Kriging interpolation method realizes spatial weighted estimation that meets the unbiasedness and optimality according to the position relationship between the estimated location sites and the known sample sites and regionalized variable spatial correlation. Traditional theoretical model shape is fixed and chosen with subjectivity, which can't reflect the changing trend and multi-scale spatial characteristics. The choice of scale and the treatment of scale effects also need to be considered. To solve the problems above, we propose a method of kriging interpolation optimized by multi-scale least squares support vector machine (LS-SVM), which provides a new idea for fitting experimental variogram. Starting from the changing trend of the actual sample data, least squares support vector machine fits experimental variogram and the results conform to the spatial changing trend of data itself. Secondly, the wavelet kernel as the LS-SVM kernel function, parameters can be adjusted according to different parts of the nuclear, which is flexible and variable. Finally, the multi-scale wavelet kernel using wavelet multi-resolution characteristics, can reflect the different details of spatial changes, to avoid the single scale LS-SVM ignoring the spatial details of the problem. Followed that, the experiment includes simulation and application. Experimental simulation mainly verifies scientific validity and accuracy by the optimized interpolation of multi-scale least squares support vector machine. Meanwhile, experimental application research of PM2.5 concentrations of temporal and spatial distribution provides the theoretical basis for city ecological protection and controlling. Final results show that kriging interpolation algorithm optimized by multi-scale least squares support vector machine is superior to the traditional method and single scale optimized kriging interpolation algorithm. It would be better to depict the variation function and reflect the different scales of spatial changes in details to further improve the accuracy of the interpolation to some extent, which is an optional kriging interpolation method.
Based on the taxi trajectory data from the city of Beijing, this study proposes a multi-scale visualization approach for trajectory OD (Origin-Destination) data. First, we extract OD points from initial trajectory raw data eliminating invalid points. Then, the distribution space of OD data is subdivided by density analysis and administrative unit aggregation. Finally, we define relevant parameters to summarize inherent OD flow pattern and customize their presentation of multi-scale visualization. In the process above, three regionalization results, which correspond to block level, business district level and district level, are obtained by setting different values of the minimal area of the aggregated region. Therefore, representations at three different scales can be outputted. The experimental results confirmed that our method could effectively achieve the reduction of trajectory big data and reveal mobility pattern, which is helpful for future decision making.
In recent years, big data of mobile phones has become a great data source for researches and applications. It has been widely used to understand the human behaviors in cyberspace space. Researching and forecasting the surfing time of mobile phone users have great significance for analyzing mobile phone users’ behaviors and patterns, designing network service, and understanding the relationship of surfing behaviors, website stickiness, users’ psychology, mobile Internet intelligent business. We proposed a mixed Markov method (Lift-Markov method. LM), combining the traditional Markov model and association rule model, to predict the surfing time period of mobile phone users. A dataset of surfing records of 4G mobile phone users collected by Hubei Mobile within twenty days is used to demonstrate the capability of predicting web-surfing time periods of users. LM method has a better prediction accuracy when it is compared with the traditional Markov model and the Most-value model. There are two main findings here: the first one is that there is obvious periodicity in surfing time periods of 37.66% mobile phone users in experimental area by Fourier transformation and periodic tests, which could help us understand the surfing characteristics of users. Also, the second one is that the average accuracy of our proposed method is better than the Markov model and the Most-value model in 10 minutes, 20 minutes, 30 minutes, 40 minutes, 50 minutes and 60 minutes intervals. LM method can perform an average accuracy of 79.75% in predicting web-surfing time on a scale of 60 minutes, better than the Markov model (74.64%) and the Most-value model (64.44%). Compared with the other two models, the accuracy distribution of the LM method is narrower, the peak value is higher, and the standard deviation is smaller, which means that the prediction accuracy of the LM method is more concentrated and stable, with good predictive performance.
