Comparison of trends in stream water quality

A method is described for assessing the strength of evidence for differences in the trends in chemical concentrations in stream water between catchments. A smoothing spline technique is used to model changes in water quality as a result of changes in flow rates, seasonal effects and an underlying trend. The method involves fitting the model twice, once with the underlying trends constrained to be identical for each river and once with the trends unconstrained. Statistical properties are assessed by simulation methods that allow for the autocorrelation in the residuals from the unconstrained fit. The method is applied to data from two streams in the Balquhidder catchments in the Highlands of Scotland and to data from the Dee, Don and Ythan rivers in the north‐east of Scotland. Copyright © 2003 John Wiley & Sons, Ltd.     

On the impact of cloudiness on the characteristics of nocturnal downslope flows

The effects of cloud cover amount and the height of cloud base on nighttime thermally induced downslope flow were investigated using analytical and numerical model approaches. The conclusions obtained with the analytical and the numerical model evaluations agreed. It was concluded that, (i) as cloud cover increases and/or the height of cloud base decreases, the depth and the intensity of nighttime thermally-induced downslope flows may decrease by a factor reaching one sixth and one tenth, respectively, in the case of overcast low cloud; (ii) when skies suddenly cloud over around midnight, the development of the downslope flow is altered in different ways: a reduction in intensity; or a cessation of further development, depending on the fraction of cloud coverage, and (iii) with a sudden clearing of overcast low cloud around midnight, the depth and the intensity of the downslope flow increases significantly.On leave from the Institute of Atmospheric Physics, Academia Sinica, Beijing, China.On leave from CSIRO, Division of Atmospheric Research, Private Bag No. 1, Mordialloc, Victoria 3195, Australia.     

Limitations on outer planet mass determinations from their mutual perturbations

The limitations on determining the masses of the outer planets from their mutual perturbations are investigated based on the magnitudes of the periodic perturbations given in general theories and the accuracy of the observational data.     

Amplification of primordial magnetic fields by anisotropic gravitational collapse

If a magnetic field is frozen into a plasma that undergoes spherical compression, then the magnetic field B varies with the plasma density ρ according to   B ∝ρ^2/3  . In the gravitational collapse of cosmological density perturbations, however, quasi-spherical evolution is very unlikely. In anisotropic collapses the magnetic field can be a much steeper function of gas density than in the isotropic case. We investigate the distribution of amplifications in realistic gravitational collapses from Gaussian initial fluctuations using the Zel'dovich approximation. Representing our results using a relation of the form   B ∝ρ^α  , we show that the median value of α can be much larger than the value  α= 2/3  resulting from spherical collapse, even if there is no initial correlation between magnetic field and principal collapse directions. These analytic arguments go some way towards understanding the results of numerical simulations.     

Multispectral imaging observations of Neptune’s cloud structure with Gemini-North

Observations of Neptune were made in September 2009 with the Gemini-North Telescope in Hawaii, using the NIFS instrument in the H-band covering the wavelength range 1.477–1.803 μm. Observations were acquired in adaptive optics mode and have a spatial resolution of approximately 0.15–0.25″.The observations were analysed with a multiple-scattering retrieval algorithm to determine the opacity of clouds at different levels in Neptune’s atmosphere. We find that the observed spectra at all locations are very well fit with a model that has two thin cloud layers, one at a pressure level of ∼2 bar all over the planet and an upper cloud whose pressure level varies from 0.02 to 0.08 bar in the bright mid-latitude region at 20–40°S to as deep as 0.2 bar near the equator. The opacity of the upper cloud is found to vary greatly with position, but the opacity of the lower cloud deck appears remarkably uniform, except for localised bright spots near 60°S and a possible slight clearing near the equator.A limb-darkening analysis of the observations suggests that the single-scattering albedo of the upper cloud particles varies from ∼0.4 in regions of low overall albedo to close to 1.0 in bright regions, while the lower cloud is consistent with particles that have a single-scattering albedo of ∼0.75 at this wavelength, similar to the value determined for the main cloud deck in Uranus’ atmosphere. The Henyey-Greenstein scattering particle asymmetry of particles in the upper cloud deck are found to be in the range g ∼ 0.6–0.7 (i.e. reasonably strongly forward scattering).Numerous bright clouds are seen near Neptune’s south pole at a range of pressure levels and at latitudes between 60 and 70°S. Discrete clouds were seen at the pressure level of the main cloud deck (∼2 bar) at 60°S on three of the six nights observed. Assuming they are the same feature we estimate the rotation rate at this latitude and pressure to be 13.2 ± 0.1 h. However, the observations are not entirely consistent with a single non-evolving cloud feature, which suggests that the cloud opacity or albedo may vary very rapidly at this level at a rate not seen in any other giant-planet atmosphere.     

