高光谱遥感影像混合像元分解研究进展

受高光谱成像仪低空间分辨率及复杂地物的影响,高光谱遥感图像存在大量混合像元。为提高地表分类精度以及满足亚像元级目标探测的需求,混合像元分解技术一直是高光谱遥感研究热点之一。本文主要对高光谱混合像元分解技术中的核心问题:端元数目估计、端元提取算法、丰度估计算法进行综述,系统地分析了各种典型算法的原理及优缺点,进一步阐述研究过程中建立高精度遥感混合反演模型与遥感产品业务化中的混合像元分解技术难题,同时针对今后混合像元分解技术发展方向,指出在继续引入新型算法理论方法基础上,结合用户应用需求,推进高光谱混合像元分解算法业务化应用,为高光谱遥感工程化应用提供支持。     

双向相位平滑伪距的硬件延迟解算研究

针对利用GPS观测数据提取TEC过程中最主要的误差来源硬件延迟问题,该文为了获取高精度TEC,在对双频观测数据处理时,改进了基于Hatch滤波的相位平滑伪距算法的使用方法,即双向平滑,取得较好的效果。研究采用了VTEC多项式和球冠谐分析模型来进行区域电离层建模及硬件延迟解算,经比较模型解算的硬件延迟与IGS发布值最大差异不超过1ns,其中VTEC多项式模型解80%差异值小于0.5ns,球冠谐函数模型解所有差异值均小于0.5ns。     

黔西北大方地区龙潭组海陆过渡相烃源岩特征及评价

黔西北大方地区龙潭组以泥页岩、煤层与砂岩交替互层发育为特点,具有典型的海陆过渡相特征。文章在剖面实测和样品采集的基础上,结合全岩和黏土矿物XRD分析、总有机碳(TOC)含量测定、岩石热解、干酪根显微组分及镜质体反射率(R_O)等分析,对研究区龙潭组烃源岩发育特征进行了研究和评价。结果表明: ① 龙潭组矿物以黏土矿物为主,具有较强的吸附能力,但黏土矿物含量与石英含量呈明显的负相关关系(R>0.8),尤其是泥页岩中70%以上的黏土含量以及较低的脆性指数增加了压裂开发的难度。 ② 龙潭组有机质丰度受岩性变化控制明显。煤层具有最高的TOC含量(平均值42.9%)和生烃潜量(平均值为2.68 mg/g),显示强大的煤层气生烃潜力;泥页岩生烃潜量均小于2 mg/g,但约有80%样品的TOC含量超过2%,页岩气资源潜力不及煤层气;粉砂岩生烃潜力最差,生烃潜量平均值为0.13 mg/g,TOC含量为1.4%~5.6%,显示一定的致密气潜力。 ③ 龙潭组有机显微组以壳质组和镜质组占主导,干酪根以Ⅲ型为主,Ⅱ₂型为辅,热演化程度高,处于高-过成熟的生干气阶段。 ④ 煤层更为发育的龙潭组中段是大方地区煤层气、页岩气和致密砂岩气联合勘探的有利层段。     

岩体多尺度异质性及其力学行为

异质性是岩体的重要属性之一,从岩体的物质组成到结构特征,再到微宏观力学行为表现,异质性现象普遍存在。本文在梳理已有研究成果的基础上,从岩体的细微观异质性、结构面异质性和宏观异质性3个不同的尺度,对异质性的产生原因、表现特征、力学效应和时空演化规律等进行阐述,并介绍了相关研究方法以及异质性在不同尺度上的差异表现。岩体异质性的研究,是岩体工程地质力学理论体系的延伸,而多尺度的研究方法也成为全面揭示岩体变形破坏机制的一种有效手段。     

青藏高原错那—沃卡裂谷中部电性结构及其动力学意义

藏南裂谷作为青藏高原最显著的伸展构造样式之一,是研究高原生长过程的重要窗口,但目前其深部成因机制仍存在较大争议。本文利用沿错那—沃卡裂谷中部的大地电磁数据,分析裂谷区域的大地电磁测深曲线特征和相位张量,并通过三维大地电磁反演获得邛多江地堑和沃卡地堑深部电性结构。三维大地电磁反演结果显示,沃卡地堑和邛多江地堑深部存在一条连续的显著高导异常,并呈现“俯冲”形态,且上覆高阻结构体,而在邛多江地堑两侧浅部则分布低阻异常。结合早期的研究结果,本文支持错那—沃卡裂谷深部的高导异常为地壳部分熔融,可能与南向地壳流相关,并研究认为在南北向持续挤压作用下,裂谷下方的弱地壳层,通过解耦上地壳和下部地壳,促进了藏南裂谷系的发育。     

电感耦合等离子体质谱-原子荧光光谱法研究上海口岸进口印度尼西亚煤炭微量元素的赋存形态特征

印度尼西亚是我国最大的煤炭进口国,本文应用电感耦合等离子体质谱、原子荧光光谱、直接测汞仪等技术分析了上海口岸31批进口印度尼西亚煤炭中的12种微量元素,结合数理统计方法研究该类煤炭中微量元素的赋存形态。结果表明,进口印尼煤炭中含有高汞煤、三级含砷煤,As、Hg的平均富集系数大于1,其迁移风险值得关注;Be、Cu、Mo、Cd、Sn、Pb含量均低于中国煤和世界煤炭的平均水平,体现出印尼煤炭低灰分的品质特征。12种微量元素和相关项目(灰分和全硫)可划分为3类:第一类归纳为黏土矿物吸附类,包括As、Be、Cr、Co、Ni、Cu、Mo、Cd、Sn、Pb、灰分;第二类归纳为硫铁矿类,包括Hg、全硫;第三类归纳为碳酸盐矿物类,包括Ba。本文研究结果对于指导进口煤炭开发、利用过程中的环境评价和洁净化处理具有一定的参考价值。     

