Habitat quality has been widely regarded as a proxy indicator for biodiversity. In the context of the continuous decline of global biodiversity and degradation of ecosystem services， it is of great significance to study regional spatio-temporal trends of habitat quality for its sustainable development. Over the past 40 years， rapid urbanization has profoundly impacted the spatial and temporal distribution and functions of habitats， leading to ecological degradation and the weakening of ecosystem service functions. Exploring the spatial and temporal dynamics and driving factors of regional habitat quality is important for ecological restoration and biodiversity conservation. Based on remote sensing monitoring data of land use in the Beijing-Tianjin-Hebei （BTH） region from 1980 to 2020 （every 10 years）， this study used the InVEST model to quantify the regional habitat quality and explore the temporal and spatial evolution characteristic， then it analyzed the response of habitat quality to land use/cover change， socio-economic development， and natural conditions by adopting a Geographically Weighted Regression （GWR） model. The results showed that Habitat Quality （HQ） in the BTH region has shown a decreasing trend during the study period， and the area of low-quality areas has increased significantly， lower-quality areas and medium-quality areas have shown a decreasing trend， while higher-quality areas and high-quality areas are relatively stable. The spatial pattern of HQ in the BTH region was that the northwest areas were relatively high and the southeast areas were relatively low. The habitat quality of eastern coastal areas and western mountainous areas has improved significantly over the past four decades， while the HQ around Beijing， Tianjin， and the southern rapidly developing cities has decreased gradually. The changes of construction land and the topographic index have the most significant impact on the changes of the HQ index. HQ changes were negatively correlated with socio-economic factors such as construction land area， GDP， and population density； and positively correlated with natural environmental factors such as topographic index， precipitation， and temperature. HQ changes were more significantly influenced by the natural environment in the northwestern part of the study area and more significantly influenced by socioeconomic factors in the southeastern plain area. The results help to reveal the response of habitat quality changes to the influencing factors， which can provide scientific basis and theoretical guidance for the implementation of sustainable land use， ecological environmental protection and management.
In order to understand the change characteristics of landscape pattern in Jiangsu Province in recent 35 years， based on the seven periods of land use data from 1980 to 2015， this paper analyzes the landscape pattern index， landscape stability and their relationship with human disturbance degree in Jiangsu Province. The results show that： （1） In terms of types， the fragmentation degree of construction land in Jiangsu Province is the highest and the shape is the most irregular， but it tends to be complete and regular with time； The patch aggregation degree and compactness of cultivated land and water area are high， but they show a decreasing trend. （2） On the whole， the overall complexity of landscape types in Jiangsu Province increases， the spatial connectivity decreases gradually， and the landscape diversity and heterogeneity increase； The area of each landscape type tends to be uniform， but the dominant landscape type still plays a strong role in controlling the whole landscape. （3） From 1980 to 2015， the landscape stability of most regions in Jiangsu Province was low， the human disturbance was high， and the change between different periods was small. （4） The correlation between human disturbance and patch density， landscape division index， Shannon’s diversity index， Shannon’s evenness index and landscape stability is the highest on 15 km×15 km scale， and the correlation with total edge contrast index and contagion index is the highest on 2 km×2 km scale. The research results provide a scientific basis for solving the problem of supply and demand of construction land， optimizing land use structure and protecting ecological environment.
Coal energy is an important component of China's energy system，and the open pit mining of coal has a more profound impact on the surrounding environment than well mining. Pingshuo Mining Area is an early open-pit coal mine developed in China， which mainly adopts the mining mode of mining while repairing， resulting in rapid land use changes in the mining area， and it is urgent to efficiently and accurately extract various types of land objects and monitor their ecological restoration. Based on multi-period high-resolution image data， this paper extracted the land use information of the study area from 2013 to 2020 through multi-scale segmentation and machine learning object-oriented classification method， and further analyzed the dynamic change of land use in Pingshuo open-pit mining area. The results show that from 2013 to 2020， the mining area moved eastward year by year， the mining area decreased by 7.84 km2， farmland decreased by 36.08 km2， woodland and grassland increased by 64.77 km2， and water body， dump and mining area decreased by less than 10 km2. The results of the analysis of land use area change in the study area show that the effect of green mine construction is remarkable. Combined with the green development policy of Pingshuo Coal Mine， this study will provide methods and data support for the evaluation of green mine construction.
