Traditional surveying and mapping pay much attention to “accuracy”, but ignore the “speed”. Modern emergency mapping needs to be “speedy and accurate”. If traditional mapping techniques are used, it will consume a lot of time, which will affect the timeliness of emergency response. A tethered UAVs can collect high-quality data and realize real-time long-term video monitoring. This paper introduces the application status, characteristic advantages, and application scenarios of the tethered UAVs. And a kind of experience and method for using tethered UAVs to collect high quality image data and video data for producing surveying and mapping data products and identifying video targets through Darknet deep learning framework is proposed. Based on many simulation experiments and practical applications, the authors hold that this method is effective in providing timely and effective surveying and mapping guarantee for disaster relief and rescue.
At present, the technology of generating high precision and high current digital elevation model (DEM) based on airborne light detection and ranging (LiDAR) data has been widely used. It is urgent to control the quality of DEM data scientifically and efficiently. In this paper, the authors introduce the three-step inspection method based on practical production experience, which includes human-machine interpretation discriminant inspection, semi-automatic inspection and automatic inspection through python. This method comprehensively covers the quality inspection of DEM generated by airborne LiDAR technology and provides the solution, thus improving the efficiency and accuracy of quality inspection and filling the blank of nonexistence of ready-made software and methods for completing the work.
Mean Shift algorithm has been widely used in image segmentation because of its fast convergence speed and good segmentation accuracy. However, when large scale remote sensing images are processed, Mean Shift algorithm has some problems, such as slow speed and low efficiency. In this paper, a parallel seamless segmentation algorithm based on Mean Shift is proposed. The algorithm is based on block parallel Mean Shift segmentation. The elimination criterion of block lines is determined by uniform coding of label images and establishing corresponding relations between label values of overlapping regions. Then, the row and column directions of the label image are stitched together. Finally, the segmented label image is vectorized to generate the final segmentation result. Compared with the original Mean Shift algorithm, the algorithm put forward in this paper can not only ensure the reliability of segmentation results but also greatly improve the efficiency of image segmentation, and can also solve the problem of large scale remote sensing image segmentation.
Owing to influences of the same spectrum with the different thing and the same thing with the different spectrum, the medium resolution remotely sensed image, Landsat8 OLI, extracts the wheat extraction with the wrong information, which leads to low accuracy. The coarse resolution image with multi-temporal trait can discriminate the wheat information from other similar land cover. In this paper, the multi-temporal trait is adopted to solve the “wrong coming or wrong going” error of the OLI classification so as to increase the wheat extraction accuracy. The experiment shows that the OLI and MODIS can extract the wheat with high consistence, so the result of MODIS can correct the error of the OLI, where the phenomenon of the same spectrum with the different thing and the same thing with the different spectrum occurs. In the region of the same thing with different spectrum, the RMSE of OLI result is 0.758, while that of the MODIS correction result is 0.142. In the region of the different thing with the same spectrum, the RMSE of OLI result is 0.901, while that of the MODIS correction result is 0.122. All the results show that the MODIS result can correct OLI result for higher wheat extraction accuracy, which can solve the phenomenon of the same spectrum with the different thing and the same thing with the different spectrum.
For the problem of degrading and blurring in remote sensing images, the classical image restoration methods have poor restoration effect due to the difficulty of estimating the blur function. In order to avoid the difficulty of estimating the blur function, the authors have studied the image restoration method based on Conditional Generative Adversarial Nets (CGAN) through depth learning. Firstly, the training database of the training network is created, and then the initial parameters of the training network are set. The network alternately learns the generator model and the discriminator model in the adversarial way. By learning the difference between the degraded image and the clear image continuously and combining the adversarial loss with the perceptual loss, the difference between them can be reduced and the image restoration can be realized. A Hybrid blur training library based on GOPRO data set is used to train the network, and is compared with other methods. The results show that this means has better restoration effect in image details and evaluation indexes. The details and texture information of the restored image are guaranteed, and the method of conditional generation antagonism network is proved to be applicable to the restoration of remote sensing image.
