为获取神东矿区地表温度长期变化趋势,提取其地表温度突变特征,以2000—2018年16 d地表温度最大值合成的MOD11A2为数据源,对神东矿区地表温度进行回归分析,拟合其时序变化趋势,并利用BFAST算法(breaks for additive seasonal and trend...为获取神东矿区地表温度长期变化趋势,提取其地表温度突变特征,以2000—2018年16 d地表温度最大值合成的MOD11A2为数据源,对神东矿区地表温度进行回归分析,拟合其时序变化趋势,并利用BFAST算法(breaks for additive seasonal and trend)提取地表温度突变的时空分布特征、最大突变发生时间和突变幅度。在此基础上,通过空间叠加统计方法分析地表温度突变与土地覆被变化之间的关系。结果表明:2000—2018年,由于神东矿区整体植被覆盖度增加,矿区内98.63%的区域地表温度呈下降趋势;露天采区地表温度突变比例高于井工采区,露天采区和井工采区内地表温度突变面积占比分别为34.66%和19.02%,归因于露天开采比井工开采对地表扰动更加剧烈;土地覆被变化引起地表温度突变,地表温度突变方向和幅度取决于土地覆被变化的类型、规模等。研究结果可为矿区生态环境治理及评价提供科学依据。展开更多
准确识别地表变化的时空信息,有助于探究地表自然环境和生态系统发展演变的规律,支撑相关的科研与行政管理工作。本文以河南某生态保护修复工程部分实施范围为研究区域,基于Google Earth Engine(GEE)云平台,以2013—2020年的98景Landsat...准确识别地表变化的时空信息,有助于探究地表自然环境和生态系统发展演变的规律,支撑相关的科研与行政管理工作。本文以河南某生态保护修复工程部分实施范围为研究区域,基于Google Earth Engine(GEE)云平台,以2013—2020年的98景Landsat8/OLI遥感影像作为数据源,应用Breaks for additive season and trend(BFAST)算法对地表变化进行了信息提取和制图。首先基于GEE云平台对Landsat8/OLI地表反射率数据集进行调用和预处理,基于CFMask算法对遥感数据集进行云影掩膜,开展光谱指数(植被指数NDVI)的计算以及时间序列数据集的构建。其次基于时序数据集与BFAST算法构建由趋势项、季节项和残差项组成的广义线性回归模型,通过最小二乘法求解模型中的未知参数集,以此进一步构建时序拟合模型,而后基于残差的Moving sums(MOSUM)方法对时序结构变化进行检测。最后从检测结果中抽取像元样点,通过与Google Earth高分辨率影像数据叠置和目视解译,开展结果验证和精度评价。结果表明,本文提出的方法在研究区的时序地表变化检测中具有较高的检测精度(总体精度为83.7%,2018—2020年分年度检测结果精度分别为86.5%、80.7%、87.7%)。本文提出的方法是遥感大数据库构建、地表生态信息近实时变化扰动识别和监测等技术的一种基础方法,能够对国土空间生态保护修复调查监测和评估预警等工作提供技术支撑和决策支持。展开更多
Annual land use land cover(LULC)change information at medium spatial resolution(i.e.at 30 m)is required in numerous subjects,such as biophysical modelling,land management and global change studies.Annual LULC informat...Annual land use land cover(LULC)change information at medium spatial resolution(i.e.at 30 m)is required in numerous subjects,such as biophysical modelling,land management and global change studies.Annual LULC information,however,is usually not available at continental or national scale due to reasons such as insufficient remote sensing data coverage or lack of computational capabilities.Here we integrate high temporal resolution and coarse spatial resolution satellite images(i.e.,Moderate Resolution Imaging Spectroradiometer(MODIS)and Global Inventory Modelling and Mapping Studies(GIMMS)normalized difference vegetation index(NDVI))with high spatial resolution datasets(China’s Land-Use/cover Datasets(CLUDs)derived from 30-meter Landsat TM/ETM+/OLI)to generate reliable annual nominal 30 m LULC maps for the whole of China between 1980 and 2015.We also test the performance of a statistical based change detection algorithm(Breaks for Additive Seasonal and Trend),originally designed for tracking forest change,in classifying all-type LULC change.As a result,a nominal 30 m annual land use/land cover datasets(CLUD-A)from 1980 to 2015 was developed for the whole China.The mapping results were assessed with a change sample dataset,a regional annual validation sample set and a three-year China sample set.Of the detected change years,75.61%matched the exact time of conversion within±1 year.Annual mapping results provided a detail process of urbanization,deforestation,afforestation,water and cropland dynamics over the past 36 years.The consistent characterization of land change dynamics for China can be further used in scientific research and to support land management for policy-makers.展开更多
