基于Landsat-8遥感影像和LiDAR测深数据的水深主被动遥

主被动遥感结合反演远海岛礁周边水深信息,不仅可以有效弥补传统测深方法覆盖范围小且费时费力的不足,也可为航运安全、海洋减灾、生态环境保护等领域提供基础资料。以夏威夷瓦胡岛周边水深反演为例,应用Landsat-8多光谱遥感数据和机载Li DAR测深数据,开展了不同密度Li DAR测深数据对水深多光谱遥感反演精度的影响分析、不同水深网格化处理方法对水深遥感反演结果的影响分析和基于少量Li DAR控制区块的大区域水深反演能力分析三方面的研究工作。结果表明:(1)Li DAR测深数据密度的改变对水深反演结果的影响不大,变化后的水深反演结果与原始的水深反演结果相比,平均相对误差变化在0.3%以内,平均绝对误差变化在0.03m以内;(2)采用均值格网处理方法的多光谱遥感水深反演精度要略高于采用中值格网处理方法的水深反演精度,具体体现在均值的平均绝对误差要比中值的低0.04~0.05 m,平均相对误差低1%~10%,反演结果的残差分布显示在0~2 m和20~25 m的水深段内均值统计法的残差分布更集中且其平均值接近于0 m,而在其它水深段二者的残差分布基本相同;(3)基于少量Li DAR控制区块的大区域遥感水深反演结果较为理想,两个检查区块的水深反演结果 R2、平均绝对误差和平均相对误差分别为:0.877,1.66 m,3.5%和0.941,1.62 m,28.4%。反演结果分段分析表明各水深段内反演的精度都比较理想,平均绝对误差除20~25 m水深段外,均低于2.5 m,平均相对误差除0~2 m,2~5 m外,均低于25%。     

基于可见光遥感测深技术的长江口南支河段河势演变规律研究

以长江口南支河段为研究区,利用LandsatETM+遥感影像反演的1999年水下地形和该河段2002年实测水深数据,结合长江口来水来沙现状和河段特征,分析了河段的冲淤变化和沙体迁移规律,提出了需采取的治理措施。研究结果表明:(1)从1999到2002年南支河段总体发生冲刷,年平均冲刷量为0.1926亿m3,河道内白茆沙、新浏河沙和新浏河沙包的沙头年平均后退467,374和421m;(2)上、下扁担沙之间由于串沟发育,两者有分离的趋势,沙体5m等深线的变动格局显示两沙体沙尾向下移动导致新桥水道上段水深加深、下段南侧水深变浅;(3)固定白茆沙和上、下扁担沙沙体,对维持南支河段航道的航运、稳定南北槽河势具有重要作用。     

Swath mapping of the seafloor and its application to deep-bottom fisheries in New Caledonia

A bottom longline fishery operated in the EEZ of New Caledonia from 1988 to 1991. Fishing focused on five seamounts with summits at depths ranging from 500 to 750 m. The target species was alfonsino, Beryx splendens. As the soundings available from marine charts were not detailed enough, the fishing masters had to make their own charts in order to set the gear in the right location. A series of 11 scientific cruises devoted to a research program on alfonsino was launched in late 1991. During the first scientific cruises, several days were spent mapping the seabed to improve the existing knowledge of the topography of three seamounts by coupling the EDO echosounder depth measurements to the GPS positions. As this procedure is slow, it is applicable to limited areas or, if a wider grid is used, to large zones to locate major structures such as ridges and seamounts. The emergence of the multibeam echosounder has greatly improved seabed mapping performance. This tool seems to fit the requirement of exploratory deep-bottom fishing particularly well, as it covers large areas while providing details of the bottom. The EM 12 was used to obtain bathymetry and imagery of the SE portion of the EEZ of New Caledonia, covering an approximate area of 70,000 km² in two weeks. For depths ranging from 500 to 4,500 m, the results were impressive. They confirmed the known major features and provided greater topographical detail, revealing fine unknown structures. They also provided information on the type of substratum, information which might influence the fishing strategy. Finally, they made it possible to obtain an accurate estimate of the exploitable area which in turn led to new stock estimates.     

