People have been working for years on geo-information map in different fields of geography since it was proposed.However,most current works on geo-information map stay at the stage of representing geographic phenomena or processes as traditional map does,while lack of knowledge representation and reasoning,so it shows no difference with traditional cartographic map.This results in difficulties in extracting information automatically from geo-information maps,so experts' knowledge and experience are required in order to acquire spatial or temporal patterns through the spatial distribution of geographic phenomena on geo-information maps.This paper proposes the concept of geographic knowledge map,which uses measurements in feature dimensions as coordinate frame instead of geographic coordinate system and displays the characters of objects and the relationships between them with different signals.In this way,geographic knowledge map can disclose spatial-temporal patterns in different feature dimensions which are less possible to be found in real world and geo-coded map.In this paper,the conceptual framework and properties of geographic knowledge map is described,and its characteristics are concluded.Geographic knowledge map is not only graphic representation of geographical knowledge,but also a formal language of knowledge representation.It represents knowledge with structured formal diagrams,has specific semantics and syntax,so it is computable.Through computation of geographic knowledge map,we can transfer the knowledge it represents into other formal language descriptions.It is a way of geographic knowledge visualization,and a method of data mining and knowledge discovery as well.

The automatic building reconstruction from LIDAR data has been a hot issue for several years.Many methods and algorithms have been put forward to reconstruct the models of simple buildings,such as of flat roof,gable roof,or other rectangular shape.But it remains an open problem for the reconstruction of complex buildings.In this paper,a new idea is introduced to process the complex buildings.It is based on the contour clustering analysis of LADAR data and a bottom-up method using data-driven processing.Contours contain the shape information of object boundary,and they are closed,complete,and having explicit topological relationships among each other.So we can use these valuable characters to guide the building reconstruction.In this paper,a concept of contours cluster is introduced,which is based on the following observations: the contours of a building are usually very similar to each other in every part of the building;the nested similar contours are defined as a cluster of similar contour.The contour cluster reflects the detailed feature of the corresponding object.Analyzing the shape differences among contours clustering,the different parts of the whole complex building can be found out.So we can say,contour clustering is very useful,and is the core of the method.The process includes 4 main steps.Firstly,the LIDAR point is pre-processed,and the Delaunay mesh is constructed with the processed LIDAR points and the initial contours are traced.Secondly,some shape features are used to distinguish the contours on buildings or on other objects.With thresholds of contour length and area,some contours of vegetation can be removed.Thirdly the topology relationship and similarity relationship between contours are analyzed.Based on these relations,the contours are clustered to form the parts of buildings.At last,the building model of different types can be reconstructed from the clusters of contours.To test the approach presented above,2 experiment data with representative building models are applied.The results show our method has following advantages:(1) the closeness of contours can effectively avoid the difficulty of edges grouping in conventional reconstruction methods;(2) using contour cluster analysis can extract different hierarchical structures of the complex building;and(3) even curved surface buildings can be correctly constructed using our method.

Traffic road plays an important role in transportation and is a major driving force of urban land use expansion.Therefore,traffic road has great impact on urban expansion and development.Also,urban expansion has new demands for the development of traffic road.The relationship between traffic road and urban land use expansion is complex and uncertain,so we need to study it with a variety of research methods and from various aspects.Currently,a large number of mathematical methods and spatial models have been effectively used to analyze the relationship between traffic road and urban land use expansion.In order to analyze the interaction model and impact of traffic road on urban expansion effectively,however,new technical methods should be explored.Rough set theory,proposed by Pawlak in 1982,is an extension of set theory in analyzing data with insufficient and incomplete knowledge and has been demonstrated to be useful in fields such as pattern recognition,machine learning,and automated knowledge acquisition.Rough set theory does not need prior information in analyzing data and can derives the classification or decision rules according to a knowledge reduction procedure while keeping its classification ability unmodified.Thus,rough set theory can mine the objective and inherent rules contained in the data.Rough set theory is still not commonly used in research of the relationship between traffic road and urban land use expansion.In this paper,urbanization of Guangdong Province is taken as an example and the relationship between traffic road and urban land use expansion is studied with rough set theory.The result shows that the decision rules of traffic road and urban land use expansion extracted with rough set theory can effectively reflect the relationship between traffic road and urban land use expansion and can provide some guidance on urbanization researches.

Cadastre is a tool which is used to manage national land resources and assets.Traditional cadastre is two-dimensional one which takes parcel as basic management unit to administer land surface right,and its theoretical basis is consistency of the right nationality of the same parcel in the vertical direction,namely the ownership rights of underground,surface and ground space belong to the identical obligee.Due to the backwardness of the past economic development,land use mainly concentrated on the two-dimensional surface and it isn't necessary to extend land use into the third dimension.Therefore,traditional two-dimensional cadastre which takes surface right as core can satisfy social demands well.In recent decades,the rapid increase in population and intensified urbanization make the land use of urban areas,especially business centers under rising pressures.In order to provide more available space,city development extends unceasingly to the third dimension.Three-dimensional land use makes the subject of property become pluralism in the vertical direction,namely underground,surface and ground space can be stratified development and are distributed to different obligees.Tendency of Three-dimensional land use brings forth challenges to traditional land management theories and methods.Three-dimensional cadastre also needs a kind of basic unit like two-dimensional land parcel to register and manage property in three-dimensional space,this kind of three-dimensional basic unit is called property volume in this paper.We will take right entities such as land and property as basis,use right space as reference,research dividing mode of land and property in three-dimensional space,analyze existing form,dividing mode and space form of two kinds of basic unit,generalize theirs basic characteristic.Based on these substantive characteristics,we give the definition and representation of property volume,and test the correctness and validity of the given property volume.

