天山天池水体季节性分层特征

于2014年6-10月,对高山深水湖泊天山天池水温、电导率、溶解氧、p H值、叶绿素a浓度和蓝绿藻细胞密度进行垂直剖面的连续监测,通过对其季节动态和垂直分层结构的分析,探讨天池水体季节性分层特征.天池出现明显水温分层的时间短(6-9月),夏季温跃层变化范围为2~18 m,而秋季温跃层不断下移,10月在18 m水深以下;受水温分层影响,天池水体溶解氧浓度、电导率、p H值、叶绿素a浓度和蓝绿藻细胞密度在垂直剖面表现出明显的季节性分层,尤其是夏季水温分层影响溶解氧浓度、叶绿素a浓度和蓝绿藻细胞密度在水体中的分布,对天池水质变化产生重要影响.天池浅水层(水深小于10 m)溶解氧浓度较高(大于8 mg/L),而深水层(水深超过18 m)溶解氧浓度9月接近4 mg/L,季节性缺氧导致底泥营养盐向上扩散,对水体水质产生不利影响.所以,应在夏、秋季节加强水质监测,以防止天池水华发生;天池叶绿素a浓度与蓝绿藻细胞密度的垂直剖面变化趋势相似,均随水深增加呈先增加后减小的趋势,但叶绿素a浓度在2~12 m水深处较高,蓝绿藻细胞密度在5~15 m水深处较高,表明5~15 m深度适合藻类生长,同时,电导率、p H值的垂直变化也说明藻类的生长情况,这为监测天池水体富营养化取样和分析提供依据.     

深井电极交叉供电对自然电位的影响

通过实验论证平凉地震台深井电阻率观测自然电位畸变由电极交叉供电造成,讨论分析不同电极供电对自然电位的影响。结果表明:对测量极供电,会影响自然电位,不会影响电阻率测值;单极供电对自然电位的影响表现为阶跃突跳,并以指数形态恢复,恢复时间约10 h,干扰峰值与供电电流、供电时间等有关;对观测数据进行校正,取得较好效果。     

东海八角亭海洋深井台地震监测

上海市地震局设计、建设的中国第1个真正意义上的海洋深井地震观测台位于东海石油平台上,井深780 m,安装2套JDF-1型地震计进行观测。通过对陆域地震台站记录的大地震进行分析比较、方位校准,采用单台地震分析方法,确定2009年5月—2011年6月记录到的136个地震参数,总结海域地震波形特征。结果表明,该海域地震活动性特征与当地地质构造一致。     

深海6000 m拖曳式瞬变电磁系统及其应用

针对我国西南印度洋合同区热液硫化物矿快速、有效以及便捷地质勘探装备的要求,研发了由甲板控制系统、万米光电复合缆、仪器舱拖体和天线拖体组成的深海6000 m拖曳式瞬变电磁系统.为了便于拖体布放和快速发现异常,选择重叠回线收发装置类型;采用理论数值计算确定了在拖曳高度不大于50 m的前提下,观测1～100ms窗范围内的二次场响应,可以发现近海底深海热液硫化物矿堆;另外,通过不同拖延深度海上试验,研究了拖曳深度对瞬变电磁的影响规律及仪器性能.大洋第30航次第二航段在西南印度洋脊热液区,应用该系统发现了明显的瞬变电磁异常.印证了系统的有效性.     

Deep soil water dynamics in an undisturbed primary forest in central Amazonia: Differences between normal years and the 2005 drought

Understanding how Amazonian rainforests deal with extended droughts is critical in the face of changing climate. This research analyze the physical properties and the soil water dynamics of a deep soil profile in an area of primary forest in central Amazonia to elucidate these processes under drought and nondrought conditions. Physical soil properties derived from soil cores exhibited a distinctive layer between 480 and 880 cm deep, characterized by higher microporosity and low plant water availability. In situ soil moisture measurements collected during the period from January 2003 through February 2006 and for depths ranging from 10 to 1,430 cm suggest that, in the study site, the top 480 cm of the soil profile satisfied most of the transpirational demands in normal climatological years. However, during exceptionally dry periods, such as the 2005 drought, root uptake occurs below 480 cm. As concluded by previous studies, most of the uptake is concentrated in the first meter of the soil profile: More than 40% of the total demand for transpiration is supplied by the top meter of soil. Because deep root uptake occurred at greater depths than normal during the 2005 drought, our results suggest that this is a fundamental mechanism to cope with prolonged droughts.     

