第四讲c(3)ArcMarine
基于GPS_3G_GIS的多功能海上台风预警和搜救系统研究
[收稿日期]2009208208 [基金项目]福建省自然科学基金项目(2009J 01294)。
[作者简介]张世良(19732),男,1996年大学毕业,硕士,讲师,现主要从事GIS 的方面的教学与研究工作。
基于GPS/3G /GIS 的多功能海上台风预警和搜救系统研究 张世良 (宁德师范高等专科学校物理系,福建宁德352100)[摘要]根据国内外台风预警和搜救系统的无线通讯方式的发展现状,针对目前台风预警和搜救系统存在的可靠性低、实时性差、功能单一等问题,设计了一种基于GPS/3G /GIS 的多功能台风预警和搜救系统。
阐述了台风预警和搜救系统的组成、功能和工作原理,重点介绍了系统终端硬件设计和软件设计,3G 通信数据传输设计,后台监控系统设计,以及对GPS 定位数据中断时的补偿技术。
[关键词]GPS ;3G;GIS ;动态视频;船舶监控[中图分类号]TP393109[文献标识码]A [文章编号]167321409(2009)042N217203随着海洋事业的不断向前发展,海上台风预警和搜救问题已越来越成为急需解决的问题,目前最常用的无线数据网络有GPRS 或CDMA11x 。
无论是CDMA11x 还是GPRS 网络,其上行带宽均在100kbp s 以下,监控画面的图像格式要求在320×240以上。
通常情况下传输的是一幅幅采集时间间隔为几秒的图像,图像连续性差,还无法传输实时连贯的画面。
即使传输QCIF 格式的连续视频画面也要通过多路捆绑才能实现[1,2]。
因此系统终端和后台监控系统只通过短消息等方式实现通信联系,有限的数据传输制约了终端功能的扩展,无法解决无线视频传输所需移动通信网络的带宽瓶颈问题,从而导致对海上台风预警和搜救始终不能普及。
为此,笔者设计了一种基于GPS/3G/GIS 的多功能台风预警和搜救系统。
1 系统总体设计系统整体结构如图1所示。
图1 系统整体结构・712・长江大学学报(自然科学版) 2009年12月第6卷第4期:理工Journal of Yangtze U niversity (N at Sci Edit) Dec 12009,Vol 16No 14:Sci &Eng 1)终端适配层 包括高速摄像头、打印机和传感器等可扩展的外部设备。
第四讲(4)ArcMarine
水文水资源学院 地理信息科学系 张晓祥 Fall, 2008
Part II: ArcMarine
GIS for Marine and Coast
The marine and coastal community is a gathering place to share the work, tools, and methods of marine researchers and professionals that use GIS. (海 洋和海岸带界在海洋研究人员和专业人员之间共享GIS的工作/工具和方法) The coastal zones are home to the majority of our global population, and the oceans and seas provide some of the earth‘s most important and dynamic elements(全世界大部分人口分布在海岸带地区,海洋也很重要). As the tools marine practitioners use to gather data have become more sophisticated, so have the methods for management of the resulting information. Managing and mapping the 71 percent of our earth that is water ห้องสมุดไป่ตู้resents a unique set of challenges.(管理和绘制占全球71%的领土是一个独一 无二的挑战). From oceanography to hydrography(从海图到水文图), navigation to defense(导 航到国防), from the coastal shoreline to the bathymetric bottom(从海岸线到海 底)—marine GIS has been adapted and utilized to assist researchers and organizations in achieving their goals. Collaborative efforts are under way to define the ArcGIS Marine Data Model.(大家合作定义ArcGIS Marine Data Model)
2014-MIKE 21-港口波浪泥沙班SW培训教程
2
开始 .............................................................................................................................. 7 2.1 介绍 ........................................................................................................................ 7 2.2 数据位置 ................................................................................................................. 8 2.3 步骤 1 – 建立一个工作区域 ..................................................................................... 8 2.4 步骤 2 – 导入模型边界线......................................................................................... 8 2.5 步骤 3 – 编辑陆地边界线........................................................................................ 11 2.6 步骤 4 – 定义计算域.............................................................................................. 13 2.7 步骤 5 – 定义开边界.............................................................................................. 14 2.8 步骤 6 – 生成网格 ................................................................................................. 15 2.9 步骤 7 – 对陆地边界进行光滑处理 ........................................................................ 16 2.10 步骤 8 – 对网格进行光滑处理 ........................................................................... 18 2.11 步骤 9 – 网格地形插值 ...................................................................................... 18 2.12 步骤 10 – 使用多边形来控制节点密度 ............................................................... 21 2.13 步骤 11 – 使用多边形来定义网格生成方案 ........................................................ 24 2.14 步骤 12 – 网格分析及改进................................................................................. 26
3D inversion of marine CSEM and MT data An approach to shallow-water problem
3D inversion of marine CSEM and MT data:An approach to shallow-water problemYutaka Sasaki 1INTRODUCTIONIn the last decade,the marine controlled-source electromagnetic (CSEM)method has become an important tool for offshore hydro-carbon exploration and reservoir appraisal (Constable and Srnka,2007).In shallow-water CSEM surveys,the signal propagation in the atmosphere (or so-called airwave)can pose a significant chal-lenge for detection and characterization of reservoirs (Weiss,2007;Andreis and MacGregor,2008).Several approaches have been pro-posed to remove this masking effect from frequency-domain CSEMdata (e.g.,Amundsen et al.,2006;Chen and Alumbaugh,2011).However,because the airwave signal is affected by seafloor resis-tivity and contains information about subseafloor structure,the most effective approach to tackling the shallow-water problem would be to carry out inversion in which the airwave component is accurately modeled (Constable,2010;Darnet et al.,2010).Several 3D inversion algorithms have been developed to recon-struct resistivity models from marine EM data.Plessix and van der Sman (2007)and Commer and Newman (2008)used staggered-grid finite-difference (FD)methods in the frequency domain to compute EM responses.The 3D inversion scheme developed by Zach et al.