CN105022045A - Multi-source data fusion-based three-dimensional ionosphere chromatographic method - Google Patents

Abstract

The invention discloses a multi-source data fusion-based three-dimensional ionosphere chromatographic method, utilizes GNSS observation data, occultation observation data on low orbit satellites, Jason-1 and Jason-2 seasat altimetry data and ionosphere altimeter data for the first time, and develops research on ionosphere electron density tomography. An inversion result is compared with observation data of an incoherent scattering radar, and a result verifies effectiveness and reliability of the three-dimensional ionosphere chromatographic method provided by the invention in fusing multi-source data to perform inversion of ionosphere electron density and superiority of the method to use of GPS observation data alone to perform inversion.

Description

A kind of three-dimensional Ionospheric Tomography method based on multisource data fusion

Technical field

The present invention relates to satellite geodetic surveying and Space environment detection field of detecting, specifically relate to a kind of three-dimensional Ionospheric Tomography method based on multisource data fusion.

Background technology

Ionospheric Tomography imaging technique GNSS signal reconstruct ionosphere two dimension, three-dimensional and even four-dimensional electron density distribution.It is particularly suitable for the large-scale dimension distribution of monitoring ionosphere space environment, and builds and operating cost relative moderate, so receive the concern of PROGRESS OF IONOSPHERIC RESEARCH IN person, and has achieved many achievements in research.But, due to the restriction of ground station and satellite spatial Numerical Distribution, cause the lack of uniformity of effective observation information coverage for inverting and distribution, so when utilizing GNSS observation data to carry out Ionospheric Tomography imaging, only carrying out improving from inversion algorithm is the ill-posedness being difficult to fundamentally solve observation equation.

Along with the appearance of the foundation of multiple satellite system, more and more ground station laying and multiple observation method, provide strong support and guarantee for improving ionospheric monitoring capability further.The GNSS at ground monitoring station observes ray normally high altitude angle, causes to utilize GNSS observation data to carry out Ionospheric Tomography inverting to have higher horizontal resolution, and causes vertical resolution lower due to the disappearance of horizontal rays.In order to head it off, propose as far back as Hajj in 1994 etc. and be used in Ionospheric Tomography imaging by Observation of Occultation data, subsequently, a lot of scholar has carried out experimental study.The LEO-based GPS observation data that Li etc. utilize GPS observation data and COSMIC low-orbit satellite to provide, the effective inverting spatial and temporal distributions structure of regional Ionosphere Over electron density.Xiao etc. utilize GPS observation data and CHAMP/GRACE low-orbit satellite occultation data, the inverting space distribution of magnetic storm times ionospheric electron density, and achieve good effect.Meanwhile, also have scholar by altimeter data fusion in tomographic inversion, Zhao etc. utilize GPS observation data and the altimeter data inverting ionospheric electron density in overhead, China Partial region effectively, thus vertical resolution is increased.Chartier etc. are worth as a setting with incoherent scattering radar observed reading, merge the vertical precision that altimeter data improve the imaging of GPS Ionospheric Tomography, effectively overcome the shortcoming that GPS tomography vertical precision is not high.

These methods all improve the precision of Ionospheric Tomography inverting to a certain extent above, but data message or skewness, cause chromatographic effect can't reach a perfect condition.

Summary of the invention

For above-mentioned Problems existing, the object of this invention is to provide a kind of three-dimensional Ionospheric Tomography method based on multisource data fusion, to make up the deficiency of observation information, thus improve Ionospheric Tomography inversion accuracy higher.

For achieving the above object, the present invention adopts following technical scheme:

Based on a three-dimensional Ionospheric Tomography method for multisource data fusion, comprise the following steps:

(1) observation data is collected: determine target area scope, the Observation of Occultation data within the scope of chosen area on GNSS observation data, low orbit satellite, Jason-1 and Jason-2 seasat survey high data and ionosonde data;

(2) will the ionosphere spatial discretization of inverting be treated, form the tomographic inversion model based on pixel base;

(3) projection matrix is calculated;

(4) the three-dimensional Ionospheric Tomography model of multi-source data is built;

(5) the three-dimensional Ionospheric Tomography model of multi-source data is resolved, inverting region ionospheric electron density.

Described step (1) specifically comprises: determine the longitude of target area, latitude and altitude range; Determine the time period of inverting; Calculate ionized layer TEC according to the observation data in region, and extract the related data of satellite and survey station coordinate.

Described step (2) is calculated as follows: be nonlinear between ionospheric electron density and ionized layer TEC, in practical inversion process, in order to inverting is convenient, usually adopts Discrete Inversion to treat the ionosphere spatial discretization of inverting.Inverse model is analysed by pixel basic unit for discretize, selected pixels target function b _jas basis function, if ray is through certain pixel, then b _jbe 1, otherwise be 0; And ionosphere is turned to three-dimensional graticule mesh by discrete in longitude, latitude and short transverse, its equation expression is:

In formula, n is the graticule mesh number of discretize, namely total pixel count; x _j(j=1 ..., n) be model parameter, the ionosphere graticule mesh electron density namely after discretize.