Remote sensing of water depths in shallow waters via artificial neural networks

Determination of the water depths in coastal zones is a common requirement for the majority of coastal engineering and coastal science applications. However, production of high quality bathymetric maps requires expensive field survey, high technology equipment and expert personnel. Remotely sensed images can be conveniently used to reduce the cost and labor needed for bathymetric measurements and to overcome the difficulties in spatial and temporal depth provision. An Artificial Neural Network (ANN) methodology is introduced in this study to derive bathymetric maps in shallow waters via remote sensing images and sample depth measurements. This methodology provides fast and practical solution for depth estimation in shallow waters, coupling temporal and spatial capabilities of remote sensing imagery with modeling flexibility of ANN. Its main advantage in practice is that it enables to directly use image reflectance values in depth estimations, without refining depth-caused scatterings from other environmental factors (e.g. bottom material and vegetation). Its function-free structure allows evaluating nonlinear relationships between multi-band images and in-situ depth measurements, therefore leads more reliable depth estimations than classical regressive approaches. The west coast of the Foca, Izmir/Turkey was used as a test bed. Aster first three band images and Quickbird pan-sharpened images were used to derive ANN based bathymetric maps of this study area. In-situ depth measurements were supplied from the General Command of Mapping, Turkey (HGK). Two models were set, one for Aster and one for Quickbird image inputs. Bathymetric maps relying solely on in-situ depth measurements were used to evaluate resultant derived bathymetric maps. The efficiency of the methodology was discussed at the end of the paper. It is concluded that the proposed methodology could decrease spatial and repetitive depth measurement requirements in bathymetric mapping especially for preliminary engineering application.     

随机森林模型在遥感水深反演中的应用

随着我国浅海测绘需求的日益增长,文中利用四波段的WorldView-2高分辨率遥感影像,选取我国南海西沙群岛中的甘泉岛和台湾南湾地区作为典型试验区,开展水深反演研究。引入随机森林算法构建了随机森林水深反演模型,并同常用的3种水深反演模型进行精度对比。结果表明,在甘泉岛和南湾地区随机森林模型反演的水深值和真实水深值的RMSE分别为0.85 m和1.59 m,MRE分别为8%和12%,均优于其他3种模型。     

舰载浅水激光测深系统研究

本文提出了近海浅水区域水深测量的激光双频相位法测量方案，采用声光调制1064nm和532nm光波分别测量水面和水底距离。在实验室进行了该系统的模拟实验，证实了该方案的可行性。     

Multibeam Bathymetry of the Håkon Mosby Mud Volcano

The Håkon Mosby Mud Volcano is a natural laboratory to study geological, geochemical, and ecological processes related to deep-water mud volcanism. High resolution bathymetry of the Håkon Mosby Mud Volcano was recorded during RV Polarstern expedition ARK XIX/3 utilizing the multibeam system Hydrosweep DS-2. Dense spacing of the survey lines and slow ship speed (5 knots) provided necessary point density to generate a regular 10 m grid. Generalization was applied to preserve and represent morphological structures appropriately. Contour lines were derived showing detailed topography at the centre of the Håkon Mosby Mud Volcano and generalized contours in the vicinity. We provide a brief introduction to the Håkon Mosby Mud Volcano area and describe in detail data recording and processing methods, as well as the morphology of the area. Accuracy assessment was made to evaluate the reliability of a 10 m resolution terrain model. Multibeam sidescan data were recorded along with depth measurements and show reflectivity variations from light grey values at the centre of the Håkon Mosby Mud Volcano to dark grey values (less reflective) at the surrounding moat.     

Using inverse theory and seabeam bathymetry to improve seismic refraction data corrections

An array consisting of ocean bottom seismometer and on-bottom hydrophones, was used to conduct a seismic experiment on 0.4 Ma crust east of the Juan de Fuca Ridge. Seismic sources were large (>50 kg) explosive charges detonated by SUS devices set to explode at 1829 or 2438 m nominal depth. The objectives of the experiment were to determine the compressional wave velocity and attenuation structures of the uppermost 500 m depth. The relative positions of shots and receivers were originally determined by treating each shot-receiver pair independently, via raytracing of various water waves. Due to the reflection of some of these water waves by the rough bottom, significant scatter resulted, preventing a determination of a physically realizable velocity-depth function. A new method is described that co-locates shot and receiver positions, including receiver depths consistent withseabeam bathymetry, using only the water waves that do not interact with the bottom. Several potential pitfalls are outlined using this method. A stable solution could only be achieved by discarding shots located well outside the array. The water path corrections were applied to the refracted arrivals, again using theseabeam bathymetry. The joint inversion location procedure, along with the use of precise gridded bathymetry, reduced the travel time scatter to a level whereby a velocity-depth function could be determined. The results, using only the hydrophone data, indicate an initial velocity at the seafloor of 2.7 km s^-1 with gradients from 4.6 s^-1 slowly decreasing to 4.1 s^-1 at 679 m depth. This velocity is similar to others conducted over very young oceanic crust, and can be interpreted as being due to a high porosity at the surface, due to cracks, fissures, and open pores, which rapidly diminish with depth.     