研究黄河三角洲耕地退化状态并明晰其驱动机制对保障区域粮食安全和促进耕地保护至关重要。本文基于2001—2021年MODIS NDVI数据,采用加性季节和趋势断点检测法BFAST(Breaks For Additive Seasonal and Trend)探讨黄河三角洲耕地退化状...研究黄河三角洲耕地退化状态并明晰其驱动机制对保障区域粮食安全和促进耕地保护至关重要。本文基于2001—2021年MODIS NDVI数据,采用加性季节和趋势断点检测法BFAST(Breaks For Additive Seasonal and Trend)探讨黄河三角洲耕地退化状态,并应用地理探测器探究各自然和人为因子及其交互作用对不同尺度区域耕地退化状态空间分异的影响与驱动机制。结果表明:(1)与线性趋势相比,非线性趋势方法检测耕地退化状态的精度更高,不仅可以检测到耕地变化的整体趋势,还可检测到长期变化中的阶段性突变信息,可较全面、准确的评估耕地退化状态;(2)BFAST检测结果显示黄河三角洲耕地31.75%呈退化状态,61.06%为改善状态,7.19%呈非显著趋势。黄河三角洲耕地在长期变化过程中多受外界干扰而发生短期突变,其中中断减少趋势占比31.31%,主要分布在孤北水库、黄河故道西侧和东南沿海,呈沿海自内陆逐渐减少且零散化的空间分布格局;中断增加趋势占比55.13%,主要分布在黄河以南、淡水河流和水库附近,呈大规模集中分布特点;(3)黄河三角洲不同尺度区域耕地退化状态空间分异以土地利用为主导,且双因子交互影响力明显高于单因子,土地利用∩高程和距海洋距离∩土地利用分别是耕地退化和改善状态的主导交互因子,人为因素主导和自然因素协同作用导致黄河三角洲耕地退化状态空间分异。展开更多
Vegetation greening has long been acknowledged,but recent studies have pointed out that vegetation greening is possibly stalled or even reversed.However,detailed analyses about greening reversal or increased browning ...Vegetation greening has long been acknowledged,but recent studies have pointed out that vegetation greening is possibly stalled or even reversed.However,detailed analyses about greening reversal or increased browning of vegetation remain scarce.In this study,we utilized the normalized difference vegetation index(NDVI)as an indicator of vegetation to investigate the trends of vegetation greening and browning(monotonic,interruption,and reversal)through the breaks for the additive season and trend(BFAST)method across China’s drylands from 1982 to 2022.It also reveals the impacts of ecological restoration programs(ERPs)and climate change on these vegetation trends.We find that the vegetation displays an obvious pattern of east-greening and west-browning in China’s drylands.Greening trends mainly exhibits monotonic greening(29.8%)and greening with setback(36.8%),whereas browning shows a greening to browning reversal(19.2%).The increase rate of greening to browning reversal is 0.0342/yr,which is apparently greater than that of greening with setback,0.0078/yr.This research highlights that,under the background of widespread vegetation greening,vegetation browning is pro-gressively increasing due to the effects of climate change.Furthermore,the ERPs have significantly increased vegetation coverage,with the increase rate in 2000-2022 being twice as much as that of 1982-1999 in reveg-etation regions.Vegetation browning in southwestern Qingzang Plateau is primarily driven by adverse climatic factors and anthropogenic disturbances,which offset the efforts of ERPs.展开更多