Abundance of Fe relative to H at 1.5 solar radii

The abundance of Fe relative to H is obtained by using resonantly scattered intensities of λ 284 of Fe XV that were measured with OSO-7 and resonantly scattered intensities of Lα of H I that were obtained by Gabriel (1971). Because of possible differences in electron densities along lines of sight for these non-simultaneous measurements and in relative calibrations, results are rather uncertain but still indicate that the average Fe abundance relative to H in the corona appears to be at least as large as a recent photospheric abundance. Some limitations in using this method for obtaining abundances are examined for future experiments with simultaneous measurements and well calibrated detectors.     

A spreadsheet method of estimating best-fit hydraulic gradients using head data from multiple wells

Hydraulic gradients from planar water tables, or piezometric surfaces, and horizontal flow regimes can be quickly and conveniently calculated from data sets involving numerous wells. The matrix-solving functions of a modem spreadsheet program (Excel) were used to determine the equation of a water-table plane, Ax + By + Cz - D = 0, and the equation coefficients were then used to determine the magnitude of the hydraulic gradient, according to gradient = square root of A2 + B2/C2, and its direction, according to alpha = arctan B/A, where alpha is the angle measured from the x-axis. A pre-prepared Excel file constructed to handle data from up to 20 wells at once is available for free downloading at www.geo.ku.edu/hydro/KUHydro.html.     

Petrogenesis of Cenezoic volcanic rocks from the Aegean island arc

The Aegean volcanic arc formed in response to northeasterly subduction of the Mediterranean sea floor beneath the Aegean Sea. The active arc lies over 250 km from the Hellenic Trench in a region which has suffered considerable extension and subsidence since the mid-Tertiary. Suites of samples from the different volcanic centres making up the arc have been studied geochemically in order to assess lateral variations and to constrain the contribution of crustal contamination and sediment subduction in their petrogenesis.Lavas from all the major volcanic centres exhibit typical calc-alkaline major-element characteristics, and show enrichment in light REE and LIL elements but low contents of HFS elements. The enrichment in light REE is greater in the eastern (Nisyros, Kos) and western (Milos, Poros, Methana, Aegina) sectors of the arc (Ce_n/Yb_n=4) than in the central Santorini sector (Ce_n/Yb_n=2). All lavas have significant negative Eu anomalies and many have slight negative Ce anomalies. Less coherence is observed in the abundances and ratios of the other LIL elements, compared with the REE, along the island chain.Whereas the effects of crystal fractionation are evident in the trace-element patterns of lavas from individual islands, and are particularly well marked for Santorini, it is clear that there are consistent differences in trace-element abundances and ratios in the lavas of the various islands which reflect compositional differences in the mantle source and/or in melting conditions. Lavas from the eastern and western sectors have much higher levels of Ba and Sr but relatively lower Th, K and Rb than those from Santorini. Although some geochemical features could be explained through involvement of a component of subducted sediment in the source regions of the volcanoes, other element abundances and ratios indicate that this component must be very small. Detailed consideration of the inter-island geochemical variations suggests a complex make-up of the underlying lithosphere, resulting from a long history of subduction. In the region of Santorini, where crustal stretching is greatest, the underlying asthenosphere may be involved in magma production.     

Buckling of multi-segment underwater pressure hull

Results of a numerical and experimental study into buckling performance of multi-segment pressure hull subjected to uniform hydrostatic pressure are discussed. Constituents of multi-segment configurations are bowed-out cylindrical shells with, and without flanges. Details about five collapse tests of laboratory scale mild steel, CNC machined models are given. Segments were about 200 mm diameter, 100 mm long and had uniform wall thickness of 3 mm. Experimental collapse pressures were in the range from 12 to 20 MPa. Numerical collapse pressures agreed well with those obtained during experiments.