P2P technology can handle the performance issues of centralized architecture,such as single point of failure and hot bottleneck.Raster geographic data service based on P2P technology has been applied widely,but there are many problems when considering how to combine P2P and vector geographic data service efficiently.P2P network model and the corresponding organization of distributed vector geographic data should be especially researched deeply.A vector-oriented P2P hybrid indexing network model,which is based on Chord and Quad structure,is proposed.This model combines the advantages of structured P2P network and hierarchical P2P network.It can not only make good use of Chord network with the merit of load balancing and hierarchical P2P network with high efficient querying,but also adapt to high dynamic network environment.In this work,Linking mechanism is proposed for the first time.This novel concept has two main characteristics: the information need to be recorded is little,which can reduce network load;the reconstructed information is complete and valid.And an organizational pattern of distributed vector geographic data is designed,which is based on Linking mechanism.This pattern forms a loosely and globally distributed topology,according to linking information on each level and topological connection information in derived geometry.This kind of distributed topology supports not only distributed index,lossless and fast reconstructing for vector geographical data,but also the extension of all kinds of vector topological analysis algorithm under distributed environment.Contrast with the experimental results of JTS,we can see that the time used for vector data reconstructing is obviously short because of fewer additional linking information,and the result of reconstruction is lossless and reliable.When the number of tiles is large,the merit to JTS is more obvious.Our results show that the theory about the organization of distributed vector geographic data based on Linking mechanism and the related algorithm are valid,high efficient and robust.And it has a good practical application foreground.

China has so many ancient villages in its long history which owes to the intricate environments and diversity ecology.At first,this paper analyses a lot domestic articles on ancient villages in recent years and draws some conclusion as following:(1) A lot of researchers introduce natural methodology to landscape research.(2) Many technicians use GIS technology to develop GIS-software for ancient village management.(3) Some new methods are used in ancient village tourism planning and there are so many successful examples in this domain.(4) Some researchers analysis traditional thought of human settlement environment,architecture arts and relationship between human and nature in order to help to solve analog problems existing in our era.Professor LIU Pei-ling brought forward a new theory which called "Gene Theory of Settlement Landscape" when he paid his attention to ancient villages some years ago.In seeking for landscape's genes,we analyze the necessary tend to building a prototype system of graphic methodology for ancient villages based on GIS.At the same time,we also probe into the context and senses on using GIS based graphic methodology to research ancient villages.Obviously,the authors prove that there are so many snags lying on this study domain when they try to develop such a prototype procedure.Of course,the authors describe these difficulties carefully and the key technical problems in building such a GIS-program of graphic methodology for ancient village particularly.Moreover,the requirement for the procedure and functional design of a prototype is discussed.The authors argue that the procedure should be comprised of five modules,DATABASE,MAP,SPATIAL,GENE,GRAPHMETHOD,respectively.By using the development toolkit of Visual C++.NET 2003,a prototype program is designed and implemented according to the research aim.Finally,we take WOMAN LANGUAGE of Jiangyong County,Hunan Province of China,which,is probably the merely women's language in the world,as an example to run the program and then to extract gene elements from ancient village landscapes.

Geometric correction and location error analysis of remote sensing image is an important aspect for image processing of remote sensing.But location error analysis for geostationary satellite imagery is little when scale factor λ is as a variable.In this paper,according to linear array scanning imaging mechanism,we completed the establishment of scan line array geometry imaging equation and location model for geostationary satellite.Then,based on the geometric location model,considering the scale factor λ as a variable,a strict quantitative analysis and theoretical derivation of location error caused by external orientation elements was finished.At the same time,the use of matrix theory and transformation' nature made the complicated process of location error and formula simplified.Finally,according to the location error formula,we have developed a prototype system for geometry location and error analysis of geostationary satellite imagery based on C++.And then,simulation experiments of geostationary satellite imagery from various angles were tried.The experimental results show that: an element error of single exterior orientation and the location error have obvious linear positive correlation;the descending order of exterior orientation elements' impact for location error to center point is φ,ω,α,β,H,κ;the descending order of exterior orientation elements' impact for location error to edge point is φ,ω,H,α,β,κ;From the image center to the edge,location error caused by φ and ω and H increases gradually,location error caused by α and β has a slow decrease;location error caused by κ has a slow increase;When six exterior orientation elements all have the critical error,from the image center to the edge,the integrated location error increases gradually;At the image edge,the rising trend of the integrated location errors is evident.The analytical method and conclusions in this paper provide support to the geometric error characteristics of geostationary satellite imagery,also is valuable to other remote sensing satellite platforms and the geometric correction of remote sensing images.

Land cover of Shanxi Province had been monitored from 1980s to 2005.Based on remote sensing and GIS technique,analysis on spatial-temporal characteristics of land cover change and driving factors was implemented.Results show that:(1)Changing area: Area of industrial and transportation land、rural settlement and a spot of grass land is increasing,on the other side,area of dry land,irrigable land,a spot forest land and bottom land etc is decreasing.(2)Changing speed: Land for urban construction had a very sharp increase slop,and the type which had second fast increase speed was marsh and rural settlement.The order of decrease speed from maximum to minimum is: bottom land,un-irrigated cropland and irrigated cropland.(3)Spatial conversion: The spatial conversion mostly occurred between cultivated land and settlements.Additionally,the increase of land for urban construction includes rural settlements which assimilated by urban expansion.The fraction of xerophilous grassland increase mainly converted from un-irrigated cropland,shrub forest mainly converted to xerophilous grassland,and bottom land mainly converted to water body.The natural environmental factors were stable,and were less important to the change of land cover.However,population,economic,regulating of agricultural structure,policy and transportation play fundamenta1 roles to land cover change.The concurrent correlations analysis finds that: the most important driving factors to Shanxi Province land cover change are rapid economic development and accelerated urbanization process.

