Terrain modeling is important in the digital terrain analysis. Real-time adaptive mesh algorithm (ROAM algorithm) is a commonly used terrain modeling method. However, applying ROAM algorithm to real-time rendering of terrain visualization requires the original data to be square and cracks often occur after rendering, so this method is greatly restricted in the practice of terrain modeling. Aiming at the above disadvantages of the ROAM algorithm, an improved ROAM algorithm based on the interpolation of inner square is proposed in this paper. The polygon is divided into an inner square and multiple edge polygons, then the vertexes of these polygons are connected sequentially to divide the polygons into multiple triangles, and according to the size of the edge triangles, whether to continue segmentation of the edge polygon is determined. Thus, the problem of limiting the data source to a square is solved, and the crack generated by the terrain rendering is eliminated by adding split points. Using the irregular terrain gray image as the elevation data for terrain simulation experiment, the results show that the improved levenshtein distance algorithm can eliminate the crack with no binding requirements for data source, reduce the algorithm complexity and improve the overall visual performance. It can meet the requirements of terrain real-time dynamic display.
The National Geographic Conditions Monitoring Project of China, based on the high technology of remote sensing and geospatial information, has dynamically acquired various kinds of geographic features. The goal of this project is to assist the government's decision making in multiple domains such as politics, economics, culture, resources and environments by statistical calculation, analyses, evaluation and application of the accurate and basic geospatial information. However, as China is such a huge country, it's particularly difficult to acquire all of the images with the resolution better than 1 meter around the time of 30th June with high quality. Temporal and spatial inconsistency in the image acquisition is quite normal which has resulted in the temporal and spatial inconsistency of water data. In this study, a method of spatial and temporal consistency optimization of surface water cover data based on detailed DEM is proposed. Region growing algorithm of raster graphics and 8 neighborhood arithmetic operators were used. By iterative calculations started from the water seeds the graphics results could obtained based on detailed DEM. Through the comparison of the region growing results with water data source of the National Geographic Conditions Monitoring, the spatial correction results could be realized, so as to achieve the purpose of spatial and temporal consistency optimization. Typical study region's water data characteristics were analyzed, and optimization processing was carried out by using this method. The results show that the spatial areas relative to the fixed monitoring time point was revised 7.99%, which meets the consistency requirement of National Geographic Conditions Monitoring’s time point. The research shows that this method can make the water statistics at the national scale reflecting the same season or the acceptable time, to avoid or reduce the errors due to the spatial and temporal inconsistency in image data sources. The method could be popularized and applied for water data's optimization, and provides objective, accurate and basic water information for the government to make the decisions.
Land cover data, which plays a significant role in national geographical condition monitoring, ecological environmental protection and some other areas, is an important resource of the information on land use. At present, land cover data is produced mainly through the interpretation of remote sensing imagery and field measurement, and some limitations still exist to a certain extent. With the rapid development and wide spread of Web2.0, internet technology and various kinds of GPS equipment, the general public have the opportunity to participate in crowd sourced mapping. Volunteer users can identify the spatial distribution and attributive characters of the land cover effectively. Therefore, the classification accuracy of land cover map can be improved in the meantime. In this paper, OpenStreetMap, the most successful item of volunteered geographic information, was taken as an example on the comparison analysis with GlobeLand 30, the newly developed land cover data produced in China with 30m resolution. Firstly, the data was preprocessed and topologically checked, and then the feature relationship was established. Finally, a confusion matrix was built to analyze the consistency between the two kinds of data. The experimental results show that the missing part of the OpenStreetMap is mainly cultivated land, and the grassland and water elements are more abundant than those of GlobeLand 30. The consistency of OpenStreetMap and GlobeLand 30 is high with a value of 75%. Forest and artificial surface have the highest accuracy, and cultivated land and water body take the second place, while grassland possesses the worst consistency. The key point is to verify and determine the land cover type within the inconsistent areas, and try to find classification errors of GlobeLand 30 so as to provide basis for further modification and optimization. Volunteered geographic information contains abundant land cover information, so it can provide great potential for the development and evaluation of land cover maps. The research methods and conclusions of this paper can provide basis for exploring the application of OpenStreetMap data for land cover mapping, and provide support for assessment and improvement of the classification accuracy of GlobeLand 30.
