A ROBUST PLS PROCEDURE

A robust partial least squares(PLS)regression algorithm is developed.This is achieved by substitutionof the univariate regression steps in the iterative PLS2 algorithm by a robust alternative.The anglebetween loading vectors from both perturbed and unperturbed solutions is used as a measure ofrobustness.By means of a perturbation study on a structure-activity data set,it is demonstrated thatthe stability of the robust method is superior to standard PLS.     

Studying the effect of partial drainage on the response of soft clays reinforced with sand column groups

Acta Geotechnica - The drained and undrained response of soft clays reinforced with granular columns has been the subject of numerous geotechnical research efforts to date. Although these studies...     

Environmental influence on phytoplankton production during summer on the KwaZulu-Natal shelf of the Agulhas ecosystem

During February 2010, studies of primary production (PP) and physiology were conducted at five selected sites in the KwaZulu-Natal (KZN) Bight of the Agulhas ecosystem as part of a programme to elucidate the influence of major physical driving forces and nutrient inputs on the structure and functioning of biological communities. These sites were located in the vicinity of the Durban lee eddy, in the midshelf region of the central part of the bight, off the Thukela Mouth, and to the north and south of Richards Bay. At four of the sites, chlorophyll a ranged from 0.10 to 1.44?mg m–3 and integrated PP ranged between 0.35 and 2.58?g C m–² d–1. The highest biomass and PP, which were comparable to those observed in a wind-driven upwelling system, were associated with a diatom community observed at the midshelf site, and varied between 0.26 and 4.27?mg m–³ and 7.22 and 9.89?g C m–² d–1, respectively. Environmental conditions at each of the sites differed substantially and appeared to be influential in initiating and controlling the development and distribution of phytoplankton biomass and production. Phytoplankton adaptation to variable environmental conditions was characterised by a decreased light-limited slope (αB) and increased rate of photosynthesis (P_m ) and light saturation (Ek) with elevated temperatures. The converse (increased α^B and decreased P_m and Ek) was observed as irradiance levels declined. Generalised additive models indicated that irradiance, temperature and biomass were important variables influencing photosynthetic parameters and photosynthetic rates.     

Charts for probabilistic design of strip footings in cohesionless soils

Design charts that enable quick determination of the probability distribution parameters related to the ultimate bearing capacity of shallow foundations resting on (c = 0) soils are developed. These charts are intended to assist foundation designers and analysts in studying the reliability of structures as related to the capacity of the foundation system. The approach presented herein provides a more reliable alternative to foundation design and analysis than the current conventional design procedure which employs the assumption of an appropriate factor-of-safety. © Rapid Science Ltd. 1998     

Factors determining the vertical profile of dimethylsulfide in the Sargasso Sea during summer

The major source of reduced sulfur in the remote marine atmosphere is the biogenic compound dimethylsulfide (DMS), which is ubiquitous in the world's oceans and released through food web interactions. Relevant fluxes and concentrations of DMS, its phytoplankton-produced precursor, dimethylsulfoniopropionate (DMSP) and related parameters were measured during an intensive Lagrangian field study in two mesoscale eddies in the Sargasso Sea during July–August 2004, a period characterized by high mixed-layer DMS and low chlorophyll—the so-called ‘DMS summer paradox’. We used a 1-D vertically variable DMS production model forced with output from a 1-D vertical mixing model to evaluate the extent to which the simulated vertical structure in DMS and DMSP was consistent with changes expected from field-determined rate measurements of individual processes, such as photolysis, microbial DMS and dissolved DMSP turnover, and air–sea gas exchange. Model numerical experiments and related parametric sensitivity analyses suggested that the vertical structure of the DMS profile in the upper 60 m was determined mainly by the interplay of the two depth-variable processes—vertical mixing and photolysis—and less by biological consumption of DMS. A key finding from the model calibration was the need to increase the DMS(P) algal exudation rate constant, which includes the effects of cell rupture due to grazing and cell lysis, to significantly higher values than previously used in other regions. This was consistent with the small algal cell size and therefore high surface area-to-volume ratio of the dominant DMSP-producing group—the picoeukaryotes.     

Upper-air wind profiles investigation for tropospheric circulation study

Summary ¶Methods used to study the atmospheric circulation are essentially based on pressure classifications or geopotential fields. However, instead of using indirect means, a more direct measure of the atmospheric flow, i.e. the wind, appears preferable. In addition, it is admitted that, when explaining the weather variability, the combination of multiple level data can bring more information than studying just single-level characteristics. Thus, vertical profiles of wind are of great interest. The purpose of the article is twofold: first, to propose a method to classify this kind of data; second, to illustrate briefly the results obtained with the daily wind profiles acquired by upper-air soundings at a French meteorological station (Nancy), over a period of 11 years. The advantages of the method are: it is automatic and can easily be transposed in other areas; it provides a new type of information about the atmospheric circulation in the vertical dimension; it allows making easily the link between the circulation in the lower and upper troposphere.Received May 8, 2002; revised January 22, 2003; accepted March 15, 2003 Published online July 30, 2003     

Potential climate-change impacts on the Chesapeake Bay

We review current understanding of the potential impact of climate change on the Chesapeake Bay. Scenarios for CO₂ emissions indicate that by the end of the 21^st century the Bay region will experience significant changes in climate forcings with respect to historical conditions, including increases in CO₂ concentrations, sea level, and water temperature of 50–160%, 0.7–1.6 m, and 2–6 °C, respectively. Also likely are increases in precipitation amount (very likely in the winter and spring), precipitation intensity, intensity of tropical and extratropical cyclones (though their frequency may decrease), and sea-level variability. The greatest uncertainty is associated with changes in annual streamflow, though it is likely that winter and spring flows will increase. Climate change alone will cause the Bay to function very differently in the future. Likely changes include: (1) an increase in coastal flooding and submergence of estuarine wetlands; (2) an increase in salinity variability on many time scales; (3) an increase in harmful algae; (4) an increase in hypoxia; (5) a reduction of eelgrass, the dominant submerged aquatic vegetation in the Bay; and (6) altered interactions among trophic levels, with subtropical fish and shellfish species ultimately being favored in the Bay. The magnitude of these changes is sensitive to the CO₂ emission trajectory, so that actions taken now to reduce CO₂ emissions will reduce climate impacts on the Bay. Research needs include improved precipitation and streamflow projections for the Bay watershed and whole-system monitoring, modeling, and process studies that can capture the likely non-linear responses of the Chesapeake Bay system to climate variability, climate change, and their interaction with other anthropogenic stressors.