2013 Vol. 38, No. 8
We modified the un-differenced precise point positioning(PPP) mathematical model considering the second-and third-order ionospheric correction.The influence of the higher-order ionospheric corrections on both GNSS observations and static PPP are analyzed with respect to aspects of site location and ionospheric environment by using a globally distributed network which is composed of 38 IGS tracking stations.Numerical experimental results show that,higher-order ionospheric influences on the observation as well as static PPP at low-latitude regions are more significant than that of mid-and high-latitude regions.For static PPP,the influence in low-latitude regions(<30°) reaches about 3-5mm,and is only a sub-millimeter in other regions(>30°).Furthermore,the impact is mainly visible in the direction of North and South direction,showing a southward migration trend,especially at the lower latitudes where the influence value is likely to be more than 3 mm,2 to 3 times that of the East and Up components.The level of ionospheric activity may also have impact on positioning results,and the influence during active periods is about 20-30% higher than that of quiet periodd.
Ionospheric delay is one of the main error sources in GNSS positioning.This paper describes a formula derived to correct higher-order ionosphere error of L3 observation and analyzed the effects on the static and kinematic positioning accuracy.Numerical results show that correcting the higher-order ionospheric error improves the PPP static and kinematic positioning accuracy and shorten the convergence time of PPP kinematic positioning.
A three-dimensional water vapor tomography algorithm is presented.The precise point positioning method is used to obtain the tropospheric slant water vapor.Then COSMIC occultation data is exploited to calculate the water vapor density to be used as initial values in the tomography iteration procedure.Afterwards,the inversion of three-dimensional water vapor is conducted using a fitting method of the Gaussian weighted distance.GPS datasets collected in Hong Kong on three days and the corresponding COSMIC occultation data are used for the algorithm test.RMS statistical value for differences between SWV derived from PPP and SWV derived by tomography is better than 0.87 mm.Comparing the water vapor density by tomography with that derived by the Radiosonde sounding data,their differences are typically below 1.1 g/m3,which demonstrates the feasibility of the presented three-dimensional tomography algorithm.
The Beidou navigation satellite system has been in trial operation since last year,and will provide services in China and its surrounding areas at the end of 2013.We first analyze the system’s satellite visibility in Wuhan,and discuss the Beidou ambiguity resolution(AR) algorithm for short baseline.Experiments were conducted using the dual-frequency GPS/Beidou data collected in Wuhan and processed with GPS and Beidou system separately.The preliminary results indicate that the current Beidou system has the ability of precise positioning capacity.For the short baseline(≤ 15 km) of the Beidou system,the success rate of single epoch ambiguity resolution reaches 80%.The final positioning precision is better than 3 cm on the horizon,and is better than 6 cm for elevation.
The practical mathematical model for gravity field recovery using satellite average accelerations derived from SST-HL mode is discussed.As the numerical differentiation will rapidly amplify the high-frequency noise in orbit-derived accelerations,the filtering method and specific data processing scheme for suppressing the high-frequency noise of satellite average accelerations are presented.As a test,an Earth’s gravity field model WHU-GOCE-SST01S（RE） up to degree and order 100 is recovered with Kaula regularization constrain from 61-day GOCE kinematic orbits and accelerometer data.The result shows that the total accuracy of WHU-GOCE-SST01S（RE） is more accurate than EIGEN-CHAMP03S model and GOCONSGCF2TIMR1 model,which validates the effectiveness of the proposed data processing method.
Two Fortran programs were compiled based on energy balance approach and short arc integral approach respectively.Both ModelENG（energy balance approach） and ModelSAC1（short arc integral approach） were recovered based on 62 days GOCE orbits,while ModelSAC2 was recovered based on 1 a GOCE orbits.The results show that short arc integral approach has a higher precision than energy balance approach.The precision of the ModelSAC2 is higher than EIGEN-champ03s and ModelGRA,recovered based on 90 days GRACE orbits.ModelSAC2 and GOCE-only model GOCONSGCF2TIMR3 have the same precision before 30 degrees.
