Non-point source pollution is an important source of water pollution, thus constituting one of the decisive factors affecting water environment. The commonly used agricultural non-point source pollution evaluation methods include statistical models and physical models of the computer. No matter what kind of modeling the researchers adopt, a variety of data acquisition types and the evaluation of the accuracy of the verification results make up the main bottleneck. In order to make people aware of the importance of satellite remote sensing technology in agricultural non-point source pollution evaluation, this paper made an application analysis of the access capability and feasibility of the satellite remote sensing technology from the angle of data types required by the study of the non-point source pollution, and also forecast the application potential of the satellite remote sensing technology in the precision verification of the agricultural non-point source pollution evaluation results.
According to the character of level structure of physical matter,this paper puts forward a multi-level detaching model for extraction of remote sensing alteration information based on several years' practice. Under the framework of this technical model,the simplification of the object for complicated extraction of remote sensing alteration information is discussed,and then the basic theory and structure of the model as well as the main technical system are discussed under the framework of the model and on the basis of an analysis of application instances. It is proved that the thinking of this model not only can simplify the problem of anomaly extraction but also is commendably applicable to the practice.
In order to invert the land surface temperature data of a whole day, the authors first used data for MODIS land surface temperature and the same passing time of the two infrared channels of FY-2C satellite to obtain the regression coefficient by linear fitting of each pixel without considering atmospheric water vapor content and land surface emissivity, and then inverted the land surface radiation temperature with the spatial resolution of 5 km per hour of the day. Based on the transit time, the maximum and minimum surface temperatures appeared, which were downscaled to the spatial resolution of 1 km. Comparing the distribution of the 5 km spatial resolution land surface radiation temperature inverted from FY-2C data with the distribution of MODIS 5 km spatial resolution land surface temperature at the same time and same scale, the authors found that their spatial distributions are similar. Finally, the authors calculated the average 1 km spatial resolution surface radiation temperature inverted from FY-2C remote sensing data and MODIS 1 km spatial resolution land surface temperature in the regions with different vegetation types in combination with the vegetation cover classification map, and the results suggest that the absolute error is 1.95 K and the relative error is 10.7%, which means that the error of the land surface radiation temperature inverted by the method and the land surface temperature is below 2 K when the main body of the land surface is covered with soil and vegetation.
Compared with the traditional pixel-based classification method, the object-oriented classification method can reach a higher accuracy. As the middle entity in the process of information extraction, the object is one of the key factors of the object-oriented classification. The quality of the segmentation is directly related to the image classification accuracy. In this paper, a segmentation approach based on the high precision historical cropland parcels is presented. In this approach, the cropland parcels are considered to be homogeneous or not based on the real-time remote sensing image, then the global contrast index is calculated by each parcel to find the best local segmentation parameters, and the improved result is finally approached. The approach was tested in an agricultural area, and the result shows that: 1) the approach can get a more homogeneous object with stable boundaries; 2) Local segmentation parameters provide a more reasonable result in which the "less-segmentation" phenomenon and the "over-segmentation" phenomenon are effectively eliminated; 3) Automatic selection of local optimal segmentation parameters greatly enhances the objectivity of this approach; 4) Global contrast index is sensitive to the "less-segmentation" phenomenon and the "over-segmentation" phenomenon, so it can lead the segmentation to produce the best result. On the other hand, it can also serve as a good index to evaluate image segmentation in agricultural areas.
Object-oriented image analysis is the current research focus in information extraction,and the image segmentation method is the core technology of the object-oriented method. The effect of the segmentation directly affects the extraction of image objects. In this paper, the authors propose an object-oriented method for image segmentation which combines the advantages of edge detection using Canny operator with the secondary developing functions provided by eCognition Developer 8.0. Tests show that the segmentation method is accurate and reliable,the segmentation result is continuous and can well solve the "flood" and "broken" phenomenon. At the same time,this method that combines the advantages of the object-oriented method can satisfactorily solve the problem of "salt and pepper" and minimizes the impact of noise on the classification so as to extract the interesting object surface features.