(2008)is based on an FD time-domain (FDTD)method that pro-duces the forward solutions at multiple frequencies simultaneously.Gribenko and Zhdanov (2007)presented an integral equation ap-proach to the forward solution.Furthermore,3D joint inversion of marine CSEM and magnetotelluric (MT)data has been used to take advantage of the complementary nature of the two data sets (Mackie et al.,2007).Common to these inversion techniques is that they use some type of conjugate gradient or quasi-Newton optimi-zation method which avoids forming the full Jacobian matrix,for example,by replacing it with the computation of sparse matrix-vector products (Rodi and Mackie,2001).Alternatively,one can take a Gauss –Newton approach,which involves the computation of the Jacobian matrix.Abubakar et al.(2011)proposed a Gauss –Newton approach in which a multiplicative cost functional is minimized by solving the normal equations using a conjugate gradient least-squares (CGLS)iterative scheme.The purpose of this paper is to demonstrate how 3D inversion of CSEM data can resolve thin resistive targets in shallow-water en-vironments and how the addition of CSEM data at different trans-mitter-receiver geometries and transmission frequencies,or the addition of MT data,can help to improve target resolution.The in-version algorithm used in this study is based on a Gauss –Newton approach and uses a staggered-grid FD method for forward model-ing.After outlining the inversion method,synthetic examples of in-dividual and joint 3D inversions of CSEM and MT data are presented for the models with seawater depths of 300and 100m.Manuscript received by the Editor 12March 2012;revised manuscript received 22July 2012;published online 13December 2012.1Kyushu University,Department of Earth Resources Engineering,Fukuoka,Japan.E-mail:sasaki@mine.kyushu-u.ac.jp.©2012Society of Exploration Geophysicists.All rights reserved.E59GEOPHYSICS,VOL.78,NO.1(JANUARY-FEBRUARY 2013);P.E59–E65,11FIGS.10.1190/GEO2012-0094.1D o w n l o a d e d 09/22/13 t o 221.226.175.186. R e d i s t r i b u t i o n s u b j e c t t o SE G l i c e n s e o r c o p y r i g h t ; s e e T e r m s o f U s e a t h t t p ://l i b r a r y .s e g .o r g /INVERSION METHODIn the inversion,the model region is divided into a set of rectan-gular blocks,and the resistivity is assumed to be constant in each block.Let m be the vector denoting the unknown model parameters (or the logarithms of the block resistivities),and let d be the vector of the data,which can be CSEM,MT data,or both types of data.The inverse problem is formulated as an optimization problem,where the objective functional to be minimized is given byϕðm Þ¼k W ½d −f ðm Þ k 2þλ2k Cm k 2:(1)Here,f ðm Þrepresents the CSEM and/or MT responses predicted bya 3D finite-difference method,C is a spatial second-order finite-difference operator to define the model roughness,and λis a reg-ularization parameter that trades off between the data misfit and model roughness.The matrix W is diagonal and represents the data weights assigned according to the standard deviation of the data noise.The CSEM data used in the inversion are the in-phase and quadrature components of the electric field for given frequen-cies and source-receiver offsets,normalized by the field amplitudes,whereas the MT data are the logarithms of apparent resistivities and the scaled phases in degree for the off-diagonal components of the complex impedance tensors (Z xy and Z yx ).A scaling factor of 2π∕180is used for the phases to account for the relative standard deviation (Sasaki and Meju,2006).The inverse problem is nonlinear and is iteratively solved with a Gauss –Newton minimization approach.The function f ðm Þis line-arized with respect to the initial model m ð0Þ,and the perturbation Δm is found such that the linearized objective functional is mini-mized.At the k th iteration,the linearized objective functional isϕðΔm Þ¼k W ðΔd −A Δm Þk 2þλ2k C ðm ðk ÞþΔm Þk 2;(2)where Δd is the difference between the observed and predicted data,and A is the Jacobian matrix with respect to the logarithm of theblock resistivity.The minimization of the objective functional is equivalent to obtaining the least-squares solution of the rectangular system of equationsWAλCf Δmg ¼ W Δd −λCm ðk Þ;(3)which can be solved employing the modified Gram –Schmidt meth-od (Björck,1996).Note that this least-squares solver does not