Arctic sea ice plays a very important role in the modulation of global climate and sea ice extent is a basic parameter for sea ice monitoring. In recent 40 years, a series of environmental and climatic issues such as degradation of Arctic natural environment, frequent extreme weather in the Northern Hemisphere and global sea-level rise are caused by continuous warming and apparent sea ice decrease in Arctic. So it′s important to know the extent, variation, trend of Arctic sea ice and its response to global climate change. The most commonly used datasets such as HadISST and OISST sea ice dataset provided long time series of changes in sea ice of the Arctic regions. However, the spatial resolution of these datasets is relatively low. There are some limits in the study of response of sea ice change in Arctic key regions to weather and climate in China. To overcome these problems and to make up the lack of passive microwave sea ice dataset provided by China, FY (Feng Yun) sea ice dataset is developed by NSMC (National Satellite Meteorological Center) on June 27th, 2011. In this dataset, the Enhanced NASA Team (NT2) algorithm is used based on the data of MWRI (Microwave Radiation Imager) sensor carried on FY satellite. In this algorithm, direct radiative transfer model is used to model MWRI brightness temperature for four surface types (ice-free ocean, new-formed ice, one-year ice and multi-years ice) and for different atmospheric conditions. Then, sea ice coverage lookup table (0% to 100% in 1% increments) is obtained based on modeled brightness temperature considering different atmospheric conditions. Sea ice coverage is confirmed by comparing observed value with modeled value. Sea ice extent is consistent with the actual situation in most Arctic regions. Although matching errors between channels and positioning errors have been corrected in FY dataset, the received echo signal is relatively weak due to the shorter antenna on MWRI. The weak echo signal makes it difficult to correctly differentiate the boundary between sea ice and near sea shore land, which greatly impact the total accuracy of the dataset and its application. In order to solve this problem, this study introduces a method of optimizing FY Arctic sea ice dataset based on NSIDC (National Snow and Ice Data Center) sea ice product. In NSIDC product, judgment matrix was created covering the entire grid and identifying each pixel as land, shore, near-shore, offshore or ocean as determined by the land/sea mask. Then, these different pixels are corrected in different degrees, respectively. Sea ice extent calculated from NSIDC product is strongly consistent to the actual situation. The accuracy of FY dataset is greatly improved. The analysis results indicated an extremely significantly positive correlation with the NSIDC product (R2 = 0.9997) during June 27th, 2011-December 31st, 2015. The maximum deviation percent of daily, monthly and annually sea ice extent is 3.5%, 1.9% and 0.9%, respectively. Also, the optimization process of FY dataset has no obvious influence on the spatial stratified heterogeneity of the dataset. The optimized FY dataset can correctly reflect Arctic sea ice extent and its variation, especially in coastline regions. It can provide reliable basic data for the study of Arctic sea ice change.
Traditional R-tree and its variants cannot support frequent location updating of moving objects. By introducing a variety of moving object indexing strategies in R*-tree, we proposed a moving object index structure-LUMR*-tree (Lazy Update Memo R*-tree), which combines lazy update and memo update/insert strategy. The LUMR*-tree can quickly complete update without changing its structure by lazy update strategy. It can also simplify an update operation to an insert operation with the Update Memo (UM), which can avoid frequently purging old entries from the index tree effectively. The removal of old entries is implemented by a garbage cleaner inside the LUMR*-tree. Thus, data items and memory size of UM are maintained dynamically, which ensures high stability and efficiency of the LUMR*-tree. Experimental results show that, the LUMR*-tree achieved significantly higher update performance at the cost of slightly poorer query performance, it can support frequent location updating of moving objects. To sum up, the LUMR*-tree has good practical values and wide application prospects.
