POPULATIONS,SAMPLES AND STATISTICAL INFERENCE IN GEOGRAPHY

This paper questions the validity and relevance of the application of procedures of classical statistical inference to population data in geography. Arguments for such procedures include the notions of temporal and spatial samples, measurement error, modifiable areal units, data vetting and stochastic processes. It is concluded that such arguments cannot be justified in terms of statistical theory.     

Graphical methods for investigating the fit of a Fisher distribution to spherical data

Summary. Ordered-value plotting (probability plotting, Q-Q plotting, etc.) is well known for its utility as a diagnostic technique with data samples on the line. It can be used to give a visual indication of goodness of fit (sometimes as a prerequisite to a more formal statistical analysis), to provide quick graphical estimates of parameters, and to disclose the presence of outliers and give a visual assessment of their import. Following a résumé of plotting procedures for distributions on the line such as normal and exponential, three procedures are given for use with spherical data assumed to follow a Fisher distribution. These are illustrated by application to two palaeomagnetic data samples.     

The representation of geoscience information for data integration

In mineral exploration, resource assessment, or natural hazard assessment, many layers of geoscience maps such as lithology, structure, geophysics, geochemistry, hydrology, slope stability, mineral deposits, and preprocessed remotely sensed data can be used as evidence to delineate potential areas for further investigation. Today's PC-based data base management systems, statistical packages, spreadsheets, image processing systems, and geographical information systems provide almost unlimited capabilities of manipulating data. Generally such manipulations make a strategic separation of spatial and nonspatial attributes, which are conveniently linked in relational data bases. The first step in integration procedures usually consists of studying the individual charateristics of map features and interrelationships, and then representing them in numerical form (statistics) for finding the areas of high potential (or impact).Data representation is a transformation of our experience of the real world into a computational domain. As such, it must comply with models and rules to provide us with useful information. Quantitative representation of spatially distributed map patterns or phenomena plays a pivotal role in integration because it also determines the types of combination rules applied to them.Three representation methods—probability measures, Dempster-Shafer belief functions, and membership functions in fuzzy sets—and their corresponding estimation procedures are presented here with analyses of the implications and of the assumptions that are required in each approach to thematic mapping. Difficulties associated with the construction of probability measures, belief functions, and membership functions are also discussed; alternative procedures to overcome these difficulties are proposed. These proposed techniques are illustrated by using a simple, artificially constructed data set.     

Guest editorial

The past decade has witnessed extensive development of measures that examine characteristics of spatial subsets (local spaces) defined with respect to a complete data set (global space). Such procedures have evolved independently in fields such as geography, GIS, cartography, remote sensing, and landscape ecology. Collectively, we label these procedures as local spatial methods. We focus on those methods that share a common goal of identifying subsets whose characteristics are statistically ‘significant’ in some way. We propose the concept of local spatial statistical analysis (LoSSA) both as an integrative structure for existing methods and as a framework that facilitates the development of new local and global statistics. By formalizing what is involved when a particular local statistic is used, LoSSA helps to reveal the key features and limitations of the procedure. These include a consideration of the nature of the spatial subsets, their spatial relationship to the complete data set, and the relationship between a given global statistic and the corresponding local statistics computed for the data set.     

AN EVALUATION OF SAS/GRAPH SOFTWARE FOR COMPUTER CARTOGRAPHY1

The SAS® computer software system, widely used and respected for its capabilities in statistical analysis and data base management, now includes a new set of graphic and cartographic procedures called SAS GRAPH?. We have used these cartographic procedures in research on mapping ethno-cultural census data from metropolitan areas in Ontario and in undergraduate and graduate courses in computer cartography. On the basis of that experience, we describe and evaluate SAS/GRAPH'S cartographic capabilities and illustrate with maps drawn by various devices.     

典型相关分析在土地利用结构研究中的应用——以环渤海地区为例

典型相关分析是一种揭示两组多元随机变量之间相关关系的统计模型方法。本文通过介绍这一多元统计方法的内涵、特点和思路,将其引入有关地学问题的分析———判别土地利用类型分布与其影响因子之间的相关关系。以环渤海地区为例,通过数据准备、操作过程和统计检验等几个方面,全面介绍了典型相关分析在实际研究中的应用     

EXAMINING LARGE DATABASES:A CHEMOMETRIC APPROACH USING PRINCIPAL COMPONENT ANALYSIS

Principal component analysis is used to examine large multivariate databases.The graphical approachto exploratory data analysis is described and illustrated with a single example of chemical compositiondata obtained on environmental dust particles.While the graphical approach to exploratory data analysishas certain advantages over the numerical procedures,the empirical approach described here should beviewed as complementary to the more robust treatments that statistical methodologies afford.     

STATISTICAL INFERENCE IN GEOGRAPHY: MODERN APPROACHES SPELL BETTER TIMES AHEAD*

Concepts relevant to statistical inference testing are discussed within a framework which contrasts conventional methodology with current trends in statistics. The latter are emphasized as having the quality of being less restrictive than the former. To show this quality, randomization testing is described and an example presented. Also described is a nonparametric technique known as QAP, which is particularly versatile. Optimism is expressed about the potentiality of such procedures for geographic research.     

