Call number: LD2668 .R4 1976 S93

Dichloro-diphenyl-trichlorethane (DDT) is a chemical used in malaria control through indoor residual spraying (IRS) and has saved numerous lives in the past six decades. DDT use is restricted/banned under the Stockholm Convention on Persistent Organic Pollutants. Passive air sampling using polyurethane foam was conducted in South Africa to evaluate the presence and trends of DDT and its metabolites. Three sampling sites were used, namely, Barberspan Nature Reserve (rural agricultural), Vanderbijlpark (urban industrial) and Molopo Nature Reserve (isolated nature reserve). Sampling was conducted for a period of one year in 2008. Back trajectories from the three sampling sites were generated using HYSPILT to determine the sources of DDT metabolites to the sampling areas. Forward trajectories were also generated to determine the movement, distribution, and fate of DDT from the areas under Indoor residual spray of DDT for malaria control in South Africa and Swaziland. Chemical analysis was conducted by the RECETOX (Mazaryk University) in the Czech Republic. DDT metabolites (o,p’-DDE, p’p’-DDE, o.p’-DDD, p,p’-DDD, o,p’-DDT p,p’-DDT) were analysed using a GC-ECD (HP 5890). Vanderbijlpark had the highest concentrations of DDT metabolites throughout the year. Barberspan had the second highest concentration and Molopo the least. Seasonal changes in concentration were much the same at the three sites. %p,p’-DDT of ΣDDT is consistent with IRS spraying months in South Africa and Swaziland. A combinations of backward and forward trajectories, together with the temporal pattern of change of the %p,p’-DDT of ΣDDT support the deduction that DDT sampled from the three study sites (to some degree) came from IRS areas in South Africa and Swaziland. The presence of DDT in Molopo Nature Reserve and Barberspan is evidence of long-range transportation over dry semi-desert areas. Back-trajectories indicate the possible source of DDT were the IRS areas in the provinces of Limpopo, Mpumalanga, and KwaZulu-Natal. Some air masses to the sampling sites came from the sprayed areas. The forward trajectories also revealed that the DDT sprayed during IRS could undergo LRT. The DDT metabolites were able to travel to neighbouring countries such as Mozambique, Namibia, Zimbabwe and Botswana. / MSc (Environmental Sciences), North-West University, Potchefstroom Campus, 2013

Dichloro-diphenyl-trichlorethane (DDT) is a chemical used in malaria control through indoor residual spraying (IRS) and has saved numerous lives in the past six decades. DDT use is restricted/banned under the Stockholm Convention on Persistent Organic Pollutants. Passive air sampling using polyurethane foam was conducted in South Africa to evaluate the presence and trends of DDT and its metabolites. Three sampling sites were used, namely, Barberspan Nature Reserve (rural agricultural), Vanderbijlpark (urban industrial) and Molopo Nature Reserve (isolated nature reserve). Sampling was conducted for a period of one year in 2008. Back trajectories from the three sampling sites were generated using HYSPILT to determine the sources of DDT metabolites to the sampling areas. Forward trajectories were also generated to determine the movement, distribution, and fate of DDT from the areas under Indoor residual spray of DDT for malaria control in South Africa and Swaziland. Chemical analysis was conducted by the RECETOX (Mazaryk University) in the Czech Republic. DDT metabolites (o,p’-DDE, p’p’-DDE, o.p’-DDD, p,p’-DDD, o,p’-DDT p,p’-DDT) were analysed using a GC-ECD (HP 5890). Vanderbijlpark had the highest concentrations of DDT metabolites throughout the year. Barberspan had the second highest concentration and Molopo the least. Seasonal changes in concentration were much the same at the three sites. %p,p’-DDT of ΣDDT is consistent with IRS spraying months in South Africa and Swaziland. A combinations of backward and forward trajectories, together with the temporal pattern of change of the %p,p’-DDT of ΣDDT support the deduction that DDT sampled from the three study sites (to some degree) came from IRS areas in South Africa and Swaziland. The presence of DDT in Molopo Nature Reserve and Barberspan is evidence of long-range transportation over dry semi-desert areas. Back-trajectories indicate the possible source of DDT were the IRS areas in the provinces of Limpopo, Mpumalanga, and KwaZulu-Natal. Some air masses to the sampling sites came from the sprayed areas. The forward trajectories also revealed that the DDT sprayed during IRS could undergo LRT. The DDT metabolites were able to travel to neighbouring countries such as Mozambique, Namibia, Zimbabwe and Botswana. / MSc (Environmental Sciences), North-West University, Potchefstroom Campus, 2013

