One of the most controversial minerals in their origin and occurrence around the world is manganese deposits.The Abu Zenima area is rated one of the most economically important places where manganese ore deposits(Mn O...One of the most controversial minerals in their origin and occurrence around the world is manganese deposits.The Abu Zenima area is rated one of the most economically important places where manganese ore deposits(Mn ODs)are located in the southwest Sinai microplate,Egypt.These deposits are confined with the Um Bogma Formation(UBF)and the reserves of this region are relatively small.In this study,optical and radar data are used in a new challenge as an attempt to reach the closest controls and setting of Mn ODs.Moreover,Frequency Ratio(FR)and Logistic Regression(LogR)predictive models are applied to integrate different geospatial thematic maps,to predict new potential resource zones for increasing the ranges of mining quarries.Landsat8 OLI,Sentinel-2A Multi Spectral Instrument and Radar(Sentinel-1B)data are combined for mapping Mn ODs locations and their relationship with geological structures and the surrounding rocks.Band ratio,Principal and Independent Component Analysis techniques and four classification algorithms were implemented to the optical’VNIR and SWIR bands.Unusually,the interferometric processing steps for Sentinel-1 data were made for understanding the tectonic features in the area.The FR and LogR models are validated during fieldwork with known Mn ODs locations.Results indicate that processed images are capable of differentiation of UBF which broadly distributed in the central and southern parts of the area.Mn ODs possibly formed by thermal events that attributed to paleo-volcanic events before the rift stage.The high accuracy of LogR model(0.902)suggests that high Mn ODs potential zones are identified within the intersected fault zones near granitic units.This integration is recommended for discriminating hydrothermally Mn ODs in other arid geographic regions.展开更多
由于热带地区的雨季时间较长,云覆盖严重,基于光学影像难以准确提取区域内的水稻种植模式。该文以泰国湄南河流域中部平原水稻种植区为例,基于Sentinel-1SAR时间序列数据,提出一种融合时序统计参数与时序曲线相似性特征的热带地区水稻...由于热带地区的雨季时间较长,云覆盖严重,基于光学影像难以准确提取区域内的水稻种植模式。该文以泰国湄南河流域中部平原水稻种植区为例,基于Sentinel-1SAR时间序列数据,提出一种融合时序统计参数与时序曲线相似性特征的热带地区水稻种植结构提取方法。首先利用年内所有可获取的Sentinel-1SAR数据,分别基于像元和基于对象构建后向散射系数时间序列曲线,提取时序特征参数;利用动态时间规整(Dynamic Time Warping,DTW)算法,计算后向散射系数时序曲线与地物标准曲线间的隶属度;将时序特征参数、时序曲线隶属度相结合,利用随机森林模型进行机器学习监督分类,提取研究区的水稻种植信息并评价分类精度。结果表明,基于Sentinel-1SAR时序特征融合的算法可以较好地提高水稻种植结构分类精度。其中,基于对象的分类算法的单季稻提取用户精度为81.46%,生产者精度为82.00%;双季稻用户精度为88.0%,生产者精度为84.08%,均优于基于像元的分类算法。研究结果可为多云多雨的热带地区水稻种植信息提取提供一种新的思路。展开更多
文摘One of the most controversial minerals in their origin and occurrence around the world is manganese deposits.The Abu Zenima area is rated one of the most economically important places where manganese ore deposits(Mn ODs)are located in the southwest Sinai microplate,Egypt.These deposits are confined with the Um Bogma Formation(UBF)and the reserves of this region are relatively small.In this study,optical and radar data are used in a new challenge as an attempt to reach the closest controls and setting of Mn ODs.Moreover,Frequency Ratio(FR)and Logistic Regression(LogR)predictive models are applied to integrate different geospatial thematic maps,to predict new potential resource zones for increasing the ranges of mining quarries.Landsat8 OLI,Sentinel-2A Multi Spectral Instrument and Radar(Sentinel-1B)data are combined for mapping Mn ODs locations and their relationship with geological structures and the surrounding rocks.Band ratio,Principal and Independent Component Analysis techniques and four classification algorithms were implemented to the optical’VNIR and SWIR bands.Unusually,the interferometric processing steps for Sentinel-1 data were made for understanding the tectonic features in the area.The FR and LogR models are validated during fieldwork with known Mn ODs locations.Results indicate that processed images are capable of differentiation of UBF which broadly distributed in the central and southern parts of the area.Mn ODs possibly formed by thermal events that attributed to paleo-volcanic events before the rift stage.The high accuracy of LogR model(0.902)suggests that high Mn ODs potential zones are identified within the intersected fault zones near granitic units.This integration is recommended for discriminating hydrothermally Mn ODs in other arid geographic regions.
文摘由于热带地区的雨季时间较长,云覆盖严重,基于光学影像难以准确提取区域内的水稻种植模式。该文以泰国湄南河流域中部平原水稻种植区为例,基于Sentinel-1SAR时间序列数据,提出一种融合时序统计参数与时序曲线相似性特征的热带地区水稻种植结构提取方法。首先利用年内所有可获取的Sentinel-1SAR数据,分别基于像元和基于对象构建后向散射系数时间序列曲线,提取时序特征参数;利用动态时间规整(Dynamic Time Warping,DTW)算法,计算后向散射系数时序曲线与地物标准曲线间的隶属度;将时序特征参数、时序曲线隶属度相结合,利用随机森林模型进行机器学习监督分类,提取研究区的水稻种植信息并评价分类精度。结果表明,基于Sentinel-1SAR时序特征融合的算法可以较好地提高水稻种植结构分类精度。其中,基于对象的分类算法的单季稻提取用户精度为81.46%,生产者精度为82.00%;双季稻用户精度为88.0%,生产者精度为84.08%,均优于基于像元的分类算法。研究结果可为多云多雨的热带地区水稻种植信息提取提供一种新的思路。