We often run into the problem of how to distribute sample points in the process of social survey.Traditional spatial sampling theory is primarily used in single-target.However,social survey questionnaires sometimes are multi-objective.And some objectives are attribute data,for example,a survey related to people's age,education and other social information.Therefore,we should make some improvements based on traditional sampling methods to meet current requirements.Depending on the idea of multi-object and spatial variation,we make a multi-objective survey in this study.In order to meet the need of sampling in the questionnaire survey of Xiamen Island's settlement,we use the sandwich spatial sampling model,balancing the cost of sampling and sampling precision.In general,we adopt three methods: simple random sampling,spatial stratified sampling and sandwich spatial sampling.After comparing of those three methods,the results showed that:(1) When prior knowledge is not sufficient,the result by means of simple random sampling is more reasonable.But when there is sufficient prior knowledge,stratified sampling in space can make more representative samples for the sampling,and also can improve the accuracy of sampling.(2) The distribution of sample points of spatial stratified sampling is primarily related to stratification.The higher the stratified level,i.e.the higher the complexity of the regional,the more the number of sample points is.(3) By comparing the results of spatial stratified sampling by integrated factors and by experts,we can see that stratified sampling is directly or indirectly related with the stratifying process,the rationality of the stratifying design affects the rationality of sampling results.So,in the process of spatial stratified sampling,the rationality of stratifying design is very important for the stratified sampling.The objective of this study is to provide a basis for the sampling in the settlement of Xiamen Island,offer a concept of "complexity" to describe the situation of various regions,and suggest several ideas on the establishment of the multi-objective spatial sampling model.

A problem involved in emergency decision-making and many city management works is to evaluate the population within some given urban areas at a random time.The unknown spatial areas and evaluation time which constitute two fundamental uncertainty features of the problem bring great difficulty for its solution.Following the overall disciplines presented in daily traveling of urban residents,we know that people basically arrange their activities around various city buildings.Another explanation for the phenomenon is residents travel between buildings because they are attracted by the functions of them.So,in this paper,a method is developed to evaluate the static population inside a given area,which uses the data of buildings for habitation and average housing floorage per urban resident according to statistics.While another step must be done for assess the dynamic population,which is to confirm the population attracted by each type of building at any time.Two new indices are introduced: building correction coefficient and building attraction rate.The former refers to the ratio of average accommodated people between other building types and residential building.A rated accommodation people for each type of buildings can be achieved using the index.The latter is the ratio of real accommodation people to rated one.Here we divide the work day or break day all into seven periods according to the general time distribution of residents' traveling activities,and confirm a building type category by reference to one used in the occurrence rate investigation of city transportation.The two indices are estimated using the traveling investigation data of residents and vehicles.Then the dynamic population is drawn based on the proposed method.A case of chemical emergency in Beijing is imagined to demonstrate the application of the method and the issues concerned with parameter fixing that is the key to the method is discussed.

Path planning is a process to find the optimal path from the starting node to the destination node.Tens of algorithms have been designed to resolve the shortest path problem,in which the Dijkstra algorithm is always thought to be the most mature and classical one.However,its higher time complexity greatly restricts its practical application.Besides,for most path planning algorithms,the road network is considered as an ordinary plane network and the spatial distribution characteristics are ignored,therefore,the path planning algorithm can not do well in real road network.In order to improve the searching efficiency of the traditional Dijkstra algorithm and to meet the requirement of real-time vehicle navigation,a path planning algorithm based on SuperMap is designed by utilizing the network editing function of SuperMap GIS platform.Firstly,with the spatial distribution characteristics,a road network is built in SuperMap.In this process,we take three-dimensional transport and one-way road network into consideration,making the road network built in SuperMap able to reproduce the real road network.For three-dimensional transport,we set the parameters of non-broken lines and make sure the roads not in the same plane have no intersections;for one-way roads,the forward resistance and reverse resistance are set with different weights.Secondly,in order to improve the searching efficiency of the traditional Dijkstra algorithm,we propose a method to restrict the searching area.According to the coordinate relationships of the starting node and the destination node,we create a hexagon area,which contains the staring node and the destination node,as the restricted searching area,and make the algorithm able to search the shortest path within the restricted searching area on the theoretical basis of the classic Dijkstra algorithm.Thirdly,based on the actual needs,the path planning algorithm with certain constraints is designed.The constraints are prohibit a certain way and must pass through a certain point.At last,we apply the algorithm proposed in this paper to urban road network and compare it with traditional Dijkstra algorithm.The results show that,due to the searching area is restricted reasonably,the two algorithms get the same shortest path,but the time spent by the algorithm proposed in this paper is much less.Besides,the algorithm proposed in this paper can do well in path planning with constraints.Thus,the improved Dijkstra algorithm based on SuperMap GIS has a strong practicability,and can be used in real-time vehicle navigation.

