Industrial decentralisation in South Africa: tool of Apartheid or spontaneous restructuring?

Industrial decentralisation (ID) has been a central component of South Africa's regional strategy for over twenty years. Until recently, though, the programme has evinced little interest amongst industrialists. The fact that ID policy has been premised upon political rather than economic considerations has been thought to lie at the root of this failure. Following recent developments, however, this view is being reassessed. The reorganisation of ID policy in 1982 and the dramatic acceleration in the rate of industrial relocations that this reorganisation appears to have induced, has led theorists to revise their thinking on both the economic and political imperatives underpinning the programme. The purpose of this paper is to review the debates that are now developing concerning ID in South Africa, and to suggest where further research and analysis might be most profitably directed.     

Seismic interaction at separation joints of an instrumented concrete bridge

A multi-span, curved, concrete box-girder bridge has been extensively instrumented by the California Strong Motion Instrumentation Program (CSMIP) in cooperation with the California Department of Transportation (Caltrans). On 28 June 1992, the bridge was shaken by the magnitude 7–5 Landers and magnitude 6–6 Big Bear earthquakes in southern California. The epicentres of these earthquakes were 50 and 29 miles (81 and 46 km) from the bridge, respectively. All 34 strong-motion sensors installed on the bridge recorded the response to these earthquakes and provided an insightful set of data. A striking aspect of the response is the presence of intermittent sharp spikes in nearly all of the acceleration records from sensors at the deck of the bridge. Among these the highest spike was 0.8g for the Landers and 1.0g for the Big Bear earthquake. The peak ground acceleration at the bridge site was only about 0.1g for both these earthquakes. With the aid of visual examination and simple analysis it is deduced that (1) the spikes were caused by forces generated at separation joints by impacts and stretching of the cable restrainers between adjacent bridge segments; (2) the forces of impacts and cable stretching are directly proportional to the size of the spikes and can be estimated by the use of a simple formula; and (3) the spikes travelled from their source to other locations on the bridge with the velocity of a compression wave propagating through concrete.     

Regional Energy Demand Responses To Climate Change: Methodology And Application To The Commonwealth Of Massachusetts

Climate is a major determinant of energy demand. Changes in climate may alter energy demand as well as energy demand patterns. This study investigates the implications of climate change for energy demand under the hypothesis that impacts are scale dependent due to region-specific climatic variables, infrastructure, socioeconomic, and energy use profiles. In this analysis we explore regional energy demand responses to climate change by assessing temperature-sensitive energy demand in the Commonwealth of Massachusetts. The study employs a two-step estimation and modeling procedure. The first step evaluates the historic temperature sensitivity of residential and commercial demand for electricity and heating fuels, using a degree-day methodology. We find that when controlling for socioeconomic factors, degree-day variables have significant explanatory power in describing historic changes in residential and commercial energy demands. In the second step, we assess potential future energy demand responses to scenarios of climate change. Model results are based on alternative climate scenarios that were specifically derived for the region on the basis of local climatological data, coupled with regional information from available global climate models. We find notable changes with respect to overall energy consumption by, and energy mix of the residential and commercial sectors in the region. On the basis of our findings, we identify several methodological issues relevant to the development of climate change impact assessments of energy demand.     

Towards evaluating cloud response to climate change using clustering technique identification of cloud regimes

Most of the discrepancies in the climate sensitivity of general circulation models (GCMs) are believed to be due to differences in cloud radiative feedback. Analysis of cloud response to climate change in different ‘regimes’ may offer a more detailed understanding of how the cloud response differs between GCMs. In which case, evaluation of simulated cloud regimes against observations in terms of both their cloud properties and frequency of occurrence will assist in assessing confidence in the cloud response to climate change in a particular GCM. In this study, we use a clustering technique on International Satellite Cloud Climatology Project (ISCCP) data and on ISCCP-like diagnostics from two versions of the Hadley Centre GCM to identify cloud regimes over four different geographical regions. The two versions of the model are evaluated against observational data and their cloud response to climate change compared within the cloud regime framework. It is found that cloud clusters produced by the more recent GCM, HadSM4, compare more favourably with observations than HadSM3. In response to climate change, although the net cloud response over particular regions is often different in the two models, in several instances the same basic processes may be seen to be operating. Overall, both changes in the frequency of occurrence of cloud regimes and changes in the properties (optical depth and cloud top height) of the cloud regimes contribute to the cloud response to climate change.     

Self-organised clustering for road extraction in classified imagery

The extraction of road networks from digital imagery is a fundamental image analysis operation. Common problems encountered in automated road extraction include high sensitivity to typical scene clutter in high-resolution imagery, and inefficiency to meaningfully exploit multispectral imagery (MSI). With a ground sample distance (GSD) of less than 2 m per pixel, roads can be broadly described as elongated regions. We propose an approach of elongated region-based analysis for 2D road extraction from high-resolution imagery, which is suitable for MSI, and is insensitive to conventional edge definition. A self-organising road map (SORM) algorithm is presented, inspired from a specialised variation of Kohonen's self-organising map (SOM) neural network algorithm. A spectrally classified high-resolution image is assumed to be the input for our analysis. Our approach proceeds by performing spatial cluster analysis as a mid-level processing technique. This allows us to improve tolerance to road clutter in high-resolution images, and to minimise the effect on road extraction of common classification errors. This approach is designed in consideration of the emerging trend towards high-resolution multispectral sensors. Preliminary results demonstrate robust road extraction ability due to the non-local approach, when presented with noisy input.     

