The matching of line features is the premise and key technology of map conflation, which also plays an important role in change detection, digital map updating and map registration. To improve the accuracy of line feature matching, multiple geometrical characteristics and topological characteristics should be considered, such as distance, length, shape, orientation, node-degree and so on. It may produce errors when the weights of the characteristic factors are determined merely by experience. Therefore, to avoid this problem, a road network matching method considering multiple geometric characteristics based on the artificial neural network is proposed. Length similarity, orientation similarity, shape similarity, distance similarity and topological similarity (which is the node-degree similarity) are the five feature similarities discussed in this paper. They are the neurons of the input layer in BP (Back Propagation) neural network. To implement this method, first of all, samples of the reference layer and the adjustment layer are selected and the values of the five similarities for these samples are calculated. Secondly, the five feature similarities of these samples incorporating the matching rate are serving as the learning modes and are used to train the BP neural network. And the connection weights between the neural network and the threshold values of neurons are adjusted automatically. Next, buffers of the reference layer's arcs are generated. Arcs within the adjustment layer's buffers are defined as the candidate arcs. At last, the five feature similarities between each arc within the reference layer and its candidate arcs are calculated. Then, these similarities are put into the BP neural network to compute the matching rates of these arcs. If the matching rate is greater than 0.8, the relevant two arcs are regarded as an accurate matching; if the matching rate is less than 0.5, the relevant two arcs are considered to be bad matching; otherwise, they are considered to be less accurate matching which needs the interference of manual assistance to determine whether the two arcs match or not. Result shows that the adoption of BP neural network makes the road networks matching more efficient and accurate, and it avoids the assignment of proper weights to different geometrical characteristics at the same time.
Presently, with the emergence of online mapping and Location Based Serveries (LBS), the continuous multi-scale representation of geo-spatial data has become a critical issue in the era of cartography and Geographic Information Science (GIS). This paper proposes a continuous transformation approach for the smooth representation of multi-scale road networks. Firstly, the topological relationships between the multi-scale road networks are reconstructed, and the feature correspondences are identified and classified into three categories (i.e., 1:1 matching features, 1:0 singleton features and M:N complex features) based on a probabilistic relaxation algorithm. For the 1:1 matching features, the corresponding points are identified based on the identical accumulated length ratio calculated from their starting vertex. The middle-scale representation of the 1:1 matching features is interpolated along the straight lines between the corresponding vertexes. For the 1:0 singleton features, considering that some road features are represented as a set of turns having approximately right angles, all the road vertexes containing an approximately right angle are detected and the edges connecting those vertexes are categorized into two characteristic sub-segments regarding as the left and right sides, which are then extended under the continuous direction. Then, the starting and ending nodes of the two characteristic sub-segment sets are selected to split the original road features into a set of divided line segments. Furthermore, the Douglas-Peucker algorithm is introduced to construct a hierarchical tree structure for each of the divided line segment, and all the divided line segments are hierarchically simplified by continuously moving the inner vertexes into the hierarchical tree structure. Moreover, the M:N complex features are visualized in gradually faded colors for a smooth representation. Finally, the continuous transformation results of different matching categories are integrated with respect to a given scale factor t. The experimental results have shown that the proposed approach can achieve a good performance for the continuous representation of the multi-scale urban road networks.
