An evaluation of Guided Regularized Random Forest for classification and regression tasks in remote sensing

New Earth observation missions and technologies are delivering large amounts of data. Processing this data requires developing and evaluating novel dimensionality reduction approaches to identify the most informative features for classification and regression tasks. Here we present an exhaustive evaluation of Guided Regularized Random Forest (GRRF), a feature selection method based on Random Forest. GRRF does not require fixing a priori the number of features to be selected or setting a threshold of the feature importance. Moreover, the use of regularization ensures that features selected by GRRF are non-redundant and representative. Our experiments based on various kinds of remote sensing images, show that GRRF selected features provides similar results to those obtained when using all the available features. However, the comparison between GRRF and standard random forest features shows substantial differences: in classification, the mean overall accuracy increases by almost 6% and, in regression, the decrease in RMSE almost reaches 2%. These results demonstrate the potential of GRRF for remote sensing image classification and regression. Especially in the context of increasingly large geodatabases that challenge the application of traditional methods.     

Properties of Galaxies in the Second Byurakan Survey. II. Characteristic Features of a Separate Sample of Galaxies

Some features of a sample of galaxies from the Second Byurakan Survey are discussed. Most of them are small galaxies with star formation. It is shown that different types of galaxies are in the Survey, with high, medium, and low levels of excitation and continua ranging from blue to relatively red. Many of the galaxies have absorption lines in their spectra; their continua are formed by stars in later spectral classes.     

AN INTEGRATED APPROACH TO AIR MASS CLASSIFICATION IN THE NORTH CENTRAL UNITED STATES

Both subjective and objective schemes for air mass classification offer advantages to climatological studies. An integrated approach is developed and applied to 850 mb level data in the north central United States. Each air mass type was initially delimited by a subjective analysis. Final limits were then determined with an objective technique that separates temperature curves into normal components based on interpretation from the subjective method. Continental air was most frequent in all seasons except summer. The methodology also allows estimation of classification error.     

Discrimination of sedimentary facies by association analysis

The ability of association analysis to discriminate sedimentary facies was tested on Purdy's modal analyses of modern sediments of the Great Bahama Bank. Purdy's data set has served in the past as a standard reference for evaluating various multivariate classification algorithms. In order to adapt Purdy's data to association analysis, the percent abundance of the 12 constituents was converted to binary form by dichotomizing each variable on its mean value. The results obtained by association analysis are virtually identical to those obtained by Purdy and other authors. The same four main sedimentary facies were discriminated; 86% of the samples were identically classified (97% when misclassified borderline cases are counted as matches); the total partition variance of the classification is only negligibly greater (4%); and the grouping of the variables yielded the same four groups. The rank order of the three division-attributes responsible for the sample classification is fines, oolites, and corals. Association analysis has been employed by other authors to differentiate meaningful facies groups in studies of ancient reef carbonates, modern reef sediments, and heavy minerals in stream sediments. In all these studies, the results were found to be compatible with those obtained by using the continuous quantitative measurements, indicating that qualitative binary data may often be sufficient for the purpose of facies discrimination in many branches of geology and that association analysis is an effective method for this purpose.     

THE EVOLUTION OF SERVICE ECONOMIES: A GEOGRAPHICAL PERSPECTIVE1

Of the several classifications of services, each emphasizes different characteristics. The International Standard Industrial Classification (ISIC) is used for a comparative study on the transformation into service economies of the U.S., Canada, Great Britain, France, West Germany, Italy and Japan. Service economies are defined as national economies in which employment in services surpassed that in manufacturing. The pace of the evolution of service economies has been rapid with North America leading, followed in order by Great Britain, France, Italy, Japan and West Germany. Service economies have important implications for urban growth, intercity competition, and international trade.     

Evaluation of approaches to calculate debris-flow parameters for hazard assessment

