River embankments are designed to defend against floods over coastal and riparian areas.It is important to early detect unexpected damages on embankments before they exacerbate.To continuously monitor the stability of...River embankments are designed to defend against floods over coastal and riparian areas.It is important to early detect unexpected damages on embankments before they exacerbate.To continuously monitor the stability of the embankments and efficiently recognize such potential damages,this study takes SAR(Synthetic Aperture Radar)derived deformation as an indicator of the embankment instability,and customizes a multi-temporal InSAR(Interferometric SAR)approach-small baseline subset.Specifically,during InSAR processing,we apply a two-step amplitude difference dispersion threshold method to extract InSAR measurement points,thus improving the point density within the embankment.We applied this method to the Kangshan Embankment(KE)using 147 Sentinel-1 acquired between 2017 and 2021.We categorized KE into Waterside Slope(WS),Embankment Top(ET),and Landside Slope(LS)using InSAR height estimation.Given the dominance of downslope movement,we developed a projection matrix from InSAR-derived deformation in the satellite line-of-sight direction onto WS and LS.The study shows that KE was generally stable during the five-year period,while WS,ET,and LS experienced different deformation processes.For instance,seasonal variation was observed from the deformation time series,especially between every April and November.We applied the principal component analysis to the time-series displacement and analyzed the results in conjunction with the rainfall data of Kangshan Township.It showed that deformation due to rainfall equals 80.93%,81.30%,and 82.46%of the total deformation for WS,ET,and LS,respectively,indicating that rainfall is one of the environmental driving factors affecting the deformations.We conclude that the proposed methodology is suited for systematic embankment monitoring and identifies major driving forces.展开更多
Because of the large number and remoteness, satellite data, including microwave data and optical imagery, have commonly been used in alpine glaciers surveys. Using remote sensing and Geographical Information System （...Because of the large number and remoteness, satellite data, including microwave data and optical imagery, have commonly been used in alpine glaciers surveys. Using remote sensing and Geographical Information System （GIS） techniques, the paper presents the results of a multitemporal satellite glacier extent mapping and glacier changes by glacier sizes in the Mt. Qomolangma region at the northern slopes of the middle Himalayas over the Tibetan Plateau. Glaciers in this region have both retreated and advanced in the past 35 years, with retreat dominating. The glacier retreat area was 3.23 km2 （or o.75 km^2 yr^-1 during 1974 and 1976, 8.68 km^2 （or 0.36 km^2 yr^-1 during 1976 and 1992, 1.44 km^2 （or 0.12 km^2 yr^-1） during 1992-2ooo. 1.14 km^2 （or 0.22 km^2 yr^-1 during 2000-2003, and 0.52 km^2 （or 0.07 km^2 yr^-1 during 2003-2008, respectively. While supra-glacier lakes on the debris-terminus of the Rongbuk Glacier were enlarged dramatically at the same time, from 0.05 km^2 in 1974 increased to 0.71 km^2 in 2008, which was more than 13 times larger in the last 35 years. In addition, glacier changes also showed spatial differences, for example, glacier retreat rate was the fastest at glacier termini between 5400 and 5700 m a.s.l than at other elevations. The result also shows that glaciers in the middle Himalayas retreat almost at a same pace with those in the western Himalayas.展开更多
This paper presents some new developments in image mining. Image mining con-siders the chain from object identification from remote sensing images through modeling,tracking on a series of images and prediction,towards...This paper presents some new developments in image mining. Image mining con-siders the chain from object identification from remote sensing images through modeling,tracking on a series of images and prediction,towards communication to stakeholders. Attention is given to image mining for vague and uncertain objects. Aspects of up-and down-scaling are addressed. We further consider in this paper both spatial interpolation and decision making. The paper is illustrated with several case studies.展开更多
基金National Natural Science Foundation of China(No.41830110)National Key Research Development Program of China(No.2018YFC1503603)+2 种基金State Scholarship Fund from the Chinese Scholarship Council(No.202206710096)Project of China Railway Corporation(No.2021-key-14,2021-major-08)The Joint Planning of Technology and Water Conservancy of Jiangxi Province(No.2022KSG01009).