Using surface reflectance data and atmosphere aerosol data measured at the Inner Mongolia radiometric site in June, 2010, this paper analyzed the surface reflectance characteristics, atmosphere aerosol properties and geography conditions and made a comparison between the Inner Mongolia test site and the Dunhuang radiometric calibration site. The results show that the relative deviation of reflectance is less than 5%, indicating that the site has good uniformity, that the aerosol optical thickness of each channel is up to 0.1~0.2 during the fine weather, that the transportation is very convenient and that the cost of the calibration experiment is low. Therefore, the Inner Mongolia radiometric site not only can be used to carry out on-orbit radiometric calibration and validation experiments as the spare radiometric calibration site, but also can improve reliability and accuracy of homemade satellite remote sensing data.
The SUSAN(Smallest Univalue Segment Assimilating Nucleus) method is used to detect gradient features from QuickBird imagery,and then the imagery is segmented using marker-controlled WT(Watershed Transform),and the segmentation result is satisfactory. The SUSAN method detects gradients well. It is not sensitive to noise and the values of the gradients are in a definite range and do not change with images,which offers convenience in selecting parameters in the later processes. The method is flexible because it is easy to choose the illumination threshold and the size of SUSAN matrix is not fixed. Based on the marker derived from both SUSAN gradients and NDVI,the gradients are modified using morphological grayscale reconstruction method,which efficiently constrains much local minima of the gradients and improves the segmentation precision.
The remote sensing estimation models for reed biomass were established by using TM data and the field investigation of reed biomass around the Daxihaizi reservoir in the lower reaches of Tarim river. The results show that the gray values of TM1~5 and TM7,the perpendicular vegetation index(PVI),the brightness vegetation index(BVI) and the normalized difference vegetation index(NDVI)are significantly related to the measured values of Phragmites australis biomass, the fitting precision of the nonlinear regression model is higher than that of the linear regression model,and NDVI has the highest precision in the regression models but TM4 has the lowest precision. The multiple stepwise regression has a higher fitting precision than ordinary regressions. The ordinary linear regressions based on PVI and NDVI and the ordinary nonlinear regression based on TM1 are the best options due to the characteristics of simplicity,reliability and convenience.
In the traditional Space-to-Earth car detection system,thermal infrared,radar or aerial image data are often used, while high-resolution satellite remote sensing data have rarely been employed. To solve this problem, this paper proposes a new method for car detection based on high resolution satellite images,which is named BIS (Background Iterative Search). Firstly, according to the local differences between the object and the background,the background is searched and removed,and the preliminary car detection is achieved based on the material properties of cars. Secondly,the dynamic twin peak threshold method is used to separate roads from non-roads,and roads are roughly extracted based on shape features. Lastly,the correct car objects are obtained by constraining the elementary ones with the derived road information. The BIS method was applied with IKONOS and QuikBird data and proved to be effective.
In this paper, the quality of HyMap data simulated with different payload indexes was evaluated in two aspects,namely,pre-processing and geoscience applications of the HyMap image data. In the aspect of simulated data pre-processing,three parameters were used to evaluate the quality of the simulated HyMap image data,i.e., average variance abnormality,histogram abnormality and correlation abnormality. In the aspect of geoscience application of the simulated HyMap data,the simulated HyMap image data were used to extract the alteration information of mineralization and mineral mapping,and the quality of simulated HyMap data with different payload indexes or with the same payload indexes but different scales was analyzed and evaluated according to the number and kinds of the information extracted and the precision of mineral mapping. Some good results have been achieved,and several valuable advices and suggestions are put forward for the study of HyMap sensor.
The high spatial resolution remote sensing images are used widely in the land cover classification; nevertheless, the traditional pixel-based classification has the weakness of relatively low accuracy. For the purpose of improving the accuracy of the high spatial resolution image classification,the textural features were extracted quickly by using the method of Gray Level Co-occurrence Matrices (GLCM),and then the ALOS image of the typical test area in Huzhou city of Zhejiang province was classified based on textural features and Support Vector Machine (SVM). The results show that image classification based on textural features and SVM can better extract surface features with precision of 90.88%. The classification precision based on SVM only is higher than that based on maximum likelihood,with the former precision being 89.96% and the latter 86.16%. Extracting land cover types quickly and accurately can provide a service for the research on appearance and spatial-temporal distribution of the agricultural non-point pollution source,and also provide scientific evidence for exploration of reasonable land use model and sustainable land utilization in Taihu basin.