以江苏省永久耕地为例,基于2001—2019年中分辨率成像光谱仪(Moderate resolution imaging spectroradiometer,MODIS)遥感影像,开展耕地生产力隐性退化遥感监测和影响因素分析。BFAST(Breaks for additive seasonal and trend)算法用于...以江苏省永久耕地为例,基于2001—2019年中分辨率成像光谱仪(Moderate resolution imaging spectroradiometer,MODIS)遥感影像,开展耕地生产力隐性退化遥感监测和影响因素分析。BFAST(Breaks for additive seasonal and trend)算法用于建模历史时期耕地生产力变化的预期行为,并以此为基准判断监测时期耕地生产力是否存在隐性退化风险。基于地理探测器,从3个准则层的8项指标变量对耕地生产力隐性退化进行了主导影响因素探测和因子交互分析。研究结果表明:江苏省存在生产力隐性退化的耕地比例为21.9%,具有显著的空间差异。西北地区的徐州市、宿迁市的耕地生产力隐性退化比例最高,分别为47.2%和43.4%,且表现出聚集性。东南地区的苏州市、无锡市和南通市的耕地生产力隐性退化比例较低,均不足10%。因子探测分析表明外流人口数量、种植业从业人员数量和农业机械化总动力3项指标对江苏省耕地生产力隐性退化的解释力最强。多因子交互耦合后,人口因素与生产条件解释力增强最为显著。耕地生产力隐性退化的地域分异类型划分为生产条件约束型、产出效益约束型和人口因素约束型。农业机械化总动力、农业产值和外流人口数量分别为3种约束类型的首要因素。从地域空间来看,人口因素约束型地区在江苏省占比最大,主要集中于苏北地区。对于不同约束类型区域分别提出加强高标准农田建设、实施惠农政策、减缓劳动力析出等相应的政策建议。展开更多
Land cover change in the semi-arid environment of the eastern Hindu Kush region is driven by anthropogenic activities and environmental change impacts. Natural hazards, such as floods presumably influenced by climatic...Land cover change in the semi-arid environment of the eastern Hindu Kush region is driven by anthropogenic activities and environmental change impacts. Natural hazards, such as floods presumably influenced by climatic change, cause abrupt change of land cover. So far, little research has been conducted to investigate the spatiotemporal aspects of this abrupt change in the valleys. In order to explore the abrupt change in land cover and floods as its possible drivers in the eastern Hindu Kush, a semi-arid mountain region characterized by complex terrain, vegetation variation, and precipitation seasonality, we analyzed long-term Landsat image time series from 1988 to 2020 using Breaks For Additive Seasonal and Trend(BFAST). Overall, BFAST effectively detected abrupt change by using Landsat-derived Modified Soil Adjusted Vegetation Index(MSAVI). The results of our study indicate that approximately 95% of the study area experienced at least one abrupt change during 1988-2020. The years 1991, 1995, 1998, 2007, and 2016 were detected as the peak years, with the peaks occurring in different seasons. The annual trend of abrupt change is decreasing for the study area. The seasonality of abrupt change at the catchment level shows an increasing trend in the spring season for the southern catchments of Panjkora and Swat. The spatial distribution patterns show that abrupt change is primarily concentrated in the floodplains indicating that flooding is the primary driver of the land cover change in the region. We also demonstrated the accurate detection of past flood events(2015) based on the two case examples of Ayun, Rumbur, and Kalash valleys. The detection of the flood events was verified by fieldwork and historical high-resolution Google Earth imagery. Finally, our study provides an example of applying Landsat time series in a dry mountain region to detect abrupt changes in land cover and analyze impact of natural hazards such as floods.展开更多
在喀斯特分布区,基岩、植被、裸地等多种地表覆盖交错分布,地物覆盖高度异质,并且呈现出短周期规律性变化和长期动态趋势变化,单一时相的影像进行土地覆盖分类精度非常有限。针对这一问题,本文提出一种顾及物候特征的多时相遥感影像分...在喀斯特分布区,基岩、植被、裸地等多种地表覆盖交错分布,地物覆盖高度异质,并且呈现出短周期规律性变化和长期动态趋势变化,单一时相的影像进行土地覆盖分类精度非常有限。针对这一问题,本文提出一种顾及物候特征的多时相遥感影像分类策略,利用具有高时间分辨率的MODIS NDVI时间序列产品作为数据源,选择BFAST(Breaks For Additive Seasonal and Trend)方法进行NDVI时间序列的物候分解,采用动态阈值法对时序分解的物候轨迹进行标记,最后将物候标记特征与原始光谱时序综合特征进行组合,选择支持向量机(SVM)分类器进行土地利用覆盖分类,并且对比了不同特征空间下的分类结果。以云南省壮族苗族自治州丘北县和砚山县为研究区进行分类实验,结果表明,BFAST模型可以有效地分解出NDVI时序中的关键物候特征,相比基于单纯光谱特征的分类,物候驱动的喀斯特断陷盆地区土地覆盖分类精度有明显的提升,在NDVI、光谱和物候组合特征空间下,土地覆盖分类精度最高,总体精度和Kappa系数分别为88.94%和0.8693,尤其在灌木林、有林地、石旮旯地与稀疏植被的区分中,SOS、POS和GSG等物候特征具有较强的可分性,表明物候特征在地物识别中的有效性。展开更多