The aim of the paper is to study the spatio-temporal dynamics of per capita GDP in the Beijing-Tianjin-Hebei Metropolitan Region(thereafter short for BTHMR).A sample of 140 county-level regions over the period 1995-2007 provides clear evidence of global and local spatial autocorrelation or autodependence as well as spatial heterogeneity.Transition probability matrices were used to verify the robustness of our conclusion.Based on techniques of exploratory spatial data analysis,several conclusions are draw as follows:(1) The economic disparity and spatial concentration of the BTHMR is enlarging,as indicated by coefficients of variation and global Moran's I.The economic disparity and spatial concentration have a positive relationship,indicating that with the increase of spatial concentration,economic disparity may also enlarge.(2) With the Moran scatter plot,the heterogeneity takes on two distinct spatial regimes.One corresponds to the HH scheme including Beijing,Tianjin and Tangshan mainly,the other to the LL scheme including mostly Chengde,Zhang Jiakou and Baoding municipalities,which are surrounding Beijing and Tianjin.(3) The annual average growth rates of per capita GDP also reveal high spatial autocorrelation.Regions with relatively high growth rates are localized close to other regions with relatively high growth rates more than if this localization is purely random,and the statistical significant HH cluster falls mainly into Beijing and Tianjin,indicating that at present,the polarization effect rather than the spread effect of growth pole is dominating.(4) The relationship between initial log per capita GDP and average growth rates is different according to various patterns of local spatial autocorrelation of the initial year.In the HH scheme,initial log per capita GDP and average growth rates have a negative relationship,as an evidence of standard β-convergence models,while for LL scheme,the local autocorrelation type of growth rates of most regions remain falling into the LL scheme,indicating that there are no β-convergence.

According to the requirements of data construction for digital urban management,combined with the actualities of the basic geographic data and the urban management data resources and the demand of basic geographic spatial data for the land and planning departments in Zhengzhou,the working principle and process of photogrammetric mapping,urban parts investigation,geo-coding census were analyzed in the process of digital urban management data construction,and the interaction in its process of task was studied.Digital urban management data construction integrative operation procedure was proposed which includes the process design and integrative implementation of the aerial survey of 1∶500 topographic maps,Digital Orthographic Map(DOM) production,urban parts measurement,geo-code survey and integrated database of data etc according to their links,and during the data process of producing,the interacting procedure was considered.This provides the assurance to the completion of large-scale urban geospatial data,urban management topics data collection and database construction in a relatively short period of time,meets the urgent needs of the basic spatial data for the digital urban management,urban cadastral measurement and so on.The research shows that the integrative data construction is a reference application way for the data constructions,which makes full use of the logic of relations and the possibility of sharing among the data.Using this project,the construction projects ministry of digital city management system in Zhengzhou City completed digital urban management data acquisition and digital database construction in a relatively short period of time,the data utilization efficiency was improved.The integrative data construction not only speeds up the process of data construction and expands the construction data contents,but also ensures the quality of the data at the same time.

All cells under the unpartitioned geo-cellular automata(GeoCA) modeling evolved according the transfer probability which is decided by generic transfer rule fixed for the entire region.However,there always exists spatial heterogeneity in most geographical phenomenon,and these spatial heterogeneities are quite often ignored in generic unpartitioned geo-cellular automata modeling due to the generic transfer rules are employed to drive all cells to evolve.The accuracy of unpartitioned GeoCA simulation result is limited because of the ignorance of spatial heterogeneities in this kind of GeoCA models.To overcome this disadvantage of unpartitioned GeoCA,a partitioned GeoCA based on spatial cluster is discussed in this paper.Under the partitioned GeoCA modeling framework,cellular space is departed into several partitions by spatial cluster,and then a set of transfer rules corresponding to each partition are calculated accordingly.And then,taking Hangzhou City as a case,the Partitioned GeoCA modeling based on spatial cluster is employed to simulate the land use change in the period between 2000 and 2005.In this case study,the study area of Hanzhou is divided into 5 partitions by using K-Means cluster algorithm,and then C5.0 decision tree is employed to obtain cellular transfer rules for each partition of the case study area.In each partition,corresponding cellular transfer rules are employed to drive GeoCA to run simulation.Finally,confusion matrix and Moran I index are calculated to evaluate accuracy of simulated results by using the GeoCA modelling.Results show that the accuracy of the partitioned GeoCA modeling approach owns a higher simulation accuracy compared with those simulated based on unpartitioned GeoCA simulation.On the other hand,when Moran I index is employed to evaluate the accuracy of simulation result,the result of partitioned GeoCA modeling embedded spatial cluster is closer to the real land use pattern.The empirical study shows that a more accurate simulation result can be achieved by using the partitioned GeoCA modeling embedded spatial cluster.

Land degradation and sand storm caused by wind erosion is one of the most serious environmental problems in Farming-Pastoral Ecotone of Northern China,and the determination of wind erosion rate and the evaluation of wind erosion hazard in regional scale are puzzling problems in wind erosion research.The quantitative evaluation of soil wind erosion hazard in Erenhot-Zhangjiakou transect could provide scientific basis for prevention and treatment of wind erosion.Though many factors influence soil wind erosion hazard,among which vegetation cover,average degree of land surface relief,degree of soil dryness and intensity of wind energy were identified as the main factors that control wind erosion.By means of remote sensing inversion and meteorological data,one factor database about the above factors was established with the help of ArcGIS software,and the evaluation model of soil erosion hazard was established by Analytic Hierarchy Process(AHP),then spatial distribution of soil erosion hazard in the study area was obtained.Finally,combined with data of land use in the study area,the background of natural environment and land use was analyzed,upon which comes into view of the spatial pattern of soil erosion hazard.Result shows that area of Severe Hazard in the study area is 1.47×10⁴km²,area of intense hazard is 6.09×10⁴km²,area of Moderate hazard is 3.47×10⁴km²,area of Slight Hazard is 3.45×10⁴km²,area of no hazard is 2.19×10⁴km².The figure of hazard shows "wind erosion hazard increases gradually from southeast to northwest in the study area",which is consistent with the vegetable and climate condition in the study area.The trend is almost the same with previous researches in this field which indicated this approach of combining GIS and AHP is effective for assessing wind erosion hazard.Moreover,the study area is located at the north path of sand storms and is the sources that strengthen sand storms,so,the result could not only be used for preventing and governing soil wind erosion,but also for taking some measures to control the sand storm.