Clarifying the response relationship between landscape pattern or land use and water quality parameters is an important prerequisite for protecting and improving water quality. The Danjiangkou Reservoir is the water source for the South-to-North Water Diversion Project. Agricultural planting and industrial production activities will affect the water quality of the reservoir area in various ways. Using Sentinel-2B remote sensing image data to classify the land use status， based on the GIS spatial analysis technology to calculate the landscape development intensity index of the comprehensive effect of land use types， combined with the water quality data of the automatic monitoring station， using redundancy analysis （RDA）， a preliminary discussion of Danjiangkou the response relationship between reservoir land use and water quality changes. The results show that the nutrients produced by fertilization and animal husbandry in agricultural planting areas， and the nutrients that enter the lake through surface runoff are the main sources of non-point source pollution of Danjiangkou Reservoir. Land use in a 500 m buffer zone has the greatest impact on water quality. In this buffer zone， cultivated land and construction land are concentrated and have a high degree of connectivity and agglomeration. The pollution to rivers is relatively high. There are relatively more woodlands and overall connectivity and the degree of aggregation is relatively high， which has a certain inhibitory effect on water pollution. In the adjacent areas around the reservoir， on the one hand， it is necessary to increase forest coverage so as to increase the intensity of vegetation to improve river surface source pollution； on the other hand， it is necessary to prevent the impact of nitrogen and phosphorus from agricultural production on water quality to reduce the entire reservoir area of non-point source pollution.
The distribution and change of impervious surface is an intuitive sign and important indicator of the urban development process， and it plays an important role in urban ecological functions， urban planning and sustainable development research. This study uses the long time series 30m impervious surface dynamic data set （GISD30） from 1985 to 2020 developed by the Institute of Aerospace Information Research Institute of the Chinese Academy of Sciences， combined with GIS spatial statistical methods， to study and analyze the temporal and spatial evolution of impervious surfaces in Shanghai over the past 35 years.The results show that ：（1） Over the past 35 years， the impervious surface area of Shanghai has increased from 878.07 km2 to 2849.90 km2， and the area has expanded to 3.25 times the original size. With the development and opening of the Pudong New Area in 1990， the urbanization process of Shanghai has accelerated significantly. The expansion of impervious surfaces in Shanghai was the most significant from 1990 to 2010. After 2010， the expansion speed and intensity of impervious surfaces began to decline significantly. （2） From the perspective of location differentiation characteristics， the rapid expansion of impervious surface area is mainly located in the suburbs of the city. Among them， the expansion rate of Pudong New Area in each study period is faster than that of other districts. （3） Based on the compactness and fractal dimension， it is found that the spatial distribution structure of impervious surfaces in the central urban area tends to be evacuated， the complexity of the overall impervious surface boundary in Shanghai is reduced， and the urban spatial form is more regular. （4） The urban development layout of “North and South Wings” in Shanghai is relatively obvious. The high-intensity expansion of the southern suburbs has driven the continuous southward shift of the impervious surface space. After 2010， the layout of the impervious surface space in Shanghai has begun to stabilize. At present， under the background of tight resource and environment constraints， Shanghai's urban development faces many challenges. This study has reference value for the effective promotion of Shanghai's urban renewal work.
The dynamics of land disturbance in a region reflect its economic development， and can thus be used as an indicator for regional planning and construction. To improve the efficiency and effectiveness of land disturbance monitoring， we use a Continuous Monitoring of Land Disturbance （COLD） algorithm to detect the changes of land disturbance and monitor the disturbance of production and construction projects in Changting County， Fujian Province， using time-series remote sensing images acquired from Landsat satellites （i.e.， Landsat 5 TM， Landsat 7 ETM+， and Landsat 8 OLI） from 2000 to 2020 years. In this paper， we investigate the influence of the length of time series， i.e.， 5-year （2016~2020）， 10-year （2011~2020）， and 21-year （2000~2020）， on predicting results for unknown years. Results show that： （1） COLD can identify land changes within a year， and cumulative changes for a time series； （2） Over the past 20 years， the land disturbance of Changting County mainly distributed in its central region， with less disturbance surrounding in mountainous areas； （3） Most of the disturbances of production and construction projects occurred in the third quarter， followed by the first quarter； （4） For cloudy and rainy areas， it is suggested to have a longer time-series （e.g.， ≥10-year） to maintain a satisfactory result for monitoring the disturbance of medium and large -scale production and construction projects in 2020.