in-volve forming the normal equation and its solution is more stable than the normal-equation approach.An advantage of the Gauss –Newton method is that the convergence is fast because the model correction size is relatively large at earlier iterations.The disadvan-tage is that,for large-scale problems,it requires more memory to solve than does the conjugate gradient-type methods.For CSEM and MT responses,the forward and sensitivity calcu-lations are based on a staggered-grid FD method formulated in terms of the secondary field.The system of equations is solved using the incomplete Cholesky biconjugate gradient (ICBCG)method,and the convergence rates are improved by incorporating the static divergence correction (Sasaki,2001).The Jacobian matrix can be obtained using the reciprocity principle (McGillivray et al.,1994);that is,the sensitivity of the EM response can be obtained by making additional forward calculations for fictitious sources placed at the receiver positions.For CSEM data,an electric dipole source is placed at each receiver location to compute the partial derivative of the electric-field response with respect to the block resistivity.For MT data,which include the orthogonal components of the electric and magnetic fields,one needs to consider electric and magnetic dipoles with two polarization directions at each receiver location.SYNTHETIC EXAMPLESThe model that is used to generate the synthetic CSEM and MT data is shown in Figure 1.A single thin reservoir with resistivity of 50ohm-m is embedded in a 1-ohm-m host sediment overlying a 10-ohm-m basement.The dimensions of the reservoir are 5km ×6km ×100m in the x -(easting),y -(northing),and z -(depth)direc-tions,and the top of the reservoir is 1km below the seafloor.The basement has an escarpment 400m high.Two 400-m-thick surface slabs with a resistivity of 4ohm-m represent hydrate accumulation zones.Thirty-nine seafloor receivers on three lines of 13each are deployed with spacings of 1km in the x -direction and 2km in the y -direction,as shown by the circles in Figure 1a .For the CSEM survey,a 200-m horizontal electric dipole is as-sumed to be towed 50m above the receivers along three lines.There are 16source locations spaced at 1-km intervals on each tow line (with a total of 48source locations).The frequencies are 0.1and 0.25Hz.At each receiver,MT data are also acquired at seven frequencies (0.01,0.025,0.05,0.1,0.25,0.5,and 1Hz).The FD grid used to generate the data is different from the one used in the inversion;it has 143×103×59cells,with a minimum cell size of 125×125×50m and a maximum size of 6.4×6.4×6.4km .Gaussian noise with a standard deviation of 6%of the elec-tric-field amplitude was added to the CSEM data.The MT impe-dance components were contaminated with 3%Gaussian noise,which translates to standard deviations of 6%for the apparent resistivity and 1.72°for the phase.Thus,the signal-to-noise ratio of MT data is equal that of the CSEM data,and the data weights in equation 1are set to unity for all data (CSEM and MTdata).Figure 1.(a)Plan view and (b)vertical cross section along the cen-tral tow line of 3D seabed model used to generate the CSEM and MT data.The depth is from the seafloor.The projections of the re-servoir and seafloor anomalous bodies are indicated by the solid and dashed rectangles in the plan view,respectively.E60SasakiD o w n l o a d e d 09/22/13 t o 221.226.175.186. R e d i s t r i b u t i o n s u b j e c t t o SE G l i c e n s e o r c o p y r i g h t ; s e e T e r m s o f U s e a t h t t p ://l i b r a r y .s e g .o r g /The number of real-valued data points is1248for inline CSEM data (at a frequency)and1092for MT data.In the inversion,the forward modeling used a grid of87×71×49 cells,and the model region was divided into4860(27×9×20) blocks.The lateral boundaries of the blocks do not coincide with those of the anomalous bodies in the true model.The regularization parameter was set to values of1.0,0.5,and0.3for the first three iterations,and0.2for the remaining iterations;these values were se-lected based on experiences of EM data inversions,and are not necessarily the optimum ones.The starting model was a2-ohm-m half-space(underlying the fixed seawater layer),unless otherwise specified.300-m water depthThe first inversion test assumes a water depthof300m.The water resistivity is0.3ohm-m.Afew of the CSEM responses at0.25Hz are shownin Figure2in the form of the profiles of theelectric-field amplitude at selected transmitter-receiver offsets along the central tow line.In thisfigure,the corresponding profiles over the samemodel,but without the reservoir,are also shown(by the dotted lines)to highlight the target re-sponse.It can be seen that the largest target re-sponse occurs at offset of5.5km;the