Uncertainty Analysis Of Predictions Research Articles

Numerical Prediction Uncertainty and Data Worth Analysis of Solute Transport in an Agricultural Clay Till Setting With Preferential Flow

AbstractWith modern agricultural practices, it is essential to quantify flow and solute transport fluxes by numerical models and associated predictions. A major challenge in modeling preferential flow settings is the ability to constrain the often numerous parameters needed to physically represent these systems. Following this, there is a lack of understanding of what parameters and observations carry the most worth for a model to reduce its prediction uncertainty. Here, first‐order second moment (FOSM) analyses were used for a heavily monitored clay till field with preferential flow to investigate which parameters and observation types contribute the most to reducing the uncertainty of bromide predictions at various depths. Using a multi‐objective regularized optimization approach, a 1‐D preferential flow model was calibrated and subjected to FOSM analyses. Key parameters contributing to the prediction uncertainty of bromide concentrations in 0.25–3 m were limited to the lower boundary condition, the mass transfer coefficient, the hydraulic conductivity of the micro‐ and macropore, the macropore porosity, and the water content at wilting point. The data with the largest worth and ability to reduce the pre‐calibration prediction uncertainty were concentration observations closest to the sought prediction depth, drain concentrations, and averaged water table measurements from the entire field. The post‐calibration prediction uncertainty was increased primarily by removing concentration observations closest to the prediction depths. While this study provided new insights into parameter importance and data worth further research is required to understand if these findings apply broadly to clay till settings (or any soil setting) with preferential flow.

Discharge estimation in compound channels with converging and diverging floodplains using an optimised Gradient Boosting Algorithm

ABSTRACT River discharge estimation is vital for effective flood management and infrastructure planning. River systems consist of a main channel and floodplains, collectively forming a compound channel, posing challenges in discharge calculation, particularly when floodplains converge or diverge. In the present study, ML algorithms such as XGBoost, CatBoost, and LightGBM were developed to predict discharge in a compound channel. PSO algorithm is applied for optimization of hyperparameters of gradient boosting models, denoted as PSO-XGBoost, PSO-LightGBM, and PSO-CatBoost. ML model discharge predictions were validated with existing empirical models and feature importance was explored using SHAP and sensitivity analysis. Results show that all three gradient-boosting algorithms effectively predict discharge in compound channels and are further enhanced by application of PSO algorithm. The R2 values for XGBoost, PSO-XGBoost, CatBoost, and PSO-CatBoost exceed 0.95, whereas they are above 0.85 for LightBoost and PSO-LightBoost. PSO-CatBoost performance is better than other models based on findings of statistical performance parameters, uncertainty analysis, reliability index, and resilience index for prediction of discharge in a compound channel with converging and diverging flood plains. The findings of this study validate the suitability of the proposed models especially optimized with PSO is recommended for predicting discharge in a compound channel.

Quantifying predictive uncertainty in damage classification for nondestructive evaluation using Bayesian approximation and deep learning

Magnetic flux leakage (MFL), a widely used nondestructive evaluation (NDE) method, for inspecting pipelines to prevent potential long-term failures. However, during field testing, uncertainties can affect the accuracy of the inspection and the decision-making process regarding damage conditions. Therefore, it is essential to identify and quantify these uncertainties to ensure the reliability of the inspection. This study focuses on the uncertainties that arise during the inverse NDE process due to the dynamic magnetization process, which is affected by the relative motion of the MFL sensor and the material being tested. Specifically, the study investigates the uncertainties caused by sensing liftoff, which can affect the output signal of the sensing system. Due to the complexity of describing the forward uncertainty propagation process, this study compared two typical machine learning (ML)-based approximate Bayesian inference methods, convolutional neural network and deep ensemble, to address the input uncertainty from the MFL response data. Besides, an autoencoder method is applied to tackle the lack of experimental data for the training model by augmenting the dataset, which is constructed with the pre-trained model based on transfer learning. Prior knowledge learned from large simulated MFL signals can fine-tune the autoencoder model which enhances the subsequent learning process on experimental MFL data with faster generalization. The augmented data from the fine-tuned autoencoder is further applied for ML-based defect size classification. This study conducted prediction accuracy and uncertainty analysis with calibration, which can evaluate the prediction performance and reveal the relation between the liftoff uncertainty and prediction accuracy. Further, to strengthen the trustworthiness of the prediction results, the decision-making process guided by uncertainty is applied to provide valuable insights into the reliability of the final prediction results. Overall, the proposed framework for uncertainty quantification offers valuable insights into the assessment of reliability in MFL-based decision-making and inverse problems.

