Improving kriging of groundwater level data using nonlinear normalizing transformations—a field application

Abstract

The present research study investigates the application of nonlinear normalizing data transformations in conjunction with ordinary kriging (OK) for the accurate prediction of groundwater level spatial variability in a sparsely-gauged basin. We investigate three established normalizing methods, Gaussian anamorphosis, trans-Gaussian kriging and the Box-Cox method to improve the estimation accuracy. The first two are applied for the first time to groundwater level data. All three methods improve the mean absolute prediction error compared to the application of OK to the non-transformed data. In addition, a modified Box-Cox transformation is proposed and applied to normalize the hydraulic heads. The modified Box-Cox transformation in conjunction with OK is found to be the optimal spatial model based on leave-one-out cross-validation. The recently established Spartan semivariogram family provides the optimal model fit to the transformed data. Finally, we present maps of the groundwater level and the kriging variance based on the optimal spatial model.

Editor D. Koutsoyiannis; Associate editor A. Montanari

Citation Varouchakis, E.A., Hristopoulos, D.T., and Karatzas, G.P., 2012. Improving kriging of groundwater level data using nonlinear normalizing transformations—a field application. Hydrological Sciences Journal, 57 (7), 1404–1419.     

Comparison of bootstrap confidence intervals for an ANN model of a karstic aquifer response

Following many applications artificial neural networks (ANNs) have found in hydrology, a question has been rising for quantification of the output uncertainty. A pre‐optimized ANN simulated the hydraulic head change at two observation wells, having as input hydrological and meteorological parameters. In order to calculate confidence intervals (CI) for the ANN output two bootstrap methods were examined namely bootstrap percentile and BCa (Bias‐Corrected and accelerated). The actual coverage of the CI was compared to the theoretical coverage for different certainty levels as a means of examining the method's reliability. The results of this work support the idea that the bootstrap methods provide a simple tool for confidence interval computation of ANNs. Comparing the two methods, the percentile requires fewer calculations and yields narrower intervals with similar actual coverage to that of BCa. Overall, the actual coverage was proved lower than desired when not modeled points were present in the data subset. Copyright © 2011 John Wiley & Sons, Ltd.     

Single close encounters in the planetary problem

We consider a large massM and two small massesm ₁ andm ₂ (m ₁ m ₂;m ₁,m ₂M). The orbit ofm ₁ is initially circular and the motion ofm ₂ hyperbolic with respect toM. The orbital elements of the small masses are strongly modified after a close, single encounter betweenm ₁ andm ₂.An approximative method, similar to the theory of stellar encounters, is used to determine the probabilities of collisions, hyperbolas, direct and retrograde ellipses, as well as the mean values of the semimajor axes and their root mean square deviation after the encounter.The results are close to those which are obtained if the massm ₂ is negligibly small, (Mm ₁m ₂;m ₂ 0), as should be also expected on general grounds.     

The effects of suspended sediment on brown trout (Salmo trutta) feeding and respiration after macrophyte control

Sediment resuspension during and after mechanical excavation of macrophytes may have a significant impact on resident fish populations. Unfortunately, little is known about the influence of this sediment on the respiratory performance and feeding abilities of fishes in New Zealand waterways. We examined the effects of suspended sediment (SS) concentrations previously observed after a large-scale macrophyte removal operation on oxygen consumption (MO₂) and feeding rates of brown trout (Salmo trutta). MO₂ at 0 mg L^?1, 150 mg L^?1, 300 mg L^?1, 450 mg L^?1 and 600 mg L^?1 of SS was measured using semi-closed respirometry. Feeding rates at the same SS concentrations were also measured using laboratory tank experiments. Results suggest that SS concentrations up to 600 mg L^?1 have no effect on MO₂. Conversely, feeding rates were significantly reduced at 450 mg L^?1 (22% reduction) and 600 mg L^?1 (31% reduction), indicating that sediment concentrations above 450 mg L^?1 may negatively affect brown trout populations.     

Spatial variability estimation and risk assessment of the aquifer level at sparsely gauged basins using geostatistical methodologies

The spatial variability evaluation of the water table level of an aquifer provides useful information in water resources management plans. Three different approaches are applied to estimate the spatial variability of the water table in the study basin. All of them are based on the Kriging methodology. The first is the classical Ordinary Kriging approach, while the second involves information from a secondary variable (surface elevation) and the application of Residual Kriging. The third calculates the probability to lie below a certain groundwater level limit that could cause significant problems in groundwater resources availability. The latter is achieved by means of Indicator Kriging. A recently developed non-linear normalization method is used to transform both data and residuals closer to normal distribution for improved prediction results. In addition, the recently developed Spartan variogram model is applied to determine the spatial dependence of the measurements. The latter proves to be the optimal model, compared to a series of models tested, which provides in combination with the Kriging methodologies the most accurate cross validation estimations. The variogram form is explained with respect to the radius of influence of the pumping wells representing the spatial impact of the pumping activity. Groundwater level and probability maps are developed providing the ability to assess the spatial variability of the groundwater level in the basin and the risk that certain locations have in terms of a safe groundwater level limit that has been set for the sustainability of the groundwater resources of the basin.     