maximumamplitude ratio is2.2around the center of the line.Figure3b shows vertical cross sections alongthe three tow lines of the resistivity model recon-structed from inversion of the0.25-Hz inlineCSEM data.The root-mean-square(rms)datamisfit is0.061at the sixth iteration(see Figure4),which is close to the prescribed noise level.A comparison of Figure3b with the true model(Figure3a)shows that the target reservoir is re-covered at almost its correct position.Moreover,there are indications of the surface resistive bodiesas well as the resistive basement.Figure3c showsthe images recovered using only the MT data afterfive iterations.The surface bodies and basementare better defined,but the reservoir is not resolvedat all.This is expected because MT data are suitedto defining regional layered structures,but are insensitive to thin resistive targets(Mackie et al.,2007).The result of simultaneously inverting the0.25-Hz CSEM and MT data is given in Figure3d.The joint inversion produces an image that is an im-provement on the image obtained using only the CSEM data. Note that the initial model used is different from the true back-ground medium.Although not shown here,further CSEM inversion tests using initial(half-space)models of0.8,3,and5ohm-m pro-duced almost the same models as that shown in Figure3b.This suggests that,for this case,the subseafloor structure can be robustly recovered if the initial model is close to the true background structure.The computation time is∼22h for CSEM data inversion and ∼4.7days for MT data inversion on a computer with a3.0-GHzFigure 2.Amplitudes of the electric fields at selected offsets,plotted as a function of the transmitter-receiver midpoint for themodel with water depth of300m.Figure4.The rms data misfit versus iteration number for the inver-sions of CSEM and MT data,and the joint inversion.The waterdepth is300m.Easting (km)–2–1Depth(km)Easting (km)–2–1Depth(km)Easting (km)–2–1Depth(km)ohm-ma)b)c)d)Easting (km)–2–1Depth(km)Easting (km)–2–1Depth(km)Easting (km)–2–1Depth(km)Easting (km)–2–1Depth(km)Easting (km)–2–1Depth(km)Easting (km)–2–1Depth(km)Easting (km)–2–1Depth(km)Easting (km)–2–1Depth(km)–6–4–20246–6–4–20246–6–4–20246–6–4–20246–6–4–20246–6–4–20246–6–4–20246–6–4–20246–6–4–20246–6–4–20246–6–4–20246–6–4–20246Easting (km)–2–1Depth(km)16842132Figure3.Resistivity images along three tow lines(Y¼−2,0,and2km),obtained from3D inversion of(b)inline CSEM data at0.25Hz,(c)MT data,and(d)CSEM and MTdata sets for the model shown in Figure1.(a)shows the true model.The water depth is300m.3D inversion of marine CSEM and MT data E61 Downloaded9/22/13to221.226.175.186.RedistributionsubjecttoSEGlicenseorcopyright;seeTermsofUseathttp://library.seg.org/Core 2processor and 4GB memory available.The MT data inver-sion takes much longer because the sensitivity calculation involves the two (orthogonal)polarization directions of the electric and mag-netic dipoles and many more (seven)frequencies.100-m water depthTo illustrate the effect of water depth on detecting the target reser-voir,inversions were performed for a model with a water depth of 100m.A few of the CSEM responses at 0.25Hz along the central line are shown in Figure 5.The maximum amplitude ratio of 1.9oc-curs at offset of 4.5km around the center of the pared to the amplitude responses for the 300-m water depth case (Figure 2),themost appreciable target signatures are limited to smaller offsets (3.5and 4.5km).Figure 6b shows the images reconstructed by inversion of the 0.25-Hz inline CSEM data.Although the data misfit converges to the noise level in about five iterations (Figure 7),the recovered resistive zone is much deeper and thicker than the actual reservoir.This reflects the fact that the airwave effect increasingly obscures the target response as the water depth decreases.The images ob-tained from MT data (Figure 6c )are almost identical to the MT inversion result for the water depth of 300m (Figure 3c ).The joint inversion result in Figure 6d is a significant improvement on CSEM data inversion alone (Figure 6b )in terms of resolving the reservoir and background structure.At first glance,the target zone seems to be better defined than the corresponding result for the water depth of 300m (Figure 3d ).However,the recovered resistive zone is still located about 200m deeper than the actual reservoir zone.The dependence of the CSEM inversion result on the initial mod-el was examined.Figure 8shows the inversion results obtained using the same set of initial models as before.The convergence of the data misfit is shown in Figure 9.Unlike in the 300-m seawater case,a dependence on the initial model is evident.Although the initial model of 3ohm-m produces almost the same images as those from the 2-ohm-m model (Figure 6b ),the inversions