Impervious surface is an important factor indicates the level of urbanization and the urban ecological environment， and it is one of the current research hotspots in urban remote sensing. Compared with humid and semi-humid areas， urban vegetation coverage in arid areas is relatively low， the similar spectum between impervious surface and barren area makes the traditional optical image-based spectral mixing analysis method and spectral index method not suitable for the impervious surface extraction of cities in arid areas. In response to this problem， a method for impervious surfaces extraction of cities from arid areas in Central Asia using synthesized multi-features of multispectral-SAR images is proposed to improve the mixclassifiation between impervious surfaces and bare soil， so as to extract impervious surface in arid area. In detials， Sentinel-2 and the dual-polarization SAR image of Sentinel-1 are selected for three Central Asia cities， Astana， Tashkent and Dushanbe. The spatial characteristics of multi-spectral and SAR images， and the polarization characteristics of SAR are feeded to LightGBM algorithm to classify and extract impervious surface. This paper compares the impervious surface extraction results of different feature combinations and different classification methods. Experimental results indicated that the multi-feature synthesis method of multispectral and SAR images proposed can effectively improve the accuracy of impervious surface extraction in arid areas， indicating the improvement in the misclassification of impervious surface and other land cover types in arid areas； the LightGBM algorithm has higher accuracy than XGBoost， HistGBT and other algorithms based on gradient boosting decision trees and random forest algorithm， and it is more suitable for extraction of impervious surface in arid area. This shows that the method based on LightGBM and the combination of multispectral and SAR multi-features can effectively extract the urban impervious surface in the arid area of Central Asia.
As the capital of Xinjiang and central Asia regional economic center in "One Belt and One Road"， Urumqi is particularly important for its rational use of land resources and healthy development of urban form. Firstly， this study analyzed the spatial and temporal evolution of the land use based on the land use classification of the main urban area of Urumqi from 1990 to 2018. The main urban area of Urumqi expanded rapidly from 1990 to 2018. Since 2010， the transportation network has been laid and the main urban area has expanded rapidly. Surrounding farmland， forest land and grassland have shrunk by 95.12 km2， 6.49 km2 and 52.37 km2. The center of gravity has been expanding eastward and northward. Then， the driving factors affecting the expansion of the main urban area， such as natural， social and economic factors， were selected. Combined with the historical characteristics of the expansion of the main urban area of Urumqi， the scenarios of the priority of the primary industry， the priority of the secondary and tertiary industries and the priority of ecology were designed. The GeoSOS-FLUS model was used to simulate and predict the scenarios of the expansion of the main urban area under different scenarios. The study found that the construction land increased by 1 142.94 km2 under the priority scenario of secondary and tertiary industries. Under the ecological priority scenario， the areas of forest land， grassland and water area with high ecological benefits were significantly increased by 281.59 km2，651.38 km2 and 7.29 km2. Under the primary priority scenario， the area of construction land and cultivated land expanded by 617.14 km2 and 611.71 km2. It is not only helpful to re-examine the rationality of Urumqi's urban expansion， but also to point out the direction of its rational and scientific urban planning and development in the future.
Global land cover products are important resources of forest type data， and systematic evaluation of forest type data in different global land cover products is important for data usage and new data production. Quality of forest type data in seven global land cover products （CCI-LC， MCD12Q1， Globeland30， GLC-FCS30， FROM-GLC10， Esri10 and ESA10） were systematic assessed from four aspects， including self-stability analysis， area comparison， spatial consistency comparison and accuracy index calculation. The results show that the self-stability of forest type data in CCI-LC is significantly higher than that in other products， the self-stability of CCI-LC from 2001 to 2019 is above 94%； Forest area in product with higher spatial resolution （10 m， 30 m） is significantly larger than that with lower spatial resolution （300 m， 500 m）. Among them， shrub area in Esri10 is far greater than that in other products； Esri10 and MCD12Q1 both have low spatial consistency with the other five products， the highest spatial consistency occurs between ESA10 and FROM-GLC10， which is about 85.3%； The overall accuracy of higher spatial resolution product is better than that with lower spatial resolution. When shrubs were not considered， the overall accuracy are 90.63%， 87.99%， and 85.22% for ESA10， FROM-GLC10 and Esri10 respectively. Overall accuracy of GLC-FCS30， CCI-LC and MCD12Q1 are all below 48% when applying a finer classification legend instead of a simplified classification legend.