High-resolution digital mapping of soil erodibility in China

Soil erodibility (K) is the intrinsic susceptibility of a soil to water erosion. Currently, its detailed and accurate spatial distribution information especially over large areas is urgently required for national and regional soil erosion assessment and soil conservation decision making. This study combined pedotransfer function with digital soil mapping techniques to develop a high-resolution map of soil erodibility across China. The First, based on a recent national soil survey, we adopted the erosion-productivity impact calculator (EPIC) to calculate soil erodibility values at 4710 soil sampling points. Then, with the caclulated values of points, we used five techniques including polygon linking (PL), ordinary kriging (OK), Cubist, extreme gradient boosting (XGBoost), and random forests (RF) to generate spatial distribution of soil erodibility. The three latter machine learning techniques modeled the quantitative relationships between soil erodibility and a set of environmental covariates. The results showed that machine learning methods exhibited much more spatial details than the PL and OK did. Among the five techniques the RF achieved the highest accuracy with R2 of 0.49 and RMSE of 0.0077 t ha h ha−1 MJ−1 mm−1 based on 10-fold cross-validation. Spatial uncertainty analysis of the RF predictions showed that high uncerntainty occurred in northwestern China and low uncertainty in the center and southeast. We found that topographical and climatic variables are major environmental factors indirectly controlling spatial variation of soil erodibility while the soil particle composition and SOC contents directly influence the variation.

Translating pumping test data into groundwater model parameters: a workflow to reveal aquifer heterogeneities and implications in regional model parameterization

Groundwater modelers frequently grapple with the challenge of integrating aquifer test interpretations into parameters used by regional models. This task is complicated by issues of upscaling, data assimilation, and the need to assign prior probability distributions to numerical model parameters in order to support model predictive uncertainty analysis. To address this, we introduce a new framework that bridges the significant scale differences between aquifer tests and regional models. This framework also accounts for loss of original datasets and the heterogeneous nature of geological media in which aquifer testing often takes place. Using a fine numerical grid, the aquifer test is reproduced in a way that allows stochastic representation of site hydraulic properties at an arbitrary level of complexity. Data space inversion is then used to endow regional model cells with upscaled, aquifer-test-constrained realizations of numerical model properties. An example application demonstrates that assimilation of historical pumping test interpretations in this manner can be done relatively quickly. Furthermore, the assimilation process has the potential to significantly influence the posterior means of decision-pertinent model predictions. However, for the examples that we discuss, posterior predictive uncertainties do not undergo significant reduction. These results highlight the need for further research.

Fusing sensitive degradation features with uncertainty analysis for RUL prediction of rotating machines

Recently, deep learning technology-based neural networks have been adopted for remaining useful life (RUL) prediction of rotating machines. However, there are still some shortcomings: (1) an individual degradation feature cannot sufficiently represent the degradation process, which has an adverse impact on the accuracy of prediction results; (2) most recurrent neural network-based prediction methods have difficulty in quantifying the uncertainty of the forecast results. In this paper, a fusing sensitive degradation features with uncertainty analysis for RUL prediction of rotating machines is proposed. Firstly, the statistical features contained in the vibration signal used to monitor the degradation of rotating equipment are extracted to construct the original feature set. Then, the weight coefficients of the monotonicity, correlation and robustness criteria are determined by the self-adjusting analytic hierarchy process. The sensitive features that describe the degradation process are selected from among the statistical features. Furthermore, the sensitive features are fed into residual networks and gated recurrent unit, and the spatial and temporal correlation of the features are considered to establish the health index (HI). Finally, the fitted HI is input into a Gaussian process regression model, and the prediction results with confidence intervals are obtained. To verify the effectiveness and superiority of the proposed method, two public bearing datasets and three model methods are used for comparative experiments.