Traditional algorithms for modeling sea waves have some problems such as poor ocean surface realistic simulation and complicated calculation procedures. In order to solve these problems, this study presented a new sea wave simulation method which combines the Smoothed Particle Hydrodynamics (SPH) algorithm with the Marching Cubes (MC) algorithm. Based on space lattices, particles were allotted into different cubes and we established one way list structure for particle swarms storing and realized fast searching for particles within the smooth core radius in the calculation procedures of wave particle physics such as velocity, accelerator, position and so on. The force of the wave particle is generally composed of three parts: gravity, pressure gradient force and viscous force. Pressure gradient force is generated by the pressure difference between the fluid. Viscous force is caused by the velocity difference between the particles. According to this analysis of particle force, this study gave the Lagrange fluid control equation used for the accelerator calculation of ocean particles. In order to simulate the collision between particles and coastal barriers, we modeled the barrier surface as TIN (Triangular Irregular Network) and simplified the collision detection as whether the particle path passed through the triangle interface within a certain time. Assuming the particle as ideal rigid body, this study introduced an acceleration variation coefficient to calculate the particle's velocity after bouncing and improved the Euler equation by using the average velocity to calculate the particle's new position. In order to enhance the realistic simulation of the ocean wave fluid at the particle allotting time in the machine cube, this study calculated the density for each cube node dynamically. Through setting the threshold of sea surface density, we also extracted the wave surface dynamically by using linear interpolation method to generate triangular irregular network and realized the ocean waves 3D surface modeling and dynamic simulation. Through simulating, it verified the effectiveness and feasibility of this algorithm and could provide certain reference for ocean environmental modeling and virtual visualization.
In the period of taxis taking shift, many city dwellers find it hard to get a taxi or even get turned away in front of vacant ones. Analyzing the spatial and temporal distribution of taxi shift change behavior can improve taxis efficiency, relieve the contradictions of supply and demand of taxi, and facilitate public travelling. This study establishes space-time sequence of taxi shift behavior and puts forward a method to mine and identify taxi shift sequence from GPS trajectory. Furthermore, we adopt taxis′ GPS trajectory data in Wuhan to perform our method. Based on the identified shift events, we analyze the space-time distribution of taxis′ shift behavior and evaluate parking resource allocation for taxis′ shift change by using intensity and density as indicators. The results show that: taxis′ shift behavior in Wuhan peaks in period of 1:00 to 4:00 and 16:00 to 17:00 and the latter peak in the afternoon partially overlaps with evening rush hours, which can introduce the difficulty in taking a taxi; the parking location for taxis taking shift is relatively uniformly distributed throughout the downtown of the city except Qingshan District, which is less prosperous than the others; Wuchang District has the strongest shift intensity and Jianghan District, whose area is the smallest among all the administrative districts, has the greatest shift density. In addition, considering the regulation issued by Wuhan traffic administration in 2012, which prohibits taxi drivers from taking shift in the evening rush hours, it is revealed that about 6.5% of the drivers have serious illegal shift behavior.
Taxi is an indispensable urban traffic mode in small cities. However, there are limited efforts focusing on explaining traffic congestion or resident commuting from a perspective of land use in small cities. This study attempts to reveal the spatio-temporal dynamics of resident trip activities from the aspect of urban functional features. Based on the GPS taxi data, we build a set of temporal GWR models on an hourly basis, which indicates that urban facilities have various effects on the pick-up and drop-off events during different daytime periods. Nine facilities, including coach station, supermarket, restaurant, residential area, karaoke, hotel, hospital, bank and administrative center, have been observed to be the critical elements to explain the ridership variations. A spatio-temporal mechanism has been proposed based on the discovery that facilities with different urban functions have different impacts on resident trip demands. In contrast to the large cities, the trip activities of residents are spatially and temporally various in the small cities. The primary traffic demands are commuting activities, commerce, entertainment and intercity transfers. More rush hours, especially the “noon rush” and “midnight rush”, are revealed in small cities. The results provide valuable insights for quantitatively predicting the taxi demand as a function of the spatio-temporal variables, which may have implications on the traffic demand management and the urban planning of small cities.
With the development of economy and the improvement of living standards, urban residents pay more attention on health and medical facilities play a more and more significant role in public facilities. The continuous urban expansion make the distribution of medical facilities cannot meet the requirements of the size of the city. Thus, it is necessary to research on how to distribute medical resource reasonably and to fulfill residents′ requirement to the most extent. In this study, based on the spatial analysis of GIS, we established the database of urban traffic network, distribution of medical facilities and the population according to relevant statistical data for the downtown of Wuhan city. We also evaluated accessibility and equity of spatial distribution of medical facilities in Wuhan. The accessibility of medical facilities in downtown Wuhan was evaluated using the evaluation method of geographical spatial accessibility and the cost model of travel time based on traffic network. The assessment shows that the convenience to medical services is generally good in Wuhan, distributing radially around Yangtze and Hanjiang River. By introducing demand index, medical institution scale, regional population and per capita disposable income, the equity of medical service system was evaluated using the Gini coefficient, spatial correlation and spatial stratified heterogeneity. It indicated that the distribution of medical institutions in Wuhan is reasonable from the aspect of population distribution. Medical institutions are mainly distributed in the area of dense population while the distribution of medical institutions is comparatively rare in suburban area. The advice was proposedto improve the distributions of medical facilities. In suburban area where traffic is underdeveloped, medical conditions should be taken more considerations in city planning in order to improve the distribution of urban medical facilities.
