Artificial neural networks for sheet sediment transport

Abstract

Sheet sediment transport was modelled by artificial neural networks (ANNs). A three-layer feed-forward artificial neural network structure was constructed and a back-propagation algorithm was used for the training of ANNs. Event-based, runoff-driven experimental sediment data were used for the training and testing of the ANNs. In training, data on slope and rainfall intensity were fed into the network as inputs and data on sediment discharge were used as target outputs. The performance of the ANNs was tested against that of the most commonly used physically-based models, whose transport capacity was based on one of the dominant variables—flow velocity (V), shear stress (SS), stream power (SP), and unit stream power (USP). The comparison results revealed that the ANNs performed as well as the physically-based models for simulating nonsteady-state sediment loads from different slopes. The performances of the ANNs and the physically-based models were also quantitatively investigated to estimate mean sediment discharges from experimental runs. The investigation results indicated that better estimations were obtained for V over mild and steep slopes, under low rainfall intensity; for USP over mild and steep slopes, under high rainfall intensity; for SP and SS over very steep slopes, under high rainfall intensity; and for ANNs over steep and very steep slopes, under very high rainfall intensities.     

Artificial neural network models for estimating regional reference evapotranspiration based on climate factors

Evapotranspiration (ET) is one of the basic components of the hydrologic cycle and is essential for estimating irrigation water requirements. In this study, an artificial neural network (ANN) model for reference evapotranspiration (ET₀) calculation was investigated. ANNs were trained and tested for arid (west), semi‐arid (middle) and sub‐humid (east) areas of the Inner Mongolia district of China. Three or four climate factors, i.e. air temperature (T), relative humidity (RH), wind speed (U) and duration of sunshine (N) from 135 meteorological stations distributed throughout the study area, were used as the inputs of the ANNs. A comparison was conducted between the estimates provided by the ANNs and by multilinear regression (MLR). The results showed that ANNs using the climatic data successfully estimated ET₀ and the ANNs simulated ET₀ better than the MLRs. The ANNs with four inputs were more accurate than those with three inputs. The errors of the ANNs with four inputs were lower (with RMSE of 0·130 mm d^?1, RE of 2·7% and R² of 0·986) in the semi‐arid area than in the other two areas, but the errors of the ANNs with three inputs were lower in the sub‐humid area (with RMSE of 0·21 mm d^?1, RE of 5·2% and R² of 0·961. For the different seasons, the results indicated that the highest errors occurred in September and the lowest in April for the ANNs with four inputs. Similarly, the errors were higher in September for the ANNs with three inputs. Copyright © 2008 John Wiley & Sons, Ltd.     

Improvement of artificial neural networks to predict daily streamflow in a semi-arid area

ABSTRACT

The application of artificial neural networks (ANNs) has been widely used recently in streamflow forecasting because of their ?exible mathematical structure. However, several researchers have indicated that using ANNs in streamflow forecasting often produces a timing lag between observed and simulated time series. In addition, ANNs under- or overestimate a number of peak flows. In this paper, we proposed three data-processing techniques to improve ANN prediction and deal with its weaknesses. The Wilson-Hilferty transformation (WH) and two methods of baseflow separation (one parameter digital filter, OPDF, and recursive digital filter, RDF) were coupled with ANNs to build three hybrid models: ANN-WH, ANN-OPDF and ANN-RDF. The network behaviour was quantitatively evaluated by examining the differences between model output and observed variables. The results show that even following the guidelines of the Wilson-Hilferty transformation, which significantly reduces the effect of local variations, it was found that the ANN-WH model has shown no significant improvement of peak flow estimation or of timing error. However, combining baseflow with streamflow and rainfall provides important information to ANN models concerning the flow process operating in the aquifer and the watershed systems. The model produced excellent performance in terms of various statistical indices where timing error was totally eradicated and peak flow estimation significantly improved.

Editor D. Koutsoyiannis; Associate editor Y. Gyasi-Agyei     

Projections of future deterioration in UK river quality are hampered by climatic uncertainty under extreme conditions