A Phenology-guided Bayesian-CNN (PB-CNN) framework for soybean yield estimation and uncertainty analysis

Large-scale crop yield estimation is important for understanding the response of agriculture production to environmental forces and management practices, and plays a critical role in insurance designing, trade decision making, and economic planning. The empirical models (e.g., deep learning models) have been increasingly utilized for estimating crop yields with the ability to take into account a range of yield predictors and complex modeling relationships. Yet empirical estimation of crop yields still faces important challenges, particularly in accommodating spatio-temporal crop phenological development patterns as well as tackling the heterogeneity of a diversity of yield predictors. The different types of uncertainties associated with empirical yield estimations have seldom been explored. The objective of this study is to develop a Phenology-guided Bayesian-Convolutional Neural Network (PB-CNN) framework for county-level crop yield estimation and uncertainty quantification, with soybean in the US Corn Belt as a case study. The PB-CNN framework comprises three key components: Phenology Imagery construction, multi-stream Bayesian-CNN modeling, as well as feature importance (i.e., yield predictor and phenological stage) and predictive uncertainty analysis (i.e., aleatoric and epistemic uncertainty). With the innovative integration of critical crop phenological stages in modeling the crop yield response to a heterogeneous set of yield predictors (i.e., satellite-based, heat-related, water-related, and soil predictors) as well as the associated uncertainties, the developed PB-CNN framework outperforms three advanced benchmark models, achieving an average RMSE of 4.622 bu/ac, an average R2 of 0.709, and an average bias of −2.057 bu/ac in estimating the county-level soybean yield of the US Corn Belt in testing years 2014–2018. Among the yield predictor groups, the satellite-based predictor group is the most critical in soybean yield estimation, followed by the water- and heat-related predictor groups. Throughout the growing season, the soybean blooming to dropping leaves phenological stages play a more crucial role in modeling the soybean yield. The soil predictor group as well as the early growing stages can improve the model estimation accuracy yet potentially brings more uncertainties into the yield estimation. The further uncertainty disentanglement indicates that the dominant uncertainty in yield estimation is the aleatoric uncertainty, mainly stemming from the fluctuations and variations inherent in the modeling input observations. The PB-CNN framework largely enhances our understanding of the complex soybean yield response to varying environmental conditions across crop phenological stages as well as associated uncertainties for more sustainable agricultural development.

Predictive uncertainty assessment in flood forecasting using quantile regression

Abstract Floods and their associated impacts are topics of concern in land development planning and management, which call for efficient flood forecasting and warning systems. The performance of flood warning systems is affected by uncertainty in water level forecasts, which is due to their inability to measure or calculate a modeled value accurately. Predictive uncertainty is an emerging type of uncertainty modeling technique that emphasizes total uncertainty quantified as a probability distribution conditioned on all available knowledge. Predictive uncertainty analysis was done using quantile regression (QR) for machine learning-based flood models – Hybrid Wavelet Artificial Neural Network model (WANN) and Hybrid Wavelet Support Vector Machine model (WSVM) for different lead times. Comparing QR models of WANN and WSVM revealed that the slope, intercept, spread of forecast, and width of confidence band of the WANN model are more for each quantile indicating more uncertainty as compared to the WSVM model. In both models, with an increase in lead time, uncertainty has shown an increasing trend as well. The performance evaluation of inference obtained from QR models was evaluated using uncertainty statistics such as prediction interval coverage probability, average relative interval length (ARIL), and mean prediction interval (MPI).

Data space inversion for efficient uncertainty quantification using an integrated surface and sub-surface hydrologic model