The EOPs are determined from the IVS observations which Sheshan Station in Shanghai and Nanshan Station in Urumqi were involved.The results show that the precision is comparable to the IERS EOP series,and the current precision of UT1 determined from single baseline observations of the two stations is in the middle level of the international.However,with the development of the technologies of software and hardware,the EOP precision for observations of the two stations is improving year by year.Besides,the determination precisions of UT1 either from the IVS intensive observations in 2011 with Sheshan jointed in or from international EOP observations with the two stations jointed in,can meet the requirements of the 100-m positioning precision of satellite of Mars to UT1.
Aiming to resolve the edge effect in the process of predicting length of day(LOD) by the least squares and autoregressive(LS+AR) model,we employed a time series analysis model to extrapolate LOD series and produce a new series.Then,we used the new series to solve the coefficients for the LS model.At last,we used the LS+AR model to predict the LOD series again.By comparing the accuracy of LOD prediction by edge-effect corrected LS+AR and that by LS+AR,we conclude that edge-effect corrected LS+AR can improve the prediction accuracy,especially for medium-term and long-term predictions.
The constellation rotation error of autonomous orbit determination based on the inter-satellite measurement can be effectively eliminated using anchor stations.But,the accuracy of orbit determination using anchor stations is poor because of the system errors caused by ionospheric delay and tropospheric delay etc.An anchor station timing difference measurement method,using the principle that the system error of a navigation satellite to ground ranging link varies very little in a short time,is proposed to eliminate the system error of measurement system and improve the accuracy of autonomous orbit determination.Simulation results show that the rotation of the constellation can be effectively restricted using the anchor station method.Inaddition,the timing difference measuring algorithm performs better than direct measurement methods.
The existing variance-covariance component estimation(VCE) theory and its defects are analyzed and briefly described.A generalization adjustment factor was developed from the adjustment factor concept,and both generalization closure error and its covariance matrix are investigated based on the generalization adjustment model.A novel VCE method including four basic function models is presented using the generalization adjustment factor.The relationship between four function models and VCE analytical or iterative solution properties is effectively revealed by the generalization adjustment factor.Triangulateration network adjustment results show that the VCE iteration solution lost fewer optimal statistical properties found in the LS criterion.The VCE analytical solution,only for the condition function model provides the optimal statistical properties meeting the LS criterion.
To reduce the influence of troposphere delay for precise positioning in hydrographic surveying,a novel estimation method of troposphere delay was proposed based on the differential estimation idea.Taking the value originating from a Saastamoinen model as the priori value,the amended value of the troposphere delay was calculated through precise point positioning.Taking the residual error of the zenith troposphere delay as the unknown parameter,the delay error was determined by Kalman filtering.Then,the troposphere delay of the base station and the marine rover were estimated respectively.These delay values were then entered into the differential estimation model as the initial values,and the precise position of the rover was determined.The result of boat and airplane experiments show that as compared to the conventional kinematic solution,the locattional precision for boats and the airplanes is improved by the differential estimation technique of troposphere delay.Additionally,the precision in the vertical direction increased about 17.6%.
An integrated approach incorporating computational fluid dynamics(CFD) and GIS technologies is explored to model and simulate the movement characteristics of released gases and the concentration distribution around buildings at the street level.Some key technologies including street canyon modeling,continuous field representation,numerical simulation and visualization of gas dispersion are presented.As a case study,a hypothetical continuous release from a point source is modeled and simulated.The results show that,the proposed approach is able to predict the 3D and dynamic dispersion process in urban streets.The time-varying concentration distributions near the ground highlight the influences of building arrangement on the dispersion.The sampling concentrations around buildings are strongly affected by the positions of sampling points relative to source,including difference in height,distance from the source,and angle from wind direction.The results also confirm that,in the streets the hazardous gas may stay longer,which constitutes a threat to public health and safety.
To express the geographical phenomenon spatio-temporal evolution patterns,many methods via graphs as the result of spatial autocorrelation analysis have been proposed.However,these methods do not effectively express spatio-temporal evolution.Thus,a centroid-based method was employed to express quantitatively the spatio-temporal evolution patterns.Malaria spatial distributions data,for Hunan Province centroid coordinates are computed of every year from 1983 to 1992,as an example.Next,the spatio-temporal malaria evolution process in Hunan Province is expressed via the malaria barycenter curve,composed of these centroid coordinates.In order to verify the results,we compared the centroid-based results and the results from a common spatial autocorrelation analysis technique.The comparison showed two results are in agreement.Finally,two-scale malaria barycenter curve are computed.This experiments showed that the multi-scale effect is not so obvious.