In this paper, the authors used a U-net model to conduct water extraction, and the result was compared with that of the random forest model. The accuracy of the U-net model was validated by using GF-1 images in Chaohu Lake Basin. The results show that both models are of high accuracy for large area of water body, but random forest model has more spots for small area of water body, and the result of U-net model is more consistent with the manual visual interpretation result. Moreover, the U-net model can effectively remove the shadows of mountains and buildings. The result indicates that U-net model performs better than random forest model with the overall accuracy of 98.69%, Kappa coefficient of 0.95, omission error of 1.90% and commission error of 1.18%. In contrast, the overall accuracy, Kappa coefficient, omission error and commission error of random forest model are about 98.05%, 0.92, 1.61% and 2.99%, respectively. In addition, the classification features for traditional machine learning model are always calculated by manual extraction. However, the inputs for U-net model are the 4 band spectrum data of GF-1 images. These data suggest that the U-net model avoids the process of manually extracting classification features and has a higher degree of automation. It should be noted that the U-net model uses more train samples with less time-consuming. It is believed that this model can significantly improve the surface water detection accuracy and can be used for the automatic renewal of a larger range of water bodies.
The distribution of combustibles in forest is one of the important factors that affect the occurrence and spread of forest fires. The purpose of this study is to combine the traditional forest survey data with point cloud data from light laser detection and ranging(LiDAR), slope and meteorological factors so as to evaluate forest fire potentials with fuel characteristic classification system(FCCS). Pu’er City of Yunnan Province was selected as the research area in this paper. An object-oriented based segmentation was performed based on the crown height model(CHM) which was produced by the airborne LiDAR data, and the overlay analysis of the provincial level inventory data of forest resources of the research area was used to determine the division unit and vegetation type according to the flammability of vegetation, which was divided into coniferous forest, broad-leaved forest, shrub and bamboo forest. On such a basis, stratified random sampling was used to form the validation dataset. Then the authors extracted the LiDAR variables and applied the multivariate stepwise regression method to analyzing the extracted variables with the reference data set to obtain the forest parameters of different vegetation types. In the end, the forest parameters together with the meteorological factors were used as inputs to the forest fire classification model (FCCS), and the fire potential of each segmentation unit was calculated by the model. Finally, the authors compiled maps of potential fire behavior, crown fire, effective combustibles and comprehensive fire hazard result. The results showed that the overall fire potential level of combustible materials in the research area is relatively low, which is consistent with the actual situation in the study area; the vertical structure of the forest is closely related to the forest fire risk potentials. Accurate estimation of forest parameters plays a very important role in the mapping of combustibles.
In view of the shortcomings such as complex algorithm design, code implementation, high cost and long period in the existent methods for extracting skeleton line from arbitrary polygons, the authors put forward an automatic method for extracting skeleton lines from arbitrary polygons based on geographic information system (GIS) space analysis. First, the electronic maps are taken as the data sources, and the polygon vector boundaries of the spatial objects are extracted and preprocessed by using the model tool of ENVI classification in the ArcToolbox toolbox with the platform support of the ENVI and the ArcGIS. Secondly, variable tools such as data processing, spatial analysis and file conversion are combined for extracting the nodes of skeleton lines and post-processing them. Then, the object oriented programming language of Python is used and combined with the ArcPy package for programming the scripts to extract the skeleton lines automatically. Furthermore, the visual modeling tool of ModelBuilder is used to build the model for extracting the skeleton lines automatically. Finally, the method is applied to the extraction of the skeleton line of the road and the building polygons boundary respectively. The experimental process and its results show that the method has characteristics of effectiveness, practicality and operability.
When coastal hyperspectral remote sensing measurement and survey are conducted, the water surface cannot be used for ground control point measurement, and hence the accurate external orientation element of the data cannot be obtained by the traditional aerial triangulation method. Therefore, how to ensure the geometric accuracy of the aerial remote sensing data is one of the key problems in measurement. In this study, the authors summarized and analyzed the geometrical correction principle and model characteristics of CASI 1500H push-broom airborne hyperspectral instrument and designed a set of geometric calibration schemes for this system. The calibration results show that the geometric accuracy of CASI 1500H hyperspectral image can still be significantly improved without control points. Using this geometric calibration method, the authors acquired CASI airborne hyperspectral data of Dajin Island and its surrounding waters. Based on these data, the authors retrieved the suspended sediment concentration in the surrounding waters of Dajin Island, and the overall accuracy was better than 70%, which can meet the need of coastal airborne remote sensing survey.