基于GMS鄱阳湖拟建枢纽对地下水影响探讨

针对拟建鄱阳湖水利枢纽工程对地下水影响问题,选择赣江三角洲为代表区,采用资料收集、野外调查、统计分析和数值模拟等方法进行研究。依据地质钻探资料、地下水与河流动态关系、区域数字高程模型,结合鄱阳湖历史形成与演化过程确定研究区边界条件,运用GMS模拟软件建立了研究区地下水三维运动模型。依据推荐的枢纽调度方案,利用数值模型计算枢纽运行后对地下水运动影响的时空变化规律。结果表明:枯水年份影响大,距离河流、坝址近的地方影响大,反之亦然。影响幅度范围0~2m,地下水径流交替因此减弱,丰水年份土壤潜育化面积增加9.3%。     

The Paleoceanography during the Time with High δ13C of Phanerozoic marine carbonates

Carbon isotopic composition of marine carbonates is a record for various important geological events in the process of earth development and evolution. The carbonates of Carboniferous, Permian and Triassic, as the transition from Paleozoic to Mesozoic-Cenozoic have very high ¹³C value. Taking this as the main point, and combined with the oxygen, strontium isotopic composition in carbonates, distribution of carbonate basin area through geologic time, the correlation of carbon isotopic composition of marine carbonates to sea level change, organic carbon burial flux, exchange of CO₂ content in atmosphere and ocean, and long cycle evolution of the earth ecosystems were approached. The results are shown as follows: ①The interval of ¹³C >3‰ during Phanerozoic was concentrated in Carboniferous, Permian and the beginning of Triassic, but the beginning of Triassic was characterized by higher frequency and larger fluctuations in ¹³C value during a short time, whereas the Carboniferous-Permian presented a continuously stable high ¹³C value, indicating a larger amount of organic carbon accumulation in this time interval. Relatively high ¹⁸O values during this time was also observed, showing a long time of glaciations and cold climate, which suggest a connection among rapid organic carbon burial, cold climate, as well as pCO₂ and pO₂ states of atmosphere. ②The over consumption of atmosphere CO₂ by green plants during the time with high ¹³C of seawater forced CO₂ being transferred from ocean to atmosphere for the balance, but the decrease in the seawater amount and water column pressure caused by the global cooling could weaken dissolution capacity of CO₂ in seawater and carbon storage of marine carbonates, and also reduce the carbonate sedimentary rate and decrease the carbonate basin area globally from Devonian to Carboniferous and Permian. During the middle-late Permian carbonate was widely replaced by siliceous sediments even though in shallow carbonate platform, which resulted in the decrease of marine invertebrates, suggesting the Permian chert event should be global. ③The Phanerozoic ⁸⁷Sr/⁸⁶Sr trend of seawater showed a sharp fall in Permian and drop to a minimum at the end of the Permian, indicting input of strontium from the submarine hydrothermal systems (mantle flux). Such process should accompany with a supplement of CO₂ from deep earth to atmosphere and ocean system, but the process associated with widespread volcanism and rises of earth’s surface temperature pricked up the mass extinction during the time of end Permian. ④Cold climate and increase of continental icecap volume, the amalgamation of northern Africa and Laurentia continentals were the main reasons responsible for the sea level drop, but the water consumption result from the significantly increased accumulation of organic carbon should also be one of the reasons for the sea level drop on the order of tens of meters. ⑤The mass extinction at the end Permian was an inevitable event in the process of earth system adjustment. It was difficult for marine invertebrates to survive because of the continuously rapid burial of organic carbon, and of the decrease of sea water amount and its dissolution ability to CO₂. At last, at the end of Paleozoic, the supplement of CO₂ to atmosphere and ocean by widely magma activities resulted in a high temperature of earth surface and intensified mass extinction.     

结合MRF与ν-SVM的SAR海冰图像分类

结合支持向量机SVM和马尔可夫随机场MRF在图像分类领域的优点,提出MRF-ν-SVC分类系统。与以往针对像素的分类不同,本系统在区域水平实现分类。首先通过双阈值准则判别区域间边缘的强弱,然后针对模糊边缘区域将基于MRF的空间语境模型改进为边缘语境模型,修正SVM原始问题中的偏差因子,从而优化最优分类超平面。实验表明,本文算法可有效提高分类精度,增强对斑点噪声的抑制能力,更好地分类特征相似的区域。     

基于双边结构张量的局部自适应图像超分辨率重建

利用超分辨率重建技术可以有效提高图像的空间分辨率,其中先验模型的选取尤为关键。在最大后验概率(MAP)的框架下引入双边结构张量测度,联合像素邻域4个方向的梯度算子提出一种局部自适应先验方法,构建图像超分辨率重建模型,并利用迭代重加权范数(iteratively reweighted norm)对其进行转换求解。基于标准的测试图像进行了实验,并将本方法与拉普拉斯先验、Huber-Markov先验以及BTV先验的重建结果进行对比,验证了该方法的有效性。