As an important cause of global warming, carbon dioxide concentration and its change has aroused worldwide concern. How to have an explicit understanding of the spatial and temporal distribution of carbon dioxide concentration is a crucial technical challenge for climate change research. In this paper, based on the in situ observation data set collected in the TanSat flight test area, the correlations between the carbon dioxide concentrations and the environmental variables are analyzed, and suitable environment variables can be selected to establish a regression equation, through which we obtain a preliminary trend of surface carbon dioxide concentrations. Then combining the multiple linear regression model and High Accuracy Surface Modelling (HASM), the carbon dioxide concentrations with a high accuracy in the entire test area are produced. The results indicate that the spatial distributions of the carbon dioxide concentrations in the study area are significantly different between three periods, and the short-wave radiation is an important factor for the regression equation. Because of the high temperature and drought condition, the highest concentration appears in the first period especially in the western area. The second period has a different distribution on the carbon dioxide concentration comparing with the previous period, as in this period the high value region moves eastward, and making the concentration high in the eastern area but low in the western area. Both of the second and third periods have similar characteristics except that the high value region in the eastern area is reduced in third period. Moreover, statistical analyses show that the mean absolute error and the mean relative error of the predicted value of the HASM model are 9.8 ppm and 2.48% respectively, which are both lower than the errors produced using the Kriging method, therefore the HASM model remains to have higher simulation accuracy in a condition of few sampling points and low sampling density. Therefore a combined method of multiple linear regression model and HASM model can be used as an effective method for simulating the spatial and temporal distribution of carbon dioxide concentration in the surface layer.
Land subsidence in Beijing has developed since 1960s. Five major subsidence areas have formed: Dongjiao Ba Lizhuang-Da Jiaoting, Dong Beijiao-Lai Guangying, Changping Shahe-Ba Xianzhuang, Daxing Yufa-Lixian, and Shunyi-Ping Gezhuang. In this study, we investigated Chaoyang, Shunyi, and Tongzhou Districts, which have experienced relatively serious subsidence, and obtained land subsidence monitoring results using data from 47 ASAR images (2004-2010) and the technology of small baseline subset interferometric synthetic aperture radar (SBAS-InSAR). Weighted by the annual average subsidence rate of SFP points and the subsidence amount of each year, we calculated the spatial distribution center of SFP points and the eigenellipse to quantatively analyze the spatiotemporal characteristics of subsidence in the study area. In 2004-2010, Beijing experienced pronounced uneven subsidence, and annual maximum subsidence increased from 104.04 to 178.83 mm. The long axis of the eigenellipse was parallel to the north-south direction and it indicated that spatial development of land subsidence was more obvious in the north-south direction than that in the east-west direction . The eigenellipse area decreased from 592.25 to 503.84 km2 in 2004-2010. This result indicated that the subsidence area decreased, but the amount of subsidence still suggested increasing subsidence in Beijing.
It is very important to study the characteristics of spatial pattern and variation of soil nutrients and analyze the effect of topographical factors on the spatial distribution of soil nutrients for the effective use and management of soil nutrients. In this paper, the combination of GIS and Geostatistics methods were applied to analyze the spatial distribution characteristics and variation pattern of soil nutrients (organic matter, available nitrogen, available phosphors and available potassium) in the agricultural land of southeast hilly area of Fuzhou. We further studied the correlation between soil nutrients content and topographical factors (topography degrees, elevation, topographic wetness index, deposition and transport index and gradient). The results showed that: the range of organic matter, available nitrogen, available phosphors and available potassium contents were between 1.10~89.5 g/kg, 1.00~461 mg/kg, 0.300~298 mg/kg, 4.00~399 mg/kg and the range of variation coefficients were 35.3~99.0%, which belonged to moderate variability. There was obviously different in the spatial abundance of soil nutrients in the cultivated land. In most of the area, the organic matter and available phosphors content were abundant, available nitrogen content was a little above average level and available potassium content was relatively scarce. The nugget coefficient of organic matter, available nitrogen, available phosphors and available potassium were 32.0%, 37.3%,50.0% and 50.0%, respectively. They were medium spatial autocorrelation, indicating that they were controlled by structure and randomness. Spatial autocorrelation scale of organic matter and available nitrogen were large. They change smoothly in each direction (0°, 45°, 90° and 135°) when the step length was less than 0.3 km.and are isotropic. The variation of effective phosphorus and available potassium was small. Their direction of change was complex and they are anisotropy. These results suggested that the government needed to strengthen guidance of fertilization. Nitrogen fertilizer amount should be maintained and the potash should be increased reasonably. The organic fertilizer and phosphate fertilization should be decreased.. In addition, in the subsequent investigation, the setup of sample points should consider density and direction and appropriately increase the sampling of effective phosphorus and available potassium while nitrogen and organic matter and alkali solution sampling can be reduced based on the study.
