查看更多>>摘要：Reliable and accurate multi-step streamflow forecasting is of vital importance for the utilization of water resources and hydropower energy system. In this paper, a spatial deep learning model, directed graph deep neural network, is proposed for multi-step streamflow forecasting. The proposed model uses spatial information capture process and feature aggregation process to exploit multi-site hydrological and meteorological information. The spatial information capture process consists of multiple convolutional layers to extract the precipitation information of meteorological stations. And the feature aggregation process uses the multi-layer perceptron to aggregate the precipitation information and the streamflow information. The proposed model is applied in a realworld case study in the upstream of Yangtze River basin. Experimental results demonstrate that the proposed model significantly outperforms artificial neural network, Long Short-Term Memory Network, Gated recurrent unit and Convolutional Neural Network in terms of forecasting accuracy. In addition to the forecast accuracy, the hidden Markov regression is employed to quantify the forecasting uncertainty given by the directed graph deep neural network. The uncertainty estimation result demonstrates that the hidden Markov regression is able to handle the heteroscedastic and non-normal forecasting uncertainty given by directed graph deep neural network.
查看更多>>摘要：Hydrometeorological extreme value analysis (EVA) is complicated with multiple-uncertain methodological choices, such as extreme event sampling (EES), stationary/non-stationary assumption (SNA), and parameter estimation (PE). In this study, a factorial extreme value analysis (FEVA) method is developed through integrating four EESs, four SNAs and two PEs into a mixed-level factorial experiment framework. FEVA can not only identify the main factors that influence the risk assessment of hydrometeorological extremes, but also reflect main and interactive effects of such complex uncertainties. FEVA is applied to Central Asia for analyzing maximum 1-day precipitation (Pmax), maximum daily temperature (Tmax) and minimum daily temperature (Tmin), where data from 32 long-term (1881-2018) observation stations are adopted and 32 scenarios are examined. Main findings are: (1) during the 20th century, the occurrences of precipitation and temperature extremes were increased in Central Asia, and minimum temperature extreme increases faster than maximum temperature extreme; (2) the dominated factor affecting the return levels of Pmax, Tmax and Tmin is EES, with average contributions of 32.9%, 47.0% and 53.5%, respectively; (3) the interaction between PE and EES has an obvious effect on the return levels, and its impacts occupy 17.8% for Pmax, 9.5% for Tmax and 4.5% for Tmin; (4) based on graphical and Goodness-ofFit (GoF) metrics methods, appropriate threshold values of [11.57, 37.63] mm for Pmax, [34.15, 44.46] degrees C for Tmax and [-40.07, -10.83] degrees C for Tmin are obtained (larger than 99th percentile of the observed value) in Central Asia, which can improve the simulation and prediction accuracy of hydrometeorological extremes.
查看更多>>摘要：The focus of this paper is to establish a general model, which is capable of dynamic simulation of discrete fractures at any position and arbitrary distribution in rectangular closed reservoir. In this article, a semianalytical model for fractured horizontal wells with random distributed fracture networks is proposed to facilitate pressure transient response using a double node method. A rectangular closed reservoir model and fracture network flow model are established respectively, and then coupled at the fracture wall. In order to deal with the flow distribution at the fracture intersection, a simple adaptive material balance method is proposed in this paper, which successfully solves the flow distribution at the intersection. Using advanced mathematical means, the model is successfully solved in Laplace space, and then the pressure response solution in real space is obtained by Stehfest numerical inversion method. The obtained solutions are verified and compared with the numerical simulation results. The calculation results show that the flow characteristics of the orthogonal fracture network can be roughly divided into six stages, namely, bi-linear flow, first linear flow, first radial flow, second linear flow, second radial flow and boundary dominated flow. However, these flow characteristics will deviate or even be missing, depending on the complexity of the fracture network. Finally, the effects of some important parameters (such as dimensionless fracture network conductivity, main fracture length, fracture network density, reservoir area and eccentric position) on pressure response and pressure field distribution are discussed in detail.