Pearson-Ⅲ distribution is widely applied in various fields such as meteorology, hydrology, agriculture and so on. Pearson-Ⅲ probability density equation has three parameters to be estimated, including the shape parameter of alpha, the scale parameter of beta and the initial parameter of alpha-0. The precision of estimating the three parameters is one of the key focuses when using Pearson-Ⅲ distribution to forecast rainstorm intensity in actual applications. Without considering the effective range, the traditional method used in estimating Pearson-Ⅲ distribution's three parameters may cause alpha-0 to be less than zero, thus making the predicted rainstorm intensity to be negative. This is impossible to happen when predicting the real physical quantities such as the rainstorm intensity, flow velocity, water depth and so on. Strict theoretical derivation has proved in this paper that the results of rainstorm intensity estimation would be nonnegative as long as alpha-0 maintained nonnegative. Based on the above conclusion, an improved algorithm aiming at fixing the drawbacks of the traditional estimation algorithm was proposed in this paper. Firstly, a series of parameters are estimated by iterative calculations based on the Least Square Method. Secondly, the sum of the squared deviations was calculated for each parameter group and each group was sorted by its value for the sum of the squared deviations. On that basis, different alpha-0 values are calculated for each group respectively to exclude groups with negative alpha-0 values. Finally, one group of parameters which has the minimum sum of the squared deviations was selected as the optimal combination for the three parameters of Pearson-Ⅲ. In the end, this algorithm was verified by estimating the rainstorm intensity in Sichuan Province. The result shows that the proposed algorithm can assure the alpha-0 to always be greater than or equal to zero and the value of rainstorm intensity to always be in a reasonable range. Furthermore, it makes more weather stations pass the distribution fit test with a 95% confidence level. In short, the improved algorithm is more reasonable and accurate than the traditional algorithm, and it provides a reference to the preparation of city rainstorm intensity formula.
Understanding the user behavior of science data sharing is a key step to implement effective and accurate service for science data sharing. This study aims to explore the user behavior of science data sharing using spatial data mining and Web usage mining techniques for the National Earth System Science Data Sharing Platform. At the stage of data preprocessing, procedures of user identification, session identification and user location identification were performed. Spatial hotspot analysis was conducted to analyze the user pageviews, sessions, and dataset visits to explore the geographical variance of user behaviors using the Getis-Ord Gi* method. FP-growth was taken to be the algorithm for mining association rules, and was performed for analyzing data visits and data downloads. Data mining results show that: (1) the user distribution of data sharing platform does not show significant correlation with the overall university population distribution in China, but shows a significant positive correlation with the population of research-oriented universities; (2) the hotspot analysis shows that regions of hotspots were clustering in Beijing, Tianjin, and northern Hebei Province for all three perspectives, whereas the cold spots geographically scattered to a greater extent, e.g. the southern coastal provinces, Henan Province, Shandong Province, Sichuan Province, etc.; (3) the association rules mining reveals a number of frequently visited item sets and rules from the valuable user pageviews. The frequently visited item sets for data downloads were well coincided with the frequently visited data. However, no conspicuous rules occurred in data downloads. Results of the spatial hotspot analysis and association rules mining detected the geographical variance of users’ interests in data and discovered the usage patterns for the frequently visited data, which can be used for designing the personalized recommendation. This study provides a method for mining web user behaviors with the combination of Web usage mining and spatial data mining techniques, which can also be applied to the data mining of websites in other fields.
To explore the applicability of the covariance localization method in the Ensemble Transform Kalman Filter (ETKF) scheme, we firstly analyze some difficulties of the covariance localization method applied to the ETKF scheme in theory. In order to solve the current problem, then we develop an approximate covariance localization method for ETKF, which is accomplished through the Schur product on ensemble perturbations, and finally we test the suitability and the effect of the approximate covariance localization method in ETKF by combining the Lorenz - 96 model. This model is often used to do performance evaluation in data assimilation. The results show that the covariance localization method cannot be directly applied to ETKF assimilation, although it can eliminate some spurious correlations in the background error covariance matrix and increase the rank of the background error covariance matrix. Because the effective object of Schur product in the covariance localization method is the background error covariance matrix, but the update equations of the ETKF only contain the ensemble perturbation matrix, excluding the background error covariance matrix. Moreover, the dimensions between the correlation coefficient matrix and the ensemble perturbation matrix are different, so an approximate covariance localization method is developed. By the experiment, it shows that the approximate covariance localization method can be applied in the ETKF, but the approximate Schur product disrupts the dynamic balances of ETKF assimilation system ,which leads to bad assimilation results. The local analysis method is widely used to solve the localization problem in data assimilation systems, so we try to apply it into the ETKF scheme. The results show that the local analysis method can be directly applied to ETKF, it can remove the spurious correlations in background error covariance matrix and obtain better assimilation results. This paper is a theoretical innovation and experimental exploration, it helps the related researchers to do further studies on the localization in the data assimilation.