The Yarlung Zangbo River, Nyangqu River and Lhasa River Region (the YNL River Region) is the political, economic and cultural center of Tibet, with large population and more developed economy. In this region, there are a variety of complex terrain units, such as mountain plateau, valley plains, and the geomorphological conditions directly or indirectly affect population distribution in this region. Study on the relationship between population distribution and complex topography can be helpful to reveal the present situation of population distribution and understand the spatial structure of population distribution. This study will play an important role in selecting a livable site, improving living environment of the farmers and herdsmen and making the policy of regional economic development. This study will have great significance on the sustainable development of the population, resources and environment in this region. Based on the STRM DEM (30 m) and 1km×1km population density raster data in 2010, we used spatial analysis of ARCGIS 10.1 and raster calculation to extracted terrain features data from DEM. This paper analyzed the population distribution in the YNL River Region, Tibet and quantified the relationship between population distribution and topographical parameters, i.e. elevation, slope, aspect and relief degree of land surface (RDLS). The results showed that: (1) The population density in 90% of the total area of the YNL River Region is less than 10 persons / km2, and 80% of the population distributes in less than 5% of the land. The population distribution is relatively concentrated. At present, there are two main population gathering area in YNL River Region, Chengguan District in Lhasa and Shigatse City in Shigatse Region. Lhasa Chengguan District is particularly significant. (2) Population shows a significant characteristic of distribution along the river. 80.46% of the population lives in the area less than 10 km to the river and there is a significant exponential correlation between population and the distance to rivers. The population density within 2 km from the river is higher than 50 persons / km2. (3) Nearly 99% of the population locates at an altitude of 4500 m or less. When the altitude exceeds 3800m, population density with elevation shows a downward trend. (4) Nearly 70% of the population distributes in the area with slope gradient less than 15°. The greater slope, the smaller population density. The slope aspect has no significant influence on population distribution. (5) RDLS has a relatively significant influence on population distribution. 85% of the population is located in the regions with RDLS less than 800 m. The population density and RDLS are in inverse exponential relationship.
Urbanization has made natural and semi-natural landscape to be gradually replaced by impervious surface, which has caused significant reduction of surface permeability in urban region. Along with these changes, profound transformations have occurred in hydrological processes, water environment, urban thermal environment, and ecological service system. Impervious surface is an important indicator of characterizing the urban expansion, which has extremely important ecological implications. We chose the multi-temporal Landsat images as our data sources, and took Qinhuai River Basin as the research area in this study. Rotation forest algorithm, which belongs to the category of ensemble learning that synthesizes the advantages of different classifiers and effectively addresses the limits of the information provided by a single image, was used to produce the nine-year land cover maps of Qinhuai River Basin. Focusing on the large watershed scale, we explored the changing process of the urban landscape pattern in the research area during the past 30 years. Impervious surface coverage dynamic analysis was used to reveal the changes of impervious surface area. Land cover change trajectory analysis was used to explore the resources of impervious surface and transformation process of land cover. Landscape metrics analysis was used to quantify the spatial and temporal changes of impervious surface pattern. We aimed to reveal the spatial-temporal changing characteristics of urban landscape configuration against the background of urbanization. The results showed that the landscape pattern has changed significantly during the past 30 years. Overall, the impervious area has increased by nearly four times. The dominance of the impervious surface has increased greatly. The analysis suggested that the turning point of urban expansion is between 2001 and 2003. Urban expansion mainly occurred in Nanjing city and Jiangning district before the turning point, while the impervious surface expansion rate of Nanjing city has greatly decreased after that. At the same time, there is a sharp rise in the expansion rate of Lishui and Jurong districts. The impervious surface within the 2001-2003 period had the highest spatial heterogeneity, which then decreased significantly after the turning point until 2015. The shape of the impervious patches has becoming simpler at the latter stage, and the impervious surface has turned from a dispersed distribution into a distribution pattern with higher connectivity. Besides, the area with high level of connectivity is mainly distributed in Nanjing city and Jiangning district.