不滑移状态下深埋圆形隧道地震响应研究

地下结构的地震响应主要取决于由地震波传播产生的土体变形与土结相互作用。剪切波传播过程中将会引起隧道衬砌的椭圆化变形,进而降低衬砌有效承载力。剪切波作用下的深埋圆形隧道可认为处于均质的纯剪状态,基于相对刚度法的拟静力解析解可充分考虑土结相互作用对隧道结构内力的影响。基于此,本文将通过有限元数值分析获得的自由场地地基变形引入不滑移状态下深埋圆形隧道内力求解公式,并结合二维和三维数值模拟途径,将动力分析结果与解析解结果进行对比分析,以评价各种解析方法的适用性和数值途径的可靠性。     

软悬挂状态下隔水管动力特性分析

建立了软悬挂状态下隔水管动力分析模型,采用有限单元法对模型进行离散,结合Newmark-β法对动态方程求解,并对模型进行验证,最后选取南海某口深水井实际参数,对软悬挂状态下隔水管动力特性进行分析。研究结果表明:软悬挂状态下隔水管底端横向位移最大,顶部铰接处弯矩最大。软悬挂状态下,隔水管横向位移与弯矩随着海底隔水管总成(LMRP)重量增加而减小,随着平台运动速度、平台航向与海流方向夹角增大而增大;回收隔水管减小悬挂长度有利于减小最大横向位移,但会造成弯矩轻微增加;采用旋转刚度适当较大的挠性接头则有利于减小隔水管挠性接头以下整体位移与弯矩。研究成果为钻井平台避台撤离时隔水管的安全提供理论指导。     

多源遥感影像深度特征融合匹配算法

针对多源遥感影像之间成像机理不同、非线性光谱辐射畸变大以及灰度梯度差异明显等所导致的匹配困难问题,提出深度特征融合匹配算法(Feature Fusion Matching Algorithm, FFM)。(1)通过构建特征图金字塔网络提取影像深度特征,使用特征连接结构将语义丰富的高层特征与定位精确的低层特征互补融合,解决多源影像同名特征难以表征的问题并提高特征向量的定位精度;(2)对原始维度1/8的特征图进行交叉变换来融合自身邻域信息与待匹配影像特征信息,通过计算特征向量间的相似性得分得到初次匹配结果,针对特征稀疏区域,提出滑动窗口自适应得分阈值检测算法来提升匹配效果;(3)将匹配结果映射至亚像素级特征图,在小窗口内计算像素间的匹配概率分布期望值来检校优化匹配结果,提高匹配点对的准确性;(4)使用PROSAC算法对匹配结果进行提纯,有效剔除误匹配的同时最大限度保留正确匹配点。试验选取6对多源遥感影像,将FFM同SuperPoint、SIFT、ContextDesc以及LoFTR算法进行对比,结果表明FFM算法在匹配点正确率、匹配点均方根误差以及分布均匀度等方面远优于其他算法。将FFM匹...     

基于LBS和深度学习的旅游景区客流量的高时频预测

为实现精准的旅游景区客流量的高时频预测,本研究构建了一套基于LBS和深度学习模型的预测方法。此方法可通过对LBS数据的转换实现预测的空间范围与时频控制,并通过方法的核心模型——基于双向循环神经网络和GRU算法构建的深度双向GRU(DBi-GRU)模型完成预测。为检验方法的有效性,研究以深圳大梅沙海滨公园为例对方法进行实验测试。实验使用拟合曲线、误差指标及DM检验3种方法评估DBi-GRU模型的预测效果。此外,实验还设置了其他五种深度学习模型作为DBi-GRU的对照模型,测试基于不同深度学习算法的模型之间的预测水平差异。实验结果表明:(1)本研究提出的DBi-GRU模型在景区客流量高时频预测中具有理想的预测效果,在高峰时段的客流量预测方面也具有较高准确性,预测效果明显优于其他深度学习模型;(2)基于双向循环网络的模型的效果普遍优于基于常规循环网络的模型。尤其是基于双向LSTM算法的模型,虽然预测的准确度略逊色于DBi-GRU模型,但在模型性能上与其的差异并不显著;(3)在相同网络参数下,GRU算法较前人采用的LSTM和RNN算法有着更高的预测准确性。本研究为客流量预测领域的研究提供了一种...     

Depth of closure: New calculation method based on sediment data

Obtaining depth of closure (DoC) in an accurate manner is a fundamental issue for coastal engineering, since good results for coastal structures and beach nourishment depend mainly on DoC. Currently, there are two methods for obtaining the DoC, mathematical formulations and profile surveys. However, these methods can incur important errors if one does not take into account the characteristics and morphology of the area, or if one does not have a sufficiently long time series. In this work the DoC is obtained from the break in the trend of the sediment with the depth, that is, in general with the increase of the depth a decrease in the size of the sediment takes place. However, at one point this tendency changes and the size increases, and then decreases again. When comparing the point where the minimum sediment size occurs before the increase, it is observed that the error incurred is small compared to other methods. If the Standard Deviation of Depth Change (SDDC) method is considered as the most accurate method, the error incurred by the proposed method is less than 7%. In addition, it can be seen that the dispersion of the sediment method always occurs outside the zone of bar movement. Whereas in the methods of profiles survey (using 2 cm precision profiles), sometimes the DoC is obtained within the active zone of bar movement. In addition, where the relative minimum of the median sediment size is found, and the sizes of 0.063 and 0.125 mm predominate in the composition of the sample. Therefore, this new method allows the precise location of the DoC to be obtained in a fast and simple way. Furthermore, this method has the advantage that it is not affected by the modifications that may be experienced by both the study area and the cross-shore beach profile.