Drought is a worldwide natural disaster that has long affected agricultural production as well as social and economic activities. Frequent droughts have been observed in the Belt and Road area, in which much of the ag...Drought is a worldwide natural disaster that has long affected agricultural production as well as social and economic activities. Frequent droughts have been observed in the Belt and Road area, in which much of the agricultural land is concentrated in fragile ecological environment. Based on the Tropical Rainfall Measuring Mission Satellite(TRMM) 3 B43 precipitation data, we used the Precipitation Abnormity Percentage drought model to study the monthly spatio-temporal distribution of drought in south region of N50° of the Belt and Road area. It was observed that drought during winter was mainly distributed in Northeast Asia, Southeast Asia, and South Asia, while it was mainly distributed in Central Asia and West Asia during summer. The occurrence of historical droughts indicates an obvious seasonal cycle. The regional variations in drought were analyzed using the Breaks for Additive Season and Trend tool(BFAST) in six sub-regions according to the spatial distribution of six economic corridors in the Belt and Road area. The average drought conditions over the 18 years show a slight decreasing trend in Northeast Asia, West Asia, North Africa, South Asia, Central and Eastern Europe, and a slight increasing trend in Central Asia. However, it was a fluctuating pattern of first increasing and then decreasing in Southeast Asia. The results indicate that the total drought area in the Belt and Road region showed a general decreasing trend at a rate of 40,260 km^2 per year from 1998 to 2015.展开更多
文摘植被是联结土壤、大气和水分的自然"纽带",在全球气候变化研究中具有"指示器"的作用。对归一化植被指数(normalized difference vegetation index,NDVI)时间序列分析,可以为相关部门的工作和决策提供更好的支持。使用MODIS NDVI数据结合BFAST(breaks for additive seasonal and trend)方法实现对老哈河流域及周边地区的植被变化监测,并确定其NDVI时间序列出现突变点的时间节点。结合气象数据以及数据本身的质量作为影响因子,分析出现突变点的主要原因。研究结果表明,降水量、相对湿度、温度、日照时数、流域蒸发量与NDVI变化趋势呈正相关,风速与NDVI变化趋势相关性很小。降水量对NDVI变化的影响具有滞后性,滞后时间与降水量大小有关。
文摘为获取神东矿区地表温度长期变化趋势,提取其地表温度突变特征,以2000—2018年16 d地表温度最大值合成的MOD11A2为数据源,对神东矿区地表温度进行回归分析,拟合其时序变化趋势,并利用BFAST算法(breaks for additive seasonal and trend)提取地表温度突变的时空分布特征、最大突变发生时间和突变幅度。在此基础上,通过空间叠加统计方法分析地表温度突变与土地覆被变化之间的关系。结果表明:2000—2018年,由于神东矿区整体植被覆盖度增加,矿区内98.63%的区域地表温度呈下降趋势;露天采区地表温度突变比例高于井工采区,露天采区和井工采区内地表温度突变面积占比分别为34.66%和19.02%,归因于露天开采比井工开采对地表扰动更加剧烈;土地覆被变化引起地表温度突变,地表温度突变方向和幅度取决于土地覆被变化的类型、规模等。研究结果可为矿区生态环境治理及评价提供科学依据。
文摘准确识别地表变化的时空信息,有助于探究地表自然环境和生态系统发展演变的规律,支撑相关的科研与行政管理工作。本文以河南某生态保护修复工程部分实施范围为研究区域,基于Google Earth Engine(GEE)云平台,以2013—2020年的98景Landsat8/OLI遥感影像作为数据源,应用Breaks for additive season and trend(BFAST)算法对地表变化进行了信息提取和制图。首先基于GEE云平台对Landsat8/OLI地表反射率数据集进行调用和预处理,基于CFMask算法对遥感数据集进行云影掩膜,开展光谱指数(植被指数NDVI)的计算以及时间序列数据集的构建。其次基于时序数据集与BFAST算法构建由趋势项、季节项和残差项组成的广义线性回归模型,通过最小二乘法求解模型中的未知参数集,以此进一步构建时序拟合模型,而后基于残差的Moving sums(MOSUM)方法对时序结构变化进行检测。最后从检测结果中抽取像元样点,通过与Google Earth高分辨率影像数据叠置和目视解译,开展结果验证和精度评价。结果表明,本文提出的方法在研究区的时序地表变化检测中具有较高的检测精度(总体精度为83.7%,2018—2020年分年度检测结果精度分别为86.5%、80.7%、87.7%)。本文提出的方法是遥感大数据库构建、地表生态信息近实时变化扰动识别和监测等技术的一种基础方法,能够对国土空间生态保护修复调查监测和评估预警等工作提供技术支撑和决策支持。