ABSTRACT

A modelling study was undertaken to quantify effects that the climate likely to prevail in the 2050s might have on water quality in two contrasting UK rivers. In so doing, it pinpointed the extent to which time series of climate model output, for some variables derived following bias correction, are fit for purpose when used as a basis for projecting future water quality. Working at daily time step, the method involved linking regional climate model (HadRM3-PPE) projections, Future Flows Hydrology (rainfall–runoff modelling) and the QUESTOR river network water quality model. In the River Thames, the number of days when temperature, dissolved oxygen, biochemical oxygen demand and phytoplankton exceeded undesirable values (>25°C, <6 mg L^?1, >4 mg L^?1 and >0.03 mg L^?1, respectively) was estimated to increase by 4.1–26.7 days per year. The changes do not reflect impacts of any possible change in land use or land management. In the River Ure, smaller increases in occurrence of undesirable water quality are likely to occur in the future (by 1.0–11.5 days per year) and some scenarios suggested no change. Results from 11 scenarios of the hydroclimatic inputs revealed considerable uncertainty around the levels of change, which prompted analysis of the sensitivity of the QUESTOR model to simulations of current climate and hydrology. Hydrological model errors were deemed of less significance than those associated with the derivation and downscaling of driving climatic variables (rainfall, air temperature and solar radiation). Errors associated with incomplete understanding of river water quality interactions with the aquatic ecosystem were found likely to be more substantial than those associated with hydrology, but less than those related to climate model inputs. These errors are largely a manifestation of uncertainty concerning the extent to which phytoplankton biomass is controlled by invertebrate grazers, particularly in mid-summer; and the degree to which this varies from year to year. The quality of data from climate models for generating flows and defining driving variables at the extremes of their distributions has been highlighted as the major source of uncertainty in water quality model outputs.

EDITOR A. Castellarin; ASSOCIATE EDITOR X. Fang     

A fusion-based neural network methodology for monthly reservoir inflow prediction using MODIS products

ABSTRACT

A model fusion approach was developed based on five artificial neural networks (ANNs) and MODIS products. Static and dynamic ANNs – the multi-layer perceptron (MLP) with one and two hidden layers, general regression neural network (GRNN), radial basis function (RBF) and nonlinear autoregressive network with exogenous inputs (NARX) – were used to predict the monthly reservoir inflow in Mollasadra Dam, Fars Province, Iran. Leaf area index and snow cover from MODIS, and rainfall and runoff data were used to identify eight different combinations to train the models. Statistical error indices and the Borda count method were used to verify and rank the identified combinations. The best results for individual ANNs were combined with MODIS products in a fusion model. The results show that using MODIS products increased the accuracy of predictions, with the MLP with two hidden layers giving the best performance. Also, the fusion model was found to be superior to the best individual ANNs.     

Predicting conductance due to upconing using neural networks

Artificial neural networks (ANNs) were developed to accurately predict highly time-variable specific conductance values in an unconfined coastal aquifer. Conductance values in the fresh water lens aquifer change in response to vertical displacements of the brackish zone and fresh water-salt water interface, which are caused by variable pumping and climate conditions. Unlike physical-based models, which require hydrologic parameter inputs, such as horizontal and vertical hydraulic conductivities, porosity, and fluid densities, ANNs can "learn" system behavior from easily measurable variables. In this study, the ANN input predictor variables were initial conductance, total precipitation, mean daily temperature, and total pumping extraction. The ANNs were used to predict salinity (specific conductance) at a single monitoring well located near a high-capacity municipal-supply well over time periods ranging from 30 d to several years. Model accuracy was compared against both measured/interpolated values and predictions were made with linear regression, and in general, excellent prediction accuracy was achieved. For example, although the average percent change of conductance over 90-d periods was 39%, the absolute mean prediction error achieved with the ANN was only 1.1%. The ANNs were also used to conduct a sensitivity analysis that quantified the importance of each of the four predictor variables on final conductance values, providing valuable insights into the dynamics of the system. The results demonstrate that the ANN technology can serve as a powerful and accurate prediction and management tool, minimizing degradation of ground water quality to the extent possible by identifying appropriate pumping policies under variable and/or changing climate conditions.     

Demand modulation of water scarcity sensitivities to secular climatic variation: theoretical insights from a computational maquette

ABSTRACT

Increased demand associated with population or economic growth, and decreased supply under some climatic shifts, obviously contribute to water scarcity. As a fresh perspective, we offer a generic theoretical treatment using a computational “maquette”, employing parameterizations to avoid assumptions about the origin and scale of climate and demand changes. The results suggest a distinct (and more subtle) point: the sensitivities of water stress to changes in both the mean and the variance of hydroclimate are modulated by demand level. Theoretical behaviours generated by the reduced-complexity model are surprisingly intricate, including profound nonlinearities and bifurcations. These may form a lower bound on the dynamical complexity of the demand–supply–scarcity nexus. Overall, the outcomes suggest that demand growth substantially intensifies and nonlinearizes water stress sensitivities to secular climate variation, and, in particular, that the interactions between demand changes and second-order hydroclimatic non-stationarity may produce non-intuitive water scarcity impacts requiring much closer study.