<p>Television (TV) is the predominant advertising medium, and recent technological advances such as digital video recorders (DVRs) and set-top boxes (STBs) have the potential to transform this industry by enabling household-specific advertising. Since exposure to TV represents a substantial share of consumer time and attention, this potential to micro-target communications represents an enormous opportunity for the TV advertising market. </p><p>This paper outlines an approach to facilitate the micro-targeting of TV advertising. We employ a unique dataset, integrating TV program and advertisement viewing at the household level with purchase data, to address the question of how advertisers can achieve better advertising targeting in the digital context. Based on this dataset, we first develop a model of household TV viewing behavior. The viewing model comprises three integrated components: TV show sampling and watching, TV show recording, and advertising viewing. All three components are motivated by the theoretical concept of flow utility, that is, the moment-by-moment enjoyment a household derives from different activities: watching a TV show, watching a TV advertisement, and other non-TV activities. This model has decent out-of-sample prediction power on show choices and time spent on each selected show. We then link household advertising exposure with purchase. Finally, the viewing model and identified advertising-sales relationship are utilized to conduct counterfactual policy experiments on advertising targeting. We consider several household-level targeting scenarios by manipulating: 1) whether the advertising purchase is made in advance; and 2) whether the objective function is to minimize costs for a given set of exposures or to maximize revenues from advertising. Results indicate micro-targeting can lower advertising costs and raise incremental revenue.</p><p>The key contributions of this paper are as follows. Theoretically, we develop an integrated model on TV show viewing, TV advertising viewing, purchasing and advertising targeting. Methodologically, we propose a new modeling framework on media consumption by explicitly accounting for the role of uncertainty, and propose targeting strategies leveraging household-level data. Substantively, we offer policy recommendations to advertisers on micro-targeting which can be of great potential.</p> / Dissertation

In radio astronomy, a key limiting factor to observations made is the available bandwidth of the system. This thesis looks at two different approaches to building ultra-broadband correlators for use in radio astronomy. The first was a 2-20GHz double-sideband complex analogue correlator that was constructed before the work of this thesis. Characterisation tests are performed and a basic calibration is attempted. Both these sets of experiments show good results, with the basic calibration successfully being able to compensate for gain difference between the lags over a reduced bandwidth range used in the testing. The second approach was the investigation into different techniques for high-speed digital sampling, capable of providing equivalent bandwidths to the analogue system. The use of FPGA high-speed serial interfaces as direct 1-bit 3.125 GS/s samplers was investigated. Single-frequency sampling showed that a signal-to-noise ratio close to the theoretical maximum across the band was achieved (≈ 0.8 effective bits). Techniques were also identified to use multiple transceivers to generate a single interleaved stream at higher effective sampling rate. Two different methods were also explored for producing greater-than 1-bit sampling. A hysteresis approach was shown not to produce the desired results and a reference based sampler in the end was adopted. Finally, the interleaving and multi-bit techniques were combined to generate a single 1.5-bit 6.25 GS/s sampler. This was seen to have reduced signal-to-noise compared to the expected values. This was believed to be caused by the poor method of RF signal injection causing cross-talk between the channels and large amounts of loss. As a comparison to the direct sampling method, an external 1-bit high-speed Hittite comparator was also examined. The single-frequency experiment was repeated with a slightly higher signal-to-noise ratio found compared to the direct sampling method. This was again believed to be due to the RF environments used. From the sampling setups a four-input, six-baseline lag correlator was constructed using the direct sampling method. The entire correlator, as well as the sampling transceivers, was incorporated into a single Xilinx Virtex 5 FPGA. This was shown to have the expected response to single-frequency, broadband and noise signals. The thesis concludes with a characterisation of the RF devices used throughout the testing procedures. Several new devices were developed through the course of the experiments with the designs being documented. All the necessary components to construct IF chains for both the analogue and digital correlators described are present. This leads to simulations being made of complete IF chains, with the expected responses shown.