Location and height of intertidal bars on macrotidal ridge and runnel beaches

Previous studies devoted to the morphology and hydrodynamics of ridge and runnel beaches highlight characteristics that deviate from those initially postulated by King and Williams (Geographical Journal, 1949, vol. 113, 70–85) and King (Beaches and Coasts, 1972, Edward Arnold). Disagreements on the morphodynamics of these macrotidal beaches include the position of the ridges relative to the mean neap and spring tide levels, the variation in the height of the ridges across the intertidal profile and, most importantly, whether the ridges are formed by swash or surf zone processes. The morphological characteristics of ridge and runnel beaches from three locations with varying wave, tidal and geomorphic settings were investigated to address these disagreements. Beach profiles from each site were analysed together with water‐level data collected from neighbouring ports. It was found that the ridges occur over the entire intertidal zone. On one site (north Lincolnshire, east England), the ridges are uniformly distributed over the intertidal beach, whereas on the two other sites (Blackpool beach, northwest England, and Leffrinckoucke beach, north France) there is some indication that the ridges appear to occur at preferential locations. Most significantly, the locations of the ridge crests were found to be unrelated to the positions on the intertidal profile where the water level is stationary for the longest time. It was further found that the highest ridges generally occur just above mid‐tide level where tidal non‐stationarity is greatest. These findings argue against the hypothesis that the ridges are formed by swash processes acting at stationary tide levels. It is tentatively suggested that the ridges are the result of a combination of swash and surf zone processes acting across the intertidal zone. Elucidation of the morphodynamic roles of these two types of processes, and other processes such as strong current flows in the runnels, requires further comprehensive field measurements complemented by numerical modelling. Copyright © 2001 John Wiley & Sons, Ltd.     

Implications of model uncertainty for the mapping of hillslope‐scale soil erosion predictions

An integrated modelling approach (MIRSED) which utilizes the process‐based soil erosion model WEPP (Water Erosion Prediction Project) is presented for the assessment of hillslope‐scale soil erosion at five sites throughout England and Wales. The methodology draws upon previous uncertainty analysis of the WEPP hillslope soil erosion model by the authors to qualify model results within an uncertainty framework. A method for incorporating model uncertainty from a range of sources is discussed as a first step towards using and learning from results produced through the GLUE (Generalized Likelihood Uncertainty Estimation) technique. Results are presented and compared to available observed data, which illustrate that levels of uncertainty are significant and must be taken into account if a meaningful understanding of output from models such as WEPP is to be achieved. Furthermore, the collection of quality, observed data is underlined for two reasons: as an essential tool in the development of soil erosion modelling and also to allow further constraint of model uncertainty. Copyright © 2001 John Wiley & Sons, Ltd.     

Lithological and spatial controls on the distribution of quartz veins in andesite- and rhyolite-hosted epithermal Au–Ag deposits of the Hauraki Goldfield, New Zealand

Vein distributions in line samples from four epithermal Au–Ag deposits of the Hauraki Goldfield were logged and quantified by vein spacing, vein density, vein thickness and percentage of vein extension. One deposit is hosted in andesite lavas (Martha Hill), one in andesite lavas and dacite porphyry, dacitic tuffs and pyroclastic breccias (Golden Cross), and two in rhyolite lavas and rhyolitic tuffs with minor andesite lavas or andesite dikes (Ohui and Wharekirauponga). The vein systems in these deposits form fault-controlled arrays of extensional veins. Vein spacing distributions are non-fractal over two to three orders of magnitude (1 mm to 5 m), and therefore fractal dimension statistics are not applicable. The coefficient of variation (C_v) of vein spacing was used as a measure of the degree of vein clustering. Rock type has a marked influence on vein spacing distributions, with veining in rhyolite lava having lower average thickness and percentage extension, but a generally higher degree of vein clustering compared with veining in andesite lava in the same deposit. Vein spacing distributions in well-jointed lithologies, mainly andesite lava, have C_v values (0.8–1.2) that are indicative of anticlustered to weakly clustered patterns, particularly in the vein stockwork of the upper part of the Golden Cross deposit. These C_v values are consistent with field observations that joints are a major control on vein spacing. In the poorly jointed dacitic and rhyolitic rocks, the veins are weakly to strongly clustered as shown by higher C_v values (1.2–2.4), and are commonly associated with normal faults. Overall, andesite lava and dacite porphyry and pyroclastics host thicker and more persistent veins than rhyolite lava and tuff. These larger veins contain significant volumes of high-grade gold mineralisation. The higher chemical reactivity to hydrothermal fluids of andesite and dacite compared with rhyolite may have aided propagation and thickening of the veins in andesite-hosted deposits. Within an individual epithermal deposit, location close to thick veins, representing major fluid conduits, commonly overrides the effect of different lithologies. Sites that are deeper and located within or adjacent to major vein structures have higher average vein thickness, percentage extension and degree of vein clustering. Systematic collection and analysis of vein spacing, thickness and density data can be used to define trends that are useful in the exploration of gold-bearing epithermal vein deposits. Received: 25 August 1998 / Accepted: 23 December 1999