Derivation of a Tidal Inundation Model to Support Environmental Research in Roebuck Bay (Western Australia)

A method for deriving mean inundation times for a large expanse of intertidal mudflats is presented. Image processing techniques are used to extract waterlines from two Landsat Thematic Mapper scenes at approximately high and low tides, respectively. Along with data based on standard tidal predictions, the waterlines are utilised to derive an initial Digital Elevation Model (DEM) for the tidal flat within a GIS environment. As the model fails to fully cover the mean spring tidal range, a routine is developed to linearly extrapolate this surface to the elevation of the mean spring low tide. Tidal interpolation equations are then utilised to generate an inundation time in hours for each cell on this extrapolated DEM. Output from the generalised model is a series of maps showing isolines of inundation for both the mean spring and neap tidal cycles, as well as sets of point data that may be integrated with existing biological sample data. These data will be used to support an ongoing geologic and biologic investigation of the bay and its ecology.     

Analysis of Difference Between Direct and Geodetic Mass Balance Measurements at South Cascade Glacier, Washington

Net mass balance has been measured since 1958 at South Cascade Glacier using the 'direct method,' e.g. area averages of snow gain and firn and ice loss at stakes. Analysis of cartographic vertical photography has allowed measurement of mass balance using the 'geodetic method' in 1970, 1975, 1977, 1979–80, and 1985–97. Water equivalent change as measured by these nearly independent methods should give similar results. During 1970–97, the direct method shows a cumulative balance of about −15 m, and the geodetic method shows a cumulative balance of about −22 m. The deviation between the two methods is fairly consistent, suggesting no gross errors in either, but rather a cumulative systematic error. It is suspected that the cumulative error is in the direct method because the geodetic method is based on a non-changing reference, the bedrock control, whereas the direct method is measured with reference to only the previous year's summer surface. Possible sources of mass loss that are missing from the direct method are basal melt, internal melt, and ablation on crevasse walls. Possible systematic measurement errors include under-estimation of the density of lost material, sinking stakes, or poorly represented areas.     

In situ uptake of mercury by the intertidal mussel,Mytilus californianus

The in situ uptake of mercury in tissues of Mytilus californianus was studied by exposing control specimens to waters at four depths near a submarine wastewater discharge for a period of 24 weeks. The response patterns observed for digestive gland, adductor muscle and gonadal tissues were strikingly different. Digestive gland concentrations increased rapidly to the heightened levels found for indigenous specimens. On the other hand, changes in the mercury content of adductor muscle were not statistically significant, and gonadal tissues generally exhibited a low rate of increase. No correlation with depth was observed.     

Physical disturbance creates bacterial dominance of benthic biological communities in tropical deltaic environments of the Gulf of Papua

Unlike many reactive continental shelf mud deposits in temperate regions, bacteria and microfauna rather than macrofauna typically dominate benthic biomass and activities over large areas of the Gulf of Papua (GoP) deltaic complex, Papua New Guinea. During mid NW monsoon periods (Jan–Feb), macrofaunal densities at Gulf stations were relatively low (), large macroinfauna were absent (upper 25 cm), and small (), surface deposit-feeding polychaetes and tubiculous amphipods were dominant, reflecting a frequently destabilized seabed and high sedimentation/erosion rates. Although frequent physical disturbance generally inhibits development of macrobenthic communities, some regions of the Gulf deposits are periodically colonized and extensively bioturbated during quiescent periods, as shown by preserved biogenic sedimentary structures. Bacterial inventories integrated over the top 20 cm were extremely variable within each sub region of the clinoform complex. A possible bimodal pattern with bathymetric depth and distance offshore may occur: lowest-inventories within the sandy, proximal Fly River delta, an open Gulf inner topset zone (10–20 m) having sites of relatively high inventories, an open Gulf mid-topset region with intermediate values and less extreme variation, and the outer topset—upper foreset zone (40–50 m) where highest values are attained (). Various measures of microbial activity, including measures proportional to the cellular rRNA content and the proportion of dividing cells, indicate extremely productive populations over the upper 1-m of the seabed throughout the Gulf of Papua region. Bacterial biomass (0–20 cm) including data of Alongi et al. (1991, 1992, 1995) varied from a low of in intertidal mud banks to a high of in the topset—foreset zone. Macrofaunal biomass did not exceed in any sampled region, ranging from 0.009±0 to with no obvious correlation with bathymetric depth (1–63 m). Meiofaunal biomass was generally an order of magnitude lower than macrofaunal biomass. Relatively elevated bacterial biomass and high turnover rates are consistent with high measured rates of benthic remineralization, presumably reflecting the rapid response time of bacteria to physical reworking, the associated entrainment of organic substrate, and flushing of metabolites. Solute exchange is also enhanced below the directly mixed surface region, possibly producing ‘far field’ stimulation of microbes in underlying deposits. Physical reworking and reoxidation of sediments between 10 and 50 m water depth maintain suboxic, nonsulfidic conditions in the upper 0.5–1 m despite active microbial communities and high benthic remineralization rates.     