Many different runout prediction methods can be applied to estimate the mobility of future debris flows during hazard assessment. The present article reviews the empirical, analytical, simple flow routing and numerical techniques. All these techniques were applied to back-calculate a debris flow, which occurred in 1982 at La Guingueta catchment, in the Eastern Pyrenees. A sensitivity analysis of input parameters was carried out, while special attention was paid to the influence of rheological parameters. We used the Voellmy fluid rheology for our analytical and numerical modelling, since this flow resistance law coincided best with field observations. The simulation results indicated that the “basal” friction coefficients rather affect the runout distance, while the “turbulence” terms mainly influence flow velocity. A comparison of the velocity computed on the fan showed that the analytical model calculated values similar to the numerical ones. The values of our rheological parameters calibrated at La Guingueta agree with data back-calculated for other debris flows. Empirical relationships represent another method to estimate total runout distance. The results confirmed that they contain an important uncertainty and they are strictly valid only for the conditions, which were the basis for their development. With regards to the simple flow routing algorithm, this methods could satisfactorily simulate the total area affected by the 1982 debris flow, but it was not able to directly calculate total runout distance and velocity. Finally, a suggestion on how different runout prediction methods can be applied to generate debris-flow hazard maps is presented. Taking into account the definition of hazard and intensity, the best choice would be to divide the resulting hazard maps into two types: “final hazard maps” and “preliminary hazard maps”. Only the use of numerical models provided final hazard maps, because they could incorporate different event magnitudes and they supplied output-values for intensity calculation. In contrast, empirical relationships and flow routing algorithms, or a combination of both, could be applied to create preliminary hazard maps. The present study only focussed on runout prediction methods. Other necessary tasks to complete the hazard assessment can be looked up in the “Guidelines for landslide susceptibility, hazard and risk zoning” included in this Special Issue.     

Risk assessment strategies for the reactivation of earth flows in the Northern Apennines (Italy)

The aim of this paper is to analyze the reactivation mechanism of ancient earth flows, with a view to gleaning information that can subsequently be utilized to formulate a risk-reduction strategy. All considerations made herein are the result of direct experience and observation of actual events which have occurred over the past few decades in the Northern Apennines. Particular attention has been paid to the analysis of the evolution of landslides during actual reactivation, acknowledging a typical, recurring succession of events that precede the failure of the slope. The hazard assessment of these large landslide bodies, which are of slope scale, constitutes a thorny problem, especially in view of the inapplicability of traditional deterministic models such as limit equilibrium stability analysis. Nevertheless, a site-specific assessment of probability of reactivation of these large and ancient earth flows is fundamental to effective land-use planning.     

Landslide susceptibility and hazard mapping in Australia for land-use planning — with reference to challenges in metropolitan suburbia

This paper provides an overview of the history and current status of landslide susceptibility and hazard mapping for land-use zoning in Australia. It also describes a case study of landslide hazard mapping in a medium density, coastal, suburban residential area of metropolitan Sydney, New South Wales, Australia, with relatively steep terrain. Issues covered include identification and mapping of existing and potential landslides, and susceptibility and hazard zoning for regulatory management and land-use planning. The method involves application of the principles contained within the AGS (2000) guideline, and as updated by the AGS (2007 a,b,c,d,e) suite of guidelines.     

Development and morphometry of sinkholes in coastal plains of Apulia, southern Italy. Preliminary sinkhole susceptibility assessment

Evolution of coastlines in karst areas may be strongly controlled by dissolution processes which favour the development of surface and subsurface landforms. The generation of caves in these environments is commonly favoured by the mixing between fresh and brackish waters. The sinkholes resulting from the upward propagation of the caves may interfere with the anthropogenic environment and cause damage to human elements (property and activities). To highlight the often underestimated importance of karst phenomena in coastal areas, we have analyzed a coastal stretch of Apulia, in southern Italy. The study area, covering an extension of about 6 km², is situated in the Ionian coast, and presents several interesting karst landforms that are generally connected to caves. Tens of sinkholes were mapped through field surveys, multi-year aerial-photographs (dating back to the 1940s) and archival research. We have performed a morphometric analysis of the sinkholes. The analysis describes the main parameters of the sinkholes (area, length, width, and depth), and the control exerted by the main discontinuity systems in the area. The detrimental effects derived from interaction between human environment and these karst landforms is also under consideration. A sinkhole susceptibility map, which may provide useful information for planners, developers and the insurance industry has eventually been produced through application of a decision tree model.     

National flood discharge mapping in Austria

This article presents the approach and the results of a study in which 30, 100 and 200 year return period flood discharges were estimated for 26,000 km of Austrian streams. Three guiding principles were adopted: combination of automatic methods and manual assessments by hydrologists to allow speedy processing and account for the local hydrological situation; combination of various sources of information including flood peak samples, rainfall data, runoff coefficients and historical flood data; and involvement of the Hydrographic Services to increase the accuracy and enhance the acceptance of results. The flood discharges for ungauged catchments were estimated by the Top-kriging approach with manual adjustment to the local flood characteristics. The adopted combination approach proved to be very efficient both in terms of the project time required and in terms of the accuracy and acceptability of the estimated flood discharges of given return periods.