A systematic approach to the input determination for neural network rainfall–runoff models

Input determination has a great influence on the performance of artificial neural network (ANN) rainfall–runoff models. To improve the performance of ANN models, a systematic approach to the input determination for ANN models is proposed. In the proposed approach, the irrelevant inputs are removed. Then an adequate ANN model, which only includes highly relevant inputs, is constructed. Unlike the trial‐and‐error procedure, the proposed approach is more systematic and avoids unnecessary trials. To demonstrate the effectiveness of the proposed approach, an application to actual typhoon events is presented. The results show that the proposed ANN model, which is constructed by the proposed approach, has advantages over those obtained by the trial‐and‐error procedure. The proposed ANN model has a simpler architecture, needs less training time, and performs better. The proposed ANN model is recommended as an alternative to existing rainfall–runoff ANN models. Copyright © 2007 John Wiley & Sons, Ltd.     

Importance of no‐rain measurements on the comparison of radar and rain gauge rain rate

This study evaluated four possible cases of comparing radar and rain gauge rain rate for the detection of mean‐field bias. These four cases, or detection designs, consider in this study are: (1) design 1‐uses all the data sets available, including zero radar rain rate and zero rain gauge rain rate, (2) design 2—uses the data sets of positive radar rain rate and zero or positive rain gauge rain rate, (3) design 3—uses the data sets of zero or positive radar rain rate and positive rain gauge rain rate and (4) design 4—uses the data sets of positive radar rain rate and positive rain gauge rain rate. A theoretical review of these four detection designs showed that only the design 1 causes no design bias, but designs 2, 3 and 4 can cause positive, negative and negative design biases, respectively. This theoretical result was also verified by applying these four designs to the rain rate field generated by a multi‐dimensional rain rate model, as well as to that of the Mt Gwanak radar in Korea. The results from both applications showed that especially the design 4, which is generally used for the detection of mean‐field bias of radar rain rate, causes a serious design bias; therefore, is inappropriate as a design for detecting the mean‐field bias of radar rain rate. Copyright © 2009 John Wiley & Sons, Ltd.     

A thermo‐mechanical model for variably saturated soils based on hypoplasticity

The paper presents a mechanical model for non‐isothermal behaviour of unsaturated soils. The model is based on an incrementally non‐linear hypoplastic model for saturated clays and can therefore tackle the non‐linear behaviour of overconsolidated soils. A hypoplastic model for non‐isothermal behaviour of saturated soils was developed and combined with the existing hypoplastic model for unsaturated soils based on the effective stress principle. Features of the soil behaviour that are included into the model, and those that are not, are clearly distinguished. The number of model parameters is kept to a minimum, and they all have a clear physical interpretation, to facilitate the model usefulness for practical applications. The step‐by‐step procedure used for the parameter calibration is described. The model is finally evaluated using a comprehensive set of experimental data for the thermo‐mechanical behaviour of an unsaturated compacted silt. Copyright © 2011 John Wiley & Sons, Ltd.     

Flood frequency analysis of Turkey using L‐moments method

The index flood procedure coupled with the L‐moments method is applied to the annual flood peaks data taken at all stream‐gauging stations in Turkey having at least 15‐year‐long records. First, screening of the data is done based on the discordancy measure (D_i) in terms of the L‐moments. Homogeneity of the total geographical area of Turkey is tested using the L‐moments based heterogeneity measure, H, computed on 500 simulations generated using the four parameter Kappa distribution. The L‐moments analysis of the recorded annual flood peaks data at 543 gauged sites indicates that Turkey as a whole is hydrologically heterogeneous, and 45 of 543 gauged sites are discordant which are discarded from further analyses. The catchment areas of these 543 sites vary from 9·9 to 75121 km² and their mean annual peak floods vary from 1·72 to 3739·5 m³ s^?1. The probability distributions used in the analyses, whose parameters are computed by the L‐moments method are the general extreme values (GEV), generalized logistic (GLO), generalized normal (GNO), Pearson type III (PE3), generalized Pareto (GPA), and five‐parameter Wakeby (WAK). Based on the L‐moment ratio diagrams and the |Z_dist|‐statistic criteria, the GEV distribution is identified as the robust distribution for the study area (498 gauged sites). Hence, for estimation of flood magnitudes of various return periods in Turkey, a regional flood frequency relationship is developed using the GEV distribution. Next, the quantiles computed at all of 543 gauged sites by the GEV and the Wakeby distributions are compared with the observed values of the same probability based on two criteria, mean absolute relative error and determination coefficient. Results of these comparisons indicate that both distributions of GEV and Wakeby, whose parameters are computed by the L‐moments method, are adequate in predicting quantile estimates. Copyright © 2011 John Wiley & Sons, Ltd.     

A discrete particle representation of hillslope hydrology: hypothesis testing in reproducing a tracer experiment at Gårdsjön,Sweden

Despite the long history of the continuum equation approach in hydrology, it is not a necessary approach to the formulation of a physically based representation of hillslope hydrology. The Multiple Interacting Pathways (MIPs) model is a discrete realization that allows hillslope response and transport to be simultaneously explored in a way that reflects the potential occurrence of preferential flows and lengths of pathways. The MIPs model uses random particle tracking methods to represent the flow of water within the subsurface alongside velocity distributions that acknowledge preferential flows and transition probability matrices, which control flow pathways. An initial realization of this model is presented here in application to a tracer experiment carried out in Gårdsjön, Sweden. The model is used as an exploratory tool, testing several hypotheses in relation to this experiment. Copyright © 2011 John Wiley & Sons, Ltd.     