The frequent occurrence of coal mine accidents has brought great losses to people's lives and threatened the safety of our country and residents. The fractal theory is a newly developed subject in recent years, and it has been turned into a powerful tool to describe the nonlinear problems. In this paper, analyses and discussions on the fractal characteristics of the temporal-spatial distribution and relativity as well as the regularity of the temporal distribution of the national catastrophic coal mine accidents from 2000 to 2014 in China have been calculated using the relevant methods of fractal theory. Firstly, based on previous studies, this paper summarized the coal mine accidents occurred between 2000 and 2014 in China and obtained the time of occurrence, latitude and longitude for each coal mine accident. Secondly, the grid method is used to measure the number of accidents in grids which cover the area of the coal mine accidents at different grid scale levels, including scales for the national wide region, the three dominating physical geographical regions and the arid and humid regions. We firstly numerate the accidents in every grid and calculate the information content for different grid scales. Then, a linear fit for the logarithms of the information content and grid scale is calculated, and as a result the absolute value of the line slope is taken as the fractal dimension. Finally, the Hurst exponent of the time series of the coal mine accidents is obtained by adopting the method of resealed range analysis (R/S Analysis). Results show that the fractal characteristics of the spatial distribution of China's coal mine accidents are objectively existent. But the fractal dimension values of the spatial distribution of the coal mine accidents in different regions are significantly different. Between 2000 and 2014, the number of total accidents and deaths peaked at 2005, and then it decreased gradually indicating that the safety situation of China's coal mine production keeps improving. The Hurst exponent of the time series obtained from the R/S analysis indicates that the fractal characteristics of the time series objectively exist and the decreasing trend is enhanced. Some opinions regarding the applications of fractal theory in coal mine accidents are put forward in the end.
With the continuous advancement of urbanization, the aggravating urban heat island effect and a series of environmental problems caused by it is becoming severe, which hurdles the progress of urbanization gradually. The urban local micro-climate study tentatively combines the burgeoning 3D modeling with micro-climate simulation, so as to help the city managers to find an effective way to improve the urban micro-climate quality through adopting the simulation data and the model generated in the study. By taking the Zhujiang New Town Western District as an example, this study builds a 3D model of the city using the CGA rules, summarizes the related sub-models of micro-climate simulation, and runs a simulation of the study area with the weather factors such as temperature, humidity, wind speed and wind direction by using the local micro-climate simulation software. At last, it integrates the results and produces a visual presentation. The micro-climate simulation result is combined with the 3D model of the study area. It gives a visual analysis on the impacts of construction height, density and layout to the local micro-climate weather distribution, and gives the corresponding development advices and planning strategy. This method has overcome the limits of previous micro-climate studies which were constrained to field measurement and depended on previous climate data. The modeling approach gives a detailed presentation of the city construction layout and structure, and integrates ENVI-met into the micro-climate simulation, thus the urban micro-climate study can be improved to incorporate the trend of integrating 3D modeling. It provides a new way of thinking and a new assistant decision-making tool for the urban planning and micro-climate simulation studies in different regions. In terms of the relationship between the structure and the layout of urban construction and urban micro-climate, it presents a more visualized result.
