基于智能手机传感器的室内行人三维定位算法

针对室内行人定位中累积误差发散及高程值缺失的问题,提出了一种基于智能手机传感器(磁力计、加速度计、陀螺仪和气压计)的室内行人三维定位算法。该算法在使用行人航位推算获得二维定位信息的基础上,融合多传感器输出解算行人方位角,结合电子罗盘提出加权方位角以提高解算精度;通过传感器输出检测行人行为状态,结合虚拟地标点图层及其匹配算法,将行人的位置匹配到相应的特殊位置,重置行人初始位置以消除累积误差;使用基于气压计的差分气压测高法求解行人的高程值,从而获得行人三维定位信息。在智能手机上进行实验,结果表明,该算法不仅能够提供准确、连续的二维定位信息,还能提高行人对楼层的分辨力,可为室内行人导航、跟踪等基于位置服务提供有效的定位信息。     

Consistency of Geoid Models,Radar Altimetry,and Hydrodynamic Modelling in the North Sea

Radar altimetry, when corrected for tides, atmospheric forcing of the sea surface, and the effects of density variations and mean and time-variable currents, provides an along-track realization of the marine geoid. In this study we investigate whether and how such an ‘altimetric-hydrodynamic’ geoid over the North Sea can serve for validating satellite-gravimetric geoids. Our results indicate that, using ERS-2 and ENVISAT along-track altimetry and water levels from the high-resolution operational circulation model BSHcmod, we do find distinct differences in RMS fits for various state-of-the art satellite-only models (beyond degree 145 for GRACE-only, and beyond degree 185 for GOCE models) and for combined geoid models, very similar as seen in GPS-levelling validations over land areas. We find that, at spectral resolution of up to about 200, an RMS fit as low as about 7 cm can be obtained for the most recent GOCE-derived models such as GOCO05S. This is slightly above what we expect from budgeting individual errors. Key to the validation is a proper treatment of the spectral mismatch between satellite-gravimetric and altimetric-hydrodynamic geoids. Comparison of data fits and error budget suggests that geoid truncation errors residual to EGM2008 (i.e. EGM2008 commission and omission error) may amount up to few cm.     

基于卫星测高资料研究全球海平面变化趋势

利用全球平均海平面的高程数据,研究基于反距离权重内插方法来处理Jason测高数据资料得出全球平均海面的数学模型,并将所得结果与利用WHU2013 MSS模型得出的结果进行了对比。此外,还通过计算全球海平面的变化值,分析了全球海平面变化的季节性特征、长期性趋势以及年际变化与厄尔尼诺事件的相关性。     

Jason-2卫星阈值重跟踪算法比较

选取Jason-2卫星在中国南海北纬15°~18.5°范围1 276个波形数据,依据Threshold算法和SeaIce算法对波形进行重跟踪,得到不同阈值水平下的距离改正结果。结果表明:当星下点轨迹靠近陆地时,距离改正值增大。然后,依据不同方法、不同阈值水平对应的距离改正值对卫星观测的海面高实施修正,得到相应的海面高结果。作为验证,选取法国国家太空研究中心(CNES)发布的海面高作为外部检核条件。将修正后的海面高与CNES公布的海面高进行较差比较,较差值的统计分析表明:SeaIce重跟踪算法在阈值水平为50%时,差值的标准差明显小于其他方法的计算结果。     

A GPS and Cold Ocean Brightness Temperature Calibration of the ERS-2 and TOPEX/Poseidon Microwave Radiometers