Hyperspectral remote sensing images can be used to carry out fine-type land cover mapping， and related technologies have been widely used in agricultural production， vegetation ecology and so on. However， most of the existing methods are pixel-oriented or object-oriented analysis methods， which lack the rationality and precision of mapping units. In view of this， the idea of geo-parcel level hyperspectral land cover mapping based on random forest is proposed， and the hyperspectral image of Matiwan Village in Xiongan New Area is used to carry out experiments and accuracy evaluation， . By integrating the type accuracy and morphological accuracy indexes， the advantages and disadvantages of mapping at different scales are analyzed. In terms of type accuracy， the results show that the effect and type accuracy of geo-parcel hyperspectral land cover mapping are the best. In terms of morphological accuracy， the result shows that the land division at geo-parcel is better than that at object level. After comprehensive comparison of the two kinds of accuracy， it is shown that the geo-parcel level hyperspectral land cover mapping can produce information results more in line with user needs and has higher potential application value.
Forest Aboveground biomass is not only an important reference index to measure the productivity of forest ecosystem， but also an important part to study the surface carbon cycle and carbon balance. Based on domestic GF-3 SAR data， the suitability of different inversion models was explored to improve the inversion accuracy of Forest Aboveground biomass. The typical coniferous forest in southwest of Xiaoshao forest farm， Yiliang County， Kunming City， Yunnan Province was selected as the research object. GF-3 SAR data was used as the data source， and the polarization backscattering coefficient and polarization decomposition characteristics of four channels of GF-3 SAR data were used as the modeling factors of forest biomass. The parametric multivariate linear regression model and the nonparametric k-Nearest Neighbor method （k-NN）， Support Vector Regression （SVR） and Random Forest （RF） models were used to retrieve the Above Ground Biomass （AGB） of forest land in the study area.Pearson correlation coefficient （R2）， Root Mean Square Error （RMSE） and total accuracy （acc.） were used to assess the accuracy of the four models： R2 of the multiple linear stepwise regression model was at 0.37， with a RMSE at 20.70 T / hm2， and the total accuracy （acc.） was 61.85%； The k-NN model had a R2 of 0.34， with a RMSE at 20.29 T / hm2 and the total accuracy acc. at 62.60%； The R2 of SVR model was 0.33， the RMSE was 20.95 T / hm2， and the total accuracy was 61.39%； The RF model R2 was 0.0.35， the RMSE was 20.40 T / hm2， with the total accuracy acc. at 62.40%. This paper draws the following conclusions. ①The multiple linear stepwise regression algorithm， which belongs to the parametric model， has the highest accuracy， and is more suitable for AGB inversion of coniferous forest with Pinus Yunnanensis as the dominant tree species in this study area. ②In the nonparametric models， the SVR inversion accuracy is slightly higher， but generally lower than the parametric model. The accuracy of the four models is generally low， which may be related to the shadow overlay caused by the topographic relief in the study area and the poor heterogeneity and representativeness of the sample plot data in the sampling survey.
Forest canopy closure is an important factor in evaluating forest resources and accurate estimation of forest canopy closure is of great significance to forest management. Based on Li-Strahler geometric optics model， we estimated the forest canopy closure using Unmanned Aerial Vehicle （UAV） LiDAR and GF-6 WFV data. And in order to find a way out of the problem of mixed pixels， a reliable method was proposed. Firstly， the sunlit background component within the coverage of UAV LiDAR was calculated by using the high-precision forest structure parameters derived from UAV LiDAR. Then， the SMACC algorithm and linear spectral decomposition model were used for mixed pixel decomposition of GF-6 WFV and to determine the optimal scene component in the study area. Finally， the forest canopy closure in the study area was estimated by Li-Strahler geometric optical model， and the accuracy was verified by the measured data of field sample plots. The results showed that the determination coefficient （R2） between the estimated canopy closure and the measured canopy closure is 0.692 8， the Root Mean Square Error （RMSE） is 0.059 4， and the overall accuracy is 93.4%. Li-Strahler geometric optical model can effectively play a role in the inversion of forest canopy closure.