In order to show the degree of change of the airflow at pedestrian breath height due to trees, trees with four different spatial distribution inside the ideal deep street canyon (H/W = 2) were simulated by CFD from the aerodynamic point of view. Tree canopy was treated as uniform porous media, and an additional source term is integrated to account for additional dissipation due to trees. Our results show that the effect of different spatial configuration of trees on the airflow varies greatly with trees' spatial distribution pattern: (1) Within the street canyon, uniform planted trees hinder the pedestrian airflow while non-uniform planted trees increase its rate. The effects of the on pedestrian airflow are very different under the four spatial distributions. The order of the obstruction effect of trees on airflow from the largest to the smallest is evenly spaced 8m (Spa8m) > evenly spaced 6 m (Spa6m) > evenly spaced 20 m (Spa20m) > not uniformly planted (Non-uniform). The corresponding average airflow enhancement index sequence is$\bar{D}_{spa8}$（-19.31%）<$\bar{D}_{spa6}$（-16.14%）<$\bar{D}_{spa20}$（-10.73%）<$\bar{D}_{non-uniform}$（1.25%）. (2) The pedestrian airflow win the street canyon with uneven-planting was 106.49% higher than that in the control case (uniform tree Planting, Spa8m). Uneven-planting scheme is the case that trees are planted with the sufficient free space at both ends of the street and no trees in the middle. It can allow the corner vortex to infiltrate into the middle of the street valley, promote the vertical vortex inside the street valley and the horizontal vortex movement at both ends, enhance turbulence and vertical exchange, effectively reduce the “tuyere effect” at both ends of the street and the “calm wind effect” in the middle of the street. It improves the wind environment of the entire street valley at the pedestrian breathing plane. (3) Trees with reasonable spatial distribution can improve the street pedestrian wind environment. The airflow at the breath height of pedestrians in the street valley is very sensitive to the local conditions, and the configuration (spatial cluster and density) of the trees will cause a strong spatial change of it. These results point out the importance of trees' spatial distribution in urban greening measures under existing urban building layout with the goal of improving the pedestrian wind environment, alleviating the spread of pollution and the disease by careful landscape design.
The relationship between the threshold of catchment area for extraction of river network from digital elevation model (DEM) and the density of river source was analyzed by nonlinear fitting method, using 3030 river basins chosen as study areas. The results showed that the relationship between threshold of catchment area and density of river source was a power function relationship, and the power exponent was approximately -1, which could be regarded as an inverse function. The constant coefficient k in fitting formula indicates the development degree. The larger the k value, the more the river sources per unit, the denser the river network, and the better the development of the watershed. Across China from north to south, and from east to west, the value of the coefficient k was gradually increased. The coefficient k has the maximum value in the area where the climate is humid, the slope is large, and the landform is broken. On the contrary, a minimum value was observed in the dry, wetted, flat area. According to the statistics, the mode of k appeared in the mountainous areas of the south Yangtze River. Two threshold demarcation points were discovered by analyzing the variation of the river source density with the increase of the catchment area. The two threshold demarcation points could be considered as the critical points of watershed area for watershed delineation from DEM, which can be used to distinguish the slope, the channel, and the channel and the river, respectively. According to the distribution interval of k, Gongqu river basin, Chenqing River basin and Tengtiao River basin were selected as the sample catchments. Based on the Digital Orthophoto Map (DOM), graphical information and hydrological information of the river network of the valley, the main river networks were extracted in the sample catchments. The physical meaning of the threshold demarcation points was tested by using the relative error rate of the river network density which was obtained from the analytical data. The results showed that the error of the boundary points was less than 10%. According to the dividing point of the relation curve between the catchment area threshold and the river source density, the confluence calculation method zonation of different river basin for geomorphologic unit hydrograph derivation could be ascertained.