Described step (3) is calculated as follows: can be expressed as the TEC measured value in every bar raypath:

In formula, m is ionized layer TEC observed reading sum, a _ijfor projection matrix element, i.e. the intercept of i-th ray in a jth graticule mesh.Consider the impact of observation noise and discretization error in measurement, and suppose that in graticule mesh, electron density is constant in certain hour section, then the ionized layer TEC measurement data on every bar ray travel path can be expressed as:

By above formula matrix representation, as follows:

y _m×1＝A _m×n·x _n×1+e _m×1(5)

In formula, y is the m dimensional vector of ionized layer TEC observed reading composition, and A is projection matrix, the row vector that m the n that namely intercept of ray in respective pixel is formed ties up, x is the n dimensional vector of unknown parameter composition, and e is the m dimensional vector of observation noise and discretization error composition.

Described step (4) is calculated as follows: utilize the Observation of Occultation data on GNSS observation data, low orbit satellite, ionosonde data and Jason-1 and Jason-2 seasat to survey high data aggregate inverting ionospheric electron density, its expression formula can be expressed as:

In formula, y _gNSS, y _cOS, y _aLTand y _iONrepresent the TEC value that the TEC observed reading of ground GNSS, the TEC value of LEO spaceborne GNSS observation, the TEC value surveying the acquisition of high satellite and altimeter obtain, A respectively _gNSS, A _cOS, A _aLTand A _iONrepresent corresponding matrix of coefficients, x represents electron density value to be asked.

In described step (5), carry out tomographic inversion by multiplication algebra restructing algorithm (MART), solve ionospheric electron density.

The technology of the present invention beneficial effect:

The present invention utilizes multisource data fusion to carry out Ionospheric Tomography inverting, compensate for the problem of inverting information deficiency, by the observation data of inversion result and incoherent scattering radar is compared, demonstrate the present invention and merge the validity and reliability of multi-source data inverting ionospheric electron density and the superiority relative to employing GPS observation data inverting separately thereof.

Accompanying drawing explanation

Fig. 1 is the Ionospheric Tomography process flow diagram of multisource data fusion in the embodiment of the present invention;

Fig. 2 is IGS research station and incoherent scattering radar station distribution plan in the embodiment of the present invention;

Fig. 3 is the comparison diagram measuring section in the embodiment of the present invention at the inverting ionospheric electron density section at 13:00UT moment Jicamarca station and incoherent scattering radar station;

Fig. 4 is the comparison diagram measuring section in the embodiment of the present invention at the inverting ionospheric electron density section at 13:00UT moment Millstone Hill station and incoherent scattering radar station;

Fig. 5 is the comparison diagram measuring section in the embodiment of the present invention at the inverting ionospheric electron density section at 21:00UT moment Jicamarca station and incoherent scattering radar station;

Fig. 6 is the comparison diagram measuring section in the embodiment of the present invention at the inverting ionospheric electron density section at 21:00UT moment Millstone Hill station and incoherent scattering radar station;

Fig. 7 is the inverting ionospheric electron density peak value at Jicamarca station in the embodiment of the present invention and comparing of electron peak density height and incoherent scattering radar measured result;

Fig. 8 is comparing of inverting ionospheric electron density peak value that in the embodiment of the present invention, Millstone Hil stands and electron peak density height and incoherent scattering radar measured result.

Embodiment

Below in conjunction with the drawings and specific embodiments, the invention will be further described, but scope of the present invention is not limited to following embodiment.

The observation data that the present invention adopts is from IGS observation grid, the observatory information chosen in reconstruction region is reconstructed, and utilize incoherent scattering radar Jicamarca (76 ° of W, 11.9 ° of S) and Millstone Hill (71.5 ° of W, the 42.6 ° of N) observation data of standing carry out independently checking.As shown in Figure 2, in figure, "●" is IGS station, and " " is incoherent scattering radar station in survey station distribution.

Step (1), determines target area scope, and the Observation of Occultation data within the scope of chosen area on GNSS observation data, low orbit satellite, Jason-1 and Jason-2 seasat survey high data and ionosonde data;

Choosing longitude range is 100 ° of W ~ 30 ° W, and latitude scope is 40 ° of S ~ 50 ° N and altitude range is the data of 100km ~ 1000km, the distributed in three dimensions situation of inverting ionospheric electron density on August 6th, 2011.Pixel separation on longitude and latitude direction is set and is respectively 2 °, short transverse is spaced apart 50km.

Step (2), will treat the ionosphere spatial discretization of inverting, form the tomographic inversion model based on pixel base;

Inverse model is analysed by pixel basic unit for discretize, selected pixels target function b _jas basis function, if ray is through certain pixel, then b _jbe 1, otherwise be 0; And ionosphere is turned to three-dimensional graticule mesh by discrete in longitude, latitude and short transverse, its equation expression is:

In formula, n is the graticule mesh number of discretize, namely total pixel count; x _j(j=1 ..., n) be model parameter, the ionosphere graticule mesh electron density namely after discretize.