that were started with the half-space models of 0.8and 5ohm-m resolve the target much better.This result suggests that using an initial mod-el closer to the true background medium does not necessarily result in better reconstruction of the target.To explain this,recall that a characteristic of Gauss –Newton-based inversions is that the model correction size is fairly large at earlier iterations,compared to con-jugate gradient-based inversions.Comparison of 1D and 3D responsesIt might be surprising that a 1000-m-deep re-servoir can be detected from the synthetic data with 6%noise for the 100-m seawater depth be-cause it is generally accepted that 3D target responses are inherently smaller than the corre-sponding 1D responses,as shown by Constable and Weiss (2006)for a deepwater case.Thus,it is worthwhile to examine how the 3D target re-sponses differ from the 1D responses.Figure 10shows the 3D responses of a model that is the same as the one used in the inversion tests but without the near-seafloor heterogeneity and base-ment uplift.Also shown are the responses of the 1D (infinitely wide)reservoir and background layered (no reservoir)models.The transmitter is positioned on the central line (y ¼0)at x ¼−2000m and the electric-field amplitudes are plotted along the line,as a function of trans-mitter-receiver offset.Figure 10b shows the broadside amplitudes plotted along a different line,y ¼2000m ,parallel to the central line where the transmitter is positioned,as a function of transmitter-receiver offset in the x -direction.Interestingly,it can be seen from the normalized responses (Figure 10c and 10d )that for inline and broadside configurations,the 3D responses have slightly higher sensitivity than the1DFigure 5.Amplitudes of the electric fields at selected offsets,plotted as a function of the transmitter-receiver midpoint for the model with water depth of 100m.–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–1–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–10D e p t h(k m )D e p t h (k m )D e p t h(k m )D e p t h (k m )D e p t h(k m )D e p t h (k m )D e pt h (k m )D e p t h (k m)D e p t h (k m )D e p t h (k m )D e p t h (k m )D e p t h (k m )–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–108421ohm-ma)b)c)d)Figure 6.Resistivity images along three tow lines (Y ¼−2,0,and 2km),obtained from 3D inversion of (b)inline CSEM data at 0.25Hz,(c)MT data,and (d)CSEM and MT data sets for the model shown in Figure 1.(a)shows the true model.The water depth is 100m.E62SasakiD o w n l o a d e d 09/22/13 t o 221.226.175.186. R e d i s t r i b u t i o n s u b j e c t t o SE G l i c e n s e o r c o p y r i g h t ; s e e T e r m s o f U s e a t h t t p ://l i b r a r y .s e g .o r g /responses to the target for offsets less than6km.This result argues that3D inversion should be used to explore resistive reservoirs in shallow-water settings.Effect of including broadside dataThe importance of including broadside data is well-known for anisotropic inversion(e.g.,Morten et al.,2010;Newman et al., 2010).Inline data are mainly sensitive to vertical resistivity, whereas broadside data are far more sensitive to horizontal resistiv-ity.For the isotropic case,the usefulness of broadside data is not obvious.Key(2009)showed for1D inversion that including broad-side data offers no appreciable improvement over inversion of inline data alone.Constable and Weiss(2006)also showed that for 1000-m seawater depth,the inline(radial)fields are20times more sensitive to a thin resistive layer than the broadside(azimuthal) fields.In contrast,a comparison of the inlineand broadside responses in Figure10shows thatbroadside data can be considerably more sensi-tive to the reservoir than inline data,dependingon the offset.The fundamental difference isthat the previous studies considered broadsidedata that are acquired along a line perpendicularto the transmitter direction,whereas the broad-side data considered here are obtained for a lineparallel to the transmitter.Figure11a shows the image recovered by inver-sion of inline and broadside data.For broadsidedata,the added Gaussian noise is based on6%of the total-field amplitude to account for the vec-tor properties of the electric fields and thus,in theinversion,the data(the x-component)are normal-ized by the total-field amplitude.The total numberof data points is3744.The starting model is againa2ohm-m half-space.The rms data misfit is0.059after five iterations.It is clear that the reser-voir zone is recovered much better than in the in-version of inline data alone(Figure6b)or thejoint inversion of inline CSEM and MT data(Figure6d).Inversion of the x-and y-componentsof electric fields(not shown)produced almost thesame result as those shown in Figure11a.Effect of using multiple