As a special physiognomy in nature, the mountainous region has peculiarities of uncertainty, complexity and multifactor, and digitization and classification should be conducted to find out its specific nature. With the support of spatial analysis and mathematical statistics, four factors were extracted and analyzed on the basis of ArcGIS application characteristics in Sichuan Province. They were absolute elevation, relative elevation, slope and terrain roughness. An integrated classification system for mountainous region was built by statistical methods to generalize these tow factors into two major factors and figure out the critical value of classification. The final classification results for mountainous and non-mountainous regions in Sichuan Province were extracted by using the ArcGIS grid calculator. The extraction results show that this method has reached the accuracy of the classification levels by the traditional method, thus laying the research foundation for realizing 'digitized mountainous region.
A single control point may appear in multiple adjacent images because the unmanned aerial vehicle (UAV) images have high overlap degree. In the aerotriangulation work, the extraction of control points on images is a complex work. The original position and orientation system (POS) data can be corrected using the UAV images and control points, and then the corrected data can be applied to the work of points extraction. In this paper, an automatic extraction image points method suitable for large overlap images was put forward, and the accuracy of points extraction was assessed according to different terrain areas. Experimental results show that the method has higher precision in the flat terrain area and can be directly applied to the fast image mosaic, aerotriangulation and other follow-up processes.
The composition of the low-altitude unmanned aerial system is described, and then the single image correction method,the after stitching correction method and the aerial triangulation method of mapping by using the UAV images are proposed according to their small scene and high resolution of the UAV images. The key points of the three different mapping methods are emphatically discussed, and the three methods are jointly analyzed and tested according to their mapping effect, accuracy and efficiency. The results demonstrate that the after stitching correction method has higher efficiency, and the aerial triangulation method has better accuracy, with the single image correction method in between, as evidenced by judging the efficiency and accuracy.
The main objective of this study is to discuss the feasibility of predicting near-surface air temperature using MODIS products. Principal component analysis of land surface temperature (LST), ALBEDO, vegetation index (NDVI), altitude (ALT) and latitude (LAT) was employed, with some principal components of the cumulative variance in the front of these principal components as independent variables. Multiple linear relationships between independent variables and meteorological observation temperatures were established. The results show that the range of RMSE is between 0.5 and 2.3. Most of RMSE vary greatly in winter months but are relatively stable in summer months. Compared with Aqua/MODIS, Terra/MODIS can get better results. RMSE of Tmin modeled by Terra nighttime MODIS is lower, and so are Tmax and T14 modeled by Terra daytime MODIS. LST is the most powerful predictor, followed by ALT, LAT, NDVI and ALBEDO. The MODIS products can therefore monitor the spatial distribution of near-surface air temperature at different times, with the optimal products selections being different.
With the spectral experiment and the simultaneous observation results of Hyperion satellite on 19 August, 2004 as the basic dataset, the authors used the uncertainty principle of spatial data to develop a "bands model" for chlorophyll-a concentration retrieval algorithm of the subsection mapping retrieval model. It is thus found that in the ranges of 10-20 μg/L and 50-100 μg/L, the retrieval error of chlorophyll-a concentration is relatively low, (approximately ±20%), whereas in the range of 20-50 μg/L, the retrieval error of chlorophyll-a concentration is relatively high, (approximately ±40%). A comparison with the traditional methods for error describing shows that the "bands model" could include more detailed and accurate information of data quality for remote sensing products.
With the purpose of effectively using multi-source remote sensing data to extract surface coverage information, this paper puts forward a rule-based substitution method to achieve land-cover information extraction accurately, which is based on principles of decision-making level multi-source remote sensing images fusion. Based on the classification experiments of freshwater wetlands in the Sanjiang plain regions, the authors verified the feasibility and applicability of this method, thus providing a new means for investigation and monitoring of land cover.