There is a spatial unit mismatch between statistical socio-economic data that based on administrative division statistics and geographic elements expressed in spatial grid units. It requires spatial processing technique to solve this mismatch. Research on the disaggregation of socio-economic data currently focuses on the indicators of population and gross domestic product. There is relatively few disaggregation of other socio-economic indicators, such as capital stock and housing which are essential input data for risk analysis. Dozens of spatial disaggregation models exist for different research objects. According to the differences in disaggregation ideas and methods of disaggregation, disaggregation models can be classified into three categories: area weighting method, statistical model method, and multi-source data fusion method. Area weighting method is simple but was criticized by its low resolution when applied on small scale studies. Statistical model method is widely used in disaggregation of large-scale socioeconomic statistical data, but needs sufficient spatial data for spatial statistics. While these methods can produce acceptable results, their actual resolution cannot be considered ideal. With the updating and appearing of new data sources for the disaggregation of socio-economic data, multi-source data fusion method has become the main disaggregation method recently. Moreover, the data and methods needed for disaggregation are continuously improved. Comparing principles, advantages and disadvantages of different disaggregation methods, we can see that diversification of socio-economic research indicators and spatial precision requirements, and wide application of big data are the development trend of disaggregation of socio-economic data. Meanwhile, the appearance of new data source is an important opportunity for improving spatial accuracy of the disaggregation. Overall, disaggregation of the socioeconomic data will be a hot subject for future study, one of the reasons for which is the increase of the research needs for high resolution grid data. Another reason is that research institutes and publishers have paid more attention to the scientific data which is reflected by new emerging scientific data journals.
Mongolia is an important part of the Mongolian Plateau. Its land cover patterns and changes are of great significance for resources, environment, ecology, and sustainable development of Northeast Asia. In order to address issues such as lack of high accuracy and up-to-date land cover products in this region, we have used images obtained through Landsat thematic mapper and object-oriented classification method to produce data products and analyze the Mongolian land cover patterns. A classification system suitable for Mongolian land cover was first proposed taking characteristics of the Mongolian landscape into consideration. Interpretation algorithms and rules of 10 kinds of natural and artificial features were studied using the object-oriented interpolation method, and a complete remote sensing interpolation technological program suitable for Mongolian land cover mapping was developed. Land cover products for Mongolia for 2010 were obtained by interpretation of individual Landsat scenes. The accuracies of classification of the datasets in the first and second class levels were 92.34% and 80.24%, respectively. The main land cover types in Mongolia consist of barren lands, grasslands, and forests. Out of these, barren lands account for the maximum area (approximately 48.64% of the total area) with concentrated and continuous characteristics, and are mainly distributed in the south and west of Mongolia. The second largest areas are covered by grasslands (42.85% of the total area) with clear regional characteristics, and mainly distributed in the north of Mongolia and near the rivers. Forest areas constitute the smallest part, accounting for only 6.63% of the total area, and are mainly distributed in the mountains of north and northwest Mongolia. The spatial distribution of the land cover presents clear regional differences and land type transitivity. The land cover changes from barren to desert steppe, to real steppe, and finally, to forests as one moves from the south to north. The desert steppe forms a clear separate belt in the central part of Mongolia.
