With the development of Earth Observation (EO) technology and the increasing geospatial sensor resources, it is necessary to efficiently manage geospatial sensor resources in Geospatial Sensor Web environments which facilitate users to plan and use these resources comprehensively, accurately and in demand. Based on the analysis of (1) features of EO emergency event and (2) demands of sensor emergency observation planning, this study aims to achieve the near real-time, accurate and comprehensive discovery of existing heterogeneous EO sensor resources. We propose an integrated management method of EO sensor resources oriented to emergency observation planning. The research contents include how to organize and build the index of EO sensors’ information, how to calculate the observation area covered for space-based sensors, and how to design the inquiry of metadata for collaborative emergency observation. Based on the above mentioned workflows that included in the proposed integrated management method, this study designed and produced an integrated management tool which is entitled GeoSensorQuery for emergency observation planning, via the utilization of Java programming language and the Open Geospatial Consortium (OGC) Sensor Web Enablement (SWE) information standard model-Sensor Model Language (SensorML). Finally, an application focusing on planning the flooding observation sensors occurred in the middle reaches of Yangtze River basin of China has been exemplified, in which the designed GeoSensorQuery prototype can meet the requirements of near real-time, accurate and comprehensive discovery of massive and heterogeneous EO sensor resources. As a conclusion, the selected example has verified the feasibility of our method.
Dual spatial clustering is an exploratory data analysis that deals with spatial contiguity and attributive similarity. Conventional spatial clustering methods cannot perform effective clustering in spatial and attribute domains simultaneously. This study employs SOFM (Self-Organizing Feature Mapping) to solve dual spatial clustering problems, and then verify the proposed method in the analysis of urban expansion structure. By modifying the algorithm of best matching neuron searching in SOFM, we manage to perform clustering in both spatial and attribute domains. The algorithm includes two independent self-organizing clustering processes. The first one includes a spatial constraint, and the other one includes an attribute constraint. The final result is generated by merging the corresponding two results that derived separately from the two processes. The analysis of the structure of urban expansion of Wuhan city is used as a case study. We feed the proposed model with the location information and the expansion degree information of newly grown urban patches, and the generated dual clustering results could clearly illustrate the spatial structure of urban expansion. As a conclusion, the self-organizing dual spatial clustering method can generate spatial continuous and attributive similar clusters with little artificial interference.
Flood is a frequently occurred disaster and it causes great harm in China. In order to effectively prevent, reduce and relief disaster, it is necessary to establish a flood event information model to represent and share flood information. By far, many scholars have studied a variety of event information models. However, the existed event information models mainly represent static information, and they are lack of the capability to handle dynamic process information. To solve this problem, a flood event model is proposed in this paper based on the flood emergence management stages and the dynamic process from the perspective of observation. To build up the model, this paper firstly describes the modelling considerations, including flood phases, observation, and ten types of primary elements for modeling. Then, we construct the flood event model as a ten tuple model based on MOF modelling framework with four layers, and encode the model with a method that maps the elements of the model to elements of the Event Pattern Markup Language (EML). Finally, we simulated an experiment for a flood case occurred in the Liangzi Lake in 2010 as an example to test the proposed model. In the experiment, the building process and the results of the proposed model at different emergence management stages are detailed. The experiment results show that the proposed model has following characteristics: (1) it establishes a model from the perspective of observation, since observation plays a decisive role in real-time disaster information acquisition. (2) The flood model was established by four emergence management stages, and the main tasks of each stage are different. (3) The flood model was modelled dynamically. The proposed ten tuple model itself is not dynamic, but the establishing process and its represented information are dynamic. The modelling process is modelled on real-time data streams, and it dynamically updates time-series data. The time series data in this model are generally the latest observations, which reveal the dynamic information of current flood. (4) The model is extensible and adaptable to ensure the suitability. As a conclusion, the model can represent flooding events and model dynamic flood information effectively, as well as provide a good approach to represent and share flood information.