A number of mathematical models are used nowadays to describe behaviour of the reallife water distribution system (WDS). It is a well known fact that, to have any meaningful use, any WDS mathematical model must be calibrated first. Here, calibration is defined as process in which a number of WDS model parameters are adjusted until the model mimics behaviour of the real WDS as closely as possible. In this thesis, WDS mathematical models that are used to model water quantity aspect only are analysed. Three hydraulic models considered here are: (1) steady-state flow model, (2) quasi-steady flow (extended period simulation) model and (3) unsteady flow model. The calibration problem analysed here is formulated as a constrained optimisation problem of weighted least square type with the objective defined in a way that enables effective incorporation of prior information on calibration parameters. WDS calibration problem is then analysed in detail, including special issues of identifiability, uniqueness and stability of the problem solution. A list of diagnostic and other statistics and analysis is presented to improve existing calibration approaches by providing partial insight into the calibration process. Calibration of WDS hydraulic models is further improved by the development of new hybrid optimisation method. Being closely related to calibration, the problem of sampling design for calibration of WDS hydraulic models is also addressed here. First, sampling design is formulated as a constrained two-objective optimisation problem. Then, two novel models are developed to solve it. The first model is based on standard, single-objective Genetic Algorithms (SOGA). The second model is based on multi-objective Genetic Algorithms (MOGA). Finally, all novel methodologies presented here are verified successfully on multiple case studies that involve both artificial and real-life WDS. At the end, relevant conclusions are drawn and suggestions for further research work are made.

published_or_final_version / Statistics / Doctoral / Doctor of Philosophy

published_or_final_version / Statistics and Actuarial Science / Doctoral / Doctor of Philosophy

published_or_final_version / abstract / Computer Science / Master / Master of Philosophy

Acceptance sampling plans by variables are a basic quality control technique. These plans provide economical procedures to determine the acceptability of batches of product. Most of these plans are based on a single quality characteristic and are of the classical type. This work concentrates on Bayesian variable acceptance sampling plans for fraction defective. Both destructive and non-destructive sampling procedures are considered. A set of decision points are estimated and employed to make decisions about the inspected lots. Techniques to dispose of the rejected lots are provided. Components of the expected total cost relative to various decisions are estimated. The sample size required to obtain the expected optimum cost is found. An untrue assumption implicit in the measurement of the quality characteristic of items sampled is that the observed dimensions are error free. The distributions, means, and variances of a set of parameters for error free and error prone sampling is listed. Computer programs written in FORTRAN 77 are developed to compute the decision points and the costs for both destructive and nondestructive testing. Precise Bays estimate of the costs and other economic parameters involve the moments of the fraction defective p raised to the kᵗʰ power. Mathematical expressions for the conditional expectations of p|x and p|ẋ are derived and a computer program to estimate these moments is provided. Producing quality items with minimum cost requires keeping a production process under control. The quality characteristic X of each item produced is determined and the sample means are plotted on an Ẋ-control chart. A production process is assumed to start in control at time t = 0 with specific values of the mean and standard deviation. The occurrence of a single or multiple cause-failures shift the process mean outside the control limits. During the search for the causes of failure, the process is either allowed to continue in operation or shut down until the assignable cause or causes are discovered. The expected duration of time during which the process is shut down and the additional time to be taken to repair the process are considered. Computer programs are provided to estimate the optimal sample size, the interval between successive samples, the control limits, the probability of type I error, the power of the chart, and the average time the process operates in the presence of an assignable cause. The parameters estimated are employed to estimate the optimal loss-cost. The economic design of Ẋ -charts assumes one quality characteristic of interest. However a product quality in most industrial products and processes is characterized by more than one quality characteristic where the application of a Ẋ -control chart for each variable is inappropriate. In this work a Hotellings T² control chart is employed to handle cases of where products are tested relative to several quality characteristics.