Based on the multi-temporal MODIS NDVI and meteorological data for the period from 2001 to 2013, we estimated the vegetation coverage in the Lake Dongting Basin. At the same time, we explored the spatial-temporal variations of vegetation coverage and the determinants driving the observed variations through dimidiate pixel principle, linear regression analysis, correlation analysis and significance testing. The results show that: (1) generally, the ground surface of the whole basin was well-covered with vegetation, i.e. the vegetation coverage for 88.63% of the total area was categorized as high and moderately high cover. Also, the vegetation coverage presented a declining trend from the west to the east inside our study area. (2) During the last 13 years, the vegetation coverage decreased slightly with an average annual speed of 0.03%. (3) The variations of vegetation coverage in the northern and eastern areas are relatively stronger than the rest of the Lake Dongting Basin. Significant to highly significant changes in the vegetation coverage was observed to be approximately 5.3% and 4.29% of the study area, respectively. Geographically, these changes mainly occurred in the vicinity of water bodies (rivers and lakes) and densely populated areas. In contrast, the rest parts of the basin, which were mainly covered by perennial vegetation, did not show any recognizable change. Taken all information together, it is suggested that both the climate variations and human activities are the major driving factors behind the spatial-temporal patterns/variations for the vegetation coverage in the study area. It is also noteworthy that the former factor tends to exert a profound impact on the vegetation coverage over large spatial and temporal scales. Meanwhile, the latter could change the vegetation coverage in disseminated areas (e.g. cities and river) within a relative short time period.
This paper takes Chang-Zhu-Tan Urban Agglomeration as study area. Based on the vegetation index and land use data interpreted from Landsat TM images and combined with the population, economic and climate data, a framework model was established. The results show that: (1) the mean values of wetland ecological security index in 2000, 2005 and 2010 for Chang-Zhu-Tan Urban Agglomeration were 0.7268, 0.7151 and 0.7196, respectively. The status of regional wetland ecological security was good, and the ecological security degree was relatively safe. In the recent decade, the overall performance of the regional ecological security index had decreased, and the decrease of the corresponding area was 21577 km2, which accounted for 22.28% of the total land area; (2) in this study area, there is obvious difference of regional ecological security, that the first-class wetland ecological safety area mainly distributed in the surrounding regions of Dongting Lake, the second-class ecological safety region distributed along the major rivers, the third-class ecological safety area mainly distributed in the border area of two or three cities, and the fourth-class wetland ecological safety area mainly distributed in Yueyang city, Xiangtan city, Changsha city and Hengyang city. (3) During the study period, the area of wetland in Chang-Zhu-Tan Urban Agglomeration has changed obviously, that the total area of wetland has decreased year after year. The fractal dimension of forest swamp, herbaceous swamp, lake, river and paddy field showed an increasing trend. The fragmentation index of ponds/rivers was significantly higher than that of other types of landscape, and the value of wetland ecosystem services was decreasing in general. Finally, the main factors affecting the ecological security of wetland were analyzed from the aspects of land use change and transition, and wetland landscape structure and function.
Citizens′ intra-city trips are often influenced by the allocation of resources and urban functional areas, such as the educational areas, entertainment areas, business areas and residential areas. Therefore, citizens′ travelling pattern can reflect the city structure and unveil the urban function zoning. Meanwhile, the widespread of GPS vehicle navigation equipment makes it possible to achieve a vast amount of vehicle trajectory. With the support of the vast vehicle trajectory data, we can analyze citizens′ travelling mode and understand the city structure. In this paper, we investigated citizens′ travelling pattern and the urban functional structure of Beijing with the taxi trajectory data of one-month period and the information of land parcels divided by major roads. To analyze the citizen′s travelling mode, we extracted the trip volume time series in every parcel and adopted a new method which could cover the proximity on both the values and the behavior to cluster the time series data. In the end, we discussed the correlation between citizens′ travelling mode and urban functions in different regions, based on Beijing′s POI data. The result showed that there were obvious differences in the travelling patterns between the weekdays and weekends. During the weekdays, there were two rush hours, which were different from the ordinary commute rush hours. Looking at the clustering results of the weekday data, the spatial distribution of different clusters basically arranged like concentric circles, and the travelling volume of every circle decreased with respect to the increasing distance to its center. The conclusions made in this research are meaningful for the analysis of citizens′ travelling mode and for assisting urban transportation planning.