Abstract. It is incumbent on decision-support hydrological modelling to make predictions of uncertain quantities in a decision-support context. In implementing decision-support modelling, data assimilation and uncertainty quantification are often the most difficult and time-consuming tasks. This is because the imposition of history-matching constraints on model parameters usually requires a large number of model runs. Data space inversion (DSI) provides a highly model-run-efficient method for predictive uncertainty quantification. It does this by evaluating covariances between model outputs used for history matching (e.g. hydraulic heads) and model predictions based on model runs that sample the prior parameter probability distribution. By directly focusing on the relationship between model outputs under historical conditions and predictions of system behaviour under future conditions, DSI avoids the need to estimate or adjust model parameters. This is advantageous when using integrated surface and sub-surface hydrologic models (ISSHMs) because these models are associated with long run times, numerical instability and ideally complex parameterization schemes that are designed to respect geological realism. This paper demonstrates that DSI provides a robust and efficient means of quantifying the uncertainties of complex model predictions. At the same time, DSI provides a basis for complementary linear analysis that allows the worth of available observations to be explored, as well as of observations which are yet to be acquired. This allows for the design of highly efficient, future data acquisition campaigns. DSI is applied in conjunction with an ISSHM representing a synthetic but realistic river–aquifer system. Predictions of interest are fast travel times and surface water infiltration. Linear and non-linear estimates of predictive uncertainty based on DSI are validated against a more traditional uncertainty quantification which requires the adjustment of a large number of parameters. A DSI-generated surrogate model is then used to investigate the effectiveness and efficiency of existing and possible future monitoring networks. The example demonstrates the benefits of using DSI in conjunction with a complex numerical model to quantify predictive uncertainty and support data worth analysis in complex hydrogeological environments.

Predictive operation optimization of multi-energy virtual power plant considering behavior uncertainty of diverse stakeholders

To address the inflexibility of traditional single-energy virtual power plants (VPPs) in accommodating high proportions of renewable energy, this study proposes a predictive optimization method for multi-energy VPPs. The method supports multi-energy complementary cooperative dispatch and participation in multiple markets, and is applicable to future multi-energy VPPs that integrate carbon capture technology, power-to-gas, and energy storage. The method takes into account the uncertain parameters of stakeholders' behavior and introduces sliding time windows to improve production stability and the feasibility of VPP dispatch schemes. The optimization scheme is obtained based on the scenario method when the risk is maximum. The case results show that the proposed method can increase the profit by 8.31% compared with that without considering the time window. After stabilization, the ramping power changes by 17.39%. It is also found that accounting for the uncertainty of stakeholders increased the maximum profit drop from 47.70% to 60.45%. The introduction of power-to-gas and carbon capture technology has effectively improved the overall economy of the VPP and reduced carbon emissions. The proposed VPP predictive optimization method and uncertainty analysis of stakeholders' behavior provide basis for operation control of multi-energy VPPs.

Uncertainty analysis for prediction of elastic properties of unidirectional bamboo fiber-reinforced composites

This paper presents a virtual framework to obtain the elastic response of unidirectional (UD) continuous bamboo fiber-reinforced composites based on the micromechanical modeling approach. The virtual framework involves the development of an algorithm for the generation of stochastic Representative Volume Elements (RVEs), taking into account the micro-structural uncertainty and variability of the bio-based bamboo fibers and their composites. The developed algorithm and model are validated with experimental characterizations and theoretical models. Different statistical micro-structural realizations with random fiber packing, fiber geometries, and variable RVE sizes are compared to investigate the effects of these parameters on the accuracy of the prediction of the elastic properties of bamboo fiber-reinforced composites. All models and loading scenarios are applied automatically through Python scripting in the Abaqus finite element (FE) solver. The uncertainty analysis of input parameters is carried out through parametric studies. This virtual micromechanical framework can be further linked in the future to macroscale models of bamboo or other natural fiber composites for multiscale prediction of the behavior of biocomposite structures under in-service loading.

Successive-Station Streamflow Prediction and Precipitation Uncertainty Analysis in the Zarrineh River Basin Using a Machine Learning Technique

Precise forecasting of streamflow is crucial for the proper supervision of water resources. The purpose of the present investigation is to predict successive-station streamflow using the Gated Recurrent Unit (GRU) model and to quantify the impact of input information (i.e., precipitation) uncertainty on the GRU model’s prediction using the Generalized Likelihood Uncertainty Estimation (GLUE) computation. The Zarrineh River basin in Lake Urmia, Iran, was nominated as the case study due to the importance of the location and its significant contribution to the lake inflow. Four stations in the basin were considered to predict successive-station streamflow from upstream to downstream. The GRU model yielded highly accurate streamflow prediction in all stations. The future precipitation data generated under the Representative Concentration Pathway (RCP) scenarios were used to estimate the effect of precipitation input uncertainty on streamflow prediction. The p-factor (inside the uncertainty interval) and r-factor (width of the uncertainty interval) indices were used to evaluate the streamflow prediction uncertainty. GLUE predicted reliable uncertainty ranges for all the stations from 0.47 to 0.57 for the r-factor and 61.6% to 89.3% for the p-factor.