On the basis of the existing GIS information generalization method,a method of inertial selection is put forward in view of correlation between point clusters in GIS.The method will allow evaluation of the correlation between geographic entities through a gravity model considering the spatial and attribute information.Experimental results show that a gravity model can solve the problem of spatial and attribute data integration and can quantitatively describe the correlation between geographic entities.Moreover,the inertial selection model can not only preserve the main spatial features of the geographic entities after generalization,but also takes the influence of correlation between geographic entities on objects selection into consideration.
The research concerning public participation GIS and volunteered geographical information was analyzed,and demands and characteristics of public use of volunteered geographical information for interactive mapping were discussed.On this basis,a prototype system for public participation web cartography was designed and implemented,and successfully applied to projects related to the Shenzhen Geo-spatial information platform.
When the locations of an agent at two times,and its maximum velocity are known,the agent’s location between both those time instances is uncertain.We present a practical method,the total probability theorem,to approximate that uncertainty.First,the minimum(average) velocity from starting point to destination can be computed,and then many discrete speed values between the minimum and maximum velocity can be chosen randomly.The random speed variable V follows the Maxwell-Boltzmann distribution that describes particle speeds,and thus the probability density function of V,p(V),becomes applicable.Second,for a discrete speed value v,we calculate the agent’s reachable range(x,y) at any time t in time geography.The range follows a uniform distribution,and so at t we may obtain p(x,y | v,t),which is the conditional probability of(x,y) given the value of the random variable V,V=v.Finally,according to the total probability theorem,the probability distribution of the agent at time t,p(x,y|t),is obtained by the equation ∑p(V=v)·p(x,y | v,t) where the parameter V takes all values.When increasing the maximum velocity,experiments show that the total probability’ variance has a good convergence and steadiness,an improvement over the existing method’ divergence.
From the point of view of the geological disaster prevention and management,an integrated object-oriented conceptual model for spatial-temporal data was proposed consistent with phenomenon ological and data characteristics of geological disasters.This model is composed of three parts,including a spatial feature model,a spatial entity model and a spatial-temporal object model.Applying this model,a unified mechanism for object state storage and spatial-temporal process expression were designed to effectively describe the status of spatial-temporal objects in geological disasters at different stages of their life cycle.
Landslide hazard is influenced by many temporal and spatial factors.Traditional spatial analytical techniques cannot easily discover new and unexpected patterns,trends,and relationships that can be hidden deep within very large diverse geographic datasets.Focusing on Three Gorges Reservoir Area,environmental and triggering factors for landslide occurrences were extracted from multi-source data.Then,quantitative landslide susceptibility indices were calculated using the trained three-layered BP neural network,and the landslide susceptibility maps were generated.Finally,success rate curve was used to verify the results of landslide susceptibility mapping,and the results showed the best accuracy of 89.75%.The validation showed sufficient agreement between the prediction results and existing landslide.Therefore,the proposed model is an efficient method for landslide intelligent prediction,and can provide a significant reference for landslide hazard prediction and assessment.
Multimodal image registration is one of the most important and difficult problems in the image registration research field.Due to the nonlinear change in the image gray scale between different modalities,the measure to determine the similarity between the image areas or image features is the main difficulty.By analyzing the characteristics of a multimodal image gradient,we propose the gradient consistency operator based on the norm-weighted angle between gradient vectors to improve the region-based image registration algorithm,and apply it to the registration between medium-wave infrared and visible images.
Boresight misalignment between a hyperspectral sensor and intergrated POS system is a key error source in image direct georeferecing(DG).Due to the poor intensity of tie/pass points nets of a nadir push-broom hyperspectral image,a rigorous bundle adjustment cannot be implemented for calibration.To solve this problem,this paper develops a simple boresight misalignment and position offset bundle adjustment calibration model using the exterior orientation elements solved by POS system with only systematic errors.Test results showed that this calibration model is sound and effective.The direct georeferencing precision with this calibration is improved significantly,with a standard deviation of approximately 1.4m(about 4 pixels) at planimetry and 4.4m(about 13 pixels) at the height in the test area.