EDITOR A. Castellarin; ASSOCIATE EDITOR N. Ilich     

A watershed-scale study of climate change impacts on groundwater recharge (Annapolis Valley,Nova Scotia,Canada)

Abstract

The potential impacts of future climate change on the evolution of groundwater recharge are examined at a local scale for a 546-km² watershed in eastern Canada. Recharge is estimated using the infiltration model Hydrologic Evaluation of Landfill Performance (HELP), with inputs derived from five climate runs generated by a regional climate model in combination with the A2 greenhouse gas emissions scenario. The model runs project an increase in annual recharge over the 2041–2070 period. On a seasonal basis, however, a marked decrease in recharge during the summer and a marked increase during the winter are observed. The results suggest that increased evapotranspiration resulting from higher temperatures does not offset the large increase in winter infiltration. In terms of individual water budget components, clear differences are obtained for the different climate change scenarios. Monthly recharge values are also found to be quite variable, even for a given climate scenario. These findings are compared with results from two regional-scale studies.

Editor D. Koutsoyiannis; Associate editor M. Besbes     

Improving generalization of artificial neural networks in rainfall–runoff modelling / Amélioration de la généralisation de réseaux de neurones artificiels pour la modélisation pluie-débit

Abstract

Artificial neural networks (ANNs) are general-purpose techniques that can be used for nonlinear data-driven rainfall–runoff modelling. The key issue to construct a good model by means of ANNs is to understand their structural features and the problems related to their construction. Indeed, the quantity and quality of data, the type of noise and the mathematical properties of the algorithm for estimating the usual large number of parameters (weights) are crucial for the generalization performances of ANNs. However, it is well known that ANNs may suffer from poor generalization properties due to the high number of parameters and non-Gaussian data noise. Therefore, in the first part of this paper, the features and problems of ANNs are discussed. Eight Avoiding Overfitting Techniques are then presented, considering that these are methods for improving the generalization of ANNs. For this reason, they have been tested on two case studies—rainfall–runoff data from two drainage basins in the south of Italy—in order to gain insight into their properties and to investigate if there is one that absolutely gives the best performance.     

Projection of future world water resources under SRES scenarios: an integrated assessment

Abstract

Changes in water resources availability, as affected by global climate warming, together with changes in water withdrawal, could influence the world water resources stress situation. In this study, we investigate how the world water resources situation will likely change under the Special Report on Emissions Scenarios (SRES) by integrating water withdrawal projections. First, the potential changes in water resources availability are investigated by a multi-model analysis of the ensemble outputs of six general circulation models (GCMs) from organizations worldwide. The analysis suggests that, while climate warming might increase water resources availability to human society, there is a large discrepancy in the size of the water resource depending on the GCM used. Secondly, the changes in water-stressed basins and the number of people living in them are evaluated by two indices at the basin scale. The numbers were projected to increase in the future and possibly to be doubled in the 2050s for the three SRES scenarios A1b, A2 and B1. Finally, the relative impacts of population growth, water use change and climate warming on world water resources are investigated using the global highly water-stressed population as an overall indicator. The results suggest that population and socio-economic development are the major drivers of growing world water resources stress. Even though water availability was projected to increase under different warming scenarios, the reduction of world water stress is very limited. The principal alternative to sustainable governance of world water resources is to improve water-use efficiency globally by effectively reducing net water withdrawal.

Editor Z.W. Kundzewicz; Associate editor D. Gerten     

The development of a regional geomagnetic daily variation model using neural networks

Global and regional geomagnetic field models give the components of the geomagnetic field as functions of position and epoch; most utilise a polynomial or Fourier series to map the input variables to the geomagnetic field values. The only temporal variation generally catered for in these models is the long term secular variation. However, there is an increasing need amongst certain users for models able to provide shorter term temporal variations, such as the geomagnetic daily variation. In this study, for the first time, artificial neural networks (ANNs) are utilised to develop a geomagnetic daily variation model. The model developed is for the southern African region; however, the method used could be applied to any other region or even globally. Besides local time and latitude, input variables considered in the daily variation model are season, sunspot number, and degree of geomagnetic activity. The ANN modelling of the geomagnetic daily variation is found to give results very similar to those obtained by the synthesis of harmonic coefficients which have been computed by the more traditional harmonic analysis of the daily variation.     