Uncertainty and spatial analysis in wheat yield prediction based on robust inclusive multiple models.

Reliable prediction of wheat yield ahead of harvest is a critical challenge for decision-makers along the supply chain. Predicting wheat yield is a real challenge for better agriculture and food security management. Modeling wheat yield is complex and challenging, so robust tools are needed. The main aim of this study is to predict wheat yield using an advanced ensemble model. A multilayer perceptron model (MLP) was combined with optimization algorithms to determine MLP parameters as the first step in the study. Several optimization algorithms were used as optimizers, including Particle Swarm Optimization (PSO), Honey Badger Algorithms (HBA), Sine-Cosine Algorithms (SCA), and Shark Algorithms (SA). Meteorological data were inserted into models.Next, the outputs of optimized MLP models were incorporated into an inclusive multiple MLP model (IMM).A new hybrid gamma test was used to determine the most appropriate input combination. A hybrid gamma test was created by coupling the HBA with GT.This paper introduces a robust IMM model, develops an MLP model using optimization algorithms, develops a new hybrid gamma test, uses Generalized Likelihood Uncertainty Estimation (GLUE) to analyze uncertainty, and presents a spatial map of wheat yield prediction. Based on the Gamma Test analysis, mean air temperature (Ta), wind speed (WS), relative humidity (RH), evapotranspiration (ET), and precipitation (P) were the most important input parameters for reliable and accurate winter wheat yield predictions. At the testing level, the IMM model decreased the mean absolute error (MAE) of the MLP-HBA, MLP-SCA, MLP-SA, MLP-PSO, and MLP models by 47%, 52%, 55%, 58%, and 61%, respectively. In the study, the uncertainty of models based on input data was significantly lower than that of the model parameters. In addition, the GLUE analysis revealed that the wheat yield predictions were more stable and confident by considering the ensemble IMM technique. The pattern of root mean square error (RMSE) maps demonstrated that higher error produces in the northeast of Urmia Lake. The developed framework provides insight into rainfed yield responses to weather conditions and is simple and inexpensive. Accurate and reliable wheat yield prediction is essential for agricultural monitoring and food policy analysis.

Improvement of empirical formulas to estimate the reduction effects by berms on irregular wave runup over a dune-berm coast

This study formulates the reduction effects of a sandy berm on irregular wave runup over a dune-berm coast. The numerical experiments by Park and Cox (2016) are closely re-examined to develop an empirical formula describing the variability of reduction effects of a sandy berm over a broad range of conditions. Based on a sequence of regression analyses, the reduction effects are expressed as a reduction factor in terms of normalized berm width, normalized surge level, and wave steepness in deep water. The comparison with the numerical experiments demonstrates that the regression formula can satisfactorily reproduce the variability of the reduction effects over the range of numerical experiments. The analysis of prediction uncertainty demonstrated that the derived formula reproduced the reduction effects observed in the numerical experiments with negligible bias and a 90% confidence interval of approximately ±20% relative error. In addition, conversion formulas between representative runup values based on different statistical definitions are derived to enable consistent comparisons between them. The proposed reduction formula is implemented into three empirical runup models that are applicable to the quick estimations of irregular wave runup on a dune-berm coast: the models by Park and Cox (2016), Stockdon et al. (2006), and Mase et al. (2013). Consistent comparisons were conducted among the empirical predictions and numerical experiments based on the statistical conversion formulas. Combined with the proposed reduction formula, all three models well reproduced the normalized 2% runup in the numerical experiments over a wide range of conditions. On the other hand, the uncertainty in the runup prediction appeared in different forms depending on the selected model. When the proposed reduction formula was implemented in the modified Park and Cox (2016) and modified Stockdon et al. (2006) models, the uncertainty was described by a log-normal distribution of the error ratio between the empirical predictions and numerical experiments. Quantitatively, these two models predicted 90% of the normalized runup on a dune within a range of relative error of less than approximately 20–30%. When the proposed reduction formula was combined with the model by Mase et al. (2013), the uncertainty followed a normal distribution of the residual error between the empirical predictions and numerical experiments. On the normalized runup, the model prediction indicated a small conservative bias (+0.05) and a root-mean-square error of 0.13.

