Use of Noise to Augment Training Data: A Neural Network Method of Mineral–Potential Mapping in Regions of Limited Known Deposit Examples

One of the main factors that affects the performance of MLP neural networks trained using the backpropagation algorithm in mineral-potential mapping isthe paucity of deposit relative to barren training patterns. To overcome this problem, random noise is added to the original training patterns in order to create additional synthetic deposit training data. Experiments on the effect of the number of deposits available for training in the Kalgoorlie Terrane orogenic gold province show that both the classification performance of a trained network and the quality of the resultant prospectivity map increasesignificantly with increased numbers of deposit patterns. Experiments are conducted to determine the optimum amount of noise using both uniform and normally distributed random noise. Through the addition of noise to the original deposit training data, the number of deposit training patterns is increased from approximately 50 to 1000. The percentage of correct classifications significantly improves for the independent test set as well as for deposit patterns in the test set. For example, using ±40% uniform random noise, the test-set classification performance increases from 67.9% and 68.0% to 72.8% and 77.1% (for test-set overall and test-set deposit patterns, respectively). Indices for the quality of the resultant prospectivity map, (i.e. D/A, D × (D/A), where D is the percentage of deposits and A is the percentage of the total area for the highest prospectivity map-class, and area under an ROC curve) also increase from 8.2, 105, 0.79 to 17.9, 226, 0.87, respectively. Increasing the size of the training-stop data set results in a further increase in classification performance to 73.5%, 77.4%, 14.7, 296, 0.87 for test-set overall and test-set deposit patterns, D/A, D × (D/A), and area under the ROC curve, respectively.     

Use of Fuzzy Membership Input Layers to Combine Subjective Geological Knowledge and Empirical Data in a Neural Network Method for Mineral-Potential Mapping

Use of GIS layers, in which the cell values represent fuzzy membership variables, is an effective method of combining subjective geological knowledge with empirical data in a neural network approach to mineral-prospectivity mapping. In this study, multilayer perceptron (MLP), neural networks are used to combine up to 17 regional exploration variables to predict the potential for orogenic gold deposits in the form of prospectivity maps in the Archean Kalgoorlie Terrane of Western Australia. Two types of fuzzy membership layers are used. In the first type of layer, the statistical relationships between known gold deposits and variables in the GIS thematic layer are used to determine fuzzy membership values. For example, GIS layers depicting solid geology and rock-type combinations of categorical data at the nearest lithological boundary for each cell are converted to fuzzy membership layers representing favorable lithologies and favorable lithological boundaries, respectively. This type of fuzzy-membership input is a useful alternative to the 1-of-N coding used for categorical inputs, particularly if there are a large number of classes. Rheological contrast at lithological boundaries is modeled using a second type of fuzzy membership layer, in which the assignment of fuzzy membership value, although based on geological field data, is subjective. The methods used here could be applied to a large range of subjective data (e.g., favorability of tectonic environment, host stratigraphy, or reactivation along major faults) currently used in regional exploration programs, but which normally would not be included as inputs in an empirical neural network approach.     

'Trees, forests and communities': some historiographical approaches to environmental history on Africa

Focusing on the contested issue of trees and forests, this paper looks at various ways in which historians have explored environmental change and human agency in Africa. It looks in particular at the colonial period and positions the case of the Cape Colony (South Africa) in a broader historiographical context. Colonial silviculture involved both the appropriation of the natural forests and the creation of exotic plantations. These policies generated a mixture of reactions from African communities and had varying effects on the environment. Colonial science and the projects it gave rise to have since been critiqued, contributing to the promotion of community-based forestry schemes that try to incorporate indigenous knowledge and ideas about land use.     

Characterisation of riparian woodlands and their potential water loss in the distal Okavango Delta, Botswana

An analysis is undertaken to develop techniques to remotely sense relative evapotranspiration outputs in the distal Okavango Delta using enhanced thematic mapper and ground based techniques to help quantify water loss. Much of this work focuses on riparian woodlands which, especially in the distal portion of the Delta, are regarded as being significant in terms of groundwater removal by transpiration. This was confirmed as vegetation cover mapping led to the identification of two riparian classes which, by association with high resistivity aeromagnetic data, were found to be rooted in near surface fresh groundwater. This paper indicates that riparian trees which remain green year long, partly sustain their growth as a result of groundwater uptake. A comparison of frequently flooded and dry floodplains with distinct riparian zones was undertaken using spectral techniques (pixel radiance values and leaf water content indices) to determine whether flooding and lateral groundwater flow stimulated growth (and therefore transpiration rate) following dormancy. Results indicate a basic similarity between the two systems with mixed evidence of assumed leaf growth. Related phenological observations in the riparian zone of the dry floodplain show that renewal of leaf growth is primarily related to rainfall, not flood events in the distal Delta. The results of this work should help effect both surface and groundwater management in the vicinity of population centres in the distal Delta.     

Anthropogenic Impacts on Summer Precipitation in Central Arizona,U.S.A.