frequenciesThe tests described above used a single frequency(0.25Hz)of CSEM data.Acquiring multiple-frequency data is common in CSEM surveying,so it is important to see how the addition of data at a different frequency might improve the resolution of the reser-voir.Figure11b shows the image obtained from inversion of inline CSEM data at two frequencies,0.25and0.1Hz,with100-m water depth.As expected,using the additional lower frequency improves target resolution significantly.This is because the offset range over which target responses are appreciable expands as the frequency is lowered(e.g.,Sasaki and Meju,2009).It can be concluded that in-verting multiple-frequency CSEM data simultaneously is essential for solving the shallow-water problem.Frequency-differencing methods were proposed by Maao and Nguyen(2010)and Chen and Alumbaugh(2011)independentlyFigure7.The rms data misfit versus iteration number for the inver-sions of CSEM and MT data,and joint inversion.The water depth is100m.–6–4–20246Easting (km)–2–1–6–4–20246Easting (km)–2–1Depth(km)Depth(km)Depth(km)Depth(km)Depth(km)Depth(km)Depth(km)Depth(km)Depth(km)Depth(km)Depth(km)Depth(km)–6–4–20246Easting (km)–2–1a)b)c)d)–6–4–20246Easting (km)–2–1–6–4–20246Easting (km)–2–1–6–4–20246Easting (km)–2–1–6–4–20246Easting (km)–2–1–6–4–20246Easting (km)–2–1–6–4–20246Easting (km)–2–1–6–4–20246Easting (km)–2–1–6–4–20246Easting (km)–2–1–6–4–20246Easting (km)–2–18421ohm-mFigure8.Resistivity images along three tow lines(y¼−2,0,and2km),obtained from3D inversions of inline CSEM data at0.25Hz,using initial half-space models of(b)0.8,(c)3,and(d)5ohm-m.(a)shows the true model.The water depth is100m.Figure9.The rms data misfit versus iteration number for the inver-sions of inline CSEM data using three different initial half-spacemodels.The water depth is100m.3D inversion of marine CSEM and MT data E63 Downloaded9/22/13to221.226.175.186.RedistributionsubjecttoSEGlicenseorcopyright;seeTermsofUseathttp://library.seg.org/as a means to improve the detectability of target reservoirs at depth.Frequency-differenced data have higher sensitivity than the original data to a resistive target at depth,provided the sensitivities of in-dividual-frequency data have opposite signs,which is often the case in shallow water (Sasaki and Meju 2009).However,the benefits of such processing are not clear when the processed data are inverted together with the original data.In fact,an experiment in which fre-quency-differenced data were added did not show any significant improvements over inverting the original data at two frequencies.A possible reason is that,although frequency differencing increases the sensitivity to some parts of the model,it lowers the sensitivity to other parts,and it also amplifies data errors if they are independent,as assumed here.CONCLUSIONSA Gauss –Newton-based 3D inversion method was applied to marine CSEM and MT data for detecting and delineating resistive reservoirs in shallow ing a model with water depth of 300m,inversion of inline CSEM data at a sin-gle frequency (0.25Hz)recovered the reservoir at almost its correct position and the basement was well separated from the target zone.For this water depth,the target signals at this frequency are significant enough to resolve the thin resistive target at a burial depth of 1km for the given background structure and noise level (6%of the electric-field amplitude).For the 100-m water depth case,although a high-resistivity zone cor-responding to the reservoir was defined,it im-aged much deeper and much thicker than the actual reservoir,suggesting that the airwave com-ponent becomes a larger part of the total signal and obscures the target response.However,the tests show that inverting inline and broadside data or inline data at different frequencies to-gether can enhance target resolution consider-ably.It was also shown that,as expected,joint inversion of CSEM and MT data can improve the overall resolution of the reconstructed images.The inversion results presented herein were obtained using only a few thousand data points and about 5000model parameters.Although such a problem is smaller than many considered in previous works,the target and overall resolu-tion of the reconstructed images are comparable.This is not surprising because the resolution at-tainable with the low-frequency (diffusive)EM fields is mainly governed by the r-ger numbers of model parameters and data points do not necessarily lead to higher resolution and higher reliability in the inversion result.ACKNOWLEDGMENTSI would like to thank the associate editor,Mark Everett,and anonymous reviewers for their com-ments and suggestions that greatly improved the clarity of the paper.REFERENCESAbubakar,A.,M.Li,G.Pan,J.Liu,and T.M.Habashy,2011,Joint MT