Mobile laser scanning can acquire lots of dense point clouds. Therefore, how to get high-quality road point clouds is a problem worthy of further study. This paper proposes a method for automatic extraction of roads from mobile laser scanning point clouds by image semantic segmentation. The authors use a four-step strategy: First, semantic segmentation images are created using 2D panoramic images. Then, fusion and matching are conducted to get rough classification results. After that, the 3D Hough transform is used to get the segmentation plane before fitting. Finally, a finely classified point cloud is obtained through local optimization operations. The authors extracted and evaluated two different points of cloud data on urban roads. The accuracy and integrity are all over 99%. The extraction quality is high enough to adapt the application requirements in practice. The method proposed by the authors can extract road point clouds in different situations and has less primitive constraints on point cloud data. It shows a significant improvement in both universality and robustness compared with other methods.
Leaf angle is an important parameter to describe the canopy structure of vegetation. The leaf angle distribution (LAD) determines the interception of vegetation canopy and is an important parameter in quantitative inversion of remote sensing. The current method of measuring the leaf angle is time-consuming, labor-intensive and subjective, with no accuracy guarantee. In this paper, image-based probability density function extraction for LAD of corn plant is proposed, which can extract LAD of corn plant quickly and accurately with low cost. Firstly, the skeleton is extracted from the image. Secondly, the information such as burrs and stems in the skeleton image is removed to obtain the leaf skeleton. Finally, the leaf angle is extracted by searching the skeleton with a search window of size 2×20. The results of precision evaluation show that the correlation coefficient between the measured value of the corn dip angle and the extracted value is 0.821 4, and the correlation coefficient between measured and extracted values of the corn leaf angle at jointing stage is 0.908 7, which suggests that the method is feasible and accurate with low cost.
In order to obtain the best atmospheric correction method for real surface reflectance of SPOT6 images in high-altitude complex terrain areas with less research, the authors used the 6S model and FLAASH model to perform atmospheric correction for the SPOT6 image covering Huangshui River basin in eastern Tibet Plateau. For the 6S model, the images were processed by AVG6S and GRD6S according to the average aerosol optical depth (AOD), altitude parameters and gradient AOD as well as altitude parameters. The calibration results were verified with the Landsat8 SR surface reflectance product. The results show that the image quality is significantly improved after atmospheric correction, and the reflection characteristics of various ground objects are more realistically reflected. Correlation analysis and a comparison with typical ground reflection spectrum curve and normalized difference vegetation index (NDVI) show that the overall performance of AVG6S is the best, whereas GRD6S performance is more prominent in urban and high mountain areas. The calibration result of the 6S model is better than that of the FLAASH model and hence the 6S model is an atmospheric correction method more suitable for high altitude region.
Atmospheric correction is an important part of remote sensing image processing. It is of theoretical significance to explore the characteristics of normalized difference vegetation index (NDVI) and its terrain gradient before and after atmospheric correction. This paper draws the following conclusions: ① The NDVI obtained without atmospheric correction is generally underestimated in space. ② The NDVI obtained without atmospheric correction cannot reflect the trend and proportional relationship of each stage. The NDVI without atmospheric correction has serious deviation, and the NDVI is more than 0.6. The absolute error is over 20%. ③ Whether the atmospheric correction is made or not affects the NDVI change trend and the numerical value of each altitude; the absolute error increases below 1 000 m, and the absolute error fluctuation decreases thereafter. ④ Whether the atmospheric correction is made or not has an effect on the slope NDVI trend and value; as the slope increases, the absolute error first rises and then falls, and the slope (45°, 50° ) absolute error is the largest. ⑤ Whether the atmospheric correction is made or not has an effect on the NDVI trend of each slope; the absolute error of west slope is the largest, and the east slope is the smallest.
High-resolution remote sensing data have rich information and hence can be used to better identify micro-lithology, structure and other ore-controlling information. In order to study the extraction of ore-controlling information of high-resolution remote sensing data in Rongle area of Tibet where the elevation is high and the bedrock is exposed, the authors interpreted the ore-controlling information of strata, lithology, rock mass, structure, contact zone and other information closely related to copper polymetallic mineralization in this area on the basis of the best band combination determined by WorldView-2 high-resolution remote sensing data. Using ASTER data, the authors employed principal component analysis (PCA) to extract the mineralization-related remote sensing alteration information of iron stain, aluminum hydroxyl group and magnesium hydroxyl group. Based on comprehensive regional copper mineralization rules, the authors constructed a remote sensing prospecting model for ore-controlling information of magmatic hydrothermal copper deposits through human-computer interaction interpretation. Five prospecting targets were delineated and good mineralization clues were found through field investigation. The results provide basic information and reference basis for promoting local prospecting process. The results show that the high-resolution remote sensing data can meet the needs of mineral resources exploration and resource evaluation in the high-altitude environment hostile areas in Western China and highlight the fast, efficient and reliable application effect of high-resolution remote sensing.