Scientific collaboration is an important way of knowledge dissemination and sharing. Researches have showed that geographic factor is one of the main factors that influencing scientific collaboration. However, most of related researches have just quantitatively described the functional relationship between collaboration strength and geographic distance from the perspective of Scientometrics. As a result, it can hardly detect the spatial characteristics and relationship of scientific collaboration. In this paper, for the purpose of mining spatial patterns in scientific collaboration network, geographical preference of scientific collaboration was studied from the view of geography. Taking the haze research network in China for example, the location information was extracted from bibliographic data and then the virtual scientific collaboration network can be mapped into geo-collaboration network by using geocoding service. Based on this, a distance-based method for community detection of scientific collaboration network was proposed to explore the spatial cluster pattern in scientific collaboration. Using modified Louvain community detection algorithm, two different variables were used as weight factor to detect communities. The results showed that, the community detection algorithm considering collaboration frequency and geographic distance can make the average geographic distance minimum and the Salton index maximum inside community, which both reflect the geographical preference and collaboration strength of scientific collaboration. This method can effectively explore the spatial pattern and relationship in scientific collaboration network, and represent geographical preference of scientific collaboration in a quantitative and qualitative way. In addition, it is a novel method of introducing geographic location and geographic distance into complex network analysis. We hope that it will not only be helpful for scientific collaboration network, but also can be applied to other complex network for geographic community detection.
With the wide application of Unmanned Aerial Vehicle （UAV) in the inspection of power transmission line, the demand for objects detection and data mining from images acquired by UAV also grows significantly. Traditional detecting methods use some classical machine learning algorithms, such as support vector machine （SVM), random forest or adaboost etc. and combine the low level features such as gradient, colors or texture to detect electrical devices. These image features must be carefully designed and changed a lot from various object kinds. Thus, they are not suitable for UAV images with complex background and multiple kinds of object. On the other hand, the disadvantages of these methods are that they cannot take advantage of the high quantity and large coverage of UAV acquired images, and cannot get a satisfactory accuracy. The recent developing Deep Learning method brings light to this problem. Convolutional neural network （CNN) performs excellently in object recognition area and outstand many other methods used in the past. Without the need of extracting images’ features, CNN becomes the many state-of-the-art methods in object recognition rapidly. In object detection, Region-based convolutional neural networks （RCNN) retrieves the region that may contain the object from the images to detect and recognize the object. However, the computation is so expensive that it cannot meet the requirement of detecting massive UAV’s images and cannot be used in practical projects. Fast R-CNN and Faster R-CNN solve this problem by changing the way of object retrieval. They use features produced by CNN network layers and apply a region proposal network layer behind to locate the object. After that, fully connected layers and softmax layer follow to classify the features corresponding to object into special kinds. Using this strategy, Fast R-CNN and Faster R-CNN save lots of time to produce region proposal and can perform object detection at nearly real time. The principle and processes of Faster R-CNN and several other object detection methods are described in this paper, and they are tested for electrical devices detection from images of the power transmission line obtained by UAV. We analyzed the influence of several key parameters to the device detection results, such as the dropout ratio, non-maximum suppression （nms) and batch size. Then, we gave some constructive advice of tun ing parameters in Faster R-CNN. We also analyzed the advantages and weakness of three advanced detection algorithms, including Deformable Part Models （DPM) and two deep learning-based methods named Spatial pyramid pooling networks （SPPNet) and Faster R-CNN. Finally, we constructed image datasets of power transmission line inspection obtained by UAV and tested the three methods above. The recall ratio and accuracy ratio of them are compared and the superiority of the Faster R-CNN is validated. Testing results showed that Faster R-CNN method can detect various electrical devices of different categories in one image simultaneously within 80 milliseconds and achieve an accuracy of 92.7% on a standard test set, which is of great significance in real-time power transmission line inspection. These results also showed the advantages of the Faster R-CNN and we apply Faster R-CNN in our practical projects to detect electrical devices.