The optimal allocation of land use is an important and effective measures of promoting the sustainable development of the land. However, existing research was lack of efficient methods in the optimization allocation for the quantitative structure and spatial layout of land use by using original mixed algorithm. Therefore, this paper combined the ant colony optimization algorithm (ACO) with the chaos model, and proposed a hybrid and self-adapt chaos ant colony optimization algorithm (CACO). After that, in order to verify the feasibility and efficiency of the CACO, the Zengcheng district of Guangzhou was selected as the study case. The CACO was utilized to solve the model of land use optimization allocation based on the evaluation of actual land sustainable use. Finally, this study made some comparative analysis of the results of the CACO and the actual land use and the results of the ACO respectively in three main aspects: the quantitative structure of land use, the spatial layout of land use and the multiple objective functions. The results showed that: firstly, the CACO can effectively solve the complex problems of multi-objective land use optimization allocation under multiple constraint conditions; secondly, compared with the ACO, the coordination between economic benefit and ecological effectiveness in the CACO was weakened slightly. The CACO rose all others objective functions’ values. For example, economic benefits increased by 7.18 billion yuan, ecological effectiveness increased by 0.33 billion yuan, social benefits increased by 1.13%, while the land conversion costs shrank by 1.15%. Thirdly, compared with the ACO, the CACO decreased the diversity and evenness index of actual land use spatial distribution within 1.30%, made the number of total land patches reduced about 8.86%, and the average patches size increased 9.77%. The level of the land intensive use was improved. Therefore, the CACO could reasonably optimize actual various land use types to appropriate spatial layout, and supply useful technical support for scientific land use planning and decisions making.
Hand, foot and mouth disease (HFMD) is a common infectious disease. Previous studies showed that multiple factors, such as meteorological, geographical, environmental and socio-economic factors were associated with HFMD. The associations between these risk factors and disease are complex. HFMD incidences present strong spatial clustering and auto-correlation. It is difficult to capture such complex non-linear associations and spatial auto-correlation using ordinary linear regression. Based on the previous studies, we proposed a Bayesian network based integrated risk approach to explore the relationship between HFMD incidence risk and the influential factors, such as meteorological parameters, land-use pattern, socio-economic status and air pollution. HFMD is a typical disease of children in Shandong Province of China and it was taken as our study case. Our approach incorporated the output of spatial clusters obtained by scanning statistics to enhance spatial reasoning of the proposed Bayesian network. This could also reduce the bias and improved the performance of the prediction. The results showed that the integrated Bayesian network model proposed achieved higher accuracy than the other methods. Also, spatial hot spots incorporated well in our model. By interpreting the marginal probability of every influential factor in the model, we analyzed the effect of these risk factors, in particular meteorological parameters, socio-economic factors and air pollution on the HFMD incidence. Our spatial Bayesian network approach is useful and the results provided important information for early-warning, prevention and control of HFMD.
The temporal and spatial pattern of urban housing price and its evolution characteristics are important indicators of measuring the equilibrium of urban real estate market. Based on the large amount of real time data of the Internet, we constructed a spatio-temporal data mining method using long time series. Firstly, a large number of real estate listing information and transaction data existing in different real estate business website had been obtained by web crawler technology. Secondly, the correlation between housing listing price and transaction price was tested using a linear regression model and the usability of ubiquitous online real estate data had been validated. Thirdly, a multi-scale grid model of mixed pixels was proposed, which was based on the description of the statistical characteristics of the real estate, and the problem of multi-source data fusion was solved. Moran’s I and Geo-detector were used to analyze the geographic spatial autocorrelation and non-homogeneity of housing listing price. The spatial raster database of long term real estate was constructed based on the combination of adjacent spatio-temporal interpolation and P-Bshade interpolation. Finally, the inner part of Beijing Six Ring Road was selected as the study area. We analyzed the spatial-temporal evolution characteristics of second-hand housing prices by grid partition algorithm. Overall, we explored the real time dynamic analysis method of real estate. The results showed that: in the first half of 2016, the growth rate of second-hand housing price was larger, and the latter half of the growth was relatively flat. The spatial distribution of second-hand housing price in Beijing was dominated by a single center pattern. At the same time, there are distributions of high island area. Dongcheng and Xicheng district were the core area of high housing prices, and the magnitude of price volatility was not consistent in different direction. The descending velocity of south of central city was the fastest. The diminishing rate of the north and northwest of central city were the slowest. The house price difference was more remarkable in the center of the city than that in the periphery region of the city.