Based on MODIS product and the pollution concentration measured near the ground, the authors analyzed the spatio-temporal characteristics of aerosol optical thickness (AOT) and the regression on AOT and air quality index (AQI) by different lengths and seasons in Xiamen City, by using geographic information system (GIS) technology and statistical regression method. The results showed that there was a distinct change in the spatio-temporal characteristics of AOT from 2000 to 2015; for example, the AOT highest monthly average 1.13 appeared in April and the lowest 0.64 appeared in January, AOT seasonal average tended to decrease from spring through summer and autumn to winter, and its yearly average showed a steady trend of slow rise. The higher values of monthly and annual average AOT were almost distributed in the coastal areas and the lower values occur in northwest and northeast regions. R 2 of regression model of power function for AQI and AOT was the highest in the five regression models with its value being 0.388 3. AQI was divided into groups with a certain step length, and the regression model with AQI and AOT was built up, which exhibited larger step length and higher R 2. According to AQI grading length 50, the precision of the forecasting AQI value and the actual value could reach 77.35%, which could on the whole meet the demand of air quality level forecast. R 2 and the precision of the four-season regression models were a little higher than those of the full year regression models, and the R 2 was the lowest in spring season, R 2in other three seasons is almost the same, with the precision up to 83.33%. With the present remote sensing technology for air pollution monitoring, the utilization of the correlation models to estimate the level of air quality seems to be feasible.
Mining intensity can reflect the centralized distribution of mineral resources exploitation and provide a basis for decision-making about mineral resources planning, overall planning of local economic development, etc. The mining occupation and destruction land of each administrative region in Tibet were surveyed by field investigation and information extraction from remote sensing data acquired in 2016 and 2017. The mining intensity and changes of each administrative region in 2016 and 2017 were analyzed on the basis of the results of remote sensing investigation. The results show that, for mining intensity, five counties with the highest intensity include Maizhokunggar County, Doilungdeqen District, Dagze County, Chengguan District and Zhongba County. From 2016 to 2017, for mining intensity changes, five counties with the highest increasing mining intensity were Zhanang County, Maizhokunggar County, Doilungdeqen District Sangzhuzi District and Nedong County, whereas the county with most weakening mining was Chengguan District.
As a key and demonstration area for the implementation of the national ecological project of returning farmland to forestry, Yan’an City is a concentrated area of ecological problems in China. It is of great significance to study the spatial and temporal differences of land ecological risks for the sustainable development of regional land and the formulation of differentiated land and resources development policies. According to the basic characteristics of land ecosystem, four risk indicators, namely vegetation coverage, percentage anomaly precipitation, land use structure risk index and soil erosion index, were selected to construct a comprehensive evaluation model of land ecological risk. Then on the basis of pixel scale, each factor index and comprehensive index of land ecological risk were calculated. In combination with exploratory spatial data analysis (ESDA), the spatial and temporal evolution of land ecological risk and spatial agglomeration effect in Yan’an City from 2000 to 2015 was analyzed. Then the corresponding suggestions for comprehensive management of land in different regions were put forward. The results are as follows: The land ecology of Yan’an City is in good condition as a whole, whereas the land comprehensive ecological risk and the risk of the four ecological factors temporally decrease on the whole; nevertheless, the area of Baota District and Luochuan County is higher in this aspect, and high comprehensive risk areas increase slightly. The land comprehensive ecological risk of Yan’an City shows a strong spatial agglomeration. Hot spots include urban hot spots located in urban construction areas and northern hot spots distributed in five districts and counties of Zichang County, Ansai District, Yanchuan County, and Wuqi County. The cold points are mainly located in Huanglong County, Yichuan County, Huangling County, Fuxian County and the southwest area of Ganquan County. Thanks to the implementation of national eco-engineering measures such as returning farmland to forestry (grassland) and closing hillsides for reforestation, the agglomeration degree of hot spots has been gradually weakening. However, the area of urban hot spots in Baota District continues to increase, and hence attention should be paid to strengthening ecological management. As cold spot areas in northwest Fuxian and western Yichuan County has been shrinking because of climate drought, attention should be paid to optimizing the allocation of water resources.
