Direct induction of compact rule-based classifiers for resource mapping

Abstract

Rule-based classifiers are used regularly with geographical information systems to map categorical attributes on the basis of a set of numeric or unordered categorical attributes. Although a variety of methods exist for inducing rule-based classifiers from training data, these tend to produce large numbers of rules when the data has noise. This paper describes a method for inducing compact rule-sets whose classification accuracy can, at least in some domains, compare favourably with that achieved by larger less succinct rule-sets produced by alternative methods. One rule is induced for each output class. The condition list for this rule represents a box in n-dimensional attribute space, formed by intersecting conditions which exclude other classes. Despite this simplicity, the classifier performed well in the test application prediction of soil classes in the Port Hills, New Zealand, on the basis of regolith type and topographic attributes obtained from a digital terrain model.     

Spatial analysis weighting algorithm using Voronoi diagrams

Data layers that represent geographical constraints in a multidimensional GIS model must be appropriately weighted to effectively account for the diversity as well as the functional and spatial interrelationships between the constraints. This paper presents a spatial analysis weighting algorithm (SAWA) using Voronoi diagrams. The basic functions of the SAWA are defined so that the spatialization of weights is done according to two approaches: a global spatialization method based on the statistical distribution of the original data and a contextual approach where neighbourhood defined by Voronoi diagrams is integrated into the weighting functions. Different simulations on artificial and real maps applied to the problem of shortest path optimisation are analysed. The results show that the effective integration of the spatial dimension in a weighting process is not only possible but also improves the optimisation of shortest paths. Research is continuing to improve the contextual phase of the algorithm.     

On topographic core-mantle coupling

Abstract

The purpose of this note is two-fold: to draw attention to a perplexing difficulty connected with topographic core-mantle coupling, and to suggest tentatively an explanation. The difficulty is an apparent conflict between the most comprehensive theory of the coupling so far attempted (Anufriev and Braginsky, 1975a, b, 1977a, b) and recent explicit calculations based on magnetic and seismic information (Speith et al., 1986). It is argued that asymmetric deviations from Anufriev and Braginsky's basically axisymmetric model of the underlying core flow are capable of resolving the difficulty.     

Historical trends in Florida temperature and precipitation

Because of its low topographic relief, unique hydrology, and the large interannual variability of precipitation, Florida is especially vulnerable to climate change. In this paper, we investigate a comprehensive collection of climate metrics to study historical trends in both averages and extremes of precipitation and temperature in the state. The data investigated consist of long‐term records (1892–2008) of precipitation and raw (unadjusted) temperature at 32 stations distributed throughout the state. To evaluate trends in climate metrics, we use an iterative pre‐whitening method, which aims to separate positive autocorrelation from trend present in time series. Results show a general decrease in wet season precipitation, most evident for the month of May and possibly tied to a delayed onset of the wet season. In contrast, there seems to be an increase in the number of wet days during the dry season, especially during November through January. We found that the number of dog days (above 26.7 °C) during the year and during the wet season has increased at many locations. For the post‐1950 period, a widespread decrease in the daily temperature range (DTR) is observed mainly because of increased daily minimum temperature (Tmin). Although we did not attempt to formally attribute these trends to natural versus anthropogenic causes, we find that the urban heat island effect is at least partially responsible for the increase in Tmin and its corresponding decrease in DTR at urbanized stations compared with nearby rural stations. In the future, a formal trend attribution study should be conducted for the region. Copyright © 2012 John Wiley & Sons, Ltd.     

Evaluation of Composts from Agricultural Wastes with Fish Pond Sediment as Bulking Agent to Improve Compost Quality

Composts with five different ratios of agricultural wastes, viz. rice straw (RS), wheat straw (WS), potato plant (PP), and mustard stover (MS) were prepared with or without fish pond bottom sediment to investigate the compost maturity and their suitability for field application. The composting process was monitored through the changes in physico‐chemical parameters and germination index (GI) at every 7 days interval of the composting process. All the composts were dark brown and smelled like forest soil within 56 days of composting, which reflected its matured status. On the basis of the physico‐chemical parameters (bulk density: 0.84 g/cm³; pH 7.05; electrical conductivity: 3.52 mS/cm; cation exchange capacity:82.4 cmol/kg; total carbon:321.4 g/kg; total nitrogen: 16.9 g/kg; As: 6.8 mg/kg; Cd: 2.96 mg/kg; Cr: 29.6 mg/kg, Cu: 243.6 mg/kg; Hg: 0.019 mg/kg; Ni: 24.3 mg/kg; Pb: 62.1 mg/kg and Zn: 812 mg/kg) and GI (89–96%), it could be concluded that RS/WS/PP/MS, 1:1:2:1 v/v/v/v with fish pond sediment produced better compost in accordance with the Indian compost standard. Application of a combined randomized block design analysis revealed that there is a significant difference in the responses of the five composts, in relation to the time of composting. Hierarchical clustering algorithm was applied with a view to form homogeneous groups of five different composts on the basis of different physico‐chemical parameters. Therefore, the ratio of waste incorporation is an important decision for composting and addition of pond sediment can improve the quality of compost.     