The noise removal of sea level anomaly (SLA) data set is crucial and important for extracting the mesoscale eddies. There are many filtering methods that have been developed for eliminating the noises in the sea level anomaly data set before extracting the mesoscale eddies. Nowadays, there are two mainstream approaches of noise removal, which are the convolution filtering and the information filtering. However, these filtering methods have some disadvantages that they could not recognize the right signal from the wrong signals. Therefore, some of the wrong or negligible signals are also taken into account by these noises removal methods. For solving this problem, an envelope surface-based denoising algorithm of sea level anomaly data is proposed before extracting the mesoscale eddies. The envelope surface-based denoising algorithm could improve the effect of noise removal by using the information stability and completeness in the separated layers. This algorithm overcomes the insufficiency of the convolution filtering method that it could not distinguish the wrong signal from the right ones. The detailed process of the envelope surface-based denoising algorithm includes three steps. First of all, the upper and lower envelope structures are used on the original data sets which have been initialized for extracting the subfields of sea level anomaly. Secondly, according to the robustness inside and among several subfields, the envelope surface-based denoising algorithm decomposes the original sea level anomaly fields into several subfields. The subfields could represent the information of the original sea level anomaly from different layers. Then the ranges and standard deviations of these subfields are adopted to recombine the information of several subfields set for shaping an information field after the noise removal. In the end, based on the information field after noise removal, the mesoscale eddies in the northwestern Pacific (22°N-50°N, 130°E-150°W) are extracted by applying the Winding-Angle (WA) algorithm. The results are compared with the mesoscale eddies extracted by the universal Kriging algorithm. From the results of this case study in the northwestern Pacific, we proved the veracity and efficiency of the envelope surface-based denoising algorithm. The veracity of the extracted mesoscale eddies could reach 91.23% in total. In contrary to the universal Kriging algorithm, the veracity of the extracted mesoscale eddies is greatly enhanced.
Three-dimensional (3D) model data of overpasses is significant for traffic navigation, landscape design, and many other applications. In this study, we explore the potential of using airborne light detection and ranging (LiDAR) data for the 3D reconstruction of large multilayer overpasses. To reduce the technical difficulty of this 3D reconstruction process, we propose a concept of “structure unit”. The “structure unit” represents a contiguous object with a consistent width, but does not include the bifurcation and/or intersection structures. A new technical framework, based on the structure units, is proposed to reconstruct the 3D models of large multi-layer overpass using the airborne LiDAR data. First, the overpass points are extracted from the raw LiDAR data by using a Reversed Iterative Mathematic Morphological (RIMM) method and inputting the area of overpass. Then, the hierarchal segmentation strategy, including the connectivity-based segmentation and the determination of structure units, is used to determinate the structure units from the overpass points. The central line of each structure is derived by the binarization and vectorization operations. And the obscured structures are detected and restored based on the central lines of the overpass. Finally, the complete 3D model of the overpass can be obtained by using the complete central line and the corresponding width value. Experiments were carried out to evaluate the validities of the proposed method on two different overpasses. The completeness rates of the 3D models of overpasses A and B are 92.77% and 94.58%, respectively. And the correctness rates of the 3D models of overpasses A and B are 98.84% and 98.97%, respectively. The experimental results indicate that the proposed method can provide satisfactory 3D models for large complex overpasses, and is capable to restore the occluded structures with high quality result.
Since the 1990s, a large number of trees in the Gudao forest area have been threatened by both the natural and human effects, which caused huge losses to the forestry development of the Yellow River Delta. Therefore, the more detailed and timely information of forests is required for improving the local forest management and ecosystem monitoring. Individual tree crown delineation has been demonstrated to be an efficient way to carry out forestry inventories at a fine scale, as the tree crown serves as the basic unit for biomass measurement, species recognition, or health monitoring, etc. The increasing availability and affordability of high spatial resolution remote sensing images provides great potential for accurately characterizing the tree crown. Current tree crown extraction methods based on the imagery were easily interfered by the background objects, especially when there were mixtures of trees, shrubs and grass. A conventional watershed segmentation approach could cause over-segmentation of the imagery because the branches in the imagery are similar to some small tree crowns in shape and scale. To overcome such issues, a new tree crown extraction method based on the object-oriented extraction and the marker-controlled watershed segmentation with the morphological filtering was presented in this paper to extract the individual tree crown from high spatial resolution image. Firstly, the tree crown area was separated from the background using the object-oriented method. After that, the morphological opening and closing operators were employed to smooth the image and remove the noises from the image. Then, we located the tree tops by detecting the local maximum. These tree tops were recognized as markers for the subsequent segmentation. Finally, the tree crowns were delineated through the marker-controlled watershed method. Based on the QuickBird image, we applied this method to the deciduous tree crown delineation of Gudao forest area in the Yellow River Delta. Both the sparse forest area and the dense forest area are tested in order to verify the universality and applicability. To evaluate the method proposed in this paper, we conducted a visual interpretation based on the original images of two plots, and the resulted manually delineated crown maps were then regarded as the reference maps. The delineation accuracies were assessed by comparing the spatial relationships of the automated delineated crowns and the reference crowns. The results showed that the overall extraction accuracy is 87.8% for the sparse forest area, and 65.5% for the dense forest area, which indicated that the tree crown extraction method proposed in this study is effective and promising.