Mangrove is a special species which can adapt to the ocean and the land at the same time. Mangrove distributes in the coastal estuarine intertidal zone about 30°N~30°S on Earth where the climate is tropical or subtropical. The distribution of mangrove is restricted by the natural conditions and the climate. Also, it is disturbed by anthropogenic activities at a certain degree. Two adjacent regions, the coast of the Guangxi Zhuang Autonomous Region in China and the coast of Northern Vietnam, where the geomorphology and the climate are similar and the level of economic development is different, were chosen as the study area. Based on Landsat TM remote sensing images, we used the supervised classification and manual interpretation methods to acquire the distribution data of mangrove in the study areas in 1988, 2000 and 2015. We comparatively analyzed the differences in spatial-temporal transformation, the change of landscape pattern and the driving forces of the two coastal mangroves. This study indicated that: (1) During 1988-2015, the area of Guangxi Mangrove is increasing. It increased 18% in 2000 than that in 1988 and it increased 75% in 2015 than that in 2000. Northern Vietnam decreased first and then increased. In 2000, mangrove area decreased by 20% compared to that in 1988 and it increased by 50% in 2015 compared to that in 2000. (2) In the landscape pattern of mangrove, the average patch area of Guangxi Mangrove is relatively smaller, and it has higher degree of fragmentation, and the patch shape is more regular, closer to the square, more severely disturbed by human activities. North Vietnam mangrove has a larger patch area, lower degree of fragmentation, and the plaques have a shape of strips and were disturbed less severely by human activities. (3) The major human driving factors are as follows: the main human driving factors are changing in Guangxi Mangrove in China. From 1988 to 2000, cultivation ponds constructed by reclamation dominated. During 2000-2015, urbanization and factory-mining construction dominated. In North Vietnam, cultivation ponds constructed by reclamation always dominated during 1988-2015.
The original ecological environment of island is fragile and lack of stability. Large-scale development and construction activities of island expose the vulnerable vegetation to a greater danger, which makes the detection and assessment of vegetation cover change especially necessary. Based on the remote sensing images of Landsat 5 and Landsat 8 in 2001, 2010 and 2014, the fractional vegetation cover (FVC) of Pingtan Island in Fujian Province was computed using the FVC calculation model proposed by Gutmand and Ignatov. Combining the FVC data with land cover change information, this study analyzed the variation in FVC of integrated experimental zone on Pingtan Island and explored its reason. The results showed that in 2001, 2010 and 2014, the middle and upper level of FVC in Pingtan Island accounted for 86.00%, 58.92%, 71.16%, respectively, which indicate that the overall vegetation in the study area is in good condition. The analysis of dynamic change showed that there is a declining trend in FVC from 2001 to 2014. The FVC decreased greatly by 53.95% from 2001 to 2010. On the contrast, the FVC rose by 47.77% from 2010 to 2014, which improved overall FVC condition substantially and offset the previously significant decrease of FVC. The comprehensive experimentation area was established in Pingtan Island. Greening projects and reasonable planning gradually improved the vegetation construction and protection of the island green land system. These reasons led to the increase of vegetation.
Forest landscape pattern analysis is the basis for the optimization allocation and planning of forest landscape. Based on the data obtained from the second class survey of forest resources, with the integration of the forest landscape quantitative spatial analysis method and the independently-developed forest landscape visualization system software VisForest, a forest landscape pattern analysis and forest landscape 3D visualization simulation system was developed. This system is integrated with the landscape index calculation model and the geographic information system analysis method. The mainframe of the system was coded on Visual Studio 2008 platform with ArcEngine component and OSG graphics rendering engine incorporated, which has realized the quantitative and visual analysis of landscape pattern. Minhou Baisha national forest farm was taken as an example in this study. According to the dominant tree species, the forest landscape was classified and the landscape composition, patch characteristics and landscape heterogeneity etc. were analyzed. The results showed that the Cunninghamia lanceolate forest, Pinus massoniana forest, Pinus elliottii forest, Schima superba forest and the non-forest land are the dominant landscape types in Minhou Baisha national forest farm. The landscape types are generally versatile and the heterogeneity level is relatively high. Moreover, as the median one of all species, Schima superba is scattered majorly in long and narrow patches with small area. Schima superba’s shape index and fractal dimension number are the largest and its patch shape is intricate and irregular. The 3D visualization simulation of forest landscape provides an intuitive and interactive platform to the research of landscape pattern.