SPIDER—an interactive statistical tool for the analysis of spatially distributed data

Interactive statistical graphics are reviewed in the contexts of spatial data and geographical information systems (GIS). GIS provide the user with an active geographical view of the data—a map that can be used as an entry point to the data base. Prototype software—SPIDER—illustrates the possibilities of using statistical graphics as further views of the data, which can be made active and thus provide alternative means of querying the data. These views can be cross-referenced by 'linking'. It is argued that such a system can provide a very rich environment for pursuing exploratory statistical analysis of spatial data.     

Data integration through region-based nominal filtering

Abstract

Resource models integrating disparate nominal or class grid-cell data can be implemented by using spatial filters. Most modelling procedures do not adequately handle noise created during the process of merging and integrating multiple grid-cell data sets. Data integration can be best accomplished in an environment where ready access to statistical and database management systems support the reclassification of noise grid-cells. These systems provide access to functionality and information which support the design of the spatial filter and the evaluation of the result of the spatial filter and the resource model.     

A new conglomerate test in palaeomagnetism

The conglomerate test is widely used in palaeomagnetism to date components of natural remanent magnetization with respect to deposition of conglomerates. It has been demonstrated, however, that this test may be positive even if the data are strongly contaminated by a secondary remanence, especially for the commonly used small number of clasts Starkey & Palmer 1970). Here we show with the aid of numerical simulations that different statistical procedures employed in this test have similar low sensitivities to remagnetization. We suggest a new conglomerate test which incorporates additional information on the direction of a secondary palaeomagnetic component which is isolated from either clasts themselves or their host rocks. Numerical simulations show that this new test is about twice as sensitive to remagnetization as the previous procedures and is robust with respect to small errors in the direction of a secondary component.     

热带气旋降水模拟研究进展

中国是遭受热带气旋灾害最为严重的国家之一,准确的降水模拟对于开展热带气旋灾害风险评估有重要意义。本文从热带气旋灾害风险评估的视角,将降水模拟分为基于极值理论的降水极值模拟、基于站点的降水时空模拟和基于热带气旋路径的降水事件模拟3大类;根据风险评估对降水模拟的需求,从模型构建、发展及其特点等方面对3类模型进行分析评述;进而提出面向风险评估的热带气旋降水模拟,应兼顾降水模拟的一般性和热带气旋暴雨模拟的特殊性,平衡处理降水模拟结果的准确性、统计量的可靠性和计算量问题。以极值理论对降水极值模拟为基础,充分发挥降水时空模拟在处理长时间降水序列中的优势,并加强热带气旋降水的理论研究,进一步完善热带气旋降水事件的模拟模型。     

CHILDREN'S PERCEPTION OF A DOWNTOWN SHOPPING CENTER

This paper offers an exploratory investigation of the effects of a child's age on three behavioral constructs (awareness space, activity space, and attitude) concerning a downtown shopping center in Bristol, England. The data are elicited from three samples of Bristol school children defined according to age and are analyzed using statistical inferential procedures. The results disclose that awareness space and activity space consistently increase with age. In addition, attitudinal responses to specific features of the shopping center exhibit a gradual change over a long-term period.     

美国俄亥俄州土壤有机碳密度空间分布(英文)

Historical database of National Soil Survey Center containing 1424 geo-referenced soil profiles was used in this study for estimating the organic carbon(SOC) for the soils of Ohio,USA.Specific objective of the study was to estimate the spatial distribution of SOC density(C stock per unit area) to 1.0-m depth for soils of Ohio using geographically weighted regression(GWR),and compare the results with that obtained from multiple linear regression(MLR).About 80% of the analytical data were used for calibration and 20% for validation.A total of 20 variables including terrain attributes,climate data,bedrock geology,and land use data were used for mapping the SOC density.Results showed that the GWR provided better estimations with the lowest(3.81 kg m 2) root mean square error(RMSE) than MLR approach.Total estimated SOC pool for soils in Ohio ranged from 727 to 742 Tg.This study demonstrates that,the local spatial statistical technique,the GWR can perform better in capturing the spatial distribution of SOC across the study region as compared to other global spatial statistical techniques such as MLR.Thus,GWR enhances the accuracy for mapping SOC density.     