The uneven distribution of water resources restricts the regional agricultural production. Studying the temporal and spatial characteristics of crop water consumption based on the virtual water methods and theory can improve the rational allocation and utilization efficiency of water resources and alleviate the water resources shortage. This paper studies the temporal and spatial variability of virtual water and the evapotranspiration (ETc) of soybean in Northeast China from the virtual water perspective. The results showed that: (1) During 1986-2012, the temperature of soybean growth season obviously increased and the average wind speed decreased significantly. The relative humidity decreased and the sunshine duration increased or decreased in different districts. The climate warming and drying trend in Northeast China have deteriorated. (2) The water demand (ETc) of soybean in Northeast China was more in southwest than that in northeast, and the water demand in the south decreased while increased in the northern region. Obvious decline of the average wind speed resulted in a decrease in the soybean ETc, and the changes of other meteorological factors caused an increase in soybean ETc, with the temperature as the greatest impact on ETc, accounting for a relative contribution of 36.9%. The factors that follow were relative humidity, sunshine hours and average wind speed. (3) The spatial distribution of soybean virtual water (SVW) was higher in west and lower in east and soybean virtual water reduced in most of the sites (80.6% of the sites) from 1986 to 2012. The areas with higher SVW value were concentrated in the west of Northeast China including most of Liaoning, western Jilin and western Heilongjiang. The virtual water content gradually reduced towards the east. Climate change made soybean virtual water rise by affecting the soybean water demand, but the changes in soybean production, especially in terms of yield increase, had caused soybean virtual water decline. The impact of climate change on soybean virtual water partly counteracted the decline caused by soybean production changes. Therefore, appropriately planning the soybean productive layout in Northeast China, adopting soybean varieties such as high temperature and drought tolerance, and adjusting field management measures such as irrigation and fertilization, according to the spatio-temporal characteristics of soybean virtual water, are effective adaptation measures to improve the utilization efficiency of soybean water resources in the context of climate change.
Ecological safety is the fundamental guarantee for the sustainable development of regional society and economy. However, geological hazards will pose serious threats to the ecological safety of the area. It has great significance to analyze the sensitivity of geological hazards in Urban Agglomeration of Fujian Delta Region. In this paper, eight indicators are selected including slope, elevation, soil types, NDVI, lithology, average annual precipitation, the distance from the main road and the geological hazards points in 5km grids, for comprehensively assessing the sensitivity of geological hazards in urban agglomeration of Fujian Delta Region based on SPCA, global Morans'I and local Morans'I. The results show that the urban agglomeration of Fujian Delta Region is moderately sensitive in the overall, and the areas of different sensitivities decrease in the order of moderately sensitive>highly sensitive>light sensitive>extremely sensitive>insensitive, accounting for 26.96%, 25.67%, 23.89%, 11.75% and 11.72% of the regain, respectively. The sensitivity of geological hazards from southeast coast to northwest inland shows a gradual transition from insensitivity to extreme sensitivity with obvious zonal characteristics. There is significant spatial heterogeneity in sensitivity of geological hazards between different counties (cities and districts) and their internal areas. The sensitivity of geological hazards has significant spatial autocorrelation, which is a significant positive correlation, showing significant spatial clustering characteristics with the high-high clustering in Dehua, Yongchun, Anxi and Nanjing, low-low clustering in Huian, Licheng, Jinjiang, Shishi. The sensitivity synthesis index of geological hazards between different land use types is in the order of forestland > grassland > unused land > farmland > wetland > construction land. The forestland and grassland are overall moderately sensitive, while wetland, farmland, construction land and unused land are light sensitive in the overall. The average annual precipitation, slope, distance from the main road, number of 5 km grid geological hazards points, lithology and soil types are the main driving factors of geological hazards in the study area.