Uncertainty in model prediction of energy savings in building retrofits: Case of thermal transmittance of windows

Energy saving in buildings is an important measure for mitigation of climate change. There exists a large potential for energy saving in buildings by improving the thermal performance of windows. For decisions on energy saving window retrofits, accurate estimation of the energy saved and its uncertainty is of importance. The ISO 15099 standard, which is normative for thermal modelling of windows within the building sector, does not give uncertainty estimates. The main novelty of this study was to provide uncertainty analysis for model prediction of the thermal transmittance of windows, in the perspective of decisions on window retrofits. For this purpose, we proposed a new simplified model, which facilitated uncertainty analysis, and still was similar to the ISO 15099 window model. The model was validated by application of a benchmark validation procedure to a set of previously performed validation experiments. Main conclusions were: (i) The model was accurate within a prediction uncertainty equal to 0.20 Wm−2K−1; (ii) The domain where the model is valid was described using existing well-documented validation experiments. This domain was restricted to windows with glazing thermal transmittance corresponding to 2-layer glazing, and to windows where the frame area is a minor part of the total window area. (iii) The prediction uncertainty was mainly determined by the measurement uncertainty in the validation experiments; (iv) If a window retrofit is based on reduction of window thermal transmittance, then this reduction has to be larger than 0.56 Wm−2K−1 in order to yield energy savings above the uncertainty limit.

Satellite Imagery to Map Topsoil Organic Carbon Content over Cultivated Areas: An Overview

There is a need to update soil maps and monitor soil organic carbon (SOC) in the upper horizons or plough layer for enabling decision support and land management, while complying with several policies, especially those favoring soil carbon storage. This review paper is dedicated to the satellite-based spectral approaches for SOC assessment that have been achieved from several satellite sensors, study scales and geographical contexts in the past decade. Most approaches relying on pure spectral models have been carried out since 2019 and have dealt with temperate croplands in Europe, China and North America at the scale of small regions, of some hundreds of km2: dry combustion and wet oxidation were the analytical determination methods used for 50% and 35% of the satellite-derived SOC studies, for which measured topsoil SOC contents mainly referred to mineral soils, typically cambisols and luvisols and to a lesser extent, regosols, leptosols, stagnosols and chernozems, with annual cropping systems with a SOC value of ~15 g·kg−1 and a range of 30 g·kg−1 in median. Most satellite-derived SOC spectral prediction models used limited preprocessing and were based on bare soil pixel retrieval after Normalized Difference Vegetation Index (NDVI) thresholding. About one third of these models used partial least squares regression (PLSR), while another third used random forest (RF), and the remaining included machine learning methods such as support vector machine (SVM). We did not find any studies either on deep learning methods or on all-performance evaluations and uncertainty analysis of spatial model predictions. Nevertheless, the literature examined here identifies satellite-based spectral information, especially derived under bare soil conditions, as an interesting approach that deserves further investigations. Future research includes considering the simultaneous analysis of imagery acquired at several dates i.e., temporal mosaicking, testing the influence of possible disturbing factors and mitigating their effects fusing mixed models incorporating non-spectral ancillary information.

A framework for forecasting the hourly nodal water demand and improving the performance of real-time hydraulic models considering model uncertainty

Abstract The real-time hydraulic model (RTHM) is a key assistive tool in water distribution system (WDS) management, and its performance directly affects assisted decision-making. This study develops a framework to improve the timeliness and accuracy of RTHMs, which includes the following five steps: flow data processing, establishing nodal water demand (NWD) prediction models, node grouping, data assimilation (DA) and uncertainty analysis. Based on the actual network data, the performance of two data processing methods and three machine learning algorithms are, respectively, compared, and the best is selected for modeling. In the establishment of the hourly NWD prediction models, massive data, including flow measurement and data of all 26 input variables on climate, time and social influencing factors are used. It is found that the time features are the most important model input parameter. Application results of actual network prove that the flow data processing method, accurate NWD prediction, node grouping and Kalman filter-based DA method reduce the uncertainty in the RTHM and improve its timeliness and accuracy, so as to obtain the real-time state estimation of the WDS. Accurate NWD estimation (especially in the high-demand period) and combining RTHM with DA have a great influence on the uncertainty reduction in water pressure estimation, although uncertainty is weakened in the propagation process.