Refined remote sensing identification of bamboo forests in complex terrain areas can help understand the temporal distribution of bamboo forests and incorporate the ecological， economic， and social values of bamboo forests. In-depth analysis and the effective use of spectral differences and textural features in bright and shadow areas are key issues of the refined identification of bamboo forest information. In this study， we modified the “Film-Based & Class-Oriented” （FB-CO） algorithm and verified the effectiveness of the improvement using Sentinel-2A MSI images. In the “Modified Film-Based & Class-Oriented” （MFB-CO） bamboo forest information remote sensing extraction algorithm， the normalized shaded vegetation index （NSVI） is used instead of single-band thresholds to segment the forestland in bright and shadow areas， and a linear regression model is utilized to enhance the shadow area information. The BPNN， SVM， and RF classifiers are introduced to extract bamboo forests. The results show that the best segmentation thresholds for forestland in bright and shadow areas based on the NSVI and NIR are 0.41 and 0.23， with an Overall Accuracy （OA） of 96.00% and 83.50%， respectively. After the enhancement of shaded area information， the fitted model R2 was greater than 0.82 for each band， the MRE was less than 5%， the mean value increased for all bands， and the standard deviation decreased. The OA of the bamboo forest extraction is 82.41% for the FB-CO algorithm and 86.51%， 88.43%， and 88.92% for the BPNN， SVM， and RF based on the MFB-CO algorithm， respectively. The latter values are better than those of the FB-CO algorithm. The results show that the MFB-CO algorithm effectively improves the extraction of bamboo forest information by enhancing the implementation of several key steps of the FB-CO algorithm， providing technical support for the refinement of bamboo forest identification.
Existing Terrestrial Laser Scanning research on understory vegetation filtering has low precision and low degree of automation. In view of the shortcomings of the current research， the study takes the cloud data of two natural forest sample plots of Terrestrial Laser Scanning as the research object， an automatic filtering method of understory vegetation based on the main direction of the point cloud is proposed. First， after preprocessing the data such as cropping， denoising， filtering and height normalization， according to the growth height of understory vegetation in the sample plot， the data is divided into upper and lower layers at a certain height. Among them， the upper layer is a tree point cloud， and the lower layer is a point cloud containing understory vegetation. Then， the features in the spatial neighborhood of the lower layer point cloud are calculated to extract the main direction of the point cloud， and the understory main trunk is extracted according to the angle between the main direction and the normal vector of the Z-axis direction， so as to filter out a large number of understory vegetation point clouds. Finally， the Euclidean distance clustering method is used to cluster the extraction results of the understory main trunks， and the understory main trunks are finely extracted to achieve complete filtration of the understory vegetation. According to the above methods， two natural forest plots were tested. The results showed that when the neighborhood value k of plot 1 was 100 and the included angle threshold t was 30°， the neighborhood value k of plot 2 was 150 and the included angle threshold t was 30°. The number of understory trunk in the two plots achieved 100% complete extraction， indicating that the filter results of understory vegetation were good. Through the discussion of the neighborhood value k and the threshold of the included angle t， it is considered that 100~150 is appropriate for k value and 30° is appropriate for t value. This method has few parameters and high degree of automation， which can provide a certain technical reference for the study of shrubs or trees.