Gridded population distribution data are increasingly expanding their uses in a wide range of fields, such as resource utilization, disaster response and relief, environment protection, and economic research. The enhancement of resolution with detailed precision is a perpetual topic in population gridding research. In this paper, a 100m gridded population dataset was established for Yan’an city by developing a method that distributed the population based on the land use data with improved settlement information. Data used here include township boundary data, township-level demographic data, land use data, TM image, town-village settlement point data, and DMSP/OLS nighttime light images. Two approaches were used to improve the detailed information of settlement distribution. First, we identify the rural settlement information from land use data and enhance the information by the village point data obtained from Google Earth. Second, we downscale the DMSP/OLS data by spatial interpolation. The sum of light emission, lit area and unlit area under different land use types in each town were counted to be used as the independent variables, and the statistical population of each town were used as the dependent variable. A stepwise regression method was adopted to simulate their relationship. Finally, the sum of night light emission, the unlit area of build-up area, the unlit area of grassland, and the lit area of farmland were put into constructing the ultimate equation. All variables were significant under the level of 0.01 and the coefficient of multiple correlation is 0.872. We estimated the population at township level for selected towns as a validation. Through using the equation, we found that the mean error between the estimation and the statistical population is lower than 5%. The above analysis suggests that the proposed modeling strategy is highly efficient. As a result, we calculate the weight and distribute the population through the equation in the formation of 100m grid, by taking township as the unit. In summary, the gridding method used in this study can obviously improve the output resolution and the distribution details. Also, the expression of the final equation is relative simple. As a conclusion, this paper has its significance in guiding the population gridding research in the county level areas like Yan’an city.
Land-use change has been a key component for global environmental change. The conversion matrix is a fundamental tool in the analysis of land-use change. In order to obtain as much information as possible from the conversion matrix, two new patterns of land-use change, which were named as displacement and substitution, were proposed. Accordingly, the total change that characterized in an individual category is then divided into net change, displacement, and substitution. The calculation formulas for the displacement and substitution were also derived. The land-use change among eight land categories of Jinjiang watershed that locate in southeast China, was taken as an example to illustrate the proposed methods. The results show that great changes have taken place in the area from 1985 to 2006. The garden area, built-up land and unused land increased remarkably, while the dry land and grassland decreased, with the waters kept almost unchanged. The swaps between the above six land categories were relatively few, and the main format of change is in the pattern of displacement. Most of the changes in irrigated land and forestland happened in their spatial location swaps. However, for irrigated land, it was mainly displaced by the forestland or dry land; and for forestland, it was mainly substituted with garden area and grassland. The results also indicate that the new patterns—displacement and substitution, are helpful for the analysis of land-use change.
Virtual plant is a potentially powerful tool for analyzing plant growth processes and the relationship between plant growth and environment. In order to provide an intuitive representation of forest growth and be adaptive to forest management at a deeper level, the forest growth model should be integrated with 3D tree architectural model. Individual tree growth simulation is the foundation of the simulation of forest ecosystem. The aim of this study was to link an empirical forest growth model and a 3D architectural model to dynamically simulate growth processes of an individual tree. Individual-level simulation of Cunninghamia lanceolata was implemented by integrating a 3D parametric model and a distance-independent individual tree growth dynamics model. The 3D parametric model was created in ParaTree, a system developed by our work group. First of all, according to the morphological characteristics of Cunninghamia lanceolata, parametric modeling method is adopted to generate a 3D model. Then the tree height, diameter at breast height, height under branch, and the whorls of branches are predicted in the growth model based on a site quality, planting density and age. Geometric descriptive parameters of 3D tree model are then updated according to the architectural parameters. Finally, the parametric curve is applied to adjust stem diameter and tree diameter, the length of the branches, branching angle and so on to make the model change with tree growth. Taking the forest resources inventory data of Wuyi forest farm of Zhangping in Fujian for an example, we simulate the growth dynamics of Cunninghamia lanceolata and express the global morphological structure characteristics at each growth stage. The growth at average of individual trees in a forest stand was represented intuitively by the means. Linking the parameters of 3D model to some existing tree empirical growth models is conducive to reuse a lot of growth models .