and CSEM data inversion using a multiplicative cost function approach:Geo-physics,76,no.3,F203–F214,doi:10.1190/1.3560898.Amundsen,L.,L.Løseth,R.Mittet,S.Ellingsrud,and B.Ursin,2006,De-composition of electromagnetic fields into upgoing and downgoing com-ponents:Geophysics,71,no.5,G211–G223,doi:10.1190/1.2245468.Andreis,D.,and L.MacGregor,2008,Controlled-source electromagnetic sounding in shallow water:Principles and applications:Geophysics,73,no.1,F21–F32,doi:10.1190/1.2815721.Björck,A.,1996,Numerical methods for least squares problem:SIAM.Chen,J.,and D.L.Alumbaugh,2011,Three methods for mitigating air-waves in shallow water marine controlled-source electromagnetic data:Geophysics,76,no.2,F89–F99,doi:10.1190/1.3536641.Figure 10.(a)Inline and (b)broadside CSEM amplitudes as a function of transmitter-receiver offset at a frequency of 0.25Hz,calculated for the model that is the same as the one in Figure 1but without the near-seafloor heterogeneity and basement uplift,and for the corresponding 1D model (with an infinitely wide reservoir).The inline and broad-side field amplitudes normalized by the response of the background layered model are shown in (c)and (d),respectively.The broadside fields are for the x -directed line se-parated by 2km in the y -direction from the transmitter.The water depth is 100m.–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–10a)b)8421ohm-m–6–4–20246Easting (km)–2–10D e p th (k m )D e p t h (k m )D e p t h (k m )D e p t h (k m )D e p t h (k m )D e p t h (k m )–6–4–20246Easting (km)–2–10–6–4–20246Easting (km)–2–10Figure 11.Resistivity images along three tow lines (y ¼−2,0,and 2km),obtained from 3D inversions of (a)inline and broadside CSEM data at 0.25Hz and (b)inline CSEM data at 0.25and 0.1Hz for the model with water depth of 100m.E64SasakiD o w n l o a d e d 09/22/13 t o 221.226.175.186. R e d i s t r i b u t i o n s u b j e c t t o SE G l i c e n s e o r c o p y r i g h t ; s e e T e r m s o f U s e a t h t t p ://l i b r a r y .s e g .o r g /。
mariner函数 -回复
mariner函数-回复什么是mariner函数？Mariner函数是一种用于R语言编程的函数，提供了一套方便和灵活的工具来处理和分析海洋科学数据。
该函数旨在帮助海洋科学家和研究人员更轻松地处理和分析海洋数据，从而提取有价值的信息和洞察力。
1. 引言海洋科学是一个复杂而庞大的领域，涉及到数量庞大的数据集。
要从这些数据中提取有用的信息并进行适当的商业决策，需要使用强大和高效的工具。
Mariner函数的目的正是这样的一个工具，提供了一整套用于处理和分析海洋科学数据的功能。
2. 安装和设置Mariner函数要使用Mariner函数，首先需要安装相应的软件包。
可以使用R语言中的install.packages()函数来安装Mariner包。
安装完成后，使用library()函数导入Mariner库：install.packages("Mariner")library(Mariner)3. 加载和读取数据Mariner函数支持各种格式的数据读取，包括文本文件、Excel文件、数据库和API等。
可以使用read.table()函数来读取文本文件，并使用read_excel()函数读取Excel文件。
在数据读取完成后，可以使用Mariner 函数中的各种功能来处理和分析数据。
4. 数据处理和清洗在处理和分析数据之前，通常需要对数据进行清洗和预处理。
Mariner函数提供了各种强大的工具来处理和清洗数据，如删除缺失值、处理重复数据、处理异常值等。
可以使用Mariner函数中的clean_data()函数来进行数据清洗，并使用filter()函数进行数据筛选和过滤。
5. 数据可视化数据可视化是理解和解释海洋科学数据的关键步骤。
Mariner函数提供了各种强大的数据可视化功能，例如绘制折线图、直方图、散点图等。
可以使用plot()函数来绘制各种类型的图形，并使用add_layers()函数添加更多的图层和特性。
地理必修三3s技术市公开课获奖课件省名师示范课获奖课件
GPS 定---位-- RS -图---象-- GIS------顾客
五.数字地球
台湾岛旳遥感影象： 能清楚旳看到全岛旳地貌。
乞力马扎罗旳积雪
美国国防部五角大楼
9.11事件前后旳纽约
中国大陆旳卫星地形图
上海市区上空旳卫星图片
４ 特点 探测范围大,速度快,投入少
遥感－－人旳视力旳延伸
二．全球定位系统（ＧＰＳ）
1 概念 利用卫星在全球范围 内导航,定位旳系统
目前已成功地应用到了涉及资源管理、自 动制图、设施管理、城市和区域旳规划、人口 和商业管理、交通运送、石油和天然气、教育、 军事等九大类别旳一百多种领域。在美国及发 达国家，地理信息系统旳应用遍及环境保护、 资源保护、灾害预测、投资评价、城市规划建 设、政府管理等众多领域。
网络数字地图
四 三者关系
在航空器上或航天器上利用一定旳技 术装备,对地表物体进行远距离旳感
知．
2 遥感技术过程
３．遥感应用：
资源调查、地质调查、城市遥感调查、海洋资源调查、 测绘、考古调查、环境监测、自然灾害防御监测和规 划管理等
案例： 长江流域洪涝灾害监测
飞机遥感图片：b图绿色表达淹没旳农田；黑 色表达河岸受淹滩地。（）
2 构成 空间部分,地面控制部分,顾客设备部分
二、全球定位系统 GPS
GPS构成
空间部分
地面监控部分
顾客设备 －－ＧＰＳ信号机
顾客设备
顾客设备
手持式GPS定位仪
3 GPS特点
全能性,全球性, 全天候,连续性 和实时性特点
GPS功能
定位 导航
我目前在哪里？ 我要去哪里？ 怎样去？
GPS应用
我国神舟号飞船
mariner函数 -回复
mariner函数-回复中括号内的主题是“mariner函数”。
以下是一篇1500-2000字的文章，对该主题进行逐步解答。
标题：深入探究mariner函数：一种用于航海导航的先进工具导语：在航海领域，准确的定位和导航对于船舶的航行至关重要。
随着科技的不断发展，人们创造了各种先进工具来协助航海员们进行航行。
其中，mariner函数在航海导航中广泛使用。
本文将深入探究mariner函数的定义、应用和工作原理，以及它在现代航海领域中的重要性。
第一部分：引言- 介绍航海导航的重要性和挑战- 引出mariner函数作为一种解决方案的背景第二部分：什么是mariner函数- 解释mariner函数的定义和概念- 阐述其核心原则和目标第三部分：mariner函数的应用- 讨论mariner函数在航海领域的常见应用场景- 说明它在船舶航线规划、定位和导航等方面的作用第四部分：mariner函数的工作原理- 详细解释mariner函数的工作原理- 探讨其基于什么样的数学和物理模型第五部分：现代航海中的mariner函数重要性- 分析现代航海技术发展对mariner函数的依赖性- 讨论它在提高航海安全性和效率方面的优势第六部分：结论- 总结mariner函数的重要性和应用价值- 展望未来其在航海导航领域中的发展潜力第一部分：引言航海导航是航海领域的关键技术之一，它为船舶航行提供了定位和导航信息。
然而，航海导航面临许多挑战，如能见度不佳、海洋环境复杂、航线规划困难等。
为了克服这些挑战，航海员们利用先进的工具和技术来提高航行的安全性和效率。
而mariner函数作为航海导航中的一种先进工具，正在为航海领域做出重要贡献。
第二部分：什么是mariner函数mariner函数是一种航海导航工具，用于计算船舶的位置和导航信息。
它基于一系列数学模型和物理原理，利用全球定位系统（GPS）等技术，提供高精度的航行数据。
其核心原则是通过测量船舶的位置、速度、方向和时间等参数，为航海员提供精确的定位和导航信息，以确保船舶航行安全。
海航技术 structural design basics -回复
海航技术structural design basics -回复海航技术是一家专注于航空结构设计的公司。
航空结构设计是航空工程领域中至关重要的一环，它涉及到飞机的各个部件及其配件之间的力学关系、材料力学性质和制造工艺等方面的知识。
在这篇文章中，我们将一步一步回答有关海航技术的结构设计基础知识。
第一部分：结构设计的背景和概述1. 结构设计的定义及重要性结构设计是指根据给定的载荷、材料和约束条件等，通过合适的结构形式和尺寸设计出能够满足特定功能和性能要求的结构系统。
航空工程中的结构设计是确保飞机具有优良的飞行性能和寿命的关键环节。
2. 单元结构和总体结构设计在结构设计的过程中，首先要进行单元结构设计。