The Chaerhan Salt Lake contains abundant inorganic salts such as sodium chloride, potassium chloride and magnesium chloride, and it is also one of the mining bases in China. In recent years, with the rapid development of industrial mining activities in the Chaerhan Salt Lake and its surrounding salty field, the water quality of the salt lake has been polluted, and the lake water area has also rapidly decreased. By calculating the normalized vegetation index (NDVI) for the study area in 2002—2018, the authors obtained vegetation changes in the study area. Using remote sensing monitoring methods and combined with various factors such as rainfall and industrial development, the authors studied evolutionary trends and driving factors of the Chaerhan Salt Lake. Some conclusions have been reached: ① The development of salty field mine is the main factor causing the degradation of the Chaerhan Salt Lake. With the increase of the mining area of salt fields, the area of natural salt lakes has been greatly reduced, the amount of water has been reduced, and the salinization of water bodies has become serious. ② The area of the Chaerhan Salt Lake is affected by precipitation. In the years of abundant rainfall, the area of the salt lake has been larger, and smaller changes occur in the area of salt lakes in less rainy years. ③ Mining of salt fields will affect the growth of vegetation. According to the field investigation, salt field mining leads to serious salinization of salt lake water, and vegetation can hardly grow in the high salinity area, so there is less vegetation around the salt lake.
Remote sensing alteration anomaly is an important indicator of ore deposits. However, the previous studies of the geological genesis of remote sensing alteration anomaly and its indicative significance are insufficient and, as a result, the interpretation of remote sensing anomalies is often uncertain. In this paper, the authors established a multi-data source and multi-method collaborative processing method that can quantitatively explain the geological genesis and indication significance of remote sensing alteration anomaly. In this method, multi-spectral remote sensing is used to analyze the distribution characteristics of remote sensing alteration anomaly, hyperspectral remote sensing is used to decompose the remote sensing alteration anomaly information development pattern of typical geological bodies, and X-diffraction-rock identification-spectral solution is used to accurately identify altered mineral types. On the basis of the above experimental results, the geological genesis and indication significance are comprehensively explained. The test conducted in Fangshankou area of Beishan shows that the types of remote sensing alteration anomalies are basically consistent with the types of altered minerals developed on the surface, and the surrounding rock alterations of different ore-forming types of deposits can be effectively reflected by the specific altered mineral information combination. According to this regularity, lots of mineralization clues were discovered in this experiment, which realized the rapid transformation of remote sensing information into geological information. The study results show that this collaborative processing method would overcome the incompleteness of the analytical results of a single data source or a single method, and could improve the credibility of remote sensing alteration anomalies in geological applications.
Analytic hierarchy process (AHP), as a comprehensive safety evaluation method combined with qualitative analysis, has been used in many fields of safety and environmental science. Problems of environmental geology in mines are affected by many factors. This study focuses on the ecological environment evaluation of the mine by using data on land occupied and damaged by mines, according to the characteristics of mining combined with relevant information of Hainan Island. Ultimately, mining environmental grade is divided into four levels. Through field examination and verification, it is found that the theoretical value is very compatible with the actual situation. The results show that the weight calculated by this method is scientific and reasonable, and the evaluation is objective. This method is worth popularizing in mine environment evaluation.
Using MODIS normal difference vegetation index (NDVI) data from 2001 to 2015 and climate data acquired from 7 meteorological stations in Hainan Island, the authors analyzed vegetation variability under the influence of climate variations and human activities based on residual analysis method, unary trend curve regression model and relative role analysis method in Hainan Island. The results showed that NDVI observed by remote sensing (MODIS NDVI) increased over the past 15 years with a rate of 0.024 per ten years in Hainan Island. The proportions of increasing vegetation and decreasing vegetation areas were 77.77% and 22.23%, respectively. As identified by remote sensing observations in Hainan Island, the relative roles of climate changes and anthropogenic activities in vegetation increase areas were 31.04% and 68.96%, while the roles of climate variations and human activities in vegetation decrease areas were 35.03% and 64.97%, respectively, indicating that human activities played a major role in vegetation changes. The area of increased vegetation mainly influenced by human activities (relative role >50%) accounted for 80.79% of the whole increase vegetation area, which was associated with the large-scale rubber plantation in these areas. In contrast, the area of reduced vegetation mainly induced by human activities (relative role > 50%) accounted for 75.59% of the reduced vegetation regions, which may be induced by the increased urbanization and urban expansion in coastal regions. On the whole, climate changes could promote the vegetation growth, whereas human activities played a greater role in the vegetation increasing than in the vegetation decreasing.