Along with the continuous development of computer technology, internet and wireless communication technology, digital map visualization, as the core of geographic information system, has been widely used. At the same time, with the increasing number of embedded software and hardware, a variety of heterogeneous platforms bring great challenges to the development of embedded geographic information system. Therefore, it is important to study the digital map visualization method for cross-platform purpose. Firstly, after comparing the differences of middleware between the embedded system and desktop system, this paper summarizes the characteristics of the embedded middleware, and analyses the platform differences and resource restrictions for digital map visualization in the embedded environment. Based on the concept of middleware, a cross-platform middleware of digital map visualization with three layers: EMD_MW was constructed. The middleware utilizes the operating system abstraction layer to prevent the heterogeneity from different embedded operating systems, and to abstract the operating system message forwarding and the graphic interfaces associated with the digital map visualization. Meanwhile, the model utilizes the framework layer to dispel the differences among data storage services, to improve the base graphic interfaces, and to provide relevant digital map display interfaces through the function layer. Then, the abstraction layer, service framework and function layer of EMD_WM were designed and realized. Finally, several engineering applications based on EMD_MW were presented. These examples indicate that, the embedded cross-platform middleware model for digital map visualization has a good cross-platform performance, and it can effectively solve the problem of digital map display on different embedded platforms.
Great changes have taken place in China's urban scale system during the last 30 years. The driving forces that cannot be ignored include following factors: the population urbanization propelled by its aggregation, the land urbanization driven by its revenue, and the economic urbanization promoted by GDP. Using the analytical tools of GIS integrating three factors of resident population, built-up area and economic output, we constructed Zipf-PLE model for a in-depth research on the spatial pattern of China's urban scale system. The result shows that: (1) in 2010, China's urban scale system is matured with completed levels of hierarchy, which is reflected by a relatively rational pyramid-shaped pattern with larger scale locates in the middle while smaller scale stays at the bottom. The urban scale system structure in the western region has the highest rationality, which presents a pyramid-shaped pattern with larger scale at the bottom and smaller scale at the top. The urban scale system structure in the eastern regions is relatively sound and contains the largest quantity of medium-sized cities, indicating a pyramid-shaped pattern with larger scale in the middle and smaller scale at both ends. The urban scale system structure in the central regions is featured by its incomplete urban system hierarchy, which lacks megacities especially, and presents a pattern with larger scale in both the middle and the bottom, while having smaller scale at the top. (2) The overall urban scale system in China at the provincial level is rational, with 90.32% of the provinces being moderately rational. Among all 27 provinces, except 4 municipalities, the urban scale system patterns of 8 provinces are inclined to disperse and the other 19 tend to concentrate. (3) Furthermore, we allocate the provincial urban scale systems in China into different categories according to their rationality rankings, which are: the high-level (with 11 provinces and municipalities including Beijing, Shanghai, Tianjin, Chongqing, Xinjiang, Heilongjiang, Guangxi, Gansu, Fujian and Jilin), the relatively high-level (with 12 provinces of Guangdong, Shanxi, Yunnan, Hunan, Guizhou, Liaoning, Jiangxi, Jiangsu, Zhejiang, Sichuan, Hebei and Henan), the medium-level (provinces of Hubei, Shandong and Anhui), the low-level (provinces of Hainan, Mongolia and Ningxia), and the irrational areas (provinces of Qinghai and Tibet).
Runoff nodes, which could be captured from every watershed on earth, are one type of the real geographical entities within the river-network. Generally, runoff nodes often locate at sites where lower grade of runoffs merging themselves into a more advanced trunk stream. It is clear that runoff nodes carry a lot of hydraulic and landform information. Although many reports have documents related to runoff nodes, the quantitative measuring researches, that concerning the geographical information of runoff nodes and taking advantages of valid theories or methods, are still rare. This may prevent people from further understanding the existence of runoff nodes in nature. Aimed to solve the aforementioned issue, this paper employs the Theory of Rough Sets (TRS) to develop a new landscape metric for perfectly characterizing the geographical features of runoff nodes. Firstly, according to current knowledge about runoff nodes, this paper extracts the potential information carried by runoff nodes. Secondly, this paper reviews some typical valid methods on how to extract runoff nodes from Digital Elevation Models (DEM). Thirdly, this paper presents the feature information index of runoff nodes (FIIRN) and illustrates the relevant computing procedures. Based on these researches and preparations, this paper carries out a case study of FIIRN for seven special regions from the Loess Plateau of China (which well represents the typical loess landscapes existed in the world). The experimental materials of DEM data are predominantly characterized with 5 meters of resolution. With the help of ArcGIS 10.1 software, this paper extracts all runoff nodes from the experimental sample regions according to their geometric features in the runoff network. Then, this paper computes the values of FIIRN using Excel 2003 and MATLAB 7.7. Surprisingly, the results show that FIIRN can effectively map the complete geographical information of a given watershed unit. Therefore, as we concluded in this paper, that FIIRN can play a valuable role for mining the information from watershed units in future. In fact, FIIRN should be considered as an optimal parameter for analyzing the morphological patterns of watershed units.