Forest disturbance affects the carbon cycle and carbon balance of forest ecosystem， and vegetation Net Primary Productivity （NPP） is an important indicator of vegetation carbon sequestration capacity. Analyzing the effects of forest disturbance on NPP is of great significance for global change research and ecosystem carrying capacity and resilience assessment. The forest coverage in the northern section of the Greater Khingan Mountain is more than 70%， which has been disturbed by human and natural factors for a long time. This region was selected as the study area in this paper. The 30 m resolution time series NPP from 2002 to 2018 was estimated based on the MuSyQ-NPP model and the estimated time series LAI with 30 m resolution， and the 30 m resolution forest disturbance time series product was used to analyze its impact on the NPP in the growth season. The results shown that in the north section of the Greater Khingan Mountain， the multi-year average of NPP is mostly distributed from 400 to 600 g C · m-2 · a-1， and shown a slow increasing trend on the whole. Fire disturbance may be the main disturbance factor in the study area. From 2002 to 2017， the annual average reduction of NPP caused by forest disturbance was 0.01 Tg C · a-1. In 2003， 2006 and 2017， NPP decreased significantly in the forest growth season after large area and high intensity disturbances， the reduced values were 0.11 Tg C， 0.03 Tg C and 0.03 Tg C， respectively. The NPP in the disturbed forest area in 2006 with medium and high intensity disturbance was higher than that in the year before the disturbed year since 2009. In conclusion， the disturbance with large area and high intensity has great effects on NPP， and the NPP in the medium-high intensity disturbance area recovers rapidly.
The accurate extraction of impervious surface is of great significance for regional population density estimation， environmental assessment， disaster prediction， hydrological model construction， urban heat island effect research and climate change analysis. Traditional large scale impervious surface extraction methods are mainly limited by the quality of remote sensing data and the selection of extraction features， and the spatial resolution of extracted impervious surface is low， which is difficult to meet the refined requirements of impervious surface at the present stage. In this paper， based on Sentinel-1 SAR and Sentinel-2 MSI remote sensing data， multiple extraction features of impervious surface were selected from three dimensions， including spectrum， texture and time sequence， to build an impervious surface extraction model based on random forest. In addition， GEE platform was used to carry out extraction experiment of 10m impervious surface in Yangtze River Delta region in 2020. The results showed that in different types of experimental areas， compared with spectral features， spectral features and time series features， the overall accuracy and Kappa coefficient of the proposed method were increased by 5%，9% and 2%，6%， respectively， and all cities with different impervious surface coverage levels had good extraction effects. The overall accuracy and Kappa coefficient of impervious surface extraction at the global scale in the Yangtze River Delta region were 93.75% and 0.88， respectively. The impervious surface area was 6 1591.38 km2， accounting for about 17% of the total area. The impervious surface extraction method proposed in this paper for 10m resolution remote sensing images is suitable for different types of areas such as mountainous areas， rural areas， urban areas and urban fringe areas. The method is simple and easy to operate， has high precision， and is suitable for cloud platform large-area computing.
Quick and accurate access to the spatial distribution of forests is of great significance for assessing the status of forest resources and ecological environment protection.Taking Pu'er City in Yunnan Province as the research area， Based on the Google Earth Engine （GEE） platform and Sentinel-2 image data，combined with the field survey data， airborne remote sensing data and terrain auxiliary data， the spectral features， texture features and topographic features were extracted. Through feature screening， the optimal feature set suitable for forest classification was obtained.Combining Simple Non-Iterative Clustering （SNIC） superpixel segmentation algorithmto explore the influence of different classification methods and characteristic variables on the classification accuracy.The results showed that the classification accuracy of the object-oriented classification method was higher than that of the pixel-based classification method， with an overall classification accuracy of 88.21% and the Kappa coefficient of 0.87. which can accurately map the forest cover of Pu 'er City. The object-oriented method can effectively alleviate the “salt and pepper phenomenon”， and feature optimization avoids the influence of redundant information on classification results and effectively improves classification efficiency. The combination of GEE platform and object-oriented method can provide large-area， high-precision forest cover remote sensing rapid mapping.
The spatiotemporal big data of urban agglomerations has the characteristics of dynamic and real-time， across time and space， and across administrative regions. The method of traditional small sample data construction brings challenges to the technical aspects of data storage management， integration， mining analysis and knowledge discovery. This research is oriented to the major needs of the construction and management of urban agglomerations for key services of spatial information. It focuses on the characteristics of event models and related elements triggered by application subject areas. Based on elements-events-themes， we have designed a spatiotemporal big data framework system for urban agglomeration construction and management with the examples of four subjects. Based on elements-events-themes， we have designed a spatiotemporal big data framework system with the examples of four subjects and explore the process of factors design， data organization and algorithm implementation of the event of the assessment of industrial land use efficiency. The spatiotemporal big data framework system proposed in this research is helpful to promote the integration application of spatiotemporal big data on Urban Agglomeration construction and management.