Spatial topological relations are the hotspot for GIS studies.We introduced rough set theory on the basis of overall analysis of area object topological relations in this text.We took area object of RCC model as equivalence class of Pawlak rough set,when not considering boundary fuzziness and boundary itself,the expressing model of those relations is identical with RCC model.We introduced β most contain factor in the variable precision rough set,and took X as x,Y as X,then X can be expressed as β lower approximation for Y in TPP(X,Y) and NTPP(X,Y) model without considering boundary fuzziness and boundary itself,and Y can be expressed as β lower approximation for X in TPPI(X,Y) and NTPPI(X,Y) model.X is certainly belong to β lower approximation for Y and Y is certainly belong to β lower approximation for X in EQ(X,Y) model.X is possibly belong to β lower approximation for Y and Y is possibly belong to β lower approximation for X in EC(X,Y) model.X is certainly not belong to β lower approximation for Y and Y is certainly not belong to β lower approximation for X in DC(X,Y) model.and these relations are particular case for β=0 in the variable precision rough set.There are common intersect parts of two area objects in PO(X,Y) model,via X∩Y≠φ,then we can divide two area objects into β coefficient including and β coefficient separation by the proportion of the two area objects.So,when the common elemental constituent of X and Y is greater than 50% of X(C(X,Y)≤β),We can obviously get conclusion for Y⊇^βX.Therefore,we reach a conclusion that the two area objects of PO(X,Y) have two set expressions Y⊇^βX and Y⊄^βX in the variable precision rough set,and we mapped POI(X,Y) and PO(X,Y) figure of area object intersect relations in the variable precision rough set.

With the increasing applications of natural language in geographical information science,resolution of geospatial information in natural language has become one of the hot issues.Geographical named entities are identifiers of geographical location information in natural language,which include a majority of popular geographical reference systems such as geographical names,addresses,postal codes,telephone numbers and other relative location descriptions.A complete classification scheme of geographical named entities may help implement resolution,storage,management,analysis and sharing of geographical information in natural language.Commonly-used classifications,i.e.classifications of geographical features,classifications of place names,and organization classifications are identified such disadvantages as over specificity of class items,without the consideration of the relationship of time and space,and the ability of representation of partial geographical entities in natural language.To overcome these problems,based on the annotation results of geographical named entities in Chinese documents,we design a classification scheme of geographical named entities(GNEC) with the consideration of their location,attributes,geographical features and temporal features.GNEC includes one main classification of geographical feature types and one subdivision classification of Chinese historical dynasties.Finally,the semantic compatibility between our proposed classification and GB/T 18521-2001,GB/T 13923-2006,Feature Type Classification for Chinese Historical Places of Harvard University and Feature Type Thesaurus of Alexandria Digital Library are analyzed qualitatively and quantitatively.It is noted that a unique geographical entity is usually described with diverse words in natural language,and sometimes it represents different physical location.Classification schemes aim to conceptualize geographical named entities.Undoubtedly,construction of ontologies based on classification schemes could solve this kind of problem(i.e.semantic ambiguity of geographical named entities) effectively.

Cartographic generalization is an important approach for the production and updating of multi-scale maps.Maps with different scales contain different amounts of spatial information.From the view of spatial information transmission,cartographic generalization may be regarded as a process of information transmission from a larger-scale map to a smaller-scale map.Currently,the scholars from the international cartographic association pay more attention to the algorithms of cartographic generalization.However,there is still a lack of quality assessment for the results by different generalization algorithms.For this purpose,the information theory is utilized in this paper to build a comprehensive transmission model for the geometrical information of individual spatial objects from a larger-scale map to a smaller-scale map,where the transmission ratio of geometrical information is utilized to measure the quality of generalization algorithm.Obviously,the description of geometric complexity for a spatial object is very key to further measure its geometrical information.In this paper,three geometrical indictors,i.e.,angle,vertical distance and length,are taken to express the complexity of individual spatial objects in the process of generalization.On the basis of this,the computation method of geometric information for individual spatial objects is developed,where mutual information as a measure is employed to depict the geometrical information.In the process of implementation,the Shannon communication model is at first given based on the information theory and three information measure indicators(i.e.,information entropy,conditional entropy and average mutual information) are described and used to measure map information.The transmission model of map information is further developed according to the process of cartographic generalization after a simple introduction of communication characteristics of cartographic generalization.On this basis,a general computational model is proposed for individual spatial objects(i.e.lines and areas) to track their geometrical information transmission with cartographic generalization,and the quality of transmission is evaluated by using three quantitative indictors,i.e.,information transmission ratio,information loss ratio,and information incrementation ratio.It is shown by a practical example that the proposed transmission model is valid and correct.

