We compared low-flow forecasts for five different cases of seasonal meteorological forcing: (1) ensemble P and PET forecasts; (2) ensemble P forecasts and observed climate mean PET; (3) observed climate mean P and ensemble PET forecasts; (4) observed climate mean P and PET and (5) zero P and ensemble PET forecasts as input for the models. The ensemble P and PET forecasts, each consisting of 40 members, reveal the forecast ranges due to the model inputs. The five cases are compared for a lead time of 90 days based on model output ranges, whereas the models are compared based on their skill of low-flow forecasts for varying lead times up to 90 days. Before forecasting, the hydrological models are calibrated and validated for a period of 30 and 20 years respectively. The smallest difference between calibration and validation performance is found for HBV, whereas the largest difference is found for ANN-E. From the results, it appears that all models are prone to over-predict runoff during low-flow periods using ensemble seasonal meteorological forcing. The largest range for 90-day low-flow forecasts is found for the GR4J model when using ensemble seasonal meteorological forecasts as input. GR4J, HBV and ANN-E under-predicted 90-day-ahead low flows in the very dry year 2003 without precipitation data. The results of the comparison of forecast skills with varying lead times show that GR4J is less skilful than ANN-E and HBV. Overall, the uncertainty from ensemble Pforecasts has a larger effect on seasonal low-flow forecasts than the uncertainty from ensemble PET forecasts and initial model conditions. Close

• 4. [Article] Low Flow Forecasting with a Lead Time of 14 days for Navigation and Energy Supply in the Rhine River

  Low flow forecasting, days or even months in advance, is particularly important to the efficient operation of power plants and freight shipment. This study presents a low flow forecasting model with a ...

  Citation × Citation Citation Abstract Copy Title: Low Flow Forecasting with a Lead Time of 14 days for Navigation and Energy Supply in the Rhine River Author: Demirel, Mehmet C., Booij, Martijn Year: 2011 Low flow forecasting, days or even months in advance, is particularly important to the efficient operation of power plants and freight shipment. This study presents a low flow forecasting model with a lead time of 14 days for the Rhine River. The forecasts inherit uncertainty sources mainly because of model parameterization. Therefore, a systematic uncertainty analysis is applied to indicate the major uncertainty sources in the results. Firstly, the Rhine basin is divided into 7 major sub-basins. Each sub-basin is modeled separately with a data-driven model and the output discharges are routed to Lobith after German-Dutch border with another data-driven model. Five pre-selected low flow indicators (basin averaged precipitation, basin averaged potential evapotranspiration; basin averaged fresh snow depths, basin averaged groundwater levels and major lake levels in the sub-basins) are used as inputs to the models. The basin discretization and the selection of indicators are based on a literature study and seasonality analysis of the discharge time series from 108 sub-basins. The correlations between indicator and low flows with varying temporal resolution and varying lags between indicator and low flows were used to identify appropriate temporal scales of the model inputs. We assume that a suitable model structure for the Rhine basin has been determined; that is, the sub-system boundaries have been specified, the important state variables and input and output fluxes to be included have been identified and selected for each sub-basin. The results in this study show that the data-driven models used for each sub-basin are capable of representing the essential characteristics of the system based on lagged and temporally averaged low flow indicators and are forecasting low flows adequately. In addition to the forecast results, the uncertainties due to specific model parameters in the calibrated data-driven model corresponding to key processes are also given. Title: Low Flow Forecasting with a Lead Time of 14 days for Navigation and Energy Supply in the Rhine River Author: Demirel, Mehmet C., Booij, Martijn Year: 2011 Low flow forecasting, days or even months in advance, is particularly important to the efficient operation of power plants and freight shipment. This study presents a low flow forecasting model with a lead time of 14 days for the Rhine River. The forecasts inherit uncertainty sources mainly because of model parameterization. Therefore, a systematic uncertainty analysis is applied to indicate the major uncertainty sources in the results. Firstly, the Rhine basin is divided into 7 major sub-basins. Each sub-basin is modeled separately with a data-driven model and the output discharges are routed to Lobith after German-Dutch border with another data-driven model. Five pre-selected low flow indicators (basin averaged precipitation, basin averaged potential evapotranspiration; basin averaged fresh snow depths, basin averaged groundwater levels and major lake levels in the sub-basins) are used as inputs to the models. The basin discretization and the selection of indicators are based on a literature study and seasonality analysis of the discharge time series from 108 sub-basins. The correlations between indicator and low flows with varying temporal resolution and varying lags between indicator and low flows were used to identify appropriate temporal scales of the model inputs. We assume that a suitable model structure for the Rhine basin has been determined; that is, the sub-system boundaries have been specified, the important state variables and input and output fluxes to be included have been identified and selected for each sub-basin. The results in this study show that the data-driven models used for each sub-basin are capable of representing the essential characteristics of the system based on lagged and temporally averaged low flow indicators and are forecasting low flows adequately. In addition to the forecast results, the uncertainties due to specific model parameters in the calibrated data-driven model corresponding to key processes are also given. Close