Sea-level change studies from altimetric satellites are reliant on range stability of the sea surface heights computed from orbital positioning and geophysically corrected data. One such correction, namely the wet tropospheric delay induced by the highly variable atmospheric water vapor content, is provided by radiometers onboard ERS-2 and TOPEX/Poseidon (T/P). In this study the long-term stability of the ERS-2 microwave radiometer (E2MR) and the T/P microwave radiometer (TMR) are investigated with the observed drift in the brightness temperatures approximated by reference to the coldest temperatures over the oceans. The E2MR stability is characterized by a gain anomaly fall in 1996 and a drift in the 23.8 GHz channel. For the TMR, investigations show that the dominant drift is about 0.2 K/year in the 18 GHz channel over the first 7-8 years but stabilizing and even decreasing slightly thereafter. In contrast, the 21 GHz and 37 GHz channels are comparatively stable. Utilizing correction formulae a modified wet tropospheric range is inferred from “small-change” analysis of the radiometric correction given on the altimetric Geophysical Data Records. The accuracy of this formulism is validated by independent comparison against GPS derived wet tropospheric delays inferred at 14 coastal IGS stations with near continuous data from September 1992 through to the present day. Comparisons between GPS results for ERS-2 and T/P show that the E2MR path delay is 14 mm short. For T/P, the spatial distribution of the wet tropospheric enhancement is further investigated to show that the nonuniformity can equate to a deviation in sea-level height change of about 0.1 mm/year compared with global average sea-level change. Finally, the altimetric range stability of T/P is revisited by comparison against time series from the global network of tide gauges. Analysis shows that the validated TMR drift correction results in a residual trend of -0.27 ± 0.11 mm/yr which is not significant at the 3σ level.     

Comparison of Sea Surface Heights Derived from Satellite Altimetry and from Ocean Bottom Pressure Gauges: The SW Pacific MOTEVAS Project

A bottom pressure gauge (BPG) was installed in proximity (3.7 km at closest approach) of Jason-1 and formerly TOPEX/Poseidon (T/P) ground track No. 238 at the Wusi site, located ∼ 10 km offshore off the west coast of Santo Island, Vanuatu, Southwest (SW) Pacific. Sea level variations are inferred from the bottom pressure, seawater temperature, and salinity, corrected for the measured surface atmospheric pressure. The expansion of the water column (steric increase in sea surface height, SSH) due to temperature and salinity changes is approximated by the equation of state. We compare time series of SSH derived from T/P Side B altimeter Geophysical Data Records (GDR) and Jason-1 Interim Geophysical Data Records (IGDR), with the gauge-inferred sea level variations. Since altimeter SSH is a geocentric measurement, whereas the gauge-inferred observation is a relative sea level measurement, SSH comparison is conducted with the means of both series removed in this study. In addition, high-rate (1-Hz) bottom pressure implied wave heights (H_1/3) are compared with the significant wave height (SWH) measured by Jason-1. Noticeable discrepancy is found in this comparison for high waves, however the differences do not contribute significantly to the difference in sea level variations observed between the altimeter and the pressure gauge. In situ atmospheric pressure measurements are also used to verify the inverse barometer (IB) and the dry troposphere corrections (DTC) used in the Jason IGDR. We observe a bias between the IGDR corrections and those derived from the local sensors. Standard deviations of the sea level differences between T/P and BPG is 52 mm and is 48 mm between Jason and BPG, indicating that both altimeters have similar performance at the Wusi site and that it is feasible to conduct long-term monitoring of altimetry at such a site.     

太平洋卫星测高重力场与地球动力学特征

通过多卫星测高数据的综合处理，获得西太平洋卫星测高重力场，进行不同尺度、深度构造动力信息的分离，探讨诸边缘海盆的地球动力学问题。测高大地水准面反映了研究区板块相互作用的特点，其高频成分可以刻画各海盆的构造特征。测高空间重力异常也可刻画陆架构造及盆地分布，由其推算出的海底地形含有大量的海底构造信息。各边缘海盆的莫霍面埋深具有往南变浅的趋势，与菲律宾海各海盆的莫霍面埋深大致相当，说明岛弧两侧的构造动力强度基本相似。大尺度地幔流应力场总体上反映了欧亚板块向东南蠕散和太平洋板块向北西扩张的特点；日本海北侧和南海巽他陆架的中尺度上地幔对流与地幔柱之间有着密切关系，西菲律宾海的上地幔对流强化了日本-琉球-台湾-菲律宾岛弧的活动强度；小尺度地幔流主要限于软流圈层内部，在各海盆分散，而在冲绳海槽和马里亚纳海槽则会聚，可与均衡重力异常类比。还讨论了大、中、小地幔流体系的特点及相互之间的关系，籍以阐明海盆及海槽演化的地球动力学过程。     