Based on the data of POI （point of interest） in Jiangsu Province and economic and social statistics data, we used the methods of the standard deviation ellipse, nuclear density, the nearest neighbor index, and multiple linear regression to discuss the spatial distribution pattern and agglomeration characteristics of urban commercial outlets at different levels and stages of development. We also revealed the influencing factors of regional differences in the wholesale and retails, residential and catering industries. It was found that the commercial outlets in Jiangsu Province had the distribution trend of moving northwest, indicating a relatively concentrated distribution pattern and the characteristics of agglomerating in the cities south of the Yangtze River and intensive layout in centers of the prefecture-level cities. The spatial distribution patterns of different types of outlets varied from one to another. In the wholesale and retail industries, the spatial clustering characteristics of outlets are the most significant, while the distribution of residential outlets is relatively balanced. The commercial structure of each city is characterized by the number of wholesale and retail outlets and restaurants, as the sales of wholesale and retail take the dominant place. The regional GDP, population size, per capita disposable income of residents, comprehensive accessibility of cities, and proportion of the tertiary industry are the main factors deciding the distribution of outlets. The influence of different factors on the distribution of commercial outlets diverse from each other. The proportion of the tertiary industry and regional GDP of Jiangsu Province have more powerful influence on the distribution of commercial outlets.
Change detection based on Landsat time series has become one of the most popular methods of remote sensing change detection. This paper reviews the status of Landsat time series change detection, including comparison of change detection algorithms, Landsat time series construction and accuracy assessment of change detection results. Major problems and challenges of performing Landsat time series change detection are presented. Landsat time series change detection algorithms can roughly be classified into three categories, i.e., trajectory fitting methods, spectral-temporal trajectory methods, and model-based methods. These algorithms are mostly developed based on forest disturbance. Only few of them were used to detect changes in other land use/land cover types (e.g. urban expansion). Their applications in other fields need further verification. In particular, the comparative study of those different algorithms should be strengthened, which would provide better guidance for users to select optimal detection methods in specific fields. These indices commonly used for Landsat time series change detection can be divided into four groups, including spectral band, vegetation index, linear transformation and their combinations. It is often suggested to combine the advantages of various indices to detect different disturbance types. Although change detection methods based on Landsat time series have developed rapidly, many challenges remain. Upon now, the lack of consistent reference data set for accuracy assessment of Landsat time series change detection is the most serious challenge. Confronted with new challenges, new approaches are needed to calibrate the time series change detection algorithms.
With the increase of economic development, water quality degradation caused by non-point source pollution has become a serious problem in the hilly region of southern China. It is hard to set up controllable experimental environment at the watershed scale due to the complexity of non-point source pollution processes. Model simulation has become an effective way to facilitate watershed management and planning. Related studies on the simulation of non-point source pollution have been conducted in this region. However, few works have been done to summarize the outcomes of shortcomings of these studies and to point out the future research directions. Firstly, this paper analyzed the physical mechanism of non-point source pollution and regional characteristics such as special land features, human activities in this region and pointed out that the simulation methods of non-point source pollution in this region should meet the following demands: (1) coupling multiple watershed processes such as hydrology, soil erosion, plant growth and the migrating and transforming of non-point source pollutants; (2) spatially fully distributed in order to express the spatial heterogeneity of non-point source pollutant loading, and describe the migration and transformation routes of pollutants explicitly in this region; (3) taking the special land features and human activities into consideration which have important effects on the process of non-point source pollution. Then, based on the above demands, this paper summarized the current studies on the aspects of migration routes modeling and the representation of special land features and human activities in this region, and analyzed the problems of existing methods for non-point source pollution modeling that applied in the hilly region of southern China. On the aspect of spatial discretization, current methods cannot accurately describe the spatial heterogeneity of non-point source pollution processes, and the modeling of pollutant transport routes is limited to semi-distributed approaches which can’t describe the exchange relationship of material and energy among adjacent spatial units at the hillslope scale. On the aspect of describing the regional characteristics, some watershed processes that are special in this region are absent in the current models. At last, future research directions were discussed on the following aspects: (1) Strengthen the description of the special landscape features, and explore the method of spatial discretization that suitable for hilly region of southern China; (2) Improve the construction of the fully-distributed migration routes of non-point source pollutants; (3) Conduct comprehensive representation of special land features and human activities in the fully-distributed non-point source pollution model. This paper aims to provide references to the simulation of non-point source pollution in the hilly region of southern China, which can then serve as an effective tool for scientific watershed management.