An integrated power load point-interval forecasting system based on information entropy and multi-objective optimization

During an era of rapid growth in electricity demand throughout society, accurate forecasting of electricity loads has become increasingly important to guarantee a stable power supply. Nevertheless, historical models do not address the structure of the data itself, and a single model cannot accurately determine the nonlinear characteristics of the data. This would not allow for accurate and stable predictions. With the aim of filling this gap, this paper proposes an innovative intelligent power load point-interval forecasting system. The system discretizes the time series, then performs efficient dimensionality reduction by fuzzification, and multi-level optimization of five benchmark deep learning models by the proposed multi-objective optimization algorithm, and finally analyzes the uncertainty of the prediction results. Experiments comparing the developed prediction system with other models were conducted on three datasets, and the prediction results were discussed for validation from multiple perspectives. The simulation results show that the proposed model has superior prediction accuracy, robustness and uncertainty analysis capability, and can provide accurate deterministic prediction information and fluctuation interval analysis to ensure the long-term safety and stability and operation of the grid.

A hybrid approach for integrated surface and subsurface hydrologic simulation of baseflow with Iterative Ensemble Smoother

Integrated surface and subsurface hydrologic models are particularly well suited for the simulation of baseflow as exchange fluxes does not rely on boundary conditions. Regional scale hydrologic models are of particular interest as they can provide guidance to stakeholders toward sustainable water resources management. However, integrated hydrologic models are rarely considered at the regional scale because their computational costs usually prevent efficient model calibration and predictive uncertainty analysis. Moreover, estimation of baseflow and each water budget component usually requires additional post-processing steps. In this paper, we thus aim here to provide a hybrid approach for the application of an integrated hydrologic model at the regional scale when low-flow processes are of primary concern. A surface water mass balance module has been developed to solve the hydrological mass balance from precipitations and temperature datasets. This module calculates potential infiltration fluxes that are used as input to the integrated modeling platform HydroGeoSphere (HGS). This approach provides a computationally tractable integrated model where low-flow processes are explicitly considered, baseflows generated in an integrated fashion, and each water budget component accessible. The model is applied to a region covering 36 900 km2 in the Southeastern part of the province of Quebec, Canada. Thanks to the computational efficiency of the approach, a rigorous mathematical parameter estimation, the Levenberg–Marquardt form of the Iterative Ensemble Smoother, can be considered to calibrate the model to baseflow of eight main rivers, snow water equivalent and evapotranspiration. An ensemble of 187 equally-probable realizations was used for history matching of observations and for the non-linear uncertainty analysis. An improved representation of baseflow can be linked to an appropriate non-linear uncertainty quantification through an efficient integrated surface and subsurface hydrologic model to support managing water resources at the regional scale.

Sensitivity and uncertainty analysis for streamflow prediction based on multiple optimization algorithms in Yalong River Basin of southwestern China

Semi-distributed model based on spatial information is effective for simulating hydrologic cycle. However, it cannot completely simulate all relevant natural processes. Uncertainty analysis is necessary for achieving high accuracy of hydrological modelling. In this study, optimization algorithms and weighted network analysis techniques were adopted to explore hydrological parameter combination characteristics as well as hydrological simulating uncertainties in Yalong River Basin (YLRB) of southwestern China based on Sequential Uncertainty Fitting version 2 (SUFI-2), Generalized Likelihood Uncertainty Estimation (GLUE) and Particle Swarm Optimization (PSO). The results indicated that: a) groundwater recession factor, effective hydraulic conductivity in channel and saturated hydraulic conductivity could significantly affect streamflow in the studying basin, b) complex parameter combination responded diversely under aforementioned three optimization algorithms. Comparatively, GLUE brought out higher autocorrelation parameter network and c) SUFI-2 and PSO performed better in terms of effective uncertainty analysis and fitting values than those of GLUE. This work could provide references and insights for sensitive parameter modification and prediction uncertainty reduction of streamflow simulation, furthermore contributing to an optimal water resource management.