Administrative unit based statistic economic and social data present the average status for certain social and economic index,and the inner uniformity is always neglected by many researchers.Based on geographic information technology,the paper introduces one method to calculate the spatial variation of industry structure of China from 1995 to 2005.The first,the second and the third industry spatial data based on 1km×1km cell were calculated by economic data discretization model given by the authors for the year of 1995,2000 and 2005 for China.Then,the spatial variations of industry structure were calculated based on another model in the year from 1995 to 2000,and from 2000 to 2005.The results show that the spatial variation analysis method introduced in the paper is much better than the statistic method in analyzing the industry structure spatial variation.The results calculated in the paper focus on the difference of industry ratio for each spatial unit(1km×1km) of China,and are easier to find the variation of industry structure for any small regions.Such as for the third industry ratio variation,from 2000 to 2005,the region with an increasing third industry ratio is much larger than that with a decreasing third industry ratio,and the region where the third industry ratio increase 0~10% is 68.5% of the total statistic region (whole China),while the region with a decreasing third industry ratio is only 11% of the total statistic region.From 1995 to 2000,however,the region where the third industry ratio decrease is much larger than that where the third industry ratio increase,the region where the third industry ratio decrease 0~10% is 84.7% of the total statistic region (whole China),the region decrease 10%~20% is 8.9% of the total statistic region,and the region where the third industry ratio increase is only 5.8% of the total statistic region.Those results are more valuable and necessary for decision-making in the filed of industry structure,and for regional scientific development evaluations.

Extreme high altitude area is the birthplace of rivers and glaciers.It is quite difficult to accurately extract information of rivers and glaciers from remote sensing images due to the strong undulating topography and arduous field verification.This study chose ASTER image as the data source to extract water body and glacier information in the core area of Mt.Qomolangma(Everest) National Nature Preserve(QNNP) by the method of object-oriented automatic classification.In order to ensure the accuracy of information extraction,the digital elevation model(DEM) and its derived data(slope,aspect),normalized difference vegetation index(NDVI),as well as other indexes that help to distinguish water bodies and glaciers from other land objects,such as Normalized Difference Snow/Ice Index(NDSII) and band math(b1-b3;b3/b4) were stacked respectively as added spectral bands,which contributed to the supplementation of the spectral characteristics of target surface objects.Then,multi-level image segmentations were carried out for each type of water body and glacier to meet their different requirements of segmentation scale.Image segmentation is the key step of object-oriented automatic classification,which requires many times of repeated experiments to find out the most suitable segmentation scale.After the accomplishment of image segmentation,the spectral characteristics,texture features and spatial structure characteristics of each target surface object were comprehensively analyzed.And then,the knowledge-based extraction principles for each type of water body and glacier were established by determining the thresholds of each index according to their histogram features.The classification accuracy was assessed based on field survey results.According to the accuracy assessment,the total accuracy of image classification is 95.14% which satisfies the accuracy requirement of object-oriented automatic classification.This study improved the model for automatic and quick extraction of water bodies and glaciers from ASTER images in extreme high altitude area.

Basic education includes early childhood education,primary education,general secondary education in China.As Beijing's rapid economic development,population growth and great improvement of people's living standards,the need for education quality and teaching resources become increasingly strong and the impartiality and balanced development of basic education have attracted more and more attention.Although balanced development of basic education rely more on policy of educational authorities and the policy-making should be done on the basis of the actual situation of development of basic education,but the public need to know how the basic education is now first,so we need an Evaluation System to do this job.Researchers had set up some evaluation systems by using statistical methods,but these index systems are not so good,because its results always were just numbers or text which made decision-makers difficult to grasp the whole.The GIS technology,with its powerful ability of analysis,management and display spatial data,can be a good solution to this problem.So this paper used GIS technology and statistical methods to set up evaluation system base on the actual situation of Beijing's basic education which includes the education resource index and the education quality index.And the other 10 elements constitute the Secondary Index(including students' average budget expenditure,students' average floor area,students' average book number,students' average education assets,senior teacher ratio,qualified teacher ratio,student promotion rate,graduation rates,outstanding students ratio,eligible students ratio).As for the method,we use Difference Coefficient.And then we use ArcEngine components and.Net technology to programming implement the index system,so we can evaluate the balance situation of basic education of Beijing based on basic geographic information database and basic education resource database,and provide decision support for Education Management Department by showing decision makers the evaluation result in form of vivid charts and thematic maps.

Applying grid into spatial information area forms the Spatial Information Grid.Spatial Information Grid can aggregate kinds of spatial information resources,perform integrative management,support resource share and provide "on-demand services"."Service on-demand" means to provide diverse services according to the requirements of users,that's to discover spatial information resources or their combinations that satisfy the users' requests,then evaluate them and choose the best ones following some goals to provide service.However,the dynamism of grid environment makes it difficult for users to get current information of the distributed spatial information resources,and reduce the ability to provide reliable "on-demand service".Directly applying the traditional spatial information services into grid cannot overcome this issue.To improve this situation,the plan to design spatial grid services and their management services based on WSRF(Web Service Resource Framework),which is "a generic and open framework for modeling and accessing stateful resources using Web services" and support the separation of states and services is given.The plan is composed of two parts,namely spatial information grid services and the management service.The spatial information grid services are deployed at the nodes that contribute spatial information resources by providing standard accessing services and at the same time organize and publish the state of the resources to be accessed or subscribed following WSRF.The key technologies of how to design the services and the WS-Resource are discussed.The management service is deployed at center node which collects,maintains and manages the information of services and resource state of child nodes and provides access interface to support the resource selection which is necessary for "on-demand service".At last a spatial data grid implementing the plan is described.The spatial data grid contains three kinds of services,they are spatial data grid service,spatial computing grid service and management service.In spatial data grid service,the interfaces are organized complying WFS and four kinds of WS-Resource(service,request,loading and spatial data) are designed.Like spatial data grid service,the interfaces are organized complying WPS and four kinds of WS-Resource(service,request,loading and spatial algorithms) in the spatial computing grid service.The management service is made up of three parts: service register management module,sources and services information management module and public thematic service.