In order to improve the estimation accuracy of fine particulate matter (PM2.5) near the surface, fine mode fraction (FMF), one of the key parameters in the PM2.5 remote sensing method (PMRS) should be improved due to its significant error (more than 0.3). Merging MODIS FMF and ground-based FMF (AERONET&SONET) using the universal kriging (UK) method can effectively improve the accuracy of MODIS FMF over land. However, the parameters (the nugget, the sill value and the range parameter) in exponential variogram function need to be obtained using long-term MODIS FMF data because of the sparse ground-based sites, which cannot meet the need of PMRS based on instantaneous remote sensing estimates. In this study, we calculate the parameters in exponential variogram model and analyze the parameters’ variation using all MODIS data over six years from December 2010 to November 2016. Results show that the seasonal variations of correlation lengths during different years are consistent with each other. Correlation lengths in summer are significantly longer than any other three seasons while the sill values show an opposite trend, suggesting that FMF in summer has a smaller variation than the other three seasons. Furthermore, the other three seasons need more ground-based data than summer when merging MODIS FMF and ground-based FMF data. To quantify the impact of parameters in exponential variogram function on FMF fusion results and achieve instantaneous FMF fusion products, we use the range parameter in winter of 2016 (control test, denoted as CRT) and the mean value of range parameter of 6 winters over 2011-2016 (comparison test, denoted as CMP) as initial values separately to merge MODIS FMF and ground-based FMF. Leave-one-out cross-validation results show that the maximum deviation between FMF fusion results and ground-based FMF in CRT is 0.198 (in CMP is 0.218), significantly decreased compared to the maximum deviation between MODIS FMF and ground-based FMF (0.552). The mean error between FMF fusion results in CRT and ground-based FMF is close to that between FMF fusion results in CMP and ground-based FMF (0.070 vs 0.080). Then, we apply the fusion results in CRT and CMP separately to estimating PM2.5 mass concentration near the surface in combination with the same auxiliary data such as relative humidity, the planet boundary layer height. The estimated PM2.5 mass concentration near the surface has a slight discrepancy with a value of 1.2 μg/m3（77.6 μg/m3 vs 78.8 μg/m3）between CRT and CMP. In addition, compared with in-situ PM2.5 mass concentration measurements, the mean error in CRT is equal to that in CMP (37.4 μg/m3 vs 37.4 μg/m3). It can be concluded that the seasonal average of range parameter for many years can be a substitute for the range parameter in the same season since the FMF fusion results and the PM2.5 estimates are insensitive to range parameter. As a result, we can obtain instantaneous FMF fusion results to improve the estimate accuracy of PMRS when we get more satellite FMF data in the future.