Nowadays, issues regarding to housing affairs can be easily dealt through online queries and transactions, which provide a brand new data source for researches about residential space. This article mainly focuses on the causes of variations in activeness degree for different second-hand housing areas in Nanjing, through web data mining and a series of kernel density estimations. Based on three typical case studies of the Hexi district, Daguang district and Confucius Temple area, this article summarizes the underlying influential mechanism for the second-hand housing activeness in Nanjing. Research result has shown that the second-hand housing in Nanjing has a structure of “one core with three blocks”, in which the Gulou and Jianye districts are considered to be the core. Meanwhile, the sales volume in districts with highly educated inhabitants, such as Xuanwu district which lies in the eastern part of Nanjing and Jiangning district which lies in the southern part, have a low degree of activeness due to the solidified and sophisticated residential space development. There are two main factors contributing to the differentiation of activeness in second-hand housing, which are summarized as the internal factor and the external factor. The multivariate regression analysis model suggests that the internal factor is consists of the housing price, the number of building stories, the percentage of greenery coverage, and the house decoration condition. The external factor is considered to be consist of the accessibility and quality of adjacent subway stations, schools and other supporting facilities and services.
Land change survey is an important survey for investigating national conditions and national strength. The purpose of land change survey are: identify the nationwide land use status and changes in the year of concern; maintain the accuracy and timeliness of national land survey data and the basic information of the comprehensive land and resource supervision platform; and meet the requirements of land and resource management as well as economic and social development. Officially released land change survey results are the basis for the implementation of planning, management, protection and reasonable utilization of land and resources, the strategic planning of national economic and social development, and other relevant special-purpose plans. There still are problems that need to be solved, such as how to ensure the authenticity and accuracy of land change survey results, how to improve the work efficiency of land change survey with new technology, and how to shorten the error modification time. The purpose of this paper is to evaluate the quality of land change survey database and analyze the causes of database errors. For a quantitative analysis of the changes of database quality in 2010, 2011 and 2012, comparative methods were employed. Methods known as Moran's I and local Moran's I were adopted to analyze the spatial dependence in observations of database errors among administrative units, and to reveal the change of spatial distribution and the anomaly characteristics of database errors in local areas. To be specific, five steps are required for quality evaluation and pattern analysis. Firstly, the descriptive chart of original errors at the county level is processed in batch transaction by software, and then is aggregated into database tables. Secondly, since the quality inspection rules were found inconsistent among 2010, 2011 and 2012, it is impossible to use a comparison analysis method, thus we preprocessed the data and acquired the common quality inspection rules for the three-year period. The third step is to use the comparison analysis method to assess the database quality from different aspects, including the defect levels, the main check items of errors and the spatial distribution. The fourth step is to define the database errors by observing through the global autocorrelation method in analyzing the land change survey of Hebei province, and obtaining the spatial distribution characteristics and the influential factors of database errors. Finally, the spatial pattern of the phenomenon was reflected and the causes of database errors were explained. The results showed that the quality of the database of Hebei province in 2012 have been improved significantly according to the number of errors, the defect levels and the spatial distributions in assessment. In addition, the distribution of Hebei’s database errors were high autocorrelated with the fluctuation of aggregation level, while the generation of local anomaly was usually derived from artificial error. It is concluded that throughout the quality analysis of the research database and its space distribution pattern in three years, we can evaluate the changes of database quality objectively and identify the causes of database errors effectively, therefore provide a forecasting and monitoring approach to new land change surveys.