In order to promote the application of the remote sensing data of HJ-1A/1B small satellite to urbanization monitoring,the authors selected Xuzhou City as the study area and chose HJ-1A/1B multispectral remote sensing images acquired in 2008 and 2010 as the data sources. After mixed pixel decomposition,the urban impervious surfaces were extracted by Linear Spectral Mixture Model (LSMM),Multiple Layer Perceptron (MLP) and Self-organizing Map (SOM) on the basis of V-I-S model. A comparison of the three methods through accuracy analysis shows that MLP is suitable for estimating the abundance of impervious surface area (ISA)from HJ-1 A/1B data, and ISA can clearly reflect the trends of urbanization.
The spectral data and numerical analyses of the Taihu Lake during the period of October 4-8, 2010 were used to analyze the influence of solar zenith θ, the ratio of scattering and absorption coefficient b/a and parameter Q upon the variation of remote sensing reflectance. According to the results, remote sensing reflectance increases with solar zenith θ, while θ changes in the range of 0°~20° or 80°~89°, which has little influence on the growth of remote sensing reflectance. Remote sensing reflectance increases with the increase of b/a, and there obviously exists a positive relation between them. When Q is large (6~7), its alteration has little effect on the variation of remote sensing reflectance. On the contrary, when Q is small (2~3), its variation has enormous impact on the growth of remote sensing reflectance. Remote sensing reflectance tends to decrease with the growth of Q.
In this paper, the location of the submerging area at different water levels (150 m,160 m,170 m and 172 m) in Danjiangkou reservoir basin was detected based on the DEM data,and the landcover below 172 m level was obtained by using object-oriented classification approach based on RapidEye and Landsat-5 TM data. The authors analyzed not only the spatial patterns of the landcover in the drawdown areas but also the impact factors. The results show that the acreage of the drawdown area at 150-160 m,160-170 m and 170-172 m level is 232.0 km2,242.6 km2 and 43.1 km2 respectively. The acreage of the cultivated land is the largest,accounting for about 50% in every drawdown area. In addition,the acreage of the bare land is larger at the low water level,while that of the forest is larger at the high water level. Dynamic scheduling and slope constitute the impact factors on the spatial pattern of landcover. Both have a significant impact on the distribution of landcover in the low elevation region,while the slope has a significant impact on the distribution of landcover in the high elevation region.
The exploration and calculation of the eco-environment response to LUCC in Guanzhong area constitute the basis for the study and application of ecological compensation and for the realization of regional sustainable development. In this paper,the method for decision tree-based multi-layer information extraction and supervised classification was employed to extract the information of land cover in Guanzhong area in 1990,2000 and 2007. Then the different kinds of general service values of ecosystems in Guanzhong area in 1990,2000 and 2007 were evaluated respectively using the method of ecological economics. Some conclusions have been drawn: 1 In 1990,2000 and 2007, the total service values of Guanzhong area's ecosystems were 55.739 billion yuan,50.729 billion yuan and 53.377 billion yuan respectively,which further demonstrated that the regional service values of ecosystems decreased abruptly due to local serious degradation and deterioration of forest and grass from 1990 to 1998,and were increasing significantly after 1998 due to implementation of returning cultivated farmland to grassland and forest land. 2 The index of ecosystem's capacity to support economy in Guanzhong area rose from 0.436 9 in 1990 to 6.430 2 in 2007,which indicated that pressure on ecosystem is increasing continuously with the development of economy,and the task of ecosystem restoration and construction remains a key link in keeping the sustainable development of regional economy.
Using 1992,2001 and 2009 TM remote sensing data,the authors estimated the vegetation coverage change status of Chengdu in the past 20 years objectively and quantitatively. The study has practical significance for adjusting the climate,resuming the vegetation and preventing the natural disasters. Normalized difference value vegetation index (NDVI) was used to estimate the vegetation coverage in different periods and draw the change grayness and change levels chart of Chengdu vegetation coverage. Data analysis results show that, from 1992 to 2009,the vegetation coverage in Chengdu was overall decreasing. The quantity of dense forest land,shrub land,high and middle grass and cultivated land was decreasing,in which the vegetation coverage decreased obviously from 1992 to 2001,the vegetation coverage remained reduced from 2001 to 2009,but the speed was slower than that from 1992 to 2001. In addition, the speckle distribution of artificial greenbelt appeared in the urban area,and the small-area climatic condition was improved continuously.