The rapid urbanization has brought about the benefit to the human society. Nevertheless, it also leads to a series of ecological problems, such as urban heat island and urban waterlogging. According to the blueprint and guidance for urban development, urban planning has a profound implication on the urban ecological quality in various ways. Therefore, a timely and precisely monitoring and assessment of ecological responses to different urban planning techniques have become an important issue for regional decision-makers. To meet this requirement, taking two sports centers that were built in 1980s and 2010s, respectively, in Fuzhou city as cases, this study utilized a recently developed Remote Sensing Ecological Index (RSEI) to assess the ecological responses of the two sports centers to their different planning manners. A Sentinel-2A image dated on June 23, 2016 was employed to compute the RSEI of two sports centers. Furthermore, three thematic indices, Normalized Difference Impervious Surface Index (NDISI), Normalized Difference Vegetation Index (NDVI) and Modified Normalized Difference Water Index (MNDWI), were utilized to extract the thematic features of impervious surface, vegetation and open water, respectively, from the used satellite images. The overall accuracies of the thematic feature extraction were greater than 90.0%. In addition, the Red Edge Position (REP) was calculated to estimate the state of vegetation growth. Based on the analysis of the main land cover information, the reasons for the differences of ecological quality between the two sports centers were then carried out. Results showed that the RSEI value of the Fujian Olympic Sports Center, built in 1980s with traditional planning manner, was 0.39, while the Fuzhou Strait Olympic Sports Center, recently-built with green ecological planning thought, had a higher value of 0.42, indicating a higher ecological quality of the Fuzhou Strait Olympic Sports Center. This owes mainly to the green ecological planning for the sports center. The green construction techniques used in the planning of the Fuzhou Strait Olympic Sports Center include the use of pervious surface, pavement without pointing joint, increase in green patches area, and reservation of wind corridors. These have effectively improved the ground wetness, reduced the land surface temperature and dryness, and thus have a great contribution to the ecological quality of the center. In addition, due to the immature status of the green plants in the Fuzhou Strait Olympic Sports Center, the vegetation growth in this center was slightly worse than that of the Fujian Olympic Sports Center, suggested by a relatively low mean NDVI value (0.522) of the center, compared with 0.562 of the Fujian Olympic Sports Center. Nevertheless, it is predictable that the NDVI of the Fuzhou Strait Olympic Sports Center could be higher than that of the Fujian Olympic Sports Center after a period of plant growth, and the ecological quality of the center would be further enhanced. On the whole, this study revealed that the adverse impact on ecological quality brought by urban construction could be effectively reduced by the green urban planning technology. Hence, it is essential to formulate and implement the green eco-environment conservation measures during the urban planning and construction so as to prevent further deterioration of the ecological environment.
Taking Shanxi Province as the research area, this paper compared the vertical accuracy of SRTM1 DEM and ASTER GDEM V2 data based on ICESat/GLA14 altimetry data. Firstly, error values for these two DEM datasets were acquired by taking ICESat/GLA14 data as the real data, and their error parameters were also calculated, such as mean error (ME), absolute mean error (AME), standard deviation (STD) and root mean square error (RMSE). Then, the error distribution of these two DEM datasets were analysed within the classes of slope, land use type and landform type. Finally, based on topographic profile method, the vertical error differences between these two DEM datasets in topographic types were analysed. The research results showed: (1) The vertical accuracy of SRTM1 DEM data is significantly higher than that of ASTER GDEM V2 data. The RMSE values of SRTM1 DEM and ASTER GDEM V2 are 6.1 m and 10.7 m, respectively. (2) Error analysis based on slope factor showed that the vertical accuracy of these two DEM datasets is affected seriously by the slope, and the error value increases with the increase of the slope value. Error analysis based on land use factor showed that the AME, STD and RMSE values of SRTM1 DEM are the lowest in paddy field, the highest in forestland, and the three error parameters of ASTER GDEM V2 are the lowest in building and the highest in forestland. Error analysis based on landform type factor showed that the AME, STD and RMSE values of SRTM1 DEM and ASTER GDEM V2 data are the lowest in the plain area, and the highest in large fluctuation mountain area. (3) On the selected topographic profiles in plain and terrace areas, the elevation value of ASTER GDEM V2 data have abnormal fluctuations. SRTM1 DEM data is too high for the estimation of valley. Overall, SRTM1 DEM is more accurate than ASTER GDEM V2 for terrain representation, which is basically consistent with ICESat/GLA14.