In order to compare the validation performances of different validation methods on the GF-1/WFV winter wheat LAI retrieval results, the authors chose Yancheng, Luohe City of Henan Province as the study area. Three methods, i.e., single point ground measurement validation, multi-point upscaling validation, and high-resolution result validation, were tested to verify the performance of winter wheat LAI inversion based on GF-1/WFV image. The results show that the RMSE obtained by the above three verification methods are 0.57,0.80 and 0.46,respectively. The correlation coefficients are 0.885, 0.508 and 0.867,respectively. The multi-point upscaling method has higher requirements for the number of sampling points and the position of sampling points. Therefore, the accuracy is low and the effect is poor in the case of fewer sampling points in this study. The other two methods have relatively high precision and applicability, and the validation method with the introduction of high-resolution image achieves higher precision, and hence this method is more suitable for the validation of LAI inversion of GF-1/WFV images.
The soil salinization in Hetao irrigation district of Inner Mongolia has exerted severe impact on the sustainable development of local agriculture and economy. Remote sensing can be applied to achieve the real-time information of soil salinization so as to monitor salinization’s future changes. The authors used the satellite images of Landsat to extract salt index (SI) and modified soil mediation vegetation index (MSAVI) and then combined them to construct modified salinization detection index (MSDI) model so as to quantitatively analyze and monitor the soil salinization in this research. After that, the soil salinization information in the study areas obtained in 2001, 2010 and 2017 was further classified and statistically analyzed, which showed an obvious diversity of MSDI mean among various alkali soil types. The result of MSDI was validated by the precision test, field investigation and the salinity of soil samples. The validation demonstrated a strong correlation of 0.856 8 between MSDI and soil salinity, a precision test accuracy of 87.5%, and a Kappa index of 0.726. The soil salinization of this area had been mitigated according to portion changes of non-alkali soil area (from 18.5% to 30.47%) and the fragmented tendency of moderately saline land since 2001. The result indicates that MSDI based on the SI-MSAVI feature space could be applied to quantitatively extract the information of soil salinization and proves to be efficient in monitoring the development of salinization in this region.
In this study, the authors used the CLUMondo model which can deeply describe the intensity of land and the land use data of 2010 and 2015 to simulate the spatial distribution pattern of land use in the three different scenarios of “natural growth”, “economic development” and “land use optimization” in coastal cities of Guangxi in 2025. Some conclusions have been reached: The CLUMondo model can effectively simulate the development status and trajectory of land system in large-scale coastal areas; under the“natural growth” scenario, the intensive and effective use of land resources in coastal cities has been slower; under the “economic development” scenario, urban and rural construction land is growing rapidly and is closely related in space. There is a sharp contradiction between regional forest and cultivated land protection and industrial construction; under the “land use optimization” scenario, the pace of regional economic construction has gradually slowed down, and the construction of regional cities has formed a trend of concentration of resources to cities and towns and concentration of farmlands. The simulation results provide a certain reference for the future land use planning and related system formulation of coastal cities in Guangxi and even the whole country.
Water transparency is the key factor of airborne LiDAR bathymetry. Turbid waters produce noise in LiDAR echo signal and weaken the laser pulse or cause a gap. Therefore, it is necessary to study water optical properties. Using MODIS Kd (490 nm) and general bathymetric chart of the oceans (GEBCO) bathymetric data, the authors calculated the maximum detectable depth in China’s coastal area based on CZMIL Nova, and classified the result into 3 types. CZMIL test data from different areas were used to verify the accuracy of the classification. The results show that a total of 211,900 km 2 sea area is suitable for the performance of bathymetric survey with airborne LiDAR. The coastal area of Wenchang to Dongfang of Hainan, Beihai and the east and west of Leizhou peninsula, Rizhao to Qindao of Shangdong and Yinzhou to Suizhong of Liaodong Bay are suitable for the performance of land and sea integrating topographic survey, with the maximum measurable depths estimated by Kd being 20~40 m, 10~20 m, 20~25 m, 10~15 m, respectively.