The area ratio of doline to Xichengshan karst depression was calculated through remote sensing interpretation to investigate the origin of Xichengshan karst depression. It was unusual that the karst depression was formed in dry climate in northern China which was not suitable for karst formation. It is considered that the karst depression might have been formed on the basis of aerolite striking, as evidenced by the comparison of karst-forming conditions, status quo and vegetation in this area with those in other areas. the karst depression might have been formed in late Paleogene period, as shown by the comparative study of altitude of the karst caves and regional planation surface.
Two methods for removing the thin cloud cover from the RapidEye remote sensing imagery in western mountain areas of China based on Photoshop environment are introduced in this paper. One is to directly adjust the three bands respectively by adjustment function of levels, and the other is band substitution. In the second method, the first step is to substitute infrared band for blue band, and then three bands are adjusted respectively in levels. Test results show that the imagery quality is enhanced obviously when the processed imagery is compared with the original one, and the efficiency in the production practice is improved.
According to the methods of RS and GIS and after geometric correction,image fusion,image mosaicking and ortho-rectification for CBERS-02B image,the remote sensing image can be overlapped with DEM generated by TIN from the relief map. On the basis of 3D remote sensing images of a REE ore exploration area in Anyuan County,three-dimensional simulation can be applied to the three-dimensional analysis so as to study the geological environment of the mining area. Practices show that 3D remote sensing technology has great practical significance for selection of geological investigation route in areas where it is difficult to conduct field work as well as typical regional reconnaissance and mining production. In addition,it can find out the distribution pattern of the geological environment with great speed and accuracy,and can also be superimposed with vector data for the buffer analysis. Finally,the method can be used to analyze and predict the scale and development tendency of the geological hazards so as to provide the basis for decision-making on eliminating and tackling the hazards.
The land use partition is of important significance for regional concerted development, and the fast land use zoning relies on the land use zoning information system. Based on visual programming language Visual Basic (VB) and GIS components Map Objects(MO), the authors developed a land use zoning system. The key modules composed of the hierarchical clustering partition module and the dominant use zoning module was combined with GIS technology, database technology and mathematical model. On such a basis, such functions as the view, the inquiry statistics, the land use zoning and the spatial analysis could be realized so as to meet the requirements of different division levels. Verification was made with Xinzhou city as an example. This system can realize the functions of fast calculation of land use zoning and the visualization expression of zoning results so as to fulfill the land managers' multi-level requirements. In addition, it can facilitate the operation and help land managers' decision-making.
Spatial autocorrelation has been applied to extensive data obtained from many research projects, and it is a common method for studying spatial distribution. Spatial autocorrelation analysis is a means for analyzing data correlation based on the spatial correlation analysis. It studies the correlation between one polygon and the nearest neighboring polygon through recognizing the similar degree of the major spatial object and other spatial objects. Using the 2010 census data and applying the spatial statistics and GIS, the authors analyzed the models of population spatial distribution of 17 prefectures in Shandong province by such means as quarters of population density, Moran's I and Local Moran's I. The population spatial correlation shows that there are no abnormal spatial areas of high density and low density in prefectures of Shandong province. The population spatial correlation shown by Moran's I and Local Moran's I indicates that the population density spatial distribution has spatial cluster, high-high cluster and low-low cluster. The general population density spatial distribution has three spatial belt-shaped regions, where the highest density is in the southwest areas, the density decreases to the lowest density in the northeast areas, and the similar population density areas are centralized in the vicinage. There are 5 "high-high" prefectures centralized in the west and south areas, 4 "low-low" prefectures centralized in the north areas, 2 "low-high" relation isolated points existent in Laiwu and Rizhao and a "high-high" isolated point existent in Weifang.
Power planning is a data-intensive and technology-intensive work, which must draw support from the information system for the auxiliary support. As there is not a unified database platform for the domestic power planning industry, the level of data sharing, visualization and automation is low. In this paper, the authors proposed to build a power planning data supporting platform based on GIS technology, and designed the technical route, overall framework, database construction, functional modules, key technologies etc. According to the design proposed in this paper, the authors integrated the planning data from China's leading power design institutes, realized the predetermined function. and established the national power planning data platform. The construction of the platform will enhance the power data standard and sharing, improve the quality and efficiency of power planning, and promote the whole informationization level of the entire industry.