Today, an abundant supply of remote senseing data with various spatial, radiative and spectral resolutions from multi-platforms has provided rich sources of information for scientific research. In order to overcome the limitation of a particular type of remote sensing data in application, and take the full advantage of other remote sensing data, image fusion technique has been frequently used to enhance the resolution of remote sensing data and perform scale transformation among images obtained from different remote sensing platforms. A proper image fusion algorithm can not only improve the refinement of details of a low-resolution multispectral image, but also preserve its spectral information. Moreover, it can utilize the complementary information, reduce the data redundancy, and enhance the interpretation ability of the images. The fusion result is influenced by many factors, such as image fusion algorithm, seasonal difference, registration error, spatial and radiation resolution difference, etc. The images used in this study have different radiation resolutions, including 8 bit, 11 bit and 16 bit. In order to reduce the influence of differences in the radiation resolution and in the resultant data dynamic range on image fusion, this paper proposed a method for the standardization of radiation resolution. Based on the unification of quantization intervals for digital number, the transformation from low to high radiation variability through multiplication of a same proportion coefficient can reduce the loss of image information, which is caused by data bits conversion in the fusion process. The results show that the proposed standardization method of radiation resolution can be applied to remote sensing data with different quantization intervals and is easy to be programmed. When using the proposed standardization method for fusing different images, the resultant fusion results are almost identical to each other, either using real number value or integer value. Whereas, when using other quantization methods, the resultant fusion image with real number value is generally better than that with integer value. The standardization of radiation resolution is a necessary step for image fusion when using reflectance-based images, because all image fusion algorithms will cause a serious spectral distortion to the fusion results.
With the development of high resolution satellites, the attitude jitter, which is a common and complex phenomenon, has become a key problem in the applications of high resolution remote sensing images. It leads to the degradation of image quality, and non-steady and lower precision for the attitude measurement data. Therefore, in this paper, a jitter detection and analysis method based on remotely sensed imagery is proposed. It took the advantage of the imagery data to detect jitter without additional involvement of high-performance sensors or a large quantity of GCPs. Two different jitter detection methods for different types of sensors are presented, which consist of: (1) jitter detection methods based on multispectral images utilizing parallax between two different bands in the image, and (2) jitter detection methods based on stereo images using triplet stereo images and rational polynomial coefficients (RPCs). With the jitter detection methods, the relative registration error is obtained. Afterwards, we can retrieve the time-dependent attitude jitter. Consequently, the frequency and amplitude of the attitude jitter can be achieved by analyzing and modeling the jitter using cosine functions. To reduce the jitter effect on geo-positioning process, the true attitude information should be estimated from the detected jitter components. An improved sensor model is introduced to take into account the compensation for the attitude jitter. The performance of the improved sensor model is assessed in direct georeferencing. Experiments are conducted with images acquired from the first civilian three-line-array stereo mapping satellite ZY-3. A periodic pattern with a frequency of about 0.65Hz has been detected. Based on the theories and methods mentioned above, the detection and analysis of satellite attitude jitter system based on remote sensing imagery has been developed using VC++. An image sub-blocking and a parallel computation scheme based on the computer performance is performed, which can improve the efficiency. Test results show that the system can efficiently detect the attitude jitter of 0.1 pixels. Meanwhile, it can model the regular jitter and correct the jitter effect on geo-positioning process. The accuracies for the plain and height directions are improved from 2.1799 m and 3.2497 m to 1.4396 m and 0.7453 m, respectively. This system has a good stability and robustness and has been successfully applied to the ZY-3 ground application system.