Solubility of hydrous ferric oxide and iron speciation in seawater

Iron solubility equilibria were investigated in seawater at 36.22‰ salinity and 25°C using several filtration and dialysis techniques. In simple filtration experiments with 0.05 μm filters and Millipore ultra-filters, ferric chlorides fluorides, sulfates, and FeOH²⁺ species were found to be insignificant relative to Fe(OH)₂⁺ at p[H⁺] = ?log [H⁺] greater than 6.0. Hydrous ferric oxide freshly precipitated from seawater yielded a solubility product of ${}^1\text{K}{}_{\mathrm{<mtext>so</mtext>}}\text{=}\text{[Fe}{}^{\mathrm{3+}}\text{][H}{}^{\mathrm{+}}\text{]}{}^{\mathrm{?3}}\text{= 4.7 · 10}{}^5$. Solubility studies based on the rates of dialysis of various seawater solutions and on the filtration of acidified seawater solutions indicated the existence of the Fe(OH)₃⁰ species. The formation constant for this species can be calculated as ${}^1\text{β}{}_3\text{=}\text{[Fe(OH)}{}_3{}^0\text{] [H}{}^{\mathrm{+}}\text{]}{}^3\text{/[Fe}{}^{\mathrm{3+}}\text{] = 2.4 · 10}{}^{\mathrm{?14}}$. The Fe(OH)₄^? species is present at concentrations which are negligible compared to Fe(OH)₂⁺ and Fe(OH)₃⁰ in the normal pH range of seawater. However, there is at least one other significant ferric complex in seawater above p[H⁺] = 8.0 (possibly with bicarbonate, carbonate, or borate ions) in addition to the Fe(OH)₂⁺ and Fe(OH)₃⁰ species.     

The extension of the parametrization of the radio source coordinates in geodetic VLBI and its impact on the time series analysis

The radio sources within the most recent celestial reference frame (CRF) catalog ICRF2 are represented by a single, time-invariant coordinate pair. The datum sources were chosen mainly according to certain statistical properties of their position time series. Yet, such statistics are not applicable unconditionally, and also ambiguous. However, ignoring systematics in the source positions of the datum sources inevitably leads to a degradation of the quality of the frame and, therefore, also of the derived quantities such as the Earth orientation parameters. One possible approach to overcome these deficiencies is to extend the parametrization of the source positions, similarly to what is done for the station positions. We decided to use the multivariate adaptive regression splines algorithm to parametrize the source coordinates. It allows a great deal of automation, by combining recursive partitioning and spline fitting in an optimal way. The algorithm finds the ideal knot positions for the splines and, thus, the best number of polynomial pieces to fit the data autonomously. With that we can correct the ICRF2 a priori coordinates for our analysis and eliminate the systematics in the position estimates. This allows us to introduce also special handling sources into the datum definition, leading to on average 30 % more sources in the datum. We find that not only the CPO can be improved by more than 10 % due to the improved geometry, but also the station positions, especially in the early years of VLBI, can benefit greatly.     

Lithospheric thickness beneath the southern Kenya Rift: implications from basalt geochemistry

Geochemical data are reported for samples from the flanks and floor of the southern Kenya Rift Valley in the Lake Magadi area, and from two central volcanoes located within the rift valley. Rift lavas include samples of Singaraini and Ol Tepesi basalts on the eastern flank, Kirikiti basalts from the western flank, and plateau trachytes from the rift valley floor. Central volcano samples are from Ol Esayeiti and Lenderut located on the eastern flank. The rift basalts are mildly ne-normative, moderately evolved (Mg#=0.39-0.62) alkali basalts and show an overall range in differentiation. Incompatible trace element abundances are moderately elevated (Nb=17-51; Zr=93-274; La=17-55 ppm) and show strongly coherent variations and constant inter-element ratios (e.g. Zr/Nb=4.2-5.5; Nb/Ta=17.5ǂ.4; (La/Sm)_n=7.3ǃ.1); isotope ratios are restricted in range (⁸⁷Sr/⁸⁶Sr=0.70393-0.70436; ¹⁴³Nd/¹⁴⁴Nd=0.51272-0.51280; ²⁰⁶Pb/²⁰⁴Pb=19.87-19.92; ²⁰⁷Pb/²⁰⁴Pb=15.68-15.70; ²⁰⁸Pb/²⁰⁴Pb=39.56-39.71). Central volcano lavas are more alkaline in character and include basanite (Ol Esayeiti; Mg# >60) and hawaiite to benmoreite (Lenderut; Mg#=0.48-0.38). Incompatible element ratio are similar to those of the rift basalts, although the chondrite normalised REE patterns are steeper (La/Sm)_n=17.4ǃ.2). ⁸⁷Sr/⁸⁶Sr (0.70358, 0.70391), ¹⁴³Nd/¹⁴⁴Nd (0.51280, 0.51267), ²⁰⁶Pb/²⁰⁴Pb (19.96,20.17), ²⁰⁷Pb/²⁰⁴Pb (15.66,15.76) and ²⁰⁸Pb/²⁰⁴Pb (39.80,40.00) ratios of Ol Esayeiti basanites are similar to the rift basalts, whereas the Lenderut lavas have unusually low¹⁴³Nd/¹⁴⁴Nd (0.512388-0.512453) ratios for their ⁸⁷Sr/⁸⁶Sr (0.70370-0.70481) ratios, and distinctly less radiogenic and variable Pb isotope compositions (²⁰⁶Pb/²⁰⁴Pb=17.93-19.01; ²⁰⁷Pb/²⁰⁴Pb=15.43-15.58; ²⁰⁸Pb/²⁰⁴Pb=37.91-39.14). An integrated model is developed in which the geochemical signature of the lavas is attributed to variable degrees of melting to depths within the garnet stability field, and in the presence of residual amphibole. The stability fields of these phases in P-T space indicates that the lavas must have formed within the sub-continental lithosphere rather than within the underlying ambient asthenosphere or a rising mantle plume. The subcontinental lithospheric mantle must therefore extend to a depth of at least 75 km beneath the Lake Magadi area, which contrasts with recent gravity models for the area, which infer that lithospheric mantle is absent beneath this section of the southern Kenya Rift.