This article explores the possibility of urbanization‐ and irrigation‐induced increases in summer precipitation totals in central Arizona. Maximum precipitation impacts are hypothesized to occur downwind of the Phoenix area in the Lower Verde basin. Results from statistical tests indicate that summer precipitation totals in the Lower Verde basin are greater than totals in nearby basins. Precipitation totals in the basin also appear to be equivalent to totals at more monsoon‐impacted stations in eastern Arizona. While this research is preliminary, the results do provide encouraging evidence of the existence of anthropogenically enhanced summer precipitation in central Arizona.     

Controls on Fe reduction and mineral formation by a subsurface bacterium

The reductive dissolution of FeIII (hydr)oxides by dissimilatory iron-reducing bacteria (DIRB) could have a large impact on sediment genesis and Fe transport. If DIRB are able to reduce FeIII in minerals of high structural order to carry out anaerobic respiration, their range could encompass virtually every O₂-free environment containing FeIII and adequate conditions for cell growth. Previous studies have established that Shewanella putrefaciens CN32, a known DIRB, will reduce crystalline Fe oxides when initially grown at high densities in a nutrient-rich broth, conditions that poorly model the environments where CN32 is found. By contrast, we grew CN32 by batch culture solely in a minimal growth medium. The stringent conditions imposed by the growth method better represent the conditions that cells are likely to encounter in their natural habitat. Furthermore, the expression of reductases necessary to carry out dissimilatory Fe reduction depends on the method of growth. It was found that under anaerobic conditions CN32 reduced hydrous ferric oxide (HFO), a poorly crystalline FeIII mineral, and did not reduce suspensions containing 4 mM FeIII in the form of poorly ordered nanometer-sized goethite (α-FeOOH), well-ordered micron-sized goethite, or nanometer-sized hematite (α-Fe₂O₃) crystallites. Transmission electron microscopy (TEM) showed that all minerals but the micron-sized goethite attached extensively to the bacteria and appeared to penetrate the outer cellular membrane. In the treatment with HFO, new FeII and FeIII minerals formed during reduction of HFO-Fe in culture medium containing 4.0 mmol/L P_i (soluble inorganic P), as observed by TEM with energy-dispersive X-ray spectroscopy, selected area electron diffraction, and X-ray diffraction. The minerals included magnetite (Fe₃O₄), goethite, green rust, and vivianite [Fe₃(PO₄)₂ · 8H₂O]. Vivianite appeared to be the stable end product and the mean coherence length was influenced by the rate of FeIII reduction. When P_i was 0.4 mol/L under otherwise identical conditions, goethite was the only mineral observed to form, and less Fe²⁺ was produced overall. Hence, the ability of DIRB to reduce Fe (hydr)oxides may be limited when the bacteria are grown under nutrient-limited conditions, and the minerals that result depend on the vigor of FeIII reduction.     

GI Science, Disasters, and Emergency Management

Societal responses to disasters begin with the initial post‐event rescue and relief operations, followed by recovery, reconstruction, and then transcend into mitigation actions including the development of pre‐impact preparedness measures, collectively known as the emergency response cycle. This paper highlights some of the applications of GI Science to the emergency response cycle, citing examples from natural hazards and from the World Trade Center disaster on 11th September 2001. More importantly, the paper describes some of the constraints on the utilization of GI Science by the practitioner community: understandable user interfaces; data quantity, quality, and integration; real‐time data and information. Finally, the paper suggests some important GI Science research areas based on the needs of the disasters and emergency management research and practitioner communities.     

Replacing carbon lost from forests: an assessment of insurance, reserves, and expiring credits

Concern over the “non-permanence” or reversibility of carbon sequestration projects has been prominent in discussions over how to develop guidelines for forest project investments under the Clean Development Mechanism (CDM) of the UNFCCC Kyoto Protocol. Accordingly, a number of approaches have been proposed that aim to help ensure that parties do not receive credit for carbon that is lost before project obligations are fulfilled. These approaches include forest carbon insurance, land reserves, and issuance of expiring credits. The potential costs of each of these different approaches are evaluated using a range of assumptions about project length, risk and discount rate, and a comparison of costs is ventured based on the estimated reduction in value of these credits compared with uninsured, and permanent credits. Obstacles to participation in the different approaches are discussed related to problems of long-term commitments, project scale, rising replacement costs, and low credit value. It is concluded that a system of expiring credits, which could be coupled with insurance or reserves, could guarantee obligations that span time-scales longer than that of conventional insurance policies while maintaining incentives for long-term sequestration.     

The Cretaceous and younger thermal history of the KTB site (Germany): apatite fission-track data from the Vorbohrung

A detailed fission-track analysis was carried out on 41 apatite samples from the 4001-m-deep pilot hole of the German Continental Deep Drilling Program (KTB). The investigations include dating by the population method and length measurements of horizontal confined tracks. Age and length information indicate that all samples above the present-day partial annealing zone cooled from temperatures above 120?°C in the Cretaceous. The top two kilometres of the profile show less variation of age with depth than was expected. Modelling of the profile indicates that the distribution of fission track ages with depth requires that the profile was disturbed during or after cooling through the partial annealing zone. Modelling of individual samples reveals that at least the upper 2?km experienced reverse faulting, effectively thickening the upper part of the rock column by up to 1000?m in the Tertiary. A present-day partial annealing zone for apatite was observed between 2000 and 4000?m (～60–110?°C). This confirms earlier in situ observations of the position of the partial annealing zone.