Using model error in response history analysis to evaluate component calibration methods

This research is part of a larger effort to better understand and quantify the epistemic model uncertainty in dynamic response-history simulations. This paper focuses on how calibration methods influence model uncertainty. Structural models in earthquake engineering are typically built up from independently calibrated component models. During component calibration, engineers often use experimental component response under quasi-static loading to find parameters that minimize the error in structural response under dynamic loading. Since the calibration and the simulation environments are different, if a calibration method wants to provide optimal parameters for simulation, it has to focus on features of the component response that are important from the perspective of global structural behavior. Relevance describes how efficiently a calibration method can focus on such important features. A framework of virtual experiments and a methodology is proposed to evaluate the influence of calibration relevance on model error in simulations. The evaluation is demonstrated through a case study with buckling-restrained braced frames (BRBF). Two calibration methods are compared in the case study. The first, highly relevant calibration method is based on stiffness and hardening characteristics of braces; the second, less relevant calibration method is based on the axial force response of braces. The highly relevant calibration method consistently identified the preferable parameter sets. In contrast, the less relevant calibration method showed poor to mediocre performance. The framework and methodology presented here are not limited to BRBF. They have the potential to facilitate and systematize the improvement of component-model calibration methods for any structural system.     

Water isotopic composition traces source and dynamics of water supply in a semi-arid agricultural landscape

Changes in seasonality and form of precipitation alter the structure and function of grassland and steppe ecosystems and pose challenges for land management and crop production in regions like the Northern Great Plains, North America. This research uses isotopic composition of water (δ¹⁸O and δ²H) to explore the sources and fate of soil water in lower-elevation agricultural areas of the Judith River watershed, in the headwaters of the Missouri River, USA. Extensive non-irrigated cereal crop production in this area occurs on well-drained soils and depends on careful water management. Our observations indicate that colder precipitation contributes isotopically distinct water to cultivated terrace soils relative to downgradient groundwaters and streams. Riparian waters also exhibit a higher fraction of contributions from colder precipitation relative to terrace groundwaters and streams. Apparent contributions from colder precipitation in terrace and riparian soil waters suggest that snowmelt is a key component of the water supply to these systems. Riparian waters also show evidence of evaporation suggesting that water spends sufficient time in some ponds and open channels in the riparian corridor to reflect fractionation by evaporation. The evolution of water isotopic composition from soils to shallow aquifers to stream corridors indicates source water partitioning as precipitation moves through this semi-arid agricultural landscape. The apparent mixing processes evident in this evolution reveal source water dynamics that are necessary to understand plant transpiration, solute processing, and contaminant leaching processes.     

A confirmation of the correlation between hydrocarbons and chlorophyll a in the upper euphotic zone

The average hydrocarbon content found in 14 water samples from the euphotic zone off Northwest Africa was 4.4 μg l.^?1 with no extreme values. The average chlorophyll $\text{a}$ content was 1.9 μg l.^?1. The data fit a model proposed in a previous paper (Zsolnay, 1977). The resulting equation was $\text{HC}\text{= ?2.7 + 4.82}\text{Ch}\text{?0.942}\text{Ch}{}^2$ and could explain 59% of the variance in the hydrocarbon distribution. This indicates a correlation that is significant at the 0.002 level.     

Accurate Acoustic Signal Parameter Estimation for Marine Geodesy Surveys

Acoustic signal parameter estimation is important for diverse marine geodesy surveys and several other applications. However, the received signal from a far-field target characterized by planar wavefront propagation is frequently affected by strong nearby interfering signals. Their presence deteriorates the performance of direction-of-arrival (DOA) estimation for far-field target. In order to enhance the reception of signal from far-field target, the near-field/far-field (NFFF) beamformer is proposed. Such a beamformer optimizes beam pattern for far-field detection by maximizing beamformer output in the direction of the far-field target with the imposed condition to eliminate interfering signals generated in near-field locations. As the interference suppression only occurs at the position of near-field interference, a possible blind zone for far-field detection in conventional methods is not created. The NFFF beamformer is applicable for coherent signals and the scenario with multi interferences. For stationary situation where interferers locations are fixed, the NFFF beamformer computations do not require time updates with associated computational load. Furthermore the proposed method can be extended to several new situations such as acoustic monitoring performed from a stationary platform subjected to water currents, waves, winds and other variables, all of them generating nearby interferences and also to different array configurations including 2D and 3D arrays.     