Statistical Evaluation of the Jason-1 Operational Sensor Data Record

The Jason-1 satellite altimeter mission represents a first step towards operational oceanography from satellite altimeter missions. An operational data product, the Operational Sensor Data Record (OSDR), provides measurements from the on-board altimeter and radiometer within 3-5 h of real time. This data product is a wind and wave product that is aimed towards near-real-time meteorological applications. A higher accuracy and more detailed data product, the Interim Geophysical Data Record (IGDR), that is better suited to detailed scientific studies of ocean topography, is available no sooner than 2-3 days from real time. The measurements reported on the OSDR primarily differ from those on the IGDR in that the OSDR reports measurements derived from on-board processing of the altimeter waveforms, while ground retracking of the waveforms is performed for the IGDR. The altimeter-derived measurements on the OSDR are validated through a statistical evaluation of the differences between data on the OSDR and IGDR. In doing so, the impact of ground retracking of the altimeter waveforms is also illustrated.     

Depth to basement and geoid expression of the Kane Fracture Zone: A comparison

Geoid data from Geosat and subsatellite basement depth profiles of the Kane Fracture Zone in the central North Atlantic were used to examine the correlation between the short-wavelength geoid (=25–100 km) and the uncompensated basement topography. The processing technique we apply allows the stacking of geoid profiles, although each repeat cycle has an unknown long-wavelength bias. We first formed the derivative of individual profiles, stacked up to 22 repeat cycles, and then integrated the average-slope profile to reconstruct the geoid height. The stacked, filtered geoid profiles have a noise level of about 7 mm in geoid height. The subsatellite basement topography was obtained from a recent compilation of structure contours on basement along the entire length of the Kane Fracture Zone. The ratio of geoid height to topography over the Kane Fracture Zone valley decreases from about 20–25 cm km^-1 over young ocean crust to 5–0 cm km^-1 over ocean crust older than 140 Ma. Both geoid and basement depth of profiles were projected perpendicular to the Kane Fracture Zone, resampled at equal intervals and then cross correlated. The cross correlation shows that the short-wavelength geoid height is well correlated with the basement topography. For 33 of the 37 examined pro-files, the horizontal mismatches are 10 km or less with an average mismatch of about 5 km. This correlation is quite good considering that the average width of the Kane Fracture Zone valley at median depth is 10–15 km. The remaining four profiles either cross the transverse ridge just east of the active Kane transform zone or overlie old crust of the M-anomaly sequence. The mismatch over the transverse ridge probably is related to a crustal density anomaly. The relatively poor correlation of geoid and basement depth in profiles of ocean crust older than 130–140 Ma reflects poor basement-depth control along subsatellite tracks.     

Impact of Ocean Mean Dynamic Topography on Satellite Data Assimilation

The response of an eddy-permitting ocean model to changes imposed by the use of different mean dynamic topographies (MDT) is analyzed in a multivariate assimilation context, allowing the evaluation of this impact, not only on the surface circulation, but also on the interior ocean representation. The assimilation scheme is a reduced-order sequential Kalman filter (SEEK). In a first set of experiments, high resolution sea surface temperature, along-track sea surface height and sea surface salinity from climatology are assimilated into a 1/3° resolution North and Tropical Atlantic version of the HYCOM model. In a second experiment, in situ profile data are assimilated in addition to the surface measurements.

The first set of experiments illustrates that important differences in the representation of the horizontal model circulation pattern are related to differences in the MDT used. The objective of assimilation is to improve the representation of the 3D ocean state. However, the imperfect representation of the mean dynamic topography appears to be an important limiting factor with regard to the degree of realism obtained in the simulated flow.

Vertical temperature and salinity profiles are key observations to drive a general circulation ocean model toward a more realistic state. The second set of experiments shows that assimilating them in addition to sea surface measurements is a far from trivial exercise. A specific difficulty is due to inconsistencies between the dynamic topography diagnosed from in situ observations and that diagnosed from sea surface height. These two fields obtained from different data sources do not contain exactly the same information. In order to overcome this difficulty, a strategy is proposed and validated.