单元结构是指飞机的各个独立结构部分，例如机翼、机身等。
然后，将这些单元结构组合起来，形成总体结构。
总体结构设计考虑到整个飞机的受力情况和整体的平衡性。
3. 结构设计的基本原理结构设计的基本原理包括受力平衡、材料力学和结构稳定性等方面。
受力平衡是指在静态平衡和动态平衡条件下，结构各部分内外力的平衡。
材料力学是指考虑材料的力学性质，如强度、刚度、疲劳寿命等。
结构稳定性是指结构在受外部力作用下保持稳定的能力。
第二部分：海航技术的结构设计流程1. 需求分析和功能要求首先，海航技术的设计师会与客户充分沟通，了解飞机的使用需求和功能要求。
这些要求包括飞行性能、载荷要求、安全要求等。
在这个阶段，设计师还会对类似的飞机进行调研和分析，确保设计符合行业标准和最佳实践。
2. 概念设计在概念设计阶段，设计师会创建多种可能的结构设计方案。
这些方案考虑到载荷分布、几何形状和材料选择等因素，并进行初步性能分析和评估。
设计师还会利用计算机辅助设计（CAD）软件进行设计绘图。
3. 详细设计和分析在详细设计和分析阶段，设计师会选择最有前景的设计方案，并进行更详细的设计和分析。
这些分析包括应力分析、振动分析、疲劳寿命分析等。
设计师还会利用计算机辅助工程（CAE）软件对结构进行建模和分析。
- 1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。
– Extract specific area of interest by defining the outlet and inlets
Watershed Delineation
• Flexible addition and removal of basin outlets
– Merge existing basins – Split basin anywhere on the stream – Add an outlet anywhere (trace the outlet stream to an existing stream) – Profile
• Interactive or batch processing • Understanding of basin character
GeoHMS Functionality (cont.)
• HEC-GeoHMS and Arc Hydro tools are tightly linked – GeoHMS computes as many Arc Hydro attributes as possible (e.g. NextDownID, JunctionID, DrainID) – DEM preprocessing (main view) is done using Arc Hydro tools – Topographic characteristics extraction is based on Arc Hydro tools – Arc Hydro tools operate on both main and project views • Option to change stream definition threshold when extracting the project view (one threshold for the main view and different thresholds for different project views)
• Rule - The point should be located within the grid cell that has an existing stream. • Import Batch Points – Places all selected points in the map into the batch point file. Id source feature classes have “Name” and “Description” attributes, they area assigned to the batch points • Delineate Batch Points – Takes points from batch point feature class and uses them to do basin subdivision. • BatchDone and SnapOn attributes to control snapping
GeoHMS Functionality
• HEC-GeoHMS – DEM preprocessing – “parent” definition (main view) – Watershed delineation – on “baby” models (project view) – Topographic characteristics extraction – Hydrologic parameter computations – Model schematization – Model input preparation (ASCII – XML, SDF) • Other GIS processing – Rainfall distribution/interpolation – LU/soils runoff coefficient mapping
Watershed Delineation (cont.)
Topographic Characteristic Extraction
• Well defined topographic descriptors
– Longest flowpath, basin centroid, stream slope, centroidal length, CN, etc. (others planned)
Simplified HMS Watershed Runoff Representation
HMS Components
• Basin model – Watershed physical description • Meteorologic model – Precipitation – Evapotranspiration – Snowmelt • Control specifications – Time control during simulation
Basin Subdivision - Method 1
• Method 1 - An existing basin can be subdivided into two basins at an existing stream.
River Profile
• Subdivide at grade break
Selected HEC-GeoHMS Details
HMS Project Setup
• Define outlet and sources • Extract applicable GIS data for ProjView
GeoHMS: Project View
Basin and Reach Manipulation
Model Schematization
• Upstream/downstream connectivity • Development of lumped-link schema
Model Input Preparation
• Native HMS ASCII input files (hms, basin, geometry, met, dss, gage, control)
Basin Model
Transform editor
• Many input forms
Baseflow method editor
Loss rate editor
Mathematical Model Choices
• Many characteristics needed to compute model parameters are spatially based
Basin Merge
• Rules
– The subbasins must share a confluence OR – The subbasins must be adjacent in an upstream and downstream manner – More than 2 subbasins are permitted.
– click on the desired point on the profile, and that point is transferred onto the map and the basin is subdivided at that point
Batch Subbasin Delineation
DEM Preprocessing (parent definition)
• • • • • • Hydrologically correct DEM Flow direction (D8 method) Flow accumulation Stream definition and segmentation Watershed pre-delineation Project (“baby”) data extraction
Hydrologic (Model) Parameter Extraction
• Loss rate method (SCS CN) – Basin CN • Direct runoff method (SCS UH) – Basin lag (TR55, SCS CN) • Reach routing – Muskingum-Cunge standard channel parameters • Precipitation – Weighted gage (standard SCS type II, 24 hour, 100 year design distribution) – Nexrad/DSS
Hydrologic Elements
• Subbasin: watershed catchments • Reach: rivers and streams • Reservoir: dams and lakes • Junction: confluence • Diversion: bifurcations and withdrawals • Source: springs and other model sinks • Sink: outlets and terminal lakes
DEM Preprocessing (cont.)
• Stream definition and segmentation
– Threshold tion
– At stream confluences
DEM Preprocessing (cont.)
HMS Control Specifications Model
• Time window • Time related specifications