In the river basin ecosystem, channel wetland is located in aquatic terrestrial ecotone. The bridge and the link between terrestrial and aquatic ecosystems play an irreplaceable role in water detention, water purification, soil-water conservation, maintaining biodiversity and ecological balance. In this paper, we used Landsat satellite images of 1986, 1996, 2000, 2006 and 2015 to extract different types of river wetland systems between the Liujiaxia and Togtoh County of inner Mongolia in the last three decades. Then, we used spatial stastics analysis, transfer matrix and centroid position change method to analyze dynamic evolution and driving factors of wetland types. The results shows that, during 1986-2015 years, channel wetland area in the study area gradually decreased from 173×104 ha to 122×104 ha (~29.0%). Wetland transformation of the study area mainly occurs between the river, nude beach, herbal-wetland and farmland. In the last thirty years, the range of active channel wetland changes far greater than non-active channel wetland. The area of active channel wetland decreased from 15.46 ×104 ha in 1986 to 10.41×104 ha in 2015, decreased by 32.7%. The evolution of the active channel wetlands mainly occurs between the natural wetland types, namely, the river-bare Beach-swamp wetland. The non-active wetland area is basically stable, and the area is between 1.84-2.28 ×104 ha. It has characteristics of transformation between the natural wetland -constructed wetland and between natural wetland - agricultural land. The centroid position change of forest wetland, canal wetland and pond wetland are more prominent compared to other wetland types. The results of the single land use dynamics shows that, due to gradually accelerating urbanization pace, antrophy of the natural wetlands, increase the weight of farmland salinity, hydroelectric station system construction caused gradually decrease in the river area. The cropland to forest policy and the grassland to cropland policy result to accelerated dynamic change of forset, pounds, river, farmland, abandonedland and bareland. Through the analysis of channel wetlands, the change of active channel wetland mainly contribute to the wetland change of whole study area. The change of non-active channel wetlands was less affected by channel wetland changes. Our results are related to temperature, water conservancy, hydropower engineering and irrigation water, urbanization degree and ice flood season, but less sensitive to precipitation.
Emperor penguin is the indicator of Antarctic ecosystems. The distribution of its colonies owns essential significance for the study of Antarctic climate. Emperor penguins are sensitive to changes of the sea ice concentration and distribution. Thus, they have become an essential species for investigation on the effect of climate changes on the Antarctic ecosystems. However, it is difficult for the traditional manual investigation to obtain comprehensive and accurate information of the population colonies. Although some researchers have devoted to find emperor penguin colonies using remote sensing imageries in recent years, but their methods require considerable human involvement and cannot be used to detect all colonies rapidly and effectively. Emperor penguins breed and rest on land-fast sea ice and live in the same area for about six months, leaving extensive yellowish brown faeces which are significantly different from the white ice and snow around them. Inspired by the normalized difference snow index (NDSI), which is used for delineating snow cover, we can establish a/some similar index(s) to extract the faeces from extensive snow cover. On the basis of the difference between the reflectance of the faeces produced by emperor penguin in blue and red band, near infrared and shortwave infrared bands, two spectral indexes (NDII, EI) are putted forward to effectively recognize the faeces produced by emperor penguins, and determine their colony locations. According to the 195 scenes appropriate and quality-good Landsat 7 ETM+ imagery in 2009, a total of 38 emperor penguins colonies are obtained, 7 colonies of which are newly discovered (Bowman Island, Dibble Glacier, Auster, Point Geologie, Cape Crozier, Brownson Islands and Rupert Coast), 2 colonies have disappeared (Amundsen Bay disappeared and Ledda Bay), and the positions of the other 25 colonies (except Thuston Glacier, Luitpold, Sanae, Gould, Ragnhild and Beaufort Island) do not change significantly. The overall accuracy of colony detection is about 94% and the performance of the colony detection is influenced by the data quality and colony size. The performance of this method improves with increasing colony population size. Although spectral attributes are chosen to identify faeces produced by emperor penguins, some misclassifications may happen. This method may miss a few small colonies of which the sizes are smaller than spatial resolution of the imagery. The failure of detecting these colonies of small size is most likely due to the mixing of non-colony terrain in the Landsat pixels. These smaller colonies may well be identified by the satellites that have higher pixel resolution, and this method can be adapted to other high spatial resolution of satellite data in the future. The distribution of emperor penguin colonies is closely related to the climatic factors, and colonies tend to gather at the regions where temperature is low and ice concentration is high. It is different for climate change occurring at different colonies. Long-time and regional observations are needed to study the relationship between climate and the changes to the distribution of colonies. With the continued rise of the air temperature and the change of ice concentration, the colonies of which the latitude are below 70 °S are facing greater threat, and emperor penguin population shows a trend to shrink to the southern pole.
