Modeling River Stage‐Discharge Relationships Using Different Neural Network Computing Techniques

One of the most important problems in hydrology is the establishment of rating curves. The statistical tools that are commonly used for river stage‐discharge relationships are regression and curve fitting. However, these techniques are not adequate in view of the complexity of the problems involved. Three different neural network techniques, i. e., multi‐layer perceptron neural network with Levenberg‐Marquardt and quasi‐Newton algorithms and radial basis neural networks, are used for the development of river stage‐discharge relationships by constructing nonlinear relationships between stage and discharge. Daily stage and flow data from three stations, Yamula, Tuzkoy and Sogutluhan, on the Kizilirmak River in Turkey were used. Regression techniques are also applied to the same data. Different input combinations including the previous stages and discharges are used. The models' results are compared using three criteria, i. e., root mean square errors, mean absolute error and the determination coefficient. The results of the comparison reveal that the neural network techniques are much more suitable for setting up stage‐discharge relationships than the regression techniques. Among the neural network methods, the radial basis neural network is found to be slightly better than the others.     

Climate change and modelling of extreme temperatures in Switzerland

This study models maximum temperatures in Switzerland monitored in twelve locations using the generalised extreme value (GEV) distribution. The parameters of the GEV distribution are determined within a Bayesian framework. We find that the parameters of the underlying distribution underwent a substantial change in the beginning of the 1980s. This change is characterised by an increase both in the level and the variability. We assess the likelihood of the heat wave of the summer 2003 using the fitted GEV distribution by accounting for the presence of a structural break. The estimation results do suggest that the heat wave of 2003 is not that statistically improbable if an appropriate methodology is used for dealing with nonstationarity.     

Transport properties in small-scale coastal flows: relative dispersion from VHF radar measurements in the Gulf of La Spezia

Lagrangian transport characteristics in the Gulf of La Spezia, a 5 × 10-km area along the western coast of Italy, are investigated using data collected from a very high frequency (VHF) radar system with 250 m and 30-min resolution and two clusters of Coastal Dynamics Experiment surface drifters during 2 weeks in the summer of 2007. The surface drifters are found to follow the temporal and spatial evolution of the finite-scale Lyapunov exponents (FSLEs) computed by the VHF radar, indicating the precision of both the radar measurements and the diagnostic FSLE in mapping accurately the transport pathways. In light of this agreement, an analysis of the relative dispersion is conducted. It is found that the average FSLE value varies within a narrow range of 4   day^-1 ￡ l ￡ 7   day^-14 \;\mbox{day}^{-1} \leq \lambda \leq 7 \;\mbox{day}^{-1} and displays an exponential regime over the entire extent of the measurements. The dynamical implication is that relative dispersion is controlled nonlocally, namely by slow, persistent, energetic mesoscale structures as opposed to the rapidly evolving high-gradient small-scale turbulent features. The value of the exponent is about an order of magnitude larger than those found in previous modeling studies and analysis of SCULP data in the Gulf of Mexico but somewhat smaller than that estimated from CLIMODE drifters in the Gulf Stream region. Scaling of the FSLE using a metric of resolved gradients of the Eulerian fields in the form of a positive Okubo–Weiss criterion is useful, but not as precise as in modeling studies. The horizontal flow convergence is found to have a small yet tangible effect on relative dispersion.     

Seasonal and spatial changes in water and sediment quality variables in Bafa Lake

The brackish Bafa Lake located in the southwestern part of Turkey is under stress because of both natural and untreated wastewater effluents. The purpose of this research is to determine spatiotemporal distributions of some physicochemical variables in water column (temperature, salinity, pH, conductivity, dissolved oxygen, NH₄–N, NO₂–N, NO₃–N, oPO₄–P, TPO₄–P, chlorophyll-a, total suspended solids) and sediment (TN, TC, TOC, TP) and their relationships at coastal stations. In the water column, nitrate and phosphate concentrations showed seasonal variations with high values recorded in winter period. Ammonium was determined as a main source of TIN component. During summer period, a large amount of total phosphorus was found as dissolved organic form. However, in the winter period, inorganic phosphate levels increased at sampling stations. N limitation was a common feature throughout the lake where P-limitation was only observed in summer period. The total phosphorus levels which showed hypereutrophic condition at the western part of the lake changed between 1.55 and 4.99 μM and did not remain in the range for uncontaminated condition. In the lake sediment, a strong relationship was found between TOC and TC levels. Generally, the mean TOC concentrations constitute small amount of TC values in the sampling stations. The results also indicated that a strong correlation exists between TOC and TN values, and TN was greatly regulated by organic sources. In the lake, TOC:TN ratios changed between 5 and 13; the ratio greater than 10 could be an indicator of algal and land plant sources mixing as an organic matter.     

Geochemistry of uranium and thorium and natural radioactivity levels of the western Anatolian plutons,Turkey

Mineralogy and Petrology - Seventy samples from major plutons (mainly granitic) of Western Anatolia (Turkey) have been analyzed by γ-ray spectrometry to determine the specific activities of...     