With the purpose of exploring the extraction of early paddy rice area distribution information from bipolar Sentinel-1A Radar image data recognition and on the basis of an analysis of backscattering coefficients of typical terrain objects, the authors employed the idea that polarization differential SAR images and polarization ratio SAR images play an important role in the classification of typical terrain objects and proposed the utilization of the normalized parameters of water body. Then, the support vector machine (SVM) classification method and the threshold classification method were used to extract the area of early paddy rice from the normalized polarimetric SAR data of single-phase and multi-temporal water body on March 10, March 22, April 3, April 15 and 15 April 15 in 2017. The results show that the threshold classification method is better than the SVM classification method. The overall accuracy of the former method is 89.01%, Kappa coefficient is 0.823 1, mapping accuracy and user accuracy of early paddy rice are 92.68% and 82.26%, respectively. The planting area is 129,000 hectares, which is basically consistent with the spatial distribution of the main early paddy rice production bases in Lingao County. It can be concluded that multi-parameter polarimetric SAR data can improve the accuracy of recognition and extraction of terrain objects. The best monitoring data for extracting early paddy rice area are multi-temporal NDVH polarimetric SAR data.
Studying the changes of evapotranspiration of degraded grassland in Inner Mongolia is conducive to understanding the water cycle of the degraded grassland ecosystem in this region and provides an important basis for the rational use of grassland water resources in this region. In this study, Duolun County in Inner Mongolia was taken as the research object. MODIS data during the flourishing period from 2001 to 2017 were used to invert the spatial distribution and variation of vegetation coverage and daily evapotranspiration in this region for nearly 17 years, and the effects of land use and grassland degradation on evapotranspiration were analyzed. Some conclusions have been reached: ① The order of vegetation coverage of different land use types is forest land >farmland>residential construction land>grassland>unused land>water area; the order of daily evapotranspiration of different land use types is water area>forest land>farmland >residential construction land>grassland>unused land. ② From 2001 to 2017, the grassland types in Duolun County are mainly of the III and IV grades; the area of low-grade grassland has a downward trend, while the area of high-quality grassland has an upward trend, which indicates that the protective measures for grassland in this area achieved certain results. ③ During 2001-2017, there was no obvious change trend in the daily evapotranspiration of all grassland grades; except for a few years, vegetation coverage and daily evapotranspiration showed consistent interannual fluctuations. ④ According to the spatial and temporal distribution pattern of vegetation coverage and daily evapotranspiration, it is concluded that there is a positive correlation between vegetation coverage and daily evapotranspiration.
The inversion of reservoir parameters and production for the oil field can grasp reservoir status and production changes in time and effectively monitor reservoir health and safety. At present, the study of reservoir parameter inversion is very insufficient in China. The authors chose Shuguang Oil Production Plant, the largest oil production plant in the Liaohe Oilfield, as the research object. Using 21 L-band ALOS/PALSAR data obtained from January 2007 to September 2010, the authors employed StaMPS to extract deformation results. Based on these deformation results, the authors used Mogi model and Finite Prolate Spheroidal model to invert and analyze reservoir parameters respectively, with the inversion results compared with those of Okada model. The results are as follows: ① The subsidence of Shuguang Oil Production Plant is remarkable. The maximum subsidence rate is -189.6 mm/year, the maximum cumulative subsidence is about 750 mm, and the subsidence area is about 28 km 2. ② Compared with Okada model and Mogi model, Finite Prolate Spheroidal model has the highest accuracy of reservoir depth inversion, and the simulated deformation results are in the best agreement with the observed deformation results, which shows that the inversion results of Finite Prolate Spheroidal model are more reliable and more suitable for the inversion of reservoir parameters in this oilfield. This study can provide scientific reference for InSAR subsidence monitoring and reservoir parameter inversion in the oilfield.
In order to explore the recognition accuracy of remote sensing technology of low-altitude UAV for surface features in agricultural areas with different forms under karst landform conditions, the authors chose three agricultural areas (each having an area size of 200 m×200 m) in Guilin City as the research object. Supported by UAV aerial images and ground survey data, the image analysis technology based on pixel and object-oriented was combined with support vector machine (SVM) algorithm, respectively, to build the remote sensing recognition model of agricultural areas under different geomorphological conditions, and the precision was comparatively studied and analyzed. The results show that the object-oriented SVM classification results retain the rough outline of the original ground features, and the plot is relatively complete, and hence this means is more suitable for the recognition of ground features in agricultural areas under karst landform conditions. Compared with the pixel based SVM classification method, the overall accuracy is higher by 6.54% , and the Kappa coefficient is higher by 0.135 . The SVM classification method based on pixel is suitable for feature recognition in agricultural areas with regular feature distribution. Compared with the object-oriented SVM classification method, the overall accuracy is higher by 2.92% and the Kappa coefficient is higher by 0.026 .