High-resolution remote sensing data has become a major source of information for natural resources monitoring, disaster assessment and ecological environmental protection and so on with the implementation of the major high-resolution earth observation system project. But the standard specification of remote sensing application is always a weak link in the development of remote sensing industrialization. On the basis of the research and analysis of the development status of standards for remote sensing applications at home and abroad, requirements for common standards for the applications system of China high-resolution earth observation system were analyzed from the actual needs of the remote sensing application business of various related industries in our country for the standard specifications. Adopting the unified modeling tool (UML) and object-oriented software analysis and design method, from the information perspective and the computing perspective, the general technical specification system framework for the applications system of China high-resolution earth observation system is proposed, including three kinds of levels including basic standards level, professional application standards level and subject application standards level, involving 7 categories of general classes, data classes, data products, application services, management, professional application and subject application, totally consisting of 42 sub categories. According to the common technical specification system framework of the application system, the research puts forward specific standards and specifications that need to be studied and formulated under each class. Among them, the basic standard is the focus of this research, and the basic standard involves 5 categories, 24 sub categories. In view of the basic standards, 48 specific technical specifications that need to be studied are put forward. In the future, the common application technical specification system of high resolution earth observation system in China will be continuously developed and perfected with the progress of the project and the formulation, implementation and test of the standard.
Compared with the unsupervised methods of residential areas extraction, the supervised methods are of relatively higher accuracy. However, the supervised methods rely on large amounts of training samples, and manually labeling residential areas is tedious and time-consuming, limiting their applications in the era of remote sensing big data. In order to improve application performances of the extraction methods based on supervised classification, it is necessary to overcome the disadvantage that training samples need to be manually labeled with high accuracy. Dense rural residential areas composed of single-family building blocks are the predominant type of rural residential areas in North China Plain. In this paper, we set extraction targets to be the dense rural residential areas, and propose a novel extraction method for high-resolution remote sensing images. The proposed method utilizes a supervised classifier but only one positive sample labeled manually is required. Firstly, four exclusion rules are designed based on the features of rural residential areas in high-resolution remote sensing images. According to the exclusion rules, all of the image blocks are classified into two categories of residential areas or non-residential areas. After the coarse classification, a certain number of the negative samples are randomly selected from the image blocks belonging to the category of non-residential areas. Then one positive sample is labeled manually, and more positive samples are collected from the image blocks in the neighborhood of the only one positive sample by performing the nearest neighbor classifier. At last, the K-Nearest Neighbor classifier is adopted to pick up image blocks which are closer to more positive samples in the feature space. The classifier filters image blocks belonging to the category of non-residential areas from the coarse extraction result, then the final extraction result is obtained. Experimental results conducted on the test images confirm that the proposed approach is both efficient and robust to images with different backgrounds.
In this study, high accuracy surface modeling (HASM) is used to simulate the spatial distribution of XCO2 with high accuracy in a wild range of regions based on OCO-2 satellite's observations. Firstly, we discussed and analyzed the feasibility of HASM simulation of the spatial distribution of XCO2 observed by OCO-2 satellite. For the OCO-2's XCO2 from September 2014 to August 2015, 90% of each month's XCO2 data were randomly selected as the input of spatial interpolation and the remaining 10% were used to verify the result of spatial interpolation.The results show that the mean absolute error of 12 months is 0.34 ppm. Therefore, HASM is suitable for simulating the spatial distribution of XCO2 observed by OCO-2 satellite. Then we use HASM to simulate monthly XCO2 distribution at the spatial resolution of 0.5°×0.5° based on OCO-2's observations from September 2014 to August 2015. Simultaneously, the corresponding TCCON (Total Carbon Column Observing Network) observations are used to verify HASM's simulations. The result shows that the mean absolute error between HASM simulations and TCCON observations is 0.81 ppm and the correlation coefficient between them is 0.88. Therefore, HASM has great advantages in simulating the spatial distribution of XCO2 observed by OCO-2 satellite.