With the rapid development of remote sensing techniques, higher precisions of the vegetation remote sensing are required. Therefore, before using the time-series data, how to select the optimal algorithms to reconstruct it has been a hot research topic. Based on the five main land cover types in Northeast China, the reconstruction quality of three commonly used algorithms that included in TIMESAT tools has been qualitatively analyzed. Then, the fidelity performance and the capability to keep main characteristics of the three algorithms on EVI with respect to different land cover types were compared. The result shows that the S-G algorithm has a better performance in reconstructing the peak and the width of the EVI curves in the growing seasons, but it is prone to keep the noise data due to excessive fittings, especially common in land cover types of steppe and shrub. AG and DL algorithms generally present similar performances and the results are much closer to the true values for land cover types of steppe, shrub and arable land. But AG algorithm is easily influenced by noises for fitting the peak of the cures, which reduces the maximum EVI and causes the decline of vegetation growth. Spatial patterns of the fidelity performance and the capability to keep main characteristics of the three algorithms are all related to the distribution of vegetation types. Finally, we found that AG is a better algorithm to be used for the land cover types of steppe and shrub, DL is better for arable land, while S-G is better for the broadleaved deciduous forest and coniferous deciduous forest.
Land cover refers to the complex of various materials of the earth's surface and its natural attributes and characteristics. Land cover data at global scale is fundamental and indispensable for global change studies, and its data quality has been the focus of a variety of international communities. In 2010, China launched the project of 30 m Global Land Cover remotely sensed products with 10 classifications for the base years of 2000 and 2010, which takes a four year period to complete, and the data quality issue will affect its subsequent applications. This paper mainly studies the accuracy assessment method for regional land cover data, as a preliminary of global validation. First, the spatial characteristics of land cover data are discussed, such as being massive, multi-dimensional, and heterogeneous. Then, a general approach based on spatial stratified sampling method for the accuracy assessment of regional land cover product is proposed to improve the traditional sampling method. Considering the spatial characteristics, this article presents the method of spatial stratified sampling: stratified sampling is conducted according to land cover type, the total sample size is calculated using probability and statistics optimal model, and the sample size of each layer is allocated according to the area ratio; while in the sample spatial allocation process, the representative of sample, which implies that the spatial correlation between samples should be low, needed to be considered. In each layer, each sample's spatial correlation is calculated by Moran's I index, and by setting a threshold value, the representative samples are chosen. The final sample set of each layer is randomly selected based on spatial analysis. The spatial sampling scheme is divided into three parts: sampling method, determination of sample size and the sample allocation. We have made an improvement in this paper, by designing a two-step scheme, which are quantitative estimation of sample size and spatial allocation. A case study of Shaanxi Province of China shows the method and process of accuracy assessment for regional land cover product. 1467 samples are selected by spatial stratified sampling in 7 strata. According to the confusion matrix, the overall accuracy is 79.96% and kappa index is 0.74. User's accuracies of all land cover types are more than 75%, while producer's accuracies fluctuate due to the low sample size caused by the low area ratio, which have little impact on the overall accuracy. Experimental results show that the proposed method is applicable to accuracy assessment of regional land cover product. Study on global land cover product will be performed as a main research direction in future.
DMSP-OLS nighttime data has been widely used for extracting urban built-up area since the 1970s. It has also been used to estimate the economic level, density of population, changes in spatial patterns of urban landscapes, spatial dynamics of fishery, and the relationships between urbanization, human health and energy consumption. Due to the light’s characteristic of overflowing, the range of city extracted by DMSP-OLS nighttime data surpasses its real boundary in urban area extraction. Therefore, an accurate light threshold determination has become a research hotspot in recent years. In this paper, we sum up the features of four commonly used methods from other dissertations. Then, a spatial comparison method based on Google Earth is adopted to determine the optimum value of light threshold for urban built-up area extraction. Because factors such as terrain and traffic have significant impacts on the urban form, and the urban form influences the value of light threshold, the results of threshold for urban built-up area extraction are divided into two categories according to the urban form. The optimum thresholds for crumb city and strip city are 43 and 47 respectively. Compared to other methods, the spatial comparison method based on Google Earth has two advantages in innovation. Firstly, the classification of threshold result is based on the urban form; meanwhile, it not only breaks the limit of statistical data, but also improves the accuracy of urban built-up area extraction. Secondly, the data sources such as DMSP-OLS nighttime data and Google Earth image used in this method are free and easy to acquire. The method picks a series of sampling points around the whole city on Google Earth as the reference data, which displays not only the results of traditional statistics, but also the spatial differences between the real boundary and the urban area extracted by DMSP-OLS nighttime data. In a word, this method is suitable for extracting urban built-up area in regions with complicated terrains and diverse urban forms.