A neural network model for predicting aquifer water level elevations

Artificial neural networks (ANNs) were developed for accurately predicting potentiometric surface elevations (monitoring well water level elevations) in a semiconfined glacial sand and gravel aquifer under variable state, pumping extraction, and climate conditions. ANNs "learn" the system behavior of interest by processing representative data patterns through a mathematical structure analogous to the human brain. In this study, the ANNs used the initial water level measurements, production well extractions, and climate conditions to predict the final water level elevations 30 d into the future at two monitoring wells. A sensitivity analysis was conducted with the ANNs that quantified the importance of the various input predictor variables on final water level elevations. Unlike traditional physical-based models, ANNs do not require explicit characterization of the physical system and related physical data. Accordingly, ANN predictions were made on the basis of more easily quantifiable, measured variables, rather than physical model input parameters and conditions. This study demonstrates that ANNs can provide both excellent prediction capability and valuable sensitivity analyses, which can result in more appropriate ground water management strategies.     

Estimation of soil water content in watershed using artificial neural networks

Soil water content (SWC) is an important factor in transfer processes between soil and air, contributing to water and energy balances, and quantifying it remains a challenge. This study uses artificial neural networks (ANNs) to analyse spatial and temporal variation of SWC in a Brazilian watershed, based on climate information, soil physical properties and topographic variables. Thirty eight input variables were tested in 200 models. The outputs were compared with 650 gravimetric moisture measurements collected at 26 points (25 field studies). The results show that it is possible to estimate SWC efficiently (Nash-Sutcliffe statistic, NS = 0.77) using topographic data, soil physical properties and rainfall. If only climate information is considered, modelling is less efficient (NS = 0.28). Using many variables does not necessarily improve performance. Alternatively, SWC can be estimated by simplified models using rainfall and topographic maps information, although the performance is less good (NS = 0.65).     

A study on the impact of climate change on streamflow at the watershed scale in the humid tropics

Abstract

Climate change is recognized to be one of the most serious challenges facing mankind today. Driven by anthropogenic activities, it is known to be a direct threat to our food and water supplies and an indirect threat to world security. Increase in the concentration of carbon dioxide and other greenhouse gases in the atmosphere will certainly affect hydrological regimes. The consequent global warming is expected to have major implications on water resources management. The objective of this research is to present a general approach for evaluating the impacts of potential climate change on streamflow in a river basin in the humid tropical zone of India. Large-scale global climate models (GCMs) are the best available tools to provide estimates of the effect of rising greenhouse gases on rainfall and temperature. However the spatial resolution of these models (250 km?×?250 km) is not compatible with that of watershed hydrological models. Hence the outputs from GCMs have to be downscaled using regional climate models (RCMs), so as to project the output of a GCM to a finer resolution (50 km?×?50 km). In the present work, the projections of a GCM for two scenarios, A2 and B2 are downscaled by a RCM to project future climate in a watershed. Projections for two important climate variables, viz. rainfall and temperature are made. These are then used as inputs for a physically-based hydrological model, SWAT, in order to evaluate the effect of climate change on streamflow and vegetative growth in a humid tropical watershed.

Citation Raneesh, K. Y. & Santosh, G. T. (2011) A study on the impact of climate change on streamflow at the watershed scale in the humid tropics. Hydrol. Sci. J. 56(6), 946–965.     

Groundwater system response at sites with different agricultural land uses: case of the Guarani Aquifer outcrop area,Brotas/SP-Brazil

A system identification approach can be incorporated in groundwater time series analysis, revealing information concerning the local hydrogeological situation. The aim of this work was to analyse water table fluctuations in an outcrop area of the Guarani Aquifer System (GAS) in Brotas/SP, Brazil, using data from a groundwater monitoring network. The water table dynamic was modelled using continuous time series models that reference the hydrogeological system upon which they operate. The model’s climatological inputs of precipitation and evapotranspiration generate impulse response (IR) functions with parameters that can be related to the physical conditions concerning the hydrological processes involved. The interpretation of the model parameters from two sets of monitoring wells selected at different land-use sites is presented, exemplifying the effect of different water table depths and the distance to the nearest drainage location. Systematic trends of water table depths were also identified from model parameters at specific periods and related to plant development, crop harvest and land-use changes.