查看更多>>摘要：Rainfall represents the main driving force of runoff and soil erosion. Previous studies usually focused on the effects of general rainfall characteristics at event scale on runoff and soil erosion, but the effects of rainfall intensity fluctuation at intra-event scale, i.e., the temporal distribution of intensity bursts across the rainfall profile, have always been difficult to be characterized. In this study, the rainfall intensity profile was disaggregated into high-intensity zone and low-intensity zone using the average rainfall intensity as threshold, and a series of indicators were proposed to describe the features of rainfall intensity fluctuation. Field observations of runoff and soil loss in six grass-cover plots under natural rainfall conditions were conducted from 2011 to 2016 in the Loess Plateau of China. The main rainfall indicators influencing runoff coefficient (RC), sediment concentration (SC) and soil loss coefficient (SLC) were identified using the principal component analysis, and runoff and soil loss regression models were developed using data from 2011 to 2013 to predict RC, SC and SLC during 2014-2016. The results indicated that the natural rainfall events displayed strong intensity fluctuations (e.g., aggregation, intermittency, and variability), with approximate 78% of rainfall amount occurring in only 27% of rainfall duration. Most of the proposed rainfall intensity fluctuation indicators were significantly correlated with RC, SC and SLC (p < 0.05). The average intensity of rainfall peaks (I-ap) and rainfall duration of high-intensity zone (RDh) were the main variables to simulate RC, and the average intensity of high-intensity zone (I-ah) and relative amplitude of rainfall intensity (R-am) controlled SC and SLC. The models considering rainfall intensity fluctuation had good performances in predicting RC, SC and SLC with higher R-2 values (0.64 vs 0.46), lower RMSE values (0.66 vs 0.76) and higher NSE values (0.54 vs 0.42) compared to the models only considering general rainfall characteristics. The prediction accuracy of SLC was higher than that of RC and that of SC was lowest. These results gain an insight into the marked influence of rainfall intensity fluctuation in generating and predicting runoff and soil loss, and shed light on exploring rainfall-erosion relationships.
Kunstmann, H.Adeyeri, O. E.Laux, P.Ishola, K. A....
14页
查看更多>>摘要：Daily precipitation, maximum and minimum air temperature series are homogenised over the Lake Chad Basin between 1979 and 2020 using two conceptually different homogenisation methods; the Adapted CaussinusMestre Algorithm for homogenising Networks of Temperature series (ACMANT) and the iterated standard normal homogeneity (CLIMATOL). Results show the existence of unnatural breakpoints for most of the station series. However, the two methods show a general improvement in the quality of climate series. The trend estimation of the homogenised series based on the modified Mann Kendall methodology shows different modifications of the trend's magnitude for different periods. Overall, CLIMATOL and ACMANT exhibit nearly similar trend patterns, suggesting the credibility of both homogenisation methods. Relative to the base period of analysis (1981-2010), the anomaly classification for the entire basin between 1979 and 2020 into dry-warm, wet-warm, dry-cold and wet-cold are misrepresented by the raw series compared to the homogenised series. Such erroneous representations resulting from inhomogeneities in raw climate series could misinform decisions akin to climate change assessment and water resources management strategies, thereby reducing the adaptive-capability of the basin's inhabitants to climate change effects. Our study demonstrates the importance of robust homogenisation of climate series to mitigate inhomogeneity errors and improve the quality of information when observations are used in climate and hydrological studies.
查看更多>>摘要：Accurate estimation of the basin-specific parameter in the Budyko framework (e.g., parameter n in the Choudhury-Yang equation) is critical to quantify precipitation partitioning into evapotranspiration (E) and runoff. However, n is difficult to estimate due to complex interactions between the water balance and various environmental factors. In this study, we identified the controlling factors of n using random forest during 1981-2015 for 30 basins in the Loess Plateau of China. We then used the long short-term memory (LSTM) network combined with an 11-year moving window to develop a model to estimate time-varying n. This model was further incorporated into the Choudhury-Yang equation to simulate E. Our results showed that correlations between parameter n and environmental factors presented obvious spatial heterogeneity. Three land use type factors (i.e., the proportions of cropland, shrubland, and built-up land area), two climatic factors (i.e., precipi-tation and potential evapotranspiration), and a water use factor (i.e., irrigation water use) were identified as the controlling factors for n. Based on these controls, the LSTM model outperformed the traditional multiple linear regression model (MLR model) in estimating time-varying n, with root mean square error (RMSE) of 0.31/0.49 and coefficient of determination (R-2) of 0.88/0.67 for the LSTM/MLR model, respectively. Moreover, compared with the original Choudhury-Yang equation (using constant n calibrated by long-term average water balance), the improved equation (using time-varying n estimated by the LSTM) better reproduced the time series of water balance-based E. This study could enhance the applicability of the Budyko framework and provide scientific guidance for water resources management.& nbsp;
查看更多>>摘要：The influence of roughness, aperture and asperity irregularity on fluid flow regimes in rough rock fractures was investigated by performing coupled triaxial water flow tests and fluid dynamic computation. Ten sets of fabri-cated curved wedges were developed to obtain different fracture surface roughness by splitting under compression. Three fluid flow regimes were identified in mated rock fractures: pre-linear flow at low flow ve-locity, linear Darcy's flow at the medium flow velocity and post-linear flow at high flow velocity. In pre-linear flow regime, the increasing rate of flow rate increased with water pressure gradient, but it decreased in post -linear flow regime. The pre-linear flow is ascribed to the slippage effect of water-fracture interfaces, while the post-linear flow is mainly due to inertial effects. The critical hydraulic gradient for fluid flow regime transition from pre-linear to linear flow increased with the increase of confining stress. The numerical modeling shows that the asperity irregularity influences the flow regime. For low-speed flow in fracture under slip boundary condi-tion, the slip flux per unit width in the fracture with triangular asperity is largest, while that of rectangular asperity is smallest. For high-speed flow in fracture with non-slip boundary condition, the rectangular asperity element causes more severe nonlinearity when compared to the fractures of trapezoidal and triangular asperity elements.