Tianjin-Hebei region is located in the North China Plain. This area involves multiple economic center of China. The serious groundwater exploitation leads to very significant land subsidence in this area， so it is of great practical significance to carry out wide area subsidence monitoring in this region. In this paper， taking 59 sentinel-1A SAR images obtained from February 2016 to November 2018 as data source， subsidence monitoring in this area was carried out using the method of Interferometric Point Target Analysis （IPTA）， and the result accuracy was verified based on a large number of leveling data. At the last， the spatial and temporal distribution characteristics of the subsidence were analyzed and explained in detail based on the verified results. As compared with the leveling data， the Root Mean Square Error （RSME） of sentinel-1A IPTA monitoring results is ±2.70 mm?a-1. The results and analysis demonstrate that there is significant uneven land subsidence in the study area. The maximum subsidence rate is as high as 218.31 mm?a-1. Further analysis indicate that the surface subsidence is mainly related to the regional geological conditions and different types of groundwater exploitation. The study further shows that sentinel-1A time series differential interferometry has high precision and excellent application potential in wide area land subsidence monitoring.
Near-infrared Reflectance of Vegetation （NIRv），a new Vegetation index，has become an effective tool for studying plant photosynthesis in recent years. Studying the relationship between NIRv and Gross Primary Productivity （GPP） at multi-time scales is of great significance for exploring global and regional scale long-time series GPP. Using Moderate Resolution Imaging Spectroradiometer （MODIS） data and site flux data，the relationship between Normalized Difference Vegetation Index （NDVI） and NIRv and GPP was analyzed and compared at 8 typical vegetation types in China from 2004 to 2006，and the ability of NDVI and NIRv to represent seasonal variation of GPP at different time scales （daily，8-day and monthly） was analyzed. The results showed that： both NDVI and Nirv could characterize GPP，and the order of simulation degree from high to low was mixed forest，grassland，evergreen needleleaf forest，cropland and evergreen broad-leaved forest. Except for evergreen broad-leaved forest，NIRv had more effective simulation in other vegetation types. The NDVI of Inner Mongolia and Xishuangbanna was in advance and lagged at the beginning and end of the growth season，and the growth season offset of NIRv was weaker than that of NDVI. With the increase of time scale，the determination coefficient R2 between NDVI and NIRv with GPP gradually increased，and the R2 between NDVI and NIRv with GPP on monthly scale was the highest. The average R2 was 0.78 and 0.81 respectively，and the difference of R2 between different sites gradually decreased. There were significant differences among different vegetation types. The R2 of NDVI-GPP and NIRv-GPP in mixed forest，evergreen needleleaf forest，grassland and cropland were higher than that in evergreen broad-leaved forest. The research results can provide an important basis for establishing the correlation between vegetation index and GPP，and further improving the accuracy of regional carbon flux estimation.
Flood disasters pose a serious threat to the national economy and the safety of people's lives. Rapid and accurate access to disaster information is of great significance for disaster relief. This study takes the flood in Xinxiang City， Henan Province in July 2021 as an example， selects two Sentinel-1A SAR images before and after the disaster， and proposes a flood detection method based on combined difference image to monitor the flooded area. Firstly， the combined difference image of intensity difference and mean is constructed by local energy weighting， which makes full use of the change information and neighborhood information of the two difference images； Then， the combined difference image is processed by bilateral filtering to smooth the edge information and eliminate noise interference； Finally， the final flood monitoring results are obtained by FCM cluster analysis. The results show that the combined difference image obtained by local energy weighting and bilateral filtering rules can more accurately reflect the change information of water bodies. The overall detection error of this method is lower than that of the single difference image method and the equal weight combined difference image method， which greatly reduces the error rate of detection results. The Kappa coefficient is as high as 0.899 6， which effectively improves the accuracy of detection results. Finally， the method in this paper is used to monitor the flooded area of Xinxiang City， which is 77.05 km2. Therefore， using the combination rules of local energy weighting and bilateral filtering to build difference images has certain advantages and potential in flood monitoring.