Burial of organic carbon in Holocene sediments of the Zhujiang (Pearl River) and Changjiang (Yangtze River) estuaries

The East Asian marginal seas are important sinks of terrigenous materials transported by large rivers. In this study two cores from the Changjiang (Yangtze River) and Zhujiang (Pearl River) estuaries and one core from the inner shelf off of Hong Kong were investigated to examine the burial of organic matter during the postglacial period and its possible links with paleoenvironmental changes. Based on a simple two end-member mixing model, the terrestrial organic matter supplied primarily from the Zhujiang and Changjiang dominates the estuarine areas while marine organic matter contributes more to the inner-mid shelf. The competing contributions of terrestrial and marine organic matter are responsible for the downcore variations of organic elemental compositions. The overall decreasing shifts of total organic carbon concentrations and total organic carbon to total nitrogen ratios in the Zhujiang estuarine sediments since 6.5 kaBP seems to support the notion that depth profiles of organic matter compositions deposited in the Zhujiang Estuary can aid in the reconstruction of monsoon history in the Holocene. Nevertheless, organic matter compositions in the Changjiang Estuary and inner shelf off of southeastern Hong Kong respond in a different and more complex way to freshwater discharges or precipitation changes (monsoon variability) in the catchments, owing to complex controls of deposition and preservation of organic matter in these estuarine and shelf environments. Caution is therefore needed in using organic elemental and isotopic compositions to decipher paleoenvironmental changes in East Asian continental shelves where intense river–sea interactions occur and sedimentary environments change drastically.     

On the importance of training methods and ensemble aggregation for runoff prediction by means of artificial neural networks

ABSTRACT

Artificial neural networks (ANNs) become widely used for runoff forecasting in numerous studies. Usually classical gradient-based methods are applied in ANN training and a single ANN model is used. To improve the modelling performance, in some papers ensemble aggregation approaches are used whilst in others, novel training methods are proposed. In this study, the usefulness of both concepts is analysed. First, the applicability of a large number of population-based metaheuristics to ANN training for runoff forecasting is tested on data collected from four catchments, namely upper Annapolis (Nova Scotia, Canada), Biala Tarnowska (Poland), upper Allier (France) and Axe Creek (Victoria, Australia). Then, the importance of the search for novel training methods is compared with the importance of the use of a very simple ANN ensemble aggregation approach. It is shown that although some metaheuristics may slightly outperform the classical gradient-based Levenberg-Marquardt algorithm for a specific catchment, none performs better for the majority of the tested ones. One may also point out a few metaheuristics that do not suit ANN training at all. On the other hand, application of even the simplest ensemble aggregation approach clearly improves the results when the ensemble members are trained by any suitable algorithms.

EDITOR D. Koutsoyiannis; ASSOCIATE EDITOR E. Toth     

Measurements and interpretation of normal mode attenuation

summary . Measurements of Q for modes of free oscillation provide the most accurate information about the anelastic properties of the whole Earth in the period range from 100 to 3000 s. We have obtained more than 230 Q measurements, by using two different techniques. Individual LaCoste—Romberg gravimeter recordings of three large earthquakes were used to observe the time rate of decay of spectral peaks corresponding to different modes. This method provided measurements of Q for 37 different modes. By stacking 211 WWSSN recordings of two deep earthquakes, we were able to measure Q for 197 modes, including many overtones which cannot be analysed using the spectra of individual recordings.