Calli, Suleyman SelimCalli, Kubra OzdemirYilmaz, M. TugrulCelik, Mehmet...
17页
查看更多>>摘要：Snow recharge is an important dominant hydrological process in the high altitude mountainous karstic aquifer systems. In general, widely used karst-dedicated hydrological models (e.g., KarstMod, Varkarst) do not include a snow routine in the model structure to avoid increasing the number of model parameters while representing the complex hydrological process. As a result, recharge process is not represented well, which questions the optimality of the results that can be obtained under available datasets. This study presents a novel pre-processing method -called SCA routine- to compensate for the missing snow routine in karst models. The proposed preprocessing method is driven by temperature, precipitation, and satellite-based snow observation datasets. The method classifies the precipitation input into three physical phases (rain, snow, and mixed) based on the temperature datasets to distribute each phase over the catchment using satellite-driven Snow-Covered Area (SCA) products. By the proposed method, the spring discharge simulations are regulated well in time and magnitude. To examine the added utility of the SCA routine, the SCA-included simulations are compared to the model performances with no routine and the classical Degree-Day method as a benchmark. To test the efficiency of our proposed method, we used a karst hydrological model (KarstMod) to simulate the karst spring discharge in a well-observed semi-arid snow-dominated karstic aquifer (Central Taurus, Turkey). Our results confirmed that the KarstMod model coupled by SCA routine ensures better model performance with a value of NSE = 0.784 than those of the classical Degree-day method (NSE = 0.760) and the model with no routine (NSE = 0.306), thus providing a physically more realistic parameter set.
查看更多>>摘要：The mismatch between the space-time scales of flash flood occurrence and those of the typical hydro meteorological monitoring networks has stimulated the development of forensic flash flood analysis, which involves post-flood indirect peak discharge estimation in ungauged channels and flood response modelling driven by weather-radar rainfall estimates. However, both approaches are potentially affected by significant uncertainties. Assessment of scale dependence of such uncertainties is important to identify how uncertainty affecting forensic flash flood analysis increases with decreasing basin size. In this work, we apply the forensic methodology to the flash flood of November 18, 2013 in Sardinia (Italy). We introduce the 'flash flood forensic consistency index' as a tool to compare the probability distribution of peak discharge uncertainties from observational and model estimates, and to determine the scale effects of the forensic analysis concept. Uncertainties in field-based peak flow estimates are evaluated through a first-order error analysis of the Taylor series approximation of the slope-conveyance method, whereas uncertainties in flash flood modelling are based on the Generalized Likelihood Uncertainty Estimation methodology using a distributed hydrologic model. Results show no significant relationship between observational and modelling uncertainties, considered independently, and basin area and channel bed slope. Conversely, when considering the interaction between the two uncertainty distributions, a relationship arises between their degree of overlap and basin size. In particular, with decreasing basin area or increasing channel bed slope, the absolute relative bias between the estimated peak flow values increases more than their relative uncertainties, decreasing the consistency index. This calls for more robust approaches for the analysis of flash flood response in small-sized rugged-relief mountain basins, which are of high interest for flood risk management.
查看更多>>摘要：Multi-hazard risk assessment may provide comprehensive analysis of the impact of multiple hazards but still needs to resolve major challenges in three aspects: (1) proper consideration of hazard inter-dependency, (2) physically based modelling of hazard interactions, and (3) fully quantitative risk assessment to show the probability of loss. Compound flooding is a typical multi-hazard problem that involves the concurrence of multiple hazard drivers, e.g. heavy rainfall, extreme river flow, and storm surge. These hazard drivers may result from the same weather system and are thus statistically inter-dependent, physically overlayed and interacted in the same region. This paper aims to address the mentioned challenges and develop an integrated assessment framework to quantify compound flood risk. The framework is constructed based on the three typical components in disaster risk assessment, i.e. hazard, vulnerability and exposure analysis. In hazard analysis, joint probability and return period distributions of the three hazard drivers of compound flooding are estimated using Copula functions with hazard dependency analysis, which are then used to generate random multi-hazard events to drive a 2D highperformance hydrodynamic model to produce probabilistic inundation maps and frequency-inundation curves. Vulnerability and exposure analysis provides damage functions of the elements at risk, which are used to quantify multi-hazard risk with the frequency-inundation curves. The framework is applied in the Greater London and its downstream Thames estuary to demonstrate its capability to analyse hazard interactions and inter-dependencies to produce fully quantitative risk assessment results such as risk curves quantifying the probability of loss and risk maps illustrating the annual expected loss of residential buildings.
NSTL
Elsevier