Typhoon disasters are important natural disasters that lead to casualties and property losses in Fujian Province， China. Based on the historic typhoon disasters data from 2009 to 2020， precipitation， wind field， terrain， soil texture， NDVI， population density and GDP， the spatiotemporal patterns were analyzed on the county scale. Besides， the influencing factors of typhoon disasters were also explored by Geodetector technique. The results show that there were 38 typhoons landfalling or affecting Fujian Province， 16 super typhoons of which caused the most serious losses， while the most serious region mainly located in coastal areas. The maximum three-day precipitation， maximum wind velocity and distance to typhoon center are the three dominant factors that influenced the affected population， deaths， -damage crops and direct economic losses. The study can be employed to quantify influencing factors and provide theoretical reference for disaster risk –reduction in Fujian Province.
In view of the current situation of the frequent landslide disasters in Yili area in recent years， which seriously threaten the safety of local residents' lives and property. The study uses SBAS-InSAR technology to focus on the typical geological disaster area of Kalayagaqi in Yili area， combined with multi-source SAR data， combined with optical remote sensing and field geological survey， a total of 61 potential landslides were identified. The GPM precipitation data of the Global Precipitation Program was introduced to analyze the temporal and spatial evolution characteristics of the key landslide hazards G3 and H56. It was found that the deformation rate of the two had a good correlation with the precipitation， and the landslide deformation reached its peak during the freeze-thaw period， followed by the rainfall period. Combined with the analysis of geological and geomorphological characteristics， it is shown that freeze-thaw， rainfall， geological conditions and human engineering activities are the main factors that induce landslide disasters in the area. The landslides in the area are mainly distributed in the slope of 10°~30°， the slope direction is mainly sunny slope （135°~315°）， and the area is less than 0.15 km2 of ginger yellow conglomerate and other interbedded weaker sandstone and conglomerate groups. The experimental results show that InSAR technology can effectively identify and monitor landslide hazards in Yili area， and can provide technical reference and scientific basis for local disaster prevention and control.
Geographically Weighted Regression （GWR） model was used to retrieve the near-surface air temperature in Chongqing by considering the spatial non-stationary characteristic between near-surface air temperature and independent variables， then the results were compared with those of Temperature-Vegetation Index （TVX） and Ordinary Least Square （OLS）. The results show that the near-surface air temperature retrieved by the three models is consistent with the spatial distribution trend of the measured data， but the accuracy verification results show that the RMSE of the GWR model is lower than that of OLS model and TVX model at different dates， and the mean RMSE between the retrieved values by TVX model and the measured values is 2.83 ℃， the mean RMSE of OLS model is 1.65 ℃， and the mean RMSE of GWR model is 1.58 ℃.In addition， the mean standard deviation and absolute value of Temperature Deference （TD） retrieved by GWR model is also lower than those of OLS model and TVX model， which shows the advantages of GWR model in retrieved near-surface Temperature in complex surface environment.
Ecological observation station provides high-throughput canopy images for vegetation growth monitoring in karst ecological fragile area， however， there are few reports on the extractionof vegetation from bare rock and vegetation mixed underlying surface in karst area. In order to provide technical support for vegetation monitoring based on ground visible images， the canopy RGB images of the rocky desertification ecological observation station were used to study segmentation algorithm and growth monitoring index applicable to karst vegetation. The results show that： （1） The differentiation degree of light green vegetation in karst area is high for color space， nonlinear combination of color channels and machine learning， but its sensitivity to bare rock and dark green vegetation is various. The vegetation segmentation effects of the three segmentation methods were significantly different under strong light in sunny day and weak light in cloudy day. The machine learning algorithm has the best segmentation effect with the accuracy is over 80% under weak light in cloudy day and over 90% under strong light in sunny day. （2） The trend of vegetation growth reflected by GLA， NDYI， NGRDI and VARI was similar. In these indices， NDYI is more sensitive to the difference of vegetation growth. The compound sine function can simulate the daily dynamic changes of these four indices， moreover， the simulation accuracy of NGRDI trend is the highest with R2 = 0.830.