基金supported by the National Key R&D Program of China(Grant Nos.2017YFA0604401 and 2019YFA0606601)the Tsinghua University Initiative Scientific Research Program(Grant No.2019Z02CAU)the Youth Innovation Promotion Association,Chinese Academy of Sciences(Grant No.Y4YR1300QM)。
文摘Annual land use land cover(LULC)change information at medium spatial resolution(i.e.at 30 m)is required in numerous subjects,such as biophysical modelling,land management and global change studies.Annual LULC information,however,is usually not available at continental or national scale due to reasons such as insufficient remote sensing data coverage or lack of computational capabilities.Here we integrate high temporal resolution and coarse spatial resolution satellite images(i.e.,Moderate Resolution Imaging Spectroradiometer(MODIS)and Global Inventory Modelling and Mapping Studies(GIMMS)normalized difference vegetation index(NDVI))with high spatial resolution datasets(China’s Land-Use/cover Datasets(CLUDs)derived from 30-meter Landsat TM/ETM+/OLI)to generate reliable annual nominal 30 m LULC maps for the whole of China between 1980 and 2015.We also test the performance of a statistical based change detection algorithm(Breaks for Additive Seasonal and Trend),originally designed for tracking forest change,in classifying all-type LULC change.As a result,a nominal 30 m annual land use/land cover datasets(CLUD-A)from 1980 to 2015 was developed for the whole China.The mapping results were assessed with a change sample dataset,a regional annual validation sample set and a three-year China sample set.Of the detected change years,75.61%matched the exact time of conversion within±1 year.Annual mapping results provided a detail process of urbanization,deforestation,afforestation,water and cropland dynamics over the past 36 years.The consistent characterization of land change dynamics for China can be further used in scientific research and to support land management for policy-makers.
文摘研究黄河三角洲耕地退化状态并明晰其驱动机制对保障区域粮食安全和促进耕地保护至关重要。本文基于2001—2021年MODIS NDVI数据,采用加性季节和趋势断点检测法BFAST(Breaks For Additive Seasonal and Trend)探讨黄河三角洲耕地退化状态,并应用地理探测器探究各自然和人为因子及其交互作用对不同尺度区域耕地退化状态空间分异的影响与驱动机制。结果表明:(1)与线性趋势相比,非线性趋势方法检测耕地退化状态的精度更高,不仅可以检测到耕地变化的整体趋势,还可检测到长期变化中的阶段性突变信息,可较全面、准确的评估耕地退化状态;(2)BFAST检测结果显示黄河三角洲耕地31.75%呈退化状态,61.06%为改善状态,7.19%呈非显著趋势。黄河三角洲耕地在长期变化过程中多受外界干扰而发生短期突变,其中中断减少趋势占比31.31%,主要分布在孤北水库、黄河故道西侧和东南沿海,呈沿海自内陆逐渐减少且零散化的空间分布格局;中断增加趋势占比55.13%,主要分布在黄河以南、淡水河流和水库附近,呈大规模集中分布特点;(3)黄河三角洲不同尺度区域耕地退化状态空间分异以土地利用为主导,且双因子交互影响力明显高于单因子,土地利用∩高程和距海洋距离∩土地利用分别是耕地退化和改善状态的主导交互因子,人为因素主导和自然因素协同作用导致黄河三角洲耕地退化状态空间分异。
基金supported by the National Natural Science Foundation of China(Grants No.41991231,42041004,and 41888101)the China University Research Talents Recruitment Program(111 project,Grant No.B13045).