文摘River embankments are designed to defend against floods over coastal and riparian areas.It is important to early detect unexpected damages on embankments before they exacerbate.To continuously monitor the stability of the embankments and efficiently recognize such potential damages,this study takes SAR(Synthetic Aperture Radar)derived deformation as an indicator of the embankment instability,and customizes a multi-temporal InSAR(Interferometric SAR)approach-small baseline subset.Specifically,during InSAR processing,we apply a two-step amplitude difference dispersion threshold method to extract InSAR measurement points,thus improving the point density within the embankment.We applied this method to the Kangshan Embankment(KE)using 147 Sentinel-1 acquired between 2017 and 2021.We categorized KE into Waterside Slope(WS),Embankment Top(ET),and Landside Slope(LS)using InSAR height estimation.Given the dominance of downslope movement,we developed a projection matrix from InSAR-derived deformation in the satellite line-of-sight direction onto WS and LS.The study shows that KE was generally stable during the five-year period,while WS,ET,and LS experienced different deformation processes.For instance,seasonal variation was observed from the deformation time series,especially between every April and November.We applied the principal component analysis to the time-series displacement and analyzed the results in conjunction with the rainfall data of Kangshan Township.It showed that deformation due to rainfall equals 80.93%,81.30%,and 82.46%of the total deformation for WS,ET,and LS,respectively,indicating that rainfall is one of the environmental driving factors affecting the deformations.We conclude that the proposed methodology is suited for systematic embankment monitoring and identifies major driving forces.
基金supported by the National Natural Science Foundation of China (40601056, 40121101)the Special Funds for Major State Basic Research Project (2009CB723901)+4 种基金the Special Science Foundation on Meteorological Project Research for Public Benefit (GYHY(QX)2007-6-18)the Survey Project on Glacier resources and their changes in China (No.2006FY110200)the Opening Fund projects of State Key Laboratory of Remote Sensing Science in the Institute of Remote Sensing Applicationsthe innovative project of Institute of Tibetan Plateau Research (ITPR),CASthrough a cooperation project between the Climate Change Institute, University of Maine supported by the National Oceanic and Atmospheric Administration (NA04OAR4600179) and the Institute of Tibetan Plateau Research (ITPR), CAS
文摘Because of the large number and remoteness, satellite data, including microwave data and optical imagery, have commonly been used in alpine glaciers surveys. Using remote sensing and Geographical Information System （GIS） techniques, the paper presents the results of a multitemporal satellite glacier extent mapping and glacier changes by glacier sizes in the Mt. Qomolangma region at the northern slopes of the middle Himalayas over the Tibetan Plateau. Glaciers in this region have both retreated and advanced in the past 35 years, with retreat dominating. The glacier retreat area was 3.23 km2 （or o.75 km^2 yr^-1 during 1974 and 1976, 8.68 km^2 （or 0.36 km^2 yr^-1 during 1976 and 1992, 1.44 km^2 （or 0.12 km^2 yr^-1） during 1992-2ooo. 1.14 km^2 （or 0.22 km^2 yr^-1 during 2000-2003, and 0.52 km^2 （or 0.07 km^2 yr^-1 during 2003-2008, respectively. While supra-glacier lakes on the debris-terminus of the Rongbuk Glacier were enlarged dramatically at the same time, from 0.05 km^2 in 1974 increased to 0.71 km^2 in 2008, which was more than 13 times larger in the last 35 years. In addition, glacier changes also showed spatial differences, for example, glacier retreat rate was the fastest at glacier termini between 5400 and 5700 m a.s.l than at other elevations. The result also shows that glaciers in the middle Himalayas retreat almost at a same pace with those in the western Himalayas.
文摘This paper presents some new developments in image mining. Image mining con-siders the chain from object identification from remote sensing images through modeling,tracking on a series of images and prediction,towards communication to stakeholders. Attention is given to image mining for vague and uncertain objects. Aspects of up-and down-scaling are addressed. We further consider in this paper both spatial interpolation and decision making. The paper is illustrated with several case studies.