e-Science is about collaborations in areas of scientific research.It will change the way that scientific research is undertaken due to its fundamental goal of minimizing the distance between scientists and information technologies.At the national level,more emphasis are put on researches and constructions of grid and other advanced infrastructure,but scientists individually are more concerned with practical tools and methods that can be integrated on personal computers.The coastal zone is a very special area on the earth where the environmental and ecological processes are very different from that in inland or deep-sea,therefore the coastal zone researches have an especially strong demand for information technology.In past years,great progresses have been made by researchers at home and abroad with regard to information systems and coastal zone management systems.However,the emergence and the fast development of e-Science have not only provided a good opportunity but also pose a very serious challenge for us.It is necessary to develop the e-Science platform in order to meet the increasing needs of distributed and synergetic research activities for the coastal zone environment and ecosystem studies.Therefore,in this paper,the concept of 'e-Coastal Science' was put forward.And the principle,framework,technique route and essential techniques for the e-Coastal Science platform were talked about.In detail,a framework that consist three macro-layers,five technique-units and three research-units was designed for the e-Coastal Science platform.The three macro-layers were IT Infrastructure,GeoScience Platform and Coastal Science Platform respectively.The five technique-units were hardware platform,database platform,simulating platform,synergetic & communicating platform and data information release platform respectively.The three research-units were inshore sea area environment monitoring and simulation,hydrology and ecosystem simulation of coastal river basin,and the integrated coastal zone management respectively.To achieve the e-Coastal Science platform,a detailed and feasible technique roadmap was designed mainly based on Duckling 1.2,ArcGIS 9.3 and professional modeling tools.

Hand-foot-mouth Disease(HFMD) is a common infectious disease which causes by a variety of enteric viruses.In recent years,outbreaks of HFMD have been repeatedly and the diseased population has obviously increased in China,these caused our government and society's widespread attention.For the present,most of the HFMD studies focus on medical field,and research on spatial and temporal distribution characteristics at the macro-scale is rare.Taking into account of this situation,this paper based on medical geography theories,with geographic information systems(GIS) as the analysis tool,studied HFMD cases(2008-2009) from the China Disease Prevention and Control Information System.The research process is as follows: firstly,edits statistical data,and calculates the number of reported cases,the average morbidity,the mortality rate and so on in unit time by province;and then inducts it into database and associates with the Administrative Map of China in shapefile format in order for spatial visualization and spatial analysis.The research obtains following conclusions:(1)the serious area of HFMD epidemic situation in 2008 concentrated in three regions: the three northern provinces/city(Hebei,Beijing and Shanxi),the three southern provinces/autonomous region(Guangdong,Guangxi and Hainan)as well as the eastern provinces/city(Zhejiang,Shanghai and Anhui);(2) the number of reported cases,morbidity and mortality have remarkable differences between various provinces: the number of reported cases concentrated in eastern and southern of China,and the Qinghai-Tibet Region and the Northwestern Region are few;(3) HFMD in China is ahead of time for one month compared with conventional epidemic peak(April to July);(4) in most provinces,the number of reported cases increased sharply in April,and started to reduce significantly in August;(5) the distribution of high incidence moves obviously to the north between May and June.The results help to provide a scientific basis for prevention and control of HFMD outbreak and epidemic,also help for further studies focusing on the correlation of HFMD and geographical factors.

Topological spatial relationship is the basis of GIS research issue,and is of great importance to spatial data storage and expression,spatial query and analysis and practical application.Spatial cognition and semantics of topological relations,spatial expression of topological relations,topological relations reasoning,and spatio-temporal topological relations are introduced as several aspects of research.Especially,formal expression of spatial topological relations has been a hot topic in GIS,among which RCC(Region Connection Calculus),Intersection Model,2D-String Model and so on are well known.Some recent studies,focusing around the complex object,expression of topological relations between uncertain objects,are mainly discussed.Since current models can hardly represent more complex point,line and region,they must be improved or extended.For uncertainty and ambiguity are some of natural characteristics of geographic objects,it is essential to describe topological relationships between uncertain objects.In addition,some recent research literatures also address the topological relationship between three-dimensional objects,moving objects,etc.Summing up the above,the representation of topological relationships is a complex issue,which related to geographic,GIS,computer graphics and other disciplines of knowledge.Although the studies of topological spatial relations have made progress,there are still some problems.It is to be enhanced in future studies that the more practically topological relation models are applied to solve problems in real works.

In the past two decades,along with the fast development of computer technology,high performance computing was applied to lots of scientific and engineering fields and resulted in significant breakthroughs.For example,successful decode of DNA in biology,oil exploration in geography etc.,all of which could not proceed so well without the help of high performance computing.Tourism businesses are developing very fast in China,more and more new tourist attractions are coming out in many places.In the meanwhile,it accumulated a large amount of data these several years.How to efficiently analyze and correctly understand these data becomes an interesting topic.When people travel,they often need to search for the optimal combination of sight spots and the corresponding best travel tour,the shortest driving directions in detail between any two points,the best bus path between any two points in a city,and so on.Now people can get from travel agency or search online in the web for tourism route,but these routes are usually not optimal and made from experiences.There are also some services for people to search for the bus path in a city,but this is often limited,and does not work for any two points in a city.For the shortest driving directions between any two points,we still do not have this kind of services in the range of whole country,though people in U.S.are used to searching online for driving directions before going out..In this paper,we try to solve all these problems.We propose several algorithms to design the optimal combination of sight spots by filtering tourist attractions,and the main method is by listing all maximal cliques.We then compute the optimal travel route on this combination.This will then help people to choose their attraction spots to visit and get the best tour.For other problems which people often encounter in travelling such as how to get the shortest driving directions between any two points,and optimal bus path when using public transportations,we discuss the existing algorithms and propose the practical solutions.All algorithms proposed in this paper are suitable for large scale settings such as applying to millions of tourism sight spots to list all optimal tours etc.The key point is to deal with very large amount of data,for which we must adopt high performance computing.In the end of the paper,computational results are discussed and concluded that our algorithms and designs are suitable for real applications,this will make it much easier for peoples' life.