EDITOR D. Koutsoyiannis

ASSOCIATE EDITOR L. Ruiz     

Evaluating the performance of ground-based and remotely sensed near real-time rainfall fields from a hydrological perspective

Abstract

The South African Weather Service (SAWS) issues routine experimental, near real-time rainfall maps from daily raingauge networks, radar networks and satellite images, as well as merged rainfall fields. These products are potentially useful for near real-time forecasting, especially in areas of fast hydrological response, and also to simulate the “now state” of various hydrological state variables such as soil moisture content, streamflow, and reservoir inflows. The purpose of this paper is to evaluate their skill as inputs to hydrological simulations and, in particular, the skill of the merged field in terms of better hydrological results relative to the individual products. Rainfall fields derived from raingauge, radar, satellite, conditioned satellite and the merged (gauge/radar/satellite) were evaluated for two selected days with relatively high amounts of rainfall, as well as for a continuous period of 90 days in the Mgeni catchment, South Africa. Streamflows simulated with the ACRU model indicate that the use of raingauge as well as merged fields of satellite/raingauge and satellite/radars/raingauge provides relatively realistic rainfall results, without much difference in their hydrological outputs, whereas the radar and raw satellite information by themselves cannot be used in operational hydrological application in their current status.

Citation Ghile, Y., Schulze, R. & Brown, C. (2010) Evaluating the performance of ground-based and remotely sensed near real-time rainfall fields from a hydrological perspective. Hydrol. Sci. J. 55(4), 497–511.     

Using neural networks for calibration of time-domain reflectometry measurements

Abstract

Time-domain reflectometry (TDR) is an electromagnetic technique for measurements of water and solute transport in soils. The relationship between the TDR-measured dielectric constant (K_a ) and bulk soil electrical conductivity ([sgrave]_a) to water content (θ_W) and solute concentration is difficult to describe physically due to the complex dielectric response of wet soil. This has led to the development of mostly empirical calibration models. In the present study, artificial neural networks (ANNs) are utilized for calculations of θ_w and soil solution electrical conductivity ([sgrave]_w) from TDR-measured K_a and [sgrave]_a in sand. The ANN model performance is compared to other existing models. The results show that the ANN performs consistently better than all other models, suggesting the suitability of ANNs for accurate TDR calibrations.     

A model approach for in- and outflow calculation of upper lake constance: —An investigation of a 60 year time span and observations about the flood of 1999

The revised empirical model for in- and outflow calculation of Upper Lake Constance has provided satisfying results supported by measured values. The given model was implemented to simulate total water inputs of the lake during the period from 1941 to 2000 with emphasis on the flood conditions of 1999. Analysis of annual water input development reveals a tendency toward slight increases until the 1960s. Thereafter, a reduction in inputs can be noted. This trend probably continues to hold true to present. Weather conditions of given individual years have caused distinct fluctuations to the water budget.Unusual meteorological conditions led to extreme flooding in early May of 1999. Daily water inputs of up to 200 mio m³ generated the highest water levels ever observed for this time of the year. Continual extraordinarily high water inputs occurring from February until July and then again from September until the end of 1999 resulted in the second largest annual total water input recorded since 1941.     

Statistical and geostatistical investigations into the effects of the Gabcikovo hydropower plant on the groundwater resources of northwest Hungary / Analyses statistiques et géostatistiques des effets de la centrale hydroélectrique de Gabcikovo sur les ressources en eaux souterraines du Nord-Ouest de la Hongrie

Abstract

Abstract The construction of the Gabcikovo hydropower plant and the diversion of the Danube River over 25 km into an artificial channel in 1992 influenced the groundwater regime of the region considerably. Statistical and geostatistical methods are used to quantify changes of different groundwater characteristics on the Hungarian side of the river based on observations in the time period 1960–2000. External drift kriging was used to interpolate groundwater levels and the other related variables. While mean groundwater levels did not change appreciably, there are significant changes in the variability. Standard deviations of the groundwater levels and the amplitude of the annual cycle decreased near the old river bed of the Danube. The water-level fluctuations of the Danube have a decreased influence on the groundwater dynamics. Interrelationships between water levels in wells have also changed.     

Hybrid modelling approach to prairie hydrology: fusing data-driven and process-based hydrological models

Abstract

Much of the prairie region in North America is characterized by relatively flat terrain with many depressions on the landscape. The hydrological response (runoff) is a combination of the conventional runoff from the contributing areas and the occasional overflow from the non-contributing areas (depressions). In this study, we promote the use of a hybrid modelling structure to predict runoff generation from prairie landscapes. More specifically, the Soil and Water Assessment Tool (SWAT) is fused with artificial neural networks (ANNs), so that SWAT and the ANN module deal with the contributing and non-contributing areas, respectively. A detailed experimental study is performed to select the best set of inputs, training algorithms and hidden neurons. The results obtained in this study suggest that the fusion of process-based and data-driven models can provide improved modelling capabilities for representing the highly nonlinear nature of the hydrological processes in prairie landscapes.

Editor D. Koutsoyiannis; Associate editor L. See