Process identification at a slow‐moving landslide in the Vorarlberg Alps

A fine‐grained slope that exhibits slow movement rates was investigated to understand how geohydrological processes contribute to a consecutive development of mass movements in the Vorarlberg Alps, Austria. For that purpose intensive hydrometeorological, hydrogeological and geotechnical observations as well as surveying of surface movement rates were conducted during 1998–2001. Subsurface water dynamics at the creeping slope turned out to be dominated by a three‐dimensional pressure system. The pressure reaction is triggered by fast infiltration of surface water and subsequent lateral water flow in the south‐western part of the hillslope. The related pressure signal was shown to propagate further downhill, causing fast reactions of the piezometric head at 5·5 m depth on a daily time scale. The observed pressure reactions might belong to a temporary hillslope water body that extends further downhill. The related buoyancy forces could be one of the driving forces for the mass movement. A physically based hydrological model was adopted to model simultaneously surface and subsurface water dynamics including evapotranspiration and runoff production. It was possible to reproduce surface runoff and observed pressure reactions in principle. However, as soil hydraulic functions were only estimated on pedotransfer functions, a quantitative comparison between observed and simulated subsurface dynamics is not feasible. Nevertheless, the results suggest that it is possible to reconstruct important spatial structures based on sparse observations in the field which allow reasonable simulations with a physically based hydrological model. Copyright © 2004 John Wiley & Sons, Ltd.     

Least‐square support vector machine applied to settlement of shallow foundations on cohesionless soils

This paper examines the potential of least‐square support vector machine (LSVVM) in the prediction of settlement of shallow foundation on cohesionless soil. In LSSVM, Vapnik's ε‐insensitive loss function has been replaced by a cost function that corresponds to a form of ridge regression. The LSSVM involves equality instead of inequality constraints and works with a least‐squares cost function. The five input variables used for the LSSVM for the prediction of settlement are footing width (B), footing length (L), footing net applied pressure (P), average standard penetration test value (N) and footing embedment depth (d). Comparison between LSSVM and some of the traditional interpretation methods are also presented. LSSVM has been used to compute error bar. The results presented in this paper clearly highlight that the LSSVM is a robust tool for prediction of settlement of shallow foundation on cohesionless soil. Copyright © 2008 John Wiley & Sons, Ltd.     

Assessment of δ13C and C/N ratios in bulk organic matter as palaeosalinity indicators in Holocene and Lateglacial isolation basin sediments,northwest Scotland

Carbon isotopes (δ¹³C) and C/N ratios from bulk organic matter have recently been used as alternative proxies for relative sea‐level (RSL) reconstruction where there are problems associated with conventional biological indictors. A previous study on a single isolation basin (Upper Loch nan Eala) in northwest Scotland has shown a clear relationship between δ¹³C, C/N ratios and palaeosalinity from Younger Dryas and Holocene aged sediments. In this paper we present results of δ¹³C and C/N ratio analyses from other isolation basins in northwest Scotland over the Holocene and the Lateglacial period in order to validate this technique. The results from the Holocene sequences support the earlier findings that this technique can be used to identify RSL change from isolation basins over the Holocene in this region. The relationship between δ¹³C, C/N ratios and RSL change is not apparent in sediments of Lateglacial age. Other environmental variables such as atmospheric CO₂ concentration, poor vegetation development and temperature influence δ¹³C values during this period. Copyright © 2007 John Wiley & Sons, Ltd.     

Holocene coastal environmental changes on the periphery of an area of glacio‐isostatic uplift: an example from Scapa Bay,Orkney, UK

The first detailed investigation of a deep, coastal, sedimentary basin in Orkney reveals a complex Holocene history of back‐barrier morphodynamics. At Scapa Bay, the sea flooded a freshwater marsh after ca. 9400 yr BP at ca. ?5.4 m OD. Before ca. 7800 BP, abundant sediment from nearby cliffs was mobilised inland into a series of gravel barriers across the valley mouth. By ca. 7500 BP, direct marine influence was restricted in the back‐barrier area, although saltmarsh persisted until ca. 5900 BP. By then, at least four gravel ridges had enclosed the backing lagoon, where freshwater inputs became dominant. As terrestrial sediments filled the basin, another freshwater marsh developed. The multiple barrier complex demonstrates progradation resulting from continuous sediment supply in a sheltered embayment. The progressively rising height of the barrier crests seawards probably resulted from a combination of factors such as barrier morphodynamics, increased storminess and long‐term rising relative sea levels. The dominant vegetation surrounding Scapa Bay changed from open grassland to scrub ca. 9400 BP, then to deciduous woodland ca. 7800 BP, and to dwarf‐shrub heath ca. 2600 BP, the latter probably a response to a combination of climate change and human activity. Copyright © 2007 John Wiley & Sons, Ltd.     

Luminescence chronology of late Holocene extreme hydrological events in the upper Penner River basin,South India

Field stratigraphy, sedimentology and optically stimulated luminescence (OSL) dating have been used to reconstruct the southwest monsoon variability in the semi‐arid region of southern India during the late Holocene. Facies architecture and OSL dating of the water‐lain sediment suggest prevalence of a weak hydrological regime around 3 ka. Following this, a progressive strengthening of monsoon occurred till 2 ka. After 2 ka and until 1 ka fluvial activity was nearly dormant, indicating weakening of the monsoon. Presence of high‐magnitude flood deposits, overbank sedimentation and pedogenesis during 1–0.6 ka indicate intensification of the southwest monsoon in the basin. The onset of aridity was associated with episodic storm surge events that are manifested in the pond sedimentation and localised aeolian accretion. This phase is bracketed between 0.5 ka and 0.2 ka. A renewed phase of monsoonal activity was observed in the form of floodplain aggradation between 180 and 90 years ago. In the past 70 years no significant change in the monsoon performance has been observed. Copyright © 2007 John Wiley & Sons, Ltd.