A hydro‐economic modelling framework for flood damage estimation and the role of riparian vegetation

A modelling framework for the quick estimate of flood inundation and the resultant damages is developed in this paper. The model, called the flood economic impact analysis system (FEIAS), can be applied to a river reach of any hydrogeological river basin. For the development of the integrated modelling framework, three models were employed: (1) a modelling scheme based on the Hydrological Simulation Program FORTRAN model that was developed for any geomorphological river basin, (2) a river flow/floodplain model, and (3) a flood loss estimation model. The first sub‐model of the flood economic impact analysis system simulates the hydrological processes for extended periods of time, and its output is used as input to a second component, the river/floodplain model. The hydraulic model MIKE 11 (quasi‐2D) is the river/floodplain model employed in this study. The simulated flood parameters from the hydraulic model MIKE 11 (quasi‐2D) are passed, at the end of each time step, to a third component, the flood loss model for the estimation of flood damage. In the present work, emphasis was given to the seasonal variation of Manning's coefficient (n), which is an important parameter for the determination of the flood inundation in hydraulic modelling. High values of Manning's coefficient for a channel indicate high flow resistance. The riparian vegetation can have a large impact on channel resistance. The modelling framework developed in this paper was used to investigate the role of riparian vegetation in reducing flood damage. Moreover, it was used to investigate the influence of cutting riparian vegetation scenarios on the flow characteristics. The proposed framework was applied to the downstream part of the Koiliaris River basin in Crete, Greece, and was tested and validated with historical data. Copyright © 2012 John Wiley & Sons, Ltd.     

Pre-capture orbits of the moon

If the mass of the Earth was not considerably larger than at present, the pre-capture orbit of the Moon was in the range 0.9–1.1 A.U. Capture occurred within several 10⁸ years after formation of the Moon.     

Olive-oil mill wastewater transport under unsaturated and saturated laboratory conditions using the geoelectrical resistivity tomography method and the FEFLOW model

An integrated approach for monitoring the vertical transport of a solute into the subsurface by using a geophysical method and a simulation model is proposed and evaluated. A medium-scale (1 m³) laboratory tank experiment was constructed to represent a real subsurface system, where an olive-oil mill wastewater (OOMW) spill might occur. High-resolution cross-hole electrical resistivity tomography (ERT) was performed to monitor the OOMW transport. Time-lapse ERT images defined the spatial geometry of the interface between the contaminated and uncontaminated soil into the unsaturated and saturated zones. Knowing the subsurface characteristics, the finite element flow and transport model FEFLOW was used for simulating the contaminant movement, utilizing the ERT results as a surrogate for concentration measurements for the calibration process. A statistical analysis of the ERT measurements and the corresponding transport model results for various time steps showed a good agreement between them. In addition, a sensitivity analysis of the most important parameters of the simulation model (unsaturated flow, saturated flow and transport) was performed. This laboratory-scale study emphasizes that the combined use of geophysical and transport-modeling approaches can be useful for small-scale field applications where contaminant concentration measurements are scarce, provided that its transferability from laboratory to field conditions is investigated thoroughly.     

Ecology of rocky reef fish of northeastern New Zealand: 50 years on

Abstract

This review examines the history and current understanding of the ecology of rocky reef fishes in northeastern New Zealand, marking the fiftieth anniversary of the Leigh Marine Laboratory. The cumulative number of publications increased steadily to c. 180 by 2012. Most attention has been on the snapper (Pagrus auratus, f. Sparidae), the largest in size and economically most important species, and on triplefins (Tripterygiidae), the smallest in size and most diverse family. A strong quantitative school emerged in the 1980s, a period when there was little research on temperate reef fishes elsewhere. A ‘bottom-up’ view of the effects of habitat structure on fish emerged, identifying the key roles of depth, topography and macroalgae. By 2000, attention shifted to using marine reserves as laboratories for ecological research. A ‘top-down’ view of fish as predators emerged, documenting cascading effects on prey communities and habitats. The two different viewpoints have not yet been integrated. Research gaps, including population connectivity and climate change, represent major challenges for the next 50 years.     

Fish recruitment into a South African temporarily open/closed temperate estuary during three different hydrological mouth phases

Quantitative recruitment of ichthyofauna into a South African temporarily open/closed estuary (TOCE) was investigated during three distinct hydrological phases: closed overwash, open outflow and tidal mouth conditions. Maximum recruitment densities occurred during outflow conditions and declined when the system became tidally inundated. Recruitment densities were lowest during overwash events. Selective diel variation was evident, with higher night-time densities for each phase. This study highlights the importance of seasonal timing and duration of open conditions and overwash events for structuring ichthyofaunal community composition of southern African TOCEs.