文摘Vegetation greening has long been acknowledged,but recent studies have pointed out that vegetation greening is possibly stalled or even reversed.However,detailed analyses about greening reversal or increased browning of vegetation remain scarce.In this study,we utilized the normalized difference vegetation index(NDVI)as an indicator of vegetation to investigate the trends of vegetation greening and browning(monotonic,interruption,and reversal)through the breaks for the additive season and trend(BFAST)method across China’s drylands from 1982 to 2022.It also reveals the impacts of ecological restoration programs(ERPs)and climate change on these vegetation trends.We find that the vegetation displays an obvious pattern of east-greening and west-browning in China’s drylands.Greening trends mainly exhibits monotonic greening(29.8%)and greening with setback(36.8%),whereas browning shows a greening to browning reversal(19.2%).The increase rate of greening to browning reversal is 0.0342/yr,which is apparently greater than that of greening with setback,0.0078/yr.This research highlights that,under the background of widespread vegetation greening,vegetation browning is pro-gressively increasing due to the effects of climate change.Furthermore,the ERPs have significantly increased vegetation coverage,with the increase rate in 2000-2022 being twice as much as that of 1982-1999 in reveg-etation regions.Vegetation browning in southwestern Qingzang Plateau is primarily driven by adverse climatic factors and anthropogenic disturbances,which offset the efforts of ERPs.
文摘以江苏省永久耕地为例,基于2001—2019年中分辨率成像光谱仪(Moderate resolution imaging spectroradiometer,MODIS)遥感影像,开展耕地生产力隐性退化遥感监测和影响因素分析。BFAST(Breaks for additive seasonal and trend)算法用于建模历史时期耕地生产力变化的预期行为,并以此为基准判断监测时期耕地生产力是否存在隐性退化风险。基于地理探测器,从3个准则层的8项指标变量对耕地生产力隐性退化进行了主导影响因素探测和因子交互分析。研究结果表明:江苏省存在生产力隐性退化的耕地比例为21.9%,具有显著的空间差异。西北地区的徐州市、宿迁市的耕地生产力隐性退化比例最高,分别为47.2%和43.4%,且表现出聚集性。东南地区的苏州市、无锡市和南通市的耕地生产力隐性退化比例较低,均不足10%。因子探测分析表明外流人口数量、种植业从业人员数量和农业机械化总动力3项指标对江苏省耕地生产力隐性退化的解释力最强。多因子交互耦合后,人口因素与生产条件解释力增强最为显著。耕地生产力隐性退化的地域分异类型划分为生产条件约束型、产出效益约束型和人口因素约束型。农业机械化总动力、农业产值和外流人口数量分别为3种约束类型的首要因素。从地域空间来看,人口因素约束型地区在江苏省占比最大,主要集中于苏北地区。对于不同约束类型区域分别提出加强高标准农田建设、实施惠农政策、减缓劳动力析出等相应的政策建议。
文摘Land cover change in the semi-arid environment of the eastern Hindu Kush region is driven by anthropogenic activities and environmental change impacts. Natural hazards, such as floods presumably influenced by climatic change, cause abrupt change of land cover. So far, little research has been conducted to investigate the spatiotemporal aspects of this abrupt change in the valleys. In order to explore the abrupt change in land cover and floods as its possible drivers in the eastern Hindu Kush, a semi-arid mountain region characterized by complex terrain, vegetation variation, and precipitation seasonality, we analyzed long-term Landsat image time series from 1988 to 2020 using Breaks For Additive Seasonal and Trend(BFAST). Overall, BFAST effectively detected abrupt change by using Landsat-derived Modified Soil Adjusted Vegetation Index(MSAVI). The results of our study indicate that approximately 95% of the study area experienced at least one abrupt change during 1988-2020. The years 1991, 1995, 1998, 2007, and 2016 were detected as the peak years, with the peaks occurring in different seasons. The annual trend of abrupt change is decreasing for the study area. The seasonality of abrupt change at the catchment level shows an increasing trend in the spring season for the southern catchments of Panjkora and Swat. The spatial distribution patterns show that abrupt change is primarily concentrated in the floodplains indicating that flooding is the primary driver of the land cover change in the region. We also demonstrated the accurate detection of past flood events(2015) based on the two case examples of Ayun, Rumbur, and Kalash valleys. The detection of the flood events was verified by fieldwork and historical high-resolution Google Earth imagery. Finally, our study provides an example of applying Landsat time series in a dry mountain region to detect abrupt changes in land cover and analyze impact of natural hazards such as floods.