Comparative study of probable maximum precipitation and isohyetal maps for mountainous regions, Pakistan

Probable maximum precipitation (PMP) is widely used by hydrologists for appraisal of probable maximum flood (PMF) used for soil and water conservation structures, and design of dam spillways. A number of methods such as empirical, statistical and dynamic are used to estimate PMP, the most favored being statistical and hydro-meteorological. In this paper, PMP estimation in mountainous regions of Pakistan is studied using statistical as well as physically based hydro-meteorological approaches. Daily precipitation, dew point, wind speed and temperature data is processed to estimate PMP for a one-day duration. Maximum precipitation for different return periods is estimated by using statistical approaches such as Gumble and Log-Pearson type-III (LP-III) distribution. Goodness of fit (GOF) test, chi-square test, correlation coefficient and coefficient of determination were applied to Gumble and LP-III distributions. Results reveal that among statistical approaches, Gumble distribution performed the best result compared to LP-III distribution. Isohyetal maps of the study area at different return periods are produced by using the GIS tool, and PMP in mountainous regions varies from 150 to 320 mm at an average value of 230.83 mm. The ratio of PMP for one-day duration to highest observed rainfall (HOR) varied from 1.08 to 1.29 with an average value of 1.18. An appropriate frequency factor (K_m) is very important which is a function of mean for observed precipitation and PMP for 1-day duration, and K_m values varies from 2.54 to 4.68. The coefficient of variability (C_v) varies from minimum value of 28% to maximum value of 43.35%. It was concluded that the statistical approach gives higher results compared to moisture maximization (MM) approach. In the hydro-meteorological approach, moisture maximization (MM) and wind moisture maximization (WMM) techniques were applied and it was concluded that wind moisture maximization approach gives higher results of PMP as compared to moisture maximization approach as well as for Hershfield technique. Therefore, it is suggested that MM approach is the most favored in the study area for PMP estimation, which leads to acceptable results, compared to WMM and statistical approaches.     

Land cover classification by an artificial neural network with ancillary information

Abstract

Remote sensing is an important source of land cover data required by many GIS users. Land cover data are typically derived from remotely–sensed data through the application of a conventional statistical classification. Such classification techniques are not, however, always appropriate, particularly as they may make untenable assumptions about the data and their output is hard, comprising only the code of the most likely class of membership. Whilst some deviation from the assumptions may be tolerated and a fuzzy output may be derived, making more information on class membership properties available, alternative classification procedures are sometimes required. Artificial neural networks are an attractive alternative to the statistical classifiers and here one is used to derive a fuzzy classification output from a remotely–sensed data set that may be post–processed with ancillary data available in a GIS to increase the accuracy with which land cover may be mapped. With the aid ancillary information on soil type and prior knowledge of class occurrence the accuracy of an artificial neural network classification was increased by 29–93 to 77–37 per cent. An artificial neural network can therefore be used generate a fuzzy classification output that may be used with other data sets in a GIS, which may not have been available to the producer of the classification, to increase the accuracy with which land cover may be classified.     

The nature, use and problems of historical archives for the temporal occurrence of landslides, with specific reference to the south coast of Britain, Ventnor, Isle of Wight

The purpose of this paper is to describe and evaluate the nature of the European historical archives which are suitable for the assessment of the temporal occurrence and forecasting within landslides studies, using the British south coast as an example. The paper is based upon the British contribution to the Environment programme EPOCH, 1991–1993.A primary requirement of a research programme on process occurrence is to determine the event frequencies on as many time and space scales as possible. Thus, the analysis of archives is, potentially, an essential preliminary to the study of the temporal occurrence of landslide events. The range of such data sources extends from isolated, fortuitously dated sites from the Quaternary assemblage, through inferred event impacts using dendrochronology or lichenometric time series to historical records of causal factors such as rainfall data and more recently, deliberately recorded packages of cumulative or continuous data.Most countries have extensive historical sources which may be of considerable value in establishing the characteristics of geomorphological processes. These include narrative in literature, prints and other artwork, terrestrial and aerial photographs, remote sensing series, newspapers, incidental statements and scientific journals and reports.These are numerous difficulties in accessing, extracting, organising, databasing and analysing such data because they are not usually collated for scientific use. Problems involve such incalculable errors as: the experience, training and conscientiousness of the observer; the editing and recording process; judging the validity of the data used and the haphazard nature of recorded events in time and space.Despite these difficulties, such data do yield a record which adds to the representative temporal sample as a level above some threshold reporting position. It therefore has potential for specific statistical analysis. An example of a reasonable temporal landslide record is the data base of the Ventnor complex on the Isle of Wight initially established in 1991 by Geomorphological Services Limited (GSL), now of Rendel Geotechnics, and supplemented by the collections of the first author. The record displays an increase in landslide events over the present century, due probably to increasing technology and awareness of hazard and the development of process geomorphology. However, the landslide record was subsequently correlated with the Ventnor precipitation series. This indicated that wet year sequences usually gave rise to significant landslide events. The increasing variability and number of rainfall events predicted by various climatic units, e.g. the Hadley Centre, may therefore indicate a fundamental increase in landslide events in the future.     

The Usefulness of a Map with Geographic Text: A Reanalysis of Experimental Data

Abstract

One research report of the usefulness of a map with geographic text appeared inconsistent with other research findings related to other graphic illustrations accompanying printed narrative. Consequently, data employed in the earlier report were reanalyzed through the use of analysis of covariance, a more powerful statistical procedure than the univariate analysis originally used. Results of this reanalysis revealed no treatment effects; that is, a map with text seemed not to be a superior condition to that of text alone. Research procedures as well as the substantive findings are discussed.