According to an analysis of support vector machine(SVM) and multiple kernel theory,an improved SVM change detection model based on multi-feature differencing kernel for remote sensing imagery was proposed.The combinations of kernel functions using spectral data and textural feature were discussed.After detailing the structure of the image differencing kernel based on multi-features,the algorithm of SVM change detection model was designed,and combined with category weights for the extraction of the spatial distribution of several change classes.Experimental results show that with the help of a multiple kernel function,the change detection model can get higher detection accuracy than the traditional methods,and also avoid determining change thresholds that are complex and uncertain.
An approach for cloud removal based on linear regression after image classification is proposed in this article.First of all,the clouds in a remote sensing image and its referenced data to be processed are detected,from which two cloud masks are built.Then,an ISODATA classification is applied to the referenced image with the cloud mask.Next,the masked part of the contaminated image is classified with the existing clusters of the referenced data using the minimum distance method.Last,the digital numbers of the cloudy areas of the contaminated image are replaced with by the prediction value of the referenced data calculated by the linear relationships determined between clusters on the referenced image and the corresponding contaminates done according to the pixel location.This algorithm is programmed to automatically detect and remove the clouds areas in Landsat images.The accuracy of cloud detection and the prediction of original values of the cloud cover are evaluated.Results show that the proposed method is effective.
Hyperspectral image linear feature extraction methods often cause information loss and distortion.In view of this,a new kernel minimum noise fraction(KMNF) transform hyperspectral image nonlinear feature extraction method is proposed that introduces a kernel method to minimum noise fraction(MNF) transform.Hyperspectral image KMNF feature extraction experiments were carried out.CUPRITE AVIRIS data experimental results show that sample number influences KMNF slightly,a small number of samples can get almost the same result as a large number of samples;KMNF feature extraction reflects the nonlinear characteristics of hyperspectral images,and endmember extraction effects based on KMNF images outweigh MNF images.
In order to improve the accuracy and efficiency of the pavement crack recognition,a precise fast pavement damage detection method is proposed based on scale-span images.In the image space domain,a pavement image is transformed into differently scaled images according to different thresholds firstly.Several damage identification scale images are selected to set up scale-span image model by superimposing according to different pavement damage image characteristics.At the same time,the road image cracks pattern identifier is designed based on support vector machine(SVM) "1 V m" to extract the fine cracks and thick cracks effectively.The test results show that the algorithm has greatly increased the accuracy and efficiency of the pavement cracks image recognition,The pavement damage degree and range can be also identified quickly and accurately.
Specular scattering,which accompanies ocean breaking waves,provides the Doppler spectrum with more power,a wider bandwidth and a larger peak frequency.These Doppler features are used to detect breaking waves,and an algorithm is proposed to obtain the decision thresholds for three parameters.Breaking Doppler spectra as well as breaking positions can be obtained by the thresholds.Detection results agree well with the breaking positions obtained by visual observation.Furthermore,the results are conducted to improve the reversion of ocean wave parameters.Errors and variances of significant wave height and mean period both drop.This allow the conclusion that this algorithm is a promising method for detecting breaking waves for coherent microwave radar.
COART,a coupled ocean and atmosphere radiative transfer model,was used to simulate the water-leaving radiance of a rough surface and upward radiance just below the surface under varying observation geometric angles.Using these simulations,we analyzed the spatial distribution characteristics of the water-leaving radiance and the transmission coefficient of the upward radiance just below the surface,and discussed the effects of rough surface on these distribution characteristics.The results shows that downward and upward irradiances just below the surface decrease as solar zenith angle(SZA) increases,with a linear relations to the cosine value of solar zenith angle.The spatial distribution characteristics of the water-leaving radiance are not involved with wind-driven rough surface.A simple cubic equation was developed to fit water-leaving radiance change to viewing zenith angle.Under the conditions of the level sea surface,the transmission coefficient of the upward radiance just below the surface does not change with the viewing azimuth angle,however,it decreases with the increasing viewing zenith angle and is linear to its square tangent value.On a rough surface,the viewing zenith angle plays a more important role in determining the transmission coefficient of the upward radiance just below the surface than viewing azimuth angle.In addition,the linear relationship also exists between transmission coefficient and the square tangent value of viewing zenith angle.