• 5. [Article] Identification of Appropriate Temporal Scales of Dominant Low Flow Indicators in the Main River, Germany

  Models incorporating the appropriate temporal scales of dominant indicators for low flows are assumed to perform better than models with arbitrary selected temporal scales. In this paper, we investigate ...

  Citation × Citation Citation Abstract Copy Title: Identification of Appropriate Temporal Scales of Dominant Low Flow Indicators in the Main River, Germany Author: Demirel, Mehmet, Booij, Martijn Year: 2010 Models incorporating the appropriate temporal scales of dominant indicators for low flows are assumed to perform better than models with arbitrary selected temporal scales. In this paper, we investigate appropriate temporal scales of dominant low flow indicators: precipitation (P), evapotranspiration (ET) and the standardized groundwater storage index (G). This analysis is done in the context of low flow forecasting with a lead time of 14 days in the Main River, a tributary of the Rhine River, located in Germany. Correlation coefficients (i.e. Pearson, Kendall and Spearman) are used to reveal the appropriate temporal scales of dominant low flow indicators at different time lags between low flows and indicators and different support scales of indicators. The results are presented for lag values and support scales, which result in correlation coefficients between low flows and dominant indicators falling into the maximum 10% percentile range. P has a maximum Spearman correlation coefficient (ρ) of 0.38 (p = 0.95) at a support scale of 336 days and a lag of zero days. ET has a maximum ρ of –0.60 (p = 0.95) at a support scale of 280 days and a lag of 56 days and G has a maximum ρ of 0.69 (p = 0.95) at a support scale of 7 days and a lag of 3 days. The identified appropriate support scales and lags can be used for low flow forecasting with a lead time of 14 days. Title: Identification of Appropriate Temporal Scales of Dominant Low Flow Indicators in the Main River, Germany Author: Demirel, Mehmet, Booij, Martijn Year: 2010 Models incorporating the appropriate temporal scales of dominant indicators for low flows are assumed to perform better than models with arbitrary selected temporal scales. In this paper, we investigate appropriate temporal scales of dominant low flow indicators: precipitation (P), evapotranspiration (ET) and the standardized groundwater storage index (G). This analysis is done in the context of low flow forecasting with a lead time of 14 days in the Main River, a tributary of the Rhine River, located in Germany. Correlation coefficients (i.e. Pearson, Kendall and Spearman) are used to reveal the appropriate temporal scales of dominant low flow indicators at different time lags between low flows and indicators and different support scales of indicators. The results are presented for lag values and support scales, which result in correlation coefficients between low flows and dominant indicators falling into the maximum 10% percentile range. P has a maximum Spearman correlation coefficient (ρ) of 0.38 (p = 0.95) at a support scale of 336 days and a lag of zero days. ET has a maximum ρ of –0.60 (p = 0.95) at a support scale of 280 days and a lag of 56 days and G has a maximum ρ of 0.69 (p = 0.95) at a support scale of 7 days and a lag of 3 days. The identified appropriate support scales and lags can be used for low flow forecasting with a lead time of 14 days. Close

• 6. [Article] Uncertainty Analysis of a Low Flow Model for the Rhine River

  It is widely recognized that hydrological models are subject to parameter uncertainty. However, little attention has been paid so far to the uncertainty in parameters of the data-driven models like weights ...