To investigate the geological background of Duyun tea planting, the authors studied the relationship between geological background and high quality Duyun tea planting efficiently and rapidly based on the theory of agricultural geology and using RS and GIS technology. Firstly, by using Sentinel-2A and Landsat8 remote sensing images and object-oriented classification method in combination with customized TEI and tea phenology information, high-precision extraction of tea planting areas was realized. Then, the spatial analysis function of GIS was used to count the distribution of various geological backgrounds in Duyun and the area of tea planting on different backgrounds accurately, and the geological background of tea planting in Duyun was investigated comprehensively. Lastly, the geochemical characteristics of rocks with different backgrounds were investigated by testing and analyzing the sampled data, and the planting map of suitable geological background area for tea planting in Duyun was compiled. The results show that the chemical elements in Duyun clastic rocks are obviously better than those in carbonate rocks. Clastic rocks are the advantageous geological background for tea planting and the important influencing factors for Duyun to produce high quality tea. The distribution area of clastic rocks accounts for 47.86% of the total area of Duyun, which means that clastic rocks possess the largest proportion of all lithologies, followed by dolomite, which accounts for 16.80%. 67.82% of Duyun tea is planted in clastic rock background, which is the basic condition for local tea production.
Taking the two-year high-resolution satellite image data obtained in 2016 and 2017 as the information source, the authors processed the data of two-year remote sensing image, established the mine remote sensing interpretation mark according to the image features, and verified some map-spots in the field of Zhejiang Province. Remote sensing investigation and dynamic monitoring of the geological environment and restoration of mines in Zhejiang Province were carried out in the natural environment and topography. Through the application research, the status and the changing trend of the mine geological environment in the mining area have been clarified, which can provide a scientific basis for the government departments to formulate the decision for the next step of mine geological environment restoration and management as well as the rational development and utilization of mineral resources.
The correlation analysis between climatic factors and vegetation indexes can not only reflect the impact of climate change on vegetation but also predict the trend of vegetation in the future. Based on the data of MODIS13A1 C6 NDVI of the Three River headwater region and combined with 1∶1 000 000 map of vegetation types and meteorological data, the authors analyzed spatial-temporal characteristics of NDVI and the relationship between vegetation indexes and climatic factors by using correlation analysis from 2000 to 2016. The results are as follows: ① NDVI increased with a rate of 0.8%/10a in Three River headwater region during 2000—2016, whereas vegetation cover increased from the northwest to southeast. ② Climate factors had a greater influence on vegetation growth in the early and middle growing season, but the correlation between NDVI and climate factors was not obvious in the later growing season. ③ The partial correlation between NDVI and climate factors in the vegetation growing season of the study area showed that the influence of the air temperature on NDVI of the alpine meadow grasslands and alpine grasslands was greater than that of precipitation in the early growing season. However, during the middle of the growing season, the precipitation had a greater impact on the growth of three different types of vegetation.
Based on Landsat data, the authors extracted the impervious surface coverage of the two sceneries in Xixian New District in 2007 and 2016 by the linear spectral mixture model decomposition method, and extracted the artificial thermal information by the surface energy balance method in the same period, and investigated the relationship between them. The results are as follows: ① From 2007 to 2016, the impervious surface expanded from 294.93 km 2 to 362.62 km 2, and gradually changed from natural surface and low coverage to medium and high coverage. ② In 2016, the regional differences of anthropogenic heat in the study area were significant. The high-value areas were concentrated in the north-central part of Fengdong New Town and around Xianyang International Airport of Airport New Town, and were scattered in the central part of Qinhan New Town, northern part of Fengxi New Town and part of Jinghe New Town. ③ The mean values of impervious coverage and anthropogenic thermal mean values of land use showed the tendency of construction land>cultivated land>woodland>water body. ④ There was a positive correlation between impervious coverage and artificial heat, with a correlation coefficient of 0.97. The rate of increase of artificial heat values with impervious coverage had the tendency of Airport New Town>Fengdong New Town>Jinghe New Town>Qinhan New Town>Fengxi New Town.