文摘在喀斯特分布区,基岩、植被、裸地等多种地表覆盖交错分布,地物覆盖高度异质,并且呈现出短周期规律性变化和长期动态趋势变化,单一时相的影像进行土地覆盖分类精度非常有限。针对这一问题,本文提出一种顾及物候特征的多时相遥感影像分类策略,利用具有高时间分辨率的MODIS NDVI时间序列产品作为数据源,选择BFAST(Breaks For Additive Seasonal and Trend)方法进行NDVI时间序列的物候分解,采用动态阈值法对时序分解的物候轨迹进行标记,最后将物候标记特征与原始光谱时序综合特征进行组合,选择支持向量机(SVM)分类器进行土地利用覆盖分类,并且对比了不同特征空间下的分类结果。以云南省壮族苗族自治州丘北县和砚山县为研究区进行分类实验,结果表明,BFAST模型可以有效地分解出NDVI时序中的关键物候特征,相比基于单纯光谱特征的分类,物候驱动的喀斯特断陷盆地区土地覆盖分类精度有明显的提升,在NDVI、光谱和物候组合特征空间下,土地覆盖分类精度最高,总体精度和Kappa系数分别为88.94%和0.8693,尤其在灌木林、有林地、石旮旯地与稀疏植被的区分中,SOS、POS和GSG等物候特征具有较强的可分性,表明物候特征在地物识别中的有效性。
基金Construction Project of China Knowledge Center for Engineering Sciences and Technology(CKCEST-2017-3-1)Cultivate Project of Institute of Geographic Sciences and Natural Resources Research,Chinese Academy of Science(TSYJS03)National University of Mongolia(P2017-2396)
文摘Drought is a worldwide natural disaster that has long affected agricultural production as well as social and economic activities. Frequent droughts have been observed in the Belt and Road area, in which much of the agricultural land is concentrated in fragile ecological environment. Based on the Tropical Rainfall Measuring Mission Satellite(TRMM) 3 B43 precipitation data, we used the Precipitation Abnormity Percentage drought model to study the monthly spatio-temporal distribution of drought in south region of N50° of the Belt and Road area. It was observed that drought during winter was mainly distributed in Northeast Asia, Southeast Asia, and South Asia, while it was mainly distributed in Central Asia and West Asia during summer. The occurrence of historical droughts indicates an obvious seasonal cycle. The regional variations in drought were analyzed using the Breaks for Additive Season and Trend tool(BFAST) in six sub-regions according to the spatial distribution of six economic corridors in the Belt and Road area. The average drought conditions over the 18 years show a slight decreasing trend in Northeast Asia, West Asia, North Africa, South Asia, Central and Eastern Europe, and a slight increasing trend in Central Asia. However, it was a fluctuating pattern of first increasing and then decreasing in Southeast Asia. The results indicate that the total drought area in the Belt and Road region showed a general decreasing trend at a rate of 40,260 km^2 per year from 1998 to 2015.