函数模型与随机模型双约束的GPS-InSAR数据融合方法建立三维形变场

针对InSAR技术研究地表三维形变时监测信息不足的问题,以GPS监测信息为先验信息,建立附有随机模型约束的地表三维形变模型。考虑到SAR卫星极轨方式运行导致LOS向观测量对南北向形变不敏感的问题,以GPS南北向形变观测值作为强约束,构建三维形变解算的函数约束条件。模拟数据与西安地区实测数据的计算结果表明,基于随机模型与函数模型共同约束的地表三维形变参数最小二乘解的精度优于仅有函数模型约束或仅有随机模型约束及无任何约束的参数解精度。     

用随机有限断层方法模拟2016-11-13新西兰Mw7.8地震

在未来大地震发震构造已知的前提下,提出用经验公式确定大地震断层长度、宽度、地震矩和滑动分布等参数,并用随机有限断层方法预测未来大地震。用随机有限断层方法模拟2016-11-13新西兰M_W7.8地震的12个基岩台站地震动时程和反应谱,并用模拟的频谱幅值与记录频谱幅值平均比值确定模拟误差。结果表明,在周期为0~10 s的范围内,模拟误差在0.92~1.08之间,不同频率模拟误差的标准差变化范围不超过1。95%的置信区间宽度随频率无明显变化,模拟结果反映了地震动记录的平均效果。用准随机方法重新获取新西兰大地震断层初始破裂点和位错滑动分布,模拟地震动并计算模拟误差随频率的变化,得到与最初模型相似的结论,从而进一步证实了本文提出的获取震源参数并用随机有限断层方法预测未来大地震的可靠性,且其特别适用于大地震远场模拟。     

基于小波神经网络耦合模型对云南星云湖富营养化气象驱动因子的分析

气象因子是影响湖泊富营养化的重要因素,而湖泊富营养化对人群健康、生态系统和社会经济等均有负面影响.本文基于统计资料及遥感数据,结合Morlet小波分析和BP多层前馈神经网络(BP神经网络)构建了不同时间尺度下的小波神经网络耦合模型,分析了19862011年云南星云湖水华强度变化与月降雨量、月平均气温、月平均风速、月日照...     

Application of a vanishing, quasi-sigma, vertical coordinate for simulation of high-speed, deep currents over the Sigsbee Escarpment in the Gulf of Mexico

Recent observations over the Sigsbee Escarpment in the Gulf of Mexico have revealed extremely energetic deep currents (near 1 m s⁻¹), which are trapped along the escarpment. Both scientific interest and engineering needs demand dynamical understanding of these extreme events, and can benefit from a numerical model designed to complement observational and theoretical investigations in this region of complicated topography. The primary objective of this study is to develop a modeling methodology capable of simulating these physical processes and apply the model to the Sigsbee Escarpment region. The very steep slope of the Sigsbee Escarpment (0.05–0.1) limits the application of ocean models with traditional terrain-following (sigma) vertical coordinates, which may represent the very complicated topography in the region adequately, can result in large truncation errors during calculation of the horizontal pressure gradient. A new vertical coordinate system, termed a vanishing quasi-sigma coordinate, is implemented in the Navy Coastal Ocean Model for application to the Sigsbee Escarpment region. Vertical coordinate surfaces for this grid have noticeably gentler slopes than a traditional sigma grid, while still following the terrain near the ocean bottom. The new vertical grid is tested with a suite of numerical experiments and compared to a classical sigma-layer model. The numerical error is substantially reduced in the model with the new vertical grid. A one-year, realistic, numerical simulation is performed to simulate strong, deep currents over the Escarpment using a very-high-resolution nested modeling approach. The model results are analyzed to demonstrate that the deep-ocean currents in the simulation replicate the prominent dynamical features of the observed intense currents in the region.     

1st0pt在沉降预测中的应用研究

预测方法有多种,找一种操作简单的方法较难。本文利用综合优化软件包1 st0pt,选用几种预测模型,通过对实测数据进行拟合,得出各种预测方法的计算模型,再利用最小二乘法原理,以预测绝对误差平方和最小为目标,给每种预测模型赋予不同的权重,组成规划方程,求出权重数列并组成新的预测模型,再对建筑物进行预测,将得到较好的预测结果。     

A study on the reference frame consistency in recent Earth gravitational models

All gravity field functionals obtained from an Earth gravitational model (EGM) depend on the underlying terrestrial reference frame (TRF), with respect to which the EGM’s spherical harmonic coefficients refer to. In order to maintain a coherent framework for the comparison of current and future EGMs, it is thus important to investigate the consistency of their inherent TRFs, especially when their use is intended for high precision studies. Following the methodology described in an earlier paper by Kleusberg (1980), the similarity transformation parameters between the associated reference frames for several EGMs (including the most recent CHAMP/GRACE models at the time of writing this paper) are estimated in the present study. Specifically, the differences between the spherical harmonic coefficients for various pairs of EGMs are parameterized through a 3D-similarity spatial transformation model that relates their underlying TRFs. From the least-squares adjustment of such a parametric model, the origin, orientation and scale stability between the EGMs’ reference frames can be identified by estimating their corresponding translation, rotation and scale factor parameters. Various aspects of the estimation procedure and its results are highlighted in the paper, including data weighting schemes, the sensitivity of the results with respect to the selected harmonic spectral band, the correlation structure and precision level of the estimated transformation parameters, the effect of the estimated differences of the EGMs’ reference frames on their height anomaly signal, and the overall feasibility of Kleusberg’s formulae for the assessment of TRF inconsistencies among global geopotential models.     

自适应空间邻域分析和瑞利-高斯分布的多时相遥感影像变化检测

提出了一种基于自适应空间邻域分析和瑞利-高斯模型(Rayleigh-Gauss models, RGM)分布的多时相遥感影像自动变化检测方法。该方法把自适应空间邻域信息和改进的差值影像与比值影像乘积变换融合 法(improved multiplying transform fusion, IMTF)结合构造差异影像, 可以有效地抑制噪声和消除多时相影像之间配准误差的影响, 具有更强的鲁棒性。在对差异影像的分割处理中, 运用瑞利和高斯模型分别模拟变化类像元和非变化类像元的分布情况, 然后估计出两类像元的概率     

综合多信息进行地震约束储层随机建模——以准噶尔盆地庄1井区J1S2^2砂组为例

针对地震约束储层随机建模中地震反演波阻抗数据体与岩相概率关系差的问题,提出综合多信息、逐级约束、多级建模的思路方法。首先以地震反演数据体约束建立地球物理特征曲线随机参数场,再以每一个地球物理特征曲线参数场为约束建立相模型,最后以优选的相模型和地球物理特征曲线参数场为约束进行储层参数建模。该方法在井距较大的开发前期评价阶段具有较好的应用效果。     

黄东海表观光学量与固有光学量经验模式研究

表观光学特性和固有光学特性是海洋光学研究领域的两大主要研究内容。文中通过2003年黄东海试验和2007年908专项海洋光学调查所获取的表观光学参数中的遥感反射率Rrs(λ)、水下遥感反射率rrs(λ)、海水透明度Zs、下行辐照度漫衰减系数Kd(490)、固有光学参数中的后向散射系数bb(λ)和水体吸收系数a(λ),建立了区域性、季节性的表观光学量和固有光学量之间的经验模式,并对模式进行了误差分析和改进。