Polar bear is one of the most important mammals in the Arctic, but its number decreased in recent years. Polar bears are sensitive to changes of the sea ice distribution and depend on sea ice as a platform for hunting, moving and reproducing. In other words, sea ice is an important part of polar bear habitat. Climate is the main factor of sea ice changes. Therefore, it is very important to understand the current situation of polar bears as well as the effect of climate on the Arctic ecosystem. Although many researchers have devoted to find polar bears habitat using aerial survey in recent years, their methods require considerable human involvement and cannot be used to detect all habitats rapidly and effectively. Thus, it is necessary to find a method to quickly assess the polar bear habitat changes. Based on the sea ice concentration products from the United States National Snow and Ice Data Center (NSIDC) and the ETOPO1 bedrock product provided by the NOAA, the inter-annual variability of sea ice concentration, open water area, sea ice retreat, sea ice advance and the length of the open water season in the Arctic were analyzed. Then, the polar bear habitat stability were analized. The results indicate that from 1989 to 2016 the sea ice concentration has decreased, open water area increased and multiyear ice decreased. Most of the multiyear ice has converted to one-year ice. The sea ice appeared later and retreated earlier, so the length of the open water season increased significantly, an increase of 72 days compared to 1992. Barents Sea is the region with the most significant changes in open water area and the length of open water season among 19 habitats, with increasing rates of 9.71×103 km2/a and 71.69 days/decade, respectively. Based on the change rates of the length of the open water season, we divide the polar bears habitats into three levels of conditions: stable, sub-stable and instable. The three stable habitats, including the Chukchi sea, Western Hudson Bay and Southern Hudson Bay are located in the lower latitudes compared with other habitats. There are 13 sub-stable habitats, including Laptev Sea, Kara Sea, East Greenland, Baffin Bay, Davis Strait, Foxe Basin, Gulf of Boothia, M’Clintock Channel, Viscount Melville, Norwegian Bay, Northern Beaufort Southern Beaufort and Lancaster Sound. The three unstable habitats are located in the north of 70°N, including Arctic Basin, Barents Sea and Kane Basin. Stable habitats are mainly in low latitudes, and unstable regions are all in high latitudes. The classification results show that the high latitude area is covered with more sea ice, but the inter-annual variation is very significant. In three unstable regions, polar bears have less time to adapt to the sea ice changes, and the inter-annual migration changes greatly, which is less favorable to the survival and development of polar bears.
Hoh Xil is located in the central Tibetan Plateau, and is the transitional zone of the natural environment of the Tibetan Plateau. In recent years, the mass balance of the glaciers in this region has a trend of positive turning to negative from west to east. However, mass change of glaciers in KangzhagRi, as the transition zone of Hoh Xil region, is unknown due to its inaccessibility and high labour costs. Glacier mass changes in KangzhaRi were determined using geodetic methods based on digital elevation models (DEMs) derived from the topographic map (1970), ASTER (2012) and Shuttle Radar Terrain Mission (SRTM) data (2000). Glacier area changes between 1970 and 2016 were derived from the topographic map, ASTER, and Landsat data. Results show that KangzhaRi has 50 glaciers with an area of 162.6±1.3 km2 in 2016. Average glacier area change was observed to be -0.08±0.02% a-1 from 1970 to 2016. Weak area shrinkage of glaciers by 0.04±0.30% a-1 during 2006-2012 and 0.01±0.38% a-1 during 2012-2016. The glaciers in this region have experienced an overall loss of 1.45±0.06 km3 in ice volume or -0.21±0.01 m water equivalent (w.e.) a-1 from 1970 to 2012. The glaciers lost mass at a rate of -0.34 ± 0.01 m w.e.a-1 during 1970-1999, while gained mass at a rate of 0.16±0.02 m w.e.a-1 during 1999-2012. Glaciers with southeastern, southern, southwestern aspect showed a slight mass loss during 1970-2012, while gained most mass during 1999-2012. However, Glaciers with southeastern and eastern aspect showed more stable in the ice cover area. Because these glaciers have a long tongue with low terminal altitudes with a little mass supply from accumulation region. Air temperature rises contribute to the loss of glacier mass during 1970-1999, while precipitation increase contributes to the gain of glacier mass during 1999-2012. Glacier area reduction from 1970s shows a trend of low to high from the west to east, and the mass balance gradually decreases from the west to east. Glacier variations in KangzhagRi were not only related to westerly circulation and monsoon circulation, but also related to local circulation. Recent mass change might be a response to the changing atmospheric circulation pattern.