Compact,concurrent, and contiguous: smart growth and 50 years of residential planning in the Toronto region

In this article, we use parcel-based land-use data to analyze 50 years of residential development in the Toronto region. We test two hypotheses: (1) Toronto’s form does not conform to conventional definitions of suburban sprawl and (2) Toronto’s suburban development shows high levels of continuity over time with relatively high densities and mixed housing types. Contrary to recent research suggesting a convergence of urban forms among North American metropolitan regions, Ontario’s robust planning system has created a distinctive, highly consistent pattern of residential development that has, for half a century, achieved many of the core goals of smart growth including relatively compact, contiguous, and concurrent development. This form continues to be automobile dependent, however, and is not producing many of the benefits ascribed to smart growth. Rather than continuing to adopt United States-inspired smart growth policies, a more ambitious set of initiatives will be required to address current regional challenges.     

SHORELINE CHANGE AT THOMPSON ISLAND,BOSTON HARBOR,MASSACHUSETTS, 1938-1977

A photogrammetric and sediment analysis is presented to illustrate the relationship between beach erosion and seacliff recession on Thompson Island, Boston Harbor, Massachusetts. Aerial photographs taken in 1938, 1952, 1963, and 1977 were measured to determine rates of shoreline change around the island. The 39-year average rate of beach erosion is 0.3 m/yr ± 2% with an average rate of cliff recession at 0.2 m/yr ± 2%. Rates of beach erosion between six orientations that reflect principal wave approach to the island were not found to be significantly different (0.05) for the 39-year period. The rank order correlation between beach erosion and cliff recession for these six orientations was moderately well correlated (0.63). To determine possible controls of cliff erosion, 58 sediment samples were collected from glacial cliffs along the shoreline. The textural composition was determined, and then tested with discriminant function analysis. Partial correlation analysis between beach erosion and cliff recession holding a surrogate for sediment size constant improved the rank order from 0.63 to 0.84. The results indicate that coarser-grained cliffs recede at faster rates, but with less erosion occurring on adjacent beaches. Alternatively, finer-grained cliffs recede at slower rates, but with greater erosion occurring on adjacent beaches.     

SOIL VARIATION ON NORTH AND SOUTH SLOPE CATENAS ON A KAME IN SOUTHEAST WISCONSIN

Soils from two catenas on north-and south-facing slopes of a kame were analyzed to determine the effect of lithology, topography, and microclimate on profile development. In loess on the north-facing slope, where microclimate favored CaCO₃ dissolution and clay translocation, pedogenesis was rapid and an A-E-Bt-C profile developed. In contrast, an A-Bt-C profile developed on the south-facing slope. In gravel where limestone/dolostone dissolution was slow, A-C profiles were found and CaCO₃ was depleted to a ≤ 25-cm depth, 20± cm less than reported for similar Michigan soils. Slopewash and runoff are inferred to have caused silt accumulations and greater infiltration at the base of ≥ 20° slopes, resulting in a thicker solum in foot-and toe-slope positions, whereas on slopes of ≤ 7° infiltration and interflow are the dominant processes, resulting in a thicker solum on the transportational midslope. The differences in soil profile development are attributed to sediment facies changes at 25- to 50-cm depth and resulting groundwater movement. [Key words: soil development, soil spatial variability, kame slope catena, Wisconsin.]     

Occurrence and significance of a cold-water carbonate pseudomorph in microbialites from a saline lake

Micron-scale pseudomorphs of calcite after ikaite were discovered in microbialites from Manito Lake, a large hypersaline lake in the Great Plains of western Canada. Although environmental conditions in the lakes of this region (seasonally cold temperatures, high salinities and elevated productivity) suggest that ikaite should be common, this is the first documentation of lacustrine ikaite in Canada and the Great Plains region of North America. The calcite pseudomorphs form a porous dendritic fabric, comprise the interiors of massive shoreline microbialite mounds and pinnacles, and are encased in centimeter-scale, laminated dolomite-aragonite rinds. Because of the limited thermodynamic stability of ikaite at or near freezing temperatures, the psuedomorphs of this mineral provide evidence of an extended cold episode during the first millennium AD. The Manito deposits also show high δ¹³C values, indicating elevated productivity and δ¹⁸O signatures are consistent with precipitation in a cold, somewhat fresher lake than present.