Possible Light-Time Effect in the Eclipsing Binary System TY Boo

The O − C diagram for the eclipsing binary system TY Boo was constructed with the new minima times observed at the Ankara University Observatory along with the collected ones from the literature. The O − C diagram shows a cyclic variation superimposed on a quadratic variation. The quadratic variation can be explained in terms of mass loss/exchange mechanism in the system while the cyclic variation is attributed to a possible light-time effect caused by a third body revolving around the close binary.     

Hydrogeochemical and isotopic investigation of the Kolan geothermal field, southeastern Turkey

The Kolan geothermal field is located 24 km northwest of the Karakoçan area of Elaz city in southeastern Turkey. The Keban crystallized limestone of Permo-Carboniferous age is the oldest unit in the area. This unit is overlain by the middle-upper Eocene Krkgeçit formation consisting of conglomerate, sandstone, limestone and marl alternations. The youngest unit in the area is the Upper Miocene-Lower Pliocene Karabakr formation consisting of basalt, andesite and tuffs. Thermal Springs in the Kolan field issue along two major fault zones. The temperature of the hot springs is between 38°C and 45°C. Thermal waters in the area are heated by the geothermal gradient. The waters are classified as hot or very hot and as brackish waters. On the basis of International Association of Hydrogeologists Classification (IAH), waters are grouped as Ca–Na-HCO₃–CO₂-bearing thermal and mineral waters. Chemical analyses indicate a chemical composition of Ca>(Na+K)>Mg - HCO₃>SO₄>Cl. Saturation calculations reveal that the waters are oversaturated with respect to calcite, dolomite and aragonite while undersaturated with respect to gypsum, anhydrite and halite minerals. On the basis of ¹⁸O, ²H and ³H isotope data, Kolan waters are of meteoric origin, recharged from high elevations and then circulated deeply.     

Chemical Weathering Rates, Erosion Rates and Mobility of Major and Trace Elements in a Boreal Granitic Till

Chemical weathering rates and erosionrates of granitic till in northern Sweden have beenestimated. The present-day chemical weathering rate iscompared with the long-term average weathering ratesince the last deglaciation approximately 8,700 yearsago. Also, the present-day release rates of major andtrace elements due to chemical weathering are comparedwith the mobility of these elements in a spodosolprofile as shown by soil water samples from the vadozezone. The estimation of the past weathering rate isbased on elemental depletion trends in a soil profile(typic haplocryod), whereas the present weatheringrate is based on elemental input/output budgets in asmall catchment (9.4 km²). The long-term averagechemical erosion rate, expressed as the sum of majorelement oxides (SiO₂, Al₂O₃, CaO,Fe₂O₃, K₂O, MgO, MnO, Na₂O,P₂O₅, TiO₂), was estimated to be4.9 gm^-2 yr^-1. The long-term base cation(Ca²⁺, Mg²⁺, Na⁺, K⁺) depletionwas 0.325 keq ha^-1 yr^-1. The currentchemical erosion rate was estimated to be2.43.0 g ^-2 yr^-1, which is at least an order ofmagnitude higher than the rate of physical erosion,and the base cation flux due to chemical weathering is0.356–0.553~keq ha^-1 yr^-1. However,0.074 keq ha^-1 yr^-1 of this flux may be related tocation exchange processes induced by atmospheric inputof acid rain. There is no evidence for any recentlyincreased weathering rate of silicates in this area.The inputs of Cd, Cu, Ni and Zn exceed the outputs,and hence, these elements are currently accumulatingin the soil.There is a distinct seasonal variation in thechemical composition of the soil water. Results fromthe soil water samples show that Na, followed by Siand S (Cl was not measured), are the most mobileelements in the spodosol profile between the E-horizonand the C-horizon, and that Al and Fe were the leastmobile elements. However, there is no simplerelationship between the mobility of an element in thespodosol profile and the current release rate due toweathering of that element. This fact may haveimplications for the validity of the comparisonbetween the past and present weathering rates asperformed in this study.     

Comets and the formation of planets

A morphological study of the physical and dynamical processes of planet formation is presented, with emphasis on the intermediary role of comet nuclei. Although guided by a particular model of the evolution of the pre-planetary solar nebula, implying the freezing-out of hydrogen in the region of the giant planets, the derivations and conclusions are of wider import, applicable to other cosmogonic models as well as to certain phases of star formation. The items evaluated physically, dynamically, or statistically comprise: (1) the total number mass of comets in Oort's cloud; (2) a re-evaluation of the diameters and masses of cometary nuclei; (3) the processes of nucleation from gravitational and Boltzmann instabilities of gaseous media to agglomerations of particulate matter as conditioned by inbuilt angular momentum; (4) the statistical-dynamical conditions and time scales of orbital interaction of comets with the planets and the consequences of disintegration.A consistent model proposes the formation of comets and planets in pre-planetary rings of the residual solar nebula, with subsequent ejection, chiefly by Jupiter, of the comets to Oort's sphere. Screening by absorbing matter is not only probable, but necessary to protect the comets from dis-integration during the process of ejection.Paper dedicated to Prof. H. C. Urey on the occasion of his 80th birthday on 29 April, 1973.This work has been currently supported by grants from the National Aeronautics and Space Administration, Goddard Space Flight Center, Greenbelt, Maryland.