The mountain altitudinal zone in Bogda Natural Heritage Site, one of the most typical representatives in the northern slope of Mount Tianshan, reflects the distribution characteristics and change rules of vegetation in the temperate desert region. To obtain the characteristics of the vegetation distribution, several steps have been designed and implemented. Firstly, the Landsat images were classified by supervised classifier for producing land cover classification results. Secondly, the vegetation coverage characteristics in Bogda Natural Heritage Site were described by a special scatterplot which integrated the land cover classification with the Normalized Difference Vegetation Index (NDVI) and Digital Elevation Model (DEM). Thirdly, the quantity ratios of each land cover type were calculated in different altitude ranges using the window-sliding method, and the boundary altitudes of two adjacent zones were identified based on the thresholding analysis of the quantity ratios. Lastly, the attribution analysis of vertical vegetation zone changes was conducted by combing the climate data (temperature and precipitation) and NDVI. The results indicated that: (1) the DEM-NDVI-Land Cover Classification Scatterplot showed the change characteristics of both NDVI and land cover classification with increasing altitude in Bogda: the NDVI changed in an inverted U-shape and the land cover classification displayed agglomeration effect in a fixed altitude range. (2) in 1989 and 2016, the upper and lower boundary altitudes of the bottom-up six vegetation zones were 1278, 1784, 2706, 3272, 3636 and 1185 m, 1759, 2730, 3293, 3690, respectively. (3) during the period of 1989-2016, the mountain altitudinal zones have an obvious response to the rising of the temperature and rainfall. The Temperate Desert Steppe Zone, shrinking downward about 93 m of its upper boundary, was the most sensitive one to the climate changes. The range of Mountain Coniferous Forest Zone expanded by 49 m to both upper and lower directions. The Mountain Meadow Zone monolithic moved up about 20m with an unchanged span, and the Alpine Snow-Ice Zone, retreating upward about 54m of its lower boundary, was affected by the global warming.
The arid zone refers to the arid climate area, which accounts for about 30% of the land area. The common characteristics are: the precipitation is low and the change rate is large, the ranges of both the daily and annual temperature are large, and the evaporation is far greater than the precipitation. Thus, the estimation of evapotranspiration in the Ugan and Kuqa Rivers Delta Oasis plays an important role in the management and allocation of water resources. At the same time, it has a great significance to understand the water cycle and hydrological process of the study area. In this paper takes the Ugan and Kuqa Rivers Delta Oasis as the study site, the daily evapotranspiration of Ugan and Kuqa Rivers Delta Oasis in 1997(August 30) and 2016(August 2) was estimated, and the spatial distribution characteristics of daily evapotranspiration, and the correlation between daily evapotranspiration and relative surface parameters were analyzed by combining Landsat TM/OLI remote sensing images with meteorological data, and using SEBAL model. The results showed that the spatial distribution of daily evapotranspiration in the Ugan and Kuqa Rivers Delta Oasis was characterized by distribution patterns of high to low from west to east, and from north to south. The high value appeares in the oasis area covered with abundant water and vegetation, and the low value is distributed in the east, southeast and west part of the study area with the low vegetation coverage. In terms of the correlation between the daily evapotranspiration and surface parameters, such as vegetation indices (NDVI, SAVI, LAI), water indices (MNDWI, NDWI, TVI, VSWI, TVDI) and land surface temperature (LST), the daily evapotranspiration showed the good correlation with every index. The Pearson correlation coefficients past the significance test at 0.01 level, especially with a best correlation with NDWI. the results indicate that water bodies are the most important factor for distribution and changes of the daily evapotranspiration in the Ugan and Kuqa Rivers Delta Oasis.