Three-dimensional visualization has become a GIS development trend,and it will inevitably lead to an explosion of spatial data,hence the study on storage and management of massive three-dimensional geospatial data is of great significance.The C/S framework,consisting of database of massive three-dimensional geospatial data for server and a large number of normal PC for clients,has become a basic idea to solve the problem.It is deeply discussed and researched how to realize massive three-dimensional data server related technologies in this paper.Based on analysis of the data dispatch of GIS application thoroughly,a protocol of application layer for high-performance network transmission is designed.The layered strategy that based on the ACE proactor framework and thread pool technology was designed to build a transmission framework for massive three-dimensional geospatial data.The I/O strategy to handle the large-customer communication and the concurrent architecture based on it are discussed.Designed server with modular ideas increases the expansivity greatly.The architecture of the object oriented framework is described in this paper,which supports multiple server policy configurations.Because it is a dynamically ordered structure,it can systematically order these strategies and can statically and dynamically change their behaviors.The framework has solved the transmission problem of multi-level cache and concurrent access problem and realized a highly efficient and stable transmission of massive three-dimensional geospatial data.

Family tree,official history and chorography jointly build the entire history datum in China.The family tree,also called pedigree,is much informative in society,ethnology,custom,geography,population,economy,people and religion.It is significant to carry the data mining and regularization work on family tree for reconstructing spatial-temporal pedigree,reproducing the civilization skeleton of China and developing the national culture as well as spirit.However,due to the lack of technology,the traditional record and expression pattern can not reflect the movement,rise and fall of the family in spatial and temporal space accurately.So Family Tree GIS conception was proposed and it is combined with GIS as its technology supporting.The spatial relationships underlying family tree are imposed and the reconstruction,integration and recurrence of the family tree is its core study content.In detail,Family tree data model and the spatio-temporal supporting framework are designed according to the characteristics of traditional family tree.And then,some key technologies used in building family-tree GIS are discussed especially such as SOA based on Web Service,XML and XML Schema technology,distributed GIS platform oriented the public and the Web 2.0,also technology of management and drawing mass vector data,technique of generation SVG based on DOT,asynchronous network calling and interaction based on AJAX,XML are discussed.Thirdly,the framework of family tree GIS was proposed and the prototype system was introduced in this paper.Finally,taking a family-tree as a study case,building,analysis and representation of the family-tree is discussed.The experiments have shown that the family tree GIS we proposed is a good approach for the study of conventional family tree,and it will be helpful for future researches on family tree.

Vector and raster are two basic formats of geospatial data used in GIS.Raster data are more suitable for spatial modeling and spatial analysis than vector data,moreover,rasterization has efficacy of scaling to some extent,and raster data is the major format used in the application of spatial scaling analysis.In most cases vector data of large scales are subject to information security and intellectual property protection.These data can be shared more publicly because their spatial resolution decreases after rasterization.As the technologies of geospatial information collection improved increasingly,raster data have become the most dominant format of data sources in geosciences.However,traditional rasterizing method is a conversion process accompanied with attribute information loss,the main reason of which is "various types existing in one grid".In view of current problems in process of rasterization,taking rasterization of land cover data of China at a scale of 1:250 000 as a case study,this paper puts forward a conversion method based on grid cell,that is on the principle of "one land cover category one grid",and combines POLYGRID command with its {LOOKUP_ TABLE} INFO files in the software environment of ArcGIS,so as to guarantee no information loss in the whole conversion process.Moreover,the study can provide a new insight and a new idea for correlative researchers to deal with and analyze geographical spatial data sets,as well as supply valuable references for the scientific work of spatial scaling analysis.

Because spatial data has the features of spatial position,unstructured,spatial relationship and vast amount etc,memory access performance on spatial data becomes one of the most important indexes for geo-spatial database evaluation.BeyonDB,researched and developed under the support of National "863" Program,is the geo-spatial database system with independent intellectual property rights.Because at present,there is no software supporting performance comparison test of geo-spatial database and in order to objectively assess its performance index,it is necessary to design,research and develop a kind of performance comparison test software of geo-spatial database.Functional objectives of performance comparison test software of geo-spatial database are firstly analyzed in the paper,i.e.to support performance comparison test between BeyonDB and international leading geo-spatial database system,meanwhile,it can simply extend into other geo-spatial database;to support performance test of loading and inquiring of spatial data;to support testing methods of both SQL sentence and test script,and test report should be automatically generated and output based on "Template" method.As to performance comparison test software of geo-spatial database,we adopt object-oriented design philosophy and method,rationally use software design pattern.General framework design of the software and detailed design of main control,function command interface,access interface of spatial data and output of test results etc.are provided in this paper.After completing the design,we put forward methods of realizing functions of the software,development tool,main interface of the software,output interface of test results and other results.At present,the software has been successfully applied to performance comparison test between BeyonDB and Oracle Spatial,BeyonDB and PostGIS,which greatly improves test efficiency and enables test results to be objective and impartial.Finally,future research emphases of the software are defined in the paper.