A hybrid artificial neural network-numerical model for ground water problems

Numerical models constitute the most advanced physical-based methods for modeling complex ground water systems. Spatial and/or temporal variability of aquifer parameters, boundary conditions, and initial conditions (for transient simulations) can be assigned across the numerical model domain. While this constitutes a powerful modeling advantage, it also presents the formidable challenge of overcoming parameter uncertainty, which, to date, has not been satisfactorily resolved, inevitably producing model prediction errors. In previous research, artificial neural networks (ANNs), developed with more accessible field data, have achieved excellent predictive accuracy over discrete stress periods at site-specific field locations in complex ground water systems. In an effort to combine the relative advantages of numerical models and ANNs, a new modeling paradigm is presented. The ANN models generate accurate predictions for a limited number of field locations. Appending them to a numerical model produces an overdetermined system of equations, which can be solved using a variety of mathematical techniques, potentially yielding more accurate numerical predictions. Mathematical theory and a simple two-dimensional example are presented to overview relevant mathematical and modeling issues. Two of the three methods for solving the overdetermined system achieved an overall improvement in numerical model accuracy for various levels of synthetic ANN errors using relatively few constrained head values (i.e., cells), which, while demonstrating promise, requires further research. This hybrid approach is not limited to ANN technology; it can be used with other approaches for improving numerical model predictions, such as regression or support vector machines (SVMs).     

A Ca ii H&K survey of γ Doradus candidates

In an investigation of the starspot hypothesis as it applies to the 'slowly variable' F-type dwarfs, we spectroscopically observed eight promising γ Doradus candidates to search for Ca  ii H&K emission. We found that there are no significant emission reversals in the cores of these resonance lines. Based on the ceiling flux calculations of the Ca  ii K line and on calculations of the Rossby number, we conclude that there is no support for the presence of strong magnetic activity and the starspot hypothesis in these objects.     

Oxygen isotope composition and geothermometry of granulite to greenschist facies metamorphic rocks: a study from the Neoproterozoic collision-related nappe system, south of São Francisco Craton, SE Brazil

Oxygen isotope studies were carried out across units of a Neoproterozoic nappe system, south of São Francisco Craton. A temperature decrease toward the base of the system is found, consistent with a previously recognized inverted metamorphic pattern. The tectonic contact of the basal unit and the reworked southern São Francisco craton show a steep temperature gradient, suggesting that low temperature thrusting acted as the dominant tectonic process. The contrasts between the δ¹⁸O values of the Três Pontas-Varginha and Carmo da Cachoeira nappes and the differences among the samples and minerals are consistent with the preservation of sedimentary isotopic composition during metamorphism. The small differences in the δ¹⁸O values between the undeformed and the deformed calc-silicate samples (1.6‰) suggest that the δ¹⁸O value of mylonitization fluids was close to that which equilibrated with the metamorphic assemblage. The distinct δ¹⁸O values of metapelitic and calc-silicate samples and the great temperature difference from one type to the other indicate that no large-scale fluid interaction processes occurred during metamorphism. Oxygen isotopic estimations of both Três Pontas-Varginha undeformed rocks and Carmo da Cachoeira unaltered equivalents indicate δ¹⁸O values of up to 18‰. Comparison between these values and those from the ‘basement’ orthogneisses (8.3–8.5‰) indicates the latter are not sources for the metapelites.     

Seismic motion and response prediction alternatives

Statistical methods are available which predict the maximum response of simple oscillators given the peak acceleration (A_p), peak velocity (V_p) or peak displacement (D_p) of seismic ground motions. An alternative parameter, namely an ordinate (or ordinates) of the Fourier amplitude spectrum of ground motion acceleration, FS(f), may in fact be a preferred predictor of peak response, especially in a frequency range close to f. Other statistical methods (attenuation laws) use distance R and other parameters such as magnitude (M), Modified Mercalli epicentral Intensity (I_o) and Modified Mercalli site Intensity (MMI or I_s) to predict spectral velocity (S_v(f)), etc. In using such approaches, it is most desirable to know the total uncertainty in the predicted peak response of the system given the starting parameter values. An extensive strong motion data set is used to study these questions, The most direct prediction models are found to be preferable (have lower prediction dispersion) but data may not be available in all regions to permit their use.     