Segmentation and morphotectonic variations along a slow-spreading center: The Mid-Atlantic Ridge (24°00′ N– 30°40′ N)

Analysis of Sea Beam bathymetry along the Mid-Atlantic Ridge between 24°00 N and 30°40 N reveals the nature and scale of the segmentation of this slow-spreading center. Except for the Atlantis Transform, there are no transform offsets along this 800-km-long portion of the plate boundary. Instead, the Mid-Atlantic Ridge is offset at intervals of 10–100 km by nontransform discontinuities, usually located at local depth maxima along the rift valley. At these discontinuities, the horizontal shear between offset ridge segments is not accommodated by a narrow, sustained transform-zone. Non-transform discontinuities along the MAR can be classified according to their morphology, which is partly controlled by the distance between the offset neovolcanic zones, and their spatial and temporal stability. Some of the non-transform discontinuities are associated with off-axis basins which integrate spatially to form discordant zones on the flanks of the spreading center. These basins may be the fossil equivalents of the terminal lows which flank the neovolcanic zone at the ends of each segment. The off-axis traces, which do not lie along small circles about the pole of opening of the two plates, reflect the migration of the discontinuities along the spreading center.The spectrum of rift valley morphologies ranges from a narrow, deep, hourglass-shaped valley to a wide valley bounded by low-relief rift mountains. A simple classification of segment morphology involves two types of segments. Long and narrow segments are found preferentially on top of the long-wavelength, along-axis bathymetric high between the Kane and Atlantis Transforms. These segments are associated with circular mantle Bouguer anomalies which are consistent with focused mantle upwelling beneath the segment mid-points. Wide, U-shaped segments in cross-section are preferentially found in the deep part of the long-wavelength, along-axis depth profile. These segments do not appear to be associated with circular mantle Bouguer anomalies, indicating perhaps a more complex pattern of mantle upwelling and/or crustal structure. Thus, the long-recognized bimodal distribution of segment morphology may be associated with different patterns of mantle upwelling and/or crustal structure. We propose that the range of observed, first-order variations in segment morphology reflects differences in the flow pattern, volume and temporal continuity of magmatic upwelling at the segment scale. However, despite large first-order differences, all segments display similar intra-segment, morphotectonic variations. We postulate that the intra-segment variability represents differences in the relative importance of volcanism and tectonism along strike away from a zone of enhanced magma upwelling within each segment. The contribution of volcanism to the morphology will be more important near the shallowest portion of the rift valley within each segment, beneath which we postulate that upwelling of magma is enhanced, than beneath the ends of the segment. Conversely, the contribution of tectonic extension to the morphology will become more important toward the spreading center discontinuities. Variations in magmatic budget along the strike of a segment will result in along-axis variations in crustal structure. Segment mid-points may coincide with regions of highest melt production and thick crust, and non-transform discontinuities with regions of lowest melt production and thin crust. This hypothesis is consistent with available seismic and gravity data.The rift valley of the Mid-Atlantic Ridge is in general an asymmetric feature. Near segment mid-points, the rift valley is usually symmetric but, away from the segment mid-points, one side of the rift valley often consists of a steep, faulted slope while the other side forms a more gradual ramp. These observations suggest that half-grabens, rather than full-grabens, are the fundamental building blocks of the rift valley. They also indicate that the pattern of faulting varies along strike at the segment scale, and may be a consequence of the three-dimensional, thermo-mechanical structure of segments associated with enhanced mantle upwelling beneath their mid-points.     

Comparison of volcanic rifts on La Palma and El Hierro, Canary Islands and the Island of Hawaii

The meso-scale (km) morphology of the well-studied volcanic rift zones on the Island of Hawaii is compared to the morphology of the lesser known rift zones of La Palma and El Hierro, Canary Islands. We find that there are both differences and similarities in their morphologic characteristics. In general, the rift zones on La Palma and El Hierro are shorter (a few tens of km in length) than those on Hawaii (ranging up to >100 km in length), perhaps reflecting both magma supply and composition. Many of the rift zones on Hawaii have well defined axial zones, both on-and offshore. In contrast, the rift zones on La Palma and El Hierro display various geometries ranging from linear ridges having smooth to irregular crests to structures with a broad fan-like morphology in plan view. The pronounced fanning may be a reflection of: 1) the stress field within the rift being insufficient to trap dikes within a narrow region, 2) dike injection and volcanism shifting laterally through time, 3) volcanoes building nearly one atop of another in the Canary Islands, superimposing the stress field of one structure on the other and thus yielding a more complex distribution of gravitational stresses, and 4) low rate of magma supply producing low magma pressures and thus randomly oriented dike injections. Irregularities and curvature along the axes of the rifts on La Palma and El Hierro may be a reflection of differences in the rate of magma production. Unlike the volcanoes on the Island of Hawaii there may be insufficient volumes of lavas erupted on La Palma and El Hierro to smooth out irregularities. The superposition of rifts from different volcanoes may also add to topographic irregularities in the Canary Islands, especially if eruption rates are low.