The accurate and quick acquisition of building height information has great significance to urban planning management and ecological environment assessment. In this paper, Geoeye-1 remotely sensed stereo images combined with the Google Earth data are used to estimate the building heights in Nanjing's central urban area by methods of single scene image and stereo images respectively, which are further validated using the field measured building heights. Firstly, after extracting the building shadow information by threshold segmentation and morphological optimization method, the building height can be calculated according to the geometric relationships between the sun, the satellite and the building. Secondly, the rational function model of stereo images is established based on the mathematical relationship between the three dimensional coordinates of the ground points and the coordinates of their corresponding image points in the image plane coordinate system, which can be further optimized using the measured ground control points data. The actual building height is then calculated by subtracting the terrain height from the height got from the rational function model. Finally, the measured building height data is used to validate the precision of the calculated results from two different methods. Furthermore, this paper compares the advantages and disadvantages of these two methods and analyzes their application precisions with respect to different building types. Results show that the method based on the stereo images is more accurate with a root mean square error within 2.8 m. The method of calculating building height through stereo images proves to be more practical because it obtains the building height information more quickly and accurately. The method of extracting building height using the shadow information from single image is suitable for tall buildings without shelter and keeping a long distance from the nearby buildings, while the method from stereo images is not affected by the surrounding buildings and is more capable for the condition of intensive building distribution with uniform heights.
Bays that are similar in their geographical conditions may experience different land use changes due to the distinct levels of development. In this paper, Shenzhen Bay in China and Danang Bay in Vietnam were chosen as the study regions. Based on the land use data in 1987 and 2015, which is interpreted from Landsat image, a dynamic degree model and a comprehensive index was used to extract the information of land exploration and utilization in Shenzhen Bay and Danang Bay. Finally, the level of economic development was estimated by analyzing the land use changes. The whole research includes four parts that are explained as follows. (1) From 1987 to 2015, all of the arable land and 51% of the aquaculture land was turned into the urban land in Shenzhen Bay region. However, in Danang Bay region, 86% of the arable land was changed into the urban land and the aquaculture land had increased with an annual rate of 0.55 square kilometers, which is opposite to Shenzhen Bay. (2) The land area transformed from water in Shenzhen Bay is 32 square kilometer, being much greater than Danang Bay, which is 5 square kilometers, explaining that the scale of sea reclamation in Shenzhen Bay is far more severe than Danang Bay. (3) The dynamic degree of urban land in Danang Bay is higher than Shenzhen Bay. Meanwhile, the land utilization comprehensive index model claims that the land utilization degree in Shenzhen Bay is much higher, which reveals that the level of development in Shenzhen Bay is obviously higher than Danang Bay. Analysis on all land use changes shows that Danang Bay is in the middle of industrialization, while Shenzhen Bay is in the late industrialization. (4) The area of bare land had increased by 3.5 times in Danang Bay, while there was no increase in Shenzhen Bay. Furthermore, the area of mangroves had increased by 2.1 times in Shenzhen Bay. All of these conclusions indicate that more attention should be paid to the growth of natural functional vegetation and the protection of ecological environment in Shenzhen Bay than in Danang Bay.