利用遥感监测和分析珠江三角洲的城市扩张过程——以东莞市为例

伴随着经济的快速增长,珠江三角洲近年来出现了迅速的城市化以及乡村城市化的过程。沿着城镇边缘和公路两旁,大片的农田被侵占来单方面地满足城市开发的需要,忽略了粮食产量减少以及环境质量恶化等问题。利用遥感可以有效地监测珠江三角洲的城市扩张过程。在遥感多时相监测的基础上通过熵的计算,可以定量地描述城市扩张的空间规律及扩散过程。     

Impact of climate warming on upper layer of the Bering Sea

The impact of climate warming on the upper layer of the Bering Sea is investigated by using a high-resolution coupled global climate model. The model is forced by increasing atmospheric CO₂ at a rate of 1% per year until CO₂ reaches double its initial value (after 70 years), after which it is held constant. In response to this forcing, the upper layer of the Bering Sea warms by about 2°C in the southeastern shelf and by a little more than 1°C in the western basin. The wintertime ventilation to the permanent thermocline weakens in the western Bering Sea. After CO₂ doubling, the southeastern shelf of the Bering Sea becomes almost ice-free in March, and the stratification of the upper layer strengthens in May and June. Changes of physical condition due to the climate warming would impact the pre-condition of spring bio-productivity in the southeastern shelf.     

The ionospheric impact of the October 2003 storm event on Wide Area Augmentation System

The United States Federal Aviation Administrations (FAA) Wide-Area Augmentation System (WAAS) for civil aircraft navigation is focused primarily on the Conterminous United States (CONUS). Other Satellite-Based Augmentation Systems (SBAS) include the European Geostationary Navigation Overlay Service (EGNOS) and the Japanese Multi-transport Satellite-based Augmentation System (MSAS). Navigation using WAAS requires accurate calibration of ionospheric delays. To provide delay corrections for single frequency global positioning system (GPS) users, the wide-area differential GPS systems depend upon accurate determination of ionospheric total electron content (TEC) along radio links. Dual-frequency transmissions from GPS satellites have been used for many years to measure and map ionospheric TEC on regional and global scales. The October 2003 solar-terrestrial events are significant not only for their dramatic scale, but also for their unique phasing of solar irradiance and geomagnetic events. During 28 October, the solar X-ray and EUV irradiances were exceptionally high while the geomagnetic activity was relatively normal. Conversely, 29–31 October was geomagnetically active while solar irradiances were relatively low. These events had the most severe impact in recent history on the CONUS region and therefore had a significant effect on the WAAS performance. To help better understand the event and its impact on WAAS, we examine in detail the WAAS reference site (WRS) data consisting of triple redundant dual-frequency GPS receivers at 25 different locations within the US. To provide ground-truth, we take advantage of the three co-located GPS receivers at each WAAS reference site. To generate ground-truth and calibrate GPS receiver and transmitter inter-frequency biases, we process the GPS data using the Global Ionospheric Mapping (GIM) software developed at the Jet Propulsion Laboratory. This software allows us to compute calibrated high resolution observations of TEC. We found ionospheric range delays up to 35 m for the day-time CONUS during quiet conditions and up to 100 m during storm time conditions. For a quiet day, we obtained WAAS planar fit slant residuals less than 2 m (0.4 m root mean square (RMS)) and less than 25 m (3.4 m RMS) for the storm day. We also investigated ionospheric gradients, averaged over distances of a few hundred kilometers. The gradients were no larger than 0.5 m over 100 km for a quiet day. For the storm day, we found gradients at the 4 m level over 100 km. Similar level gradients are typically observed in the low-latitude region for quiet or storm conditions.