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.     

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.     

Drainage in finite-sized unsaturated zones

After the initiation of gravity drainage, water is often assumed to be either (a) draining under unit gradient, or (b) at capillary/gravity equilibrium. Both of these simplifications can be useful, but the regimes of validity of each assumption must be delineated. Water pressures are measured versus time and distance as water drains out of a 1.6 m long sand column to determine the relative effects of capillary and gravitational forces during drainage. For medium sized sands (0.15–0.3 mm in diameter), the capillary pressure is constant in space in a large region of the column for over 12 days, and the water continues to flow under unit gradient for relatively long time scales. Similar results are seen for finer sands, but with a much faster approach to equilibrium. Numerical simulations and analytical estimates are presented and compare favorably to the measurements. Together, the experimental, theoretical and analytical results are used to calculate when capillary/gravity equilibrium is reached as a function of porous media properties and length of the unsaturated zone. The ratio of the length of the unsaturated zone to the bubbling pressure is a key parameter in determining the drainage regime, and that even for relatively short unsaturated zones the equilibrium time scale can be on the order of years.     

Continuous wavelet-like filter for a spherical surface and its application to localized admittance function on Mars

We have developed a 2D isotropic continuous wavelet-like transform for a spherical surface. The transform is simply defined as the surface convolution between the original field and a kernel, based on the zeroth-order Bessel function with a spherical correction. This spherical correction violates the geometric similarity for the various scales of the kernels, which becomes more apparent at longer wavelengths. We found numerically that this transform is practically equivalent to a Gaussian bandpass filter in the spherical harmonic domain. We have applied this wavelet-like transform on the recently acquired Martian gravity and topography fields. Using a ratio constructed locally from these two fields, we have constructed a map describing the lateral variations of the localized admittance function on Mars.     

Web-Based interrogation of large-scale geophysical data sets from handheld devices

This paper describes the issues related to using handheld devices to interrogate wirelessly large-scale geophysical data over the Web using a client–server configuration. A modification of our current Web-based data interrogation program addresses specifically the limitations of current handheld devices with wireless access, such as processor speed, screen size and resolution, memory, and network speeds. The system uses a combination of Java, Python, and C in order to run remotely from the handheld device, while incorporating the power and memory capacity of a large workstation server.     

Geochemistry of eucrites: genesis of basaltic eucrites, and Hf and Ta as petrogenetic indicators for altered antarctic eucrites

We performed instrumental neutron activation analysis on a large suite of antarctic and nonantarctic eucrites, including unbrecciated, brecciated, and polymict eucrites and cumulate and noncumulate eucrites. We evaluate the use of Hf and Ta, two highly incompatible elements, as sensitive indicators of partial melting or fractional crystallization processes. Comparison with rare earth element (REE) data from nonantarctic and antarctic eucrites shows that Hf and Ta are unaffected by the terrestrial alteration that has modified the REE contents and patterns of some antarctic eucrites. The major host phases for Hf and Ta—zircon, baddeleyite, ilmenite, and titanite—are much less susceptible to terrestrial alteration than the phosphate hosts of REEs. The host phases for Hf and Ta are minor or trace phases, so sample heterogeneity is a serious concern for obtaining representative compositions. The trace lithophile and siderophile element contents of noncumulate eucrites do not allow for a single, simple model for the petrogenesis of the howardite-eucrite-diogenite suite. Fractional crystallization models cannot reproduce the compositional relationship between eucrites of the main group-Nuevo Laredo trend and those of the Stannern trend. Equilibrium crystallization models cannot explain the trace element diversity observed among diogenites. Partial melting models cannot explain the W variations among eucrites, unless source regions had different metal contents. We suggest that slight variations in oxygen fugacity of eucrite source regions during partial melting can explain the W variations without requiring different metal contents. This hypothesis may fail to account for eucrite Co contents, however.     

Understanding Climatic Impacts, Vulnerabilities, and Adaptation in the United States: Building a Capacity for Assessment

Based on the experience of the U.S. National Assessment, we propose a program of research and analysis to advance capability for assessment of climate impacts, vulnerabilities, and adaptation options. We identify specific priorities for scientific research on the responses of ecological and socioeconomic systems to climate and other stresses; for improvement in the climatic inputs to impact assessments; and for further development of assessment methods to improve their practical utility to decision-makers. Finally, we propose a new institutional model for assessment, based principally on regional efforts that integrate observations, research, data, applications, and assessment on climate and linked environmental-change issues. The proposed program will require effective collaboration between scientists, resource managers, and other stakeholders, all of whose expertise is needed to define and prioritize key regional issues, characterize relevant uncertainties, and assess potential responses. While both scientifically and organizationally challenging, such an integrated program holds the best promise of advancing our capacity to manage resources and the economy adaptively under a changing climate.     

Science and the desertification debate

Desertification is an environmental issue that has major environmental and social dimensions, and which has been controversial in a number of respects. The recent UN Convention on Desertification and Drought places considerable emphasis on the social dimensions of the problem and the role local peoples and NGOs have to play in tackling it. When desertification first became regarded as a major issue in the 1970s, science was seen to have a major role in understanding and solving the problem, but in some circles has been regarded as failing to contribute usefully. This paper explores the nature of scientific contributions, why science may have been seen to fail to serve the needs of policy-makers, and why scientific findings may have been misinterpreted in broader circles dealing with the issue. It concludes by identifying areas where science has still a valuable role to play in an issue that is now perhaps regarded as more political and social than scientific.