  Citation × Citation Citation Abstract Copy Title: Uncertainty Analysis of a Low Flow Model for the Rhine River Author: Demirel, Mehmet C., Booij, Martijn Year: 2011 It is widely recognized that hydrological models are subject to parameter uncertainty. However, little attention has been paid so far to the uncertainty in parameters of the data-driven models like weights in neural networks. This study aims at applying a structured uncertainty analysis to a data-driven model for low flow forecasting with a lead time of 14 days in the Rhine River. In the modeling phase, the Rhine basin is divided into seven major sub-basins. Each sub-basin is modeled separately with a data-driven model and the output discharges were routed to Lobith with another data-driven model. Basin averaged precipitation, basin averaged potential evapotranspiration, basin averaged fresh snow depths, basin averaged groundwater levels and major lake levels in the sub-basins are selected as low flow indicators and used as inputs to the models. The basin discretization and the selection of low flow indicators were not arbitrary since the dominant processes were considered by applying seasonality analysis to discharge time series from 108 sub-basins. Moreover, the correlations between indicators and low flows with varying temporal resolution and varying lags were used to identify appropriate temporal scales of the model inputs. The structure of the model can inherit uncertainty too due to many factors, including the lack of a robust hydrological theory at the spatial scale of the seven sub-basins. However, the parameter uncertainty is assumed to be the largest uncertainty source compared to other uncertainty sources. The effects of the input uncertainty were not assessed since averaging over sub-basins significantly reduces the measurement uncertainties. The model parameter sets were estimated using inverse modeling. The uncertainty of each weight is expressed as a probability distribution. Sensitivity analysis was applied for reducing the dimension and size of parameter space before uncertainty analysis. Finally, Monte Carlo Simulation was used to estimate the posterior distributions of the model outputs. The results in this study provide the effects of uncertainties in low flow model parameters on the model outputs. It has also been shown that the explicit assessment of uncertainties in the data-driven model parameters can lead to significant improvements in the information supply for low flow forecasting. Title: Uncertainty Analysis of a Low Flow Model for the Rhine River Author: Demirel, Mehmet C., Booij, Martijn Year: 2011 It is widely recognized that hydrological models are subject to parameter uncertainty. However, little attention has been paid so far to the uncertainty in parameters of the data-driven models like weights in neural networks. This study aims at applying a structured uncertainty analysis to a data-driven model for low flow forecasting with a lead time of 14 days in the Rhine River. In the modeling phase, the Rhine basin is divided into seven major sub-basins. Each sub-basin is modeled separately with a data-driven model and the output discharges were routed to Lobith with another data-driven model. Basin averaged precipitation, basin averaged potential evapotranspiration, basin averaged fresh snow depths, basin averaged groundwater levels and major lake levels in the sub-basins are selected as low flow indicators and used as inputs to the models. The basin discretization and the selection of low flow indicators were not arbitrary since the dominant processes were considered by applying seasonality analysis to discharge time series from 108 sub-basins. Moreover, the correlations between indicators and low flows with varying temporal resolution and varying lags were used to identify appropriate temporal scales of the model inputs. The structure of the model can inherit uncertainty too due to many factors, including the lack of a robust hydrological theory at the spatial scale of the seven sub-basins. However, the parameter uncertainty is assumed to be the largest uncertainty source compared to other uncertainty sources. The effects of the input uncertainty were not assessed since averaging over sub-basins significantly reduces the measurement uncertainties. The model parameter sets were estimated using inverse modeling. The uncertainty of each weight is expressed as a probability distribution. Sensitivity analysis was applied for reducing the dimension and size of parameter space before uncertainty analysis. Finally, Monte Carlo Simulation was used to estimate the posterior distributions of the model outputs. The results in this study provide the effects of uncertainties in low flow model parameters on the model outputs. It has also been shown that the explicit assessment of uncertainties in the data-driven model parameters can lead to significant improvements in the information supply for low flow forecasting. Close