Step (3), calculates projection matrix;

Can be expressed as the TEC measured value in every bar raypath:

In formula, m is ionized layer TEC observed reading sum, a _ijfor projection matrix element, i.e. the intercept of i-th ray in a jth graticule mesh.Consider the impact of observation noise and discretization error in measurement, and suppose that in graticule mesh, electron density is constant in certain hour section, then the ionized layer TEC measurement data on every bar ray travel path can be expressed as:

By above formula matrix representation, as follows:

y _m×1＝A _m×n·x _n×1+e _m×1(5)

In formula, y is the m dimensional vector of ionized layer TEC observed reading composition, and A is projection matrix, the row vector that m the n that namely intercept of ray in respective pixel is formed ties up, x is the n dimensional vector of unknown parameter composition, and e is the m dimensional vector of observation noise and discretization error composition.

Step (4), builds the three-dimensional Ionospheric Tomography model of multi-source data;

Utilize the Observation of Occultation data on GNSS observation data, low orbit satellite, ionosonde data and Jason-1 and Jason-2 seasat to survey high data aggregate inverting ionospheric electron density, its expression formula can be expressed as:

In formula, y _gNSS, y _cOS, y _aLTand y _iONrepresent the TEC value that the TEC observed reading of ground GNSS, the TEC value of LEO spaceborne GNSS observation, the TEC value surveying the acquisition of high satellite and altimeter obtain, A respectively _gNSS, A _cOS, A _aLTand A _iONrepresent corresponding matrix of coefficients, x represents electron density value to be asked.

Step (5), resolves the three-dimensional Ionospheric Tomography model of multi-source data, inverting region ionospheric electron density.

Carry out tomographic inversion by multiplication algebra restructing algorithm, solve ionospheric electron density.Fig. 3-6 show not to be measured section based on the ionospheric electron density section of multi-source data and the inverting of single GPS observation data with incoherent scattering radar station in the same time and compares.Fig. 3 and Fig. 4 is respectively and compares at the section that 13:00UT moment Jicamarca stands and Millstone Hill stands, Fig. 5 and Fig. 6 is respectively and compares at the section that 21:00UT moment Jicamarca stands and Millstone Hill stands.Therefrom can find out, for two different moment and different research station, based on the ionospheric electron density section of multi-source data inverting on the whole more close to the measurement result at incoherent scattering radar station, and vertical resolution improves greatly, this method effectively compensate for the problem of CIT technology limited perspective.Confirm that result based on multi-source data inverting is relative to the superiority of single gps data and reliability.Fig. 7-8 give comparing based on the ionosphere peak electron density of multi-source data and the inverting of single GPS observation data and peak height and incoherent scattering radar measured result.Fig. 7 and Fig. 8 is respectively Jicamarca station and MillstoneHill and stands the comparison of peak electron density in a day.Therefrom can find out, based on the ionospheric electron density peak bulk of multi-source data inverting more close to the measurement result at incoherent scattering radar station.

By above technical scheme, confirm the reliability based on the inverting of multi-source data Ionospheric Tomography and superiority.

The concrete meaning of all pa-rameter symbols that the present invention relates to is:

for ionospheric electron density;

N is the graticule mesh number of discretize, namely total pixel count;

X _j(j=1 ..., n) be model parameter, the ionosphere graticule mesh electron density namely after discretize;

TEC is ionosphere total electron content;

M is ionized layer TEC observed reading sum;

A _ijfor projection matrix element, i.e. the intercept of i-th ray in a jth graticule mesh;

ε _ifor observation noise;

A is projection matrix;

E is the m dimensional vector of observation noise and discretization error composition;

Y _gNSSfor the TEC observed reading of ground GNSS;

Y _cOSfor the TEC value of the spaceborne GNSS observation of LEO;

Y _aLTfor surveying the TEC value that high satellite obtains;

Y _iONfor the TEC value that altimeter obtains;

A _gNSSfor ground GNSS projection matrix;

A _cOSfor the spaceborne GNSS projection matrix of LEO;

A _aLTfor surveying high satellite projection matrix;

A _iONfor altimeter projection matrix;

X is electron density value to be asked.

Above embodiment only for illustration of the present invention, and is not limitation of the present invention, and all any amendments done within the spirit and principles in the present invention, equivalent replacement and improvement etc., all should be included within protection scope of the present invention.

Claims (6)

1., based on a three-dimensional Ionospheric Tomography method for multisource data fusion, it is characterized in that, comprise the following steps:

(1) observation data is collected: determine target area scope, the Observation of Occultation data within the scope of chosen area on GNSS observation data, low orbit satellite, Jason-1 and Jason-2 seasat survey high data and ionosonde data;

(2) will the ionosphere spatial discretization of inverting be treated, form the tomographic inversion model based on pixel base;

(3) projection matrix is calculated;

(4) the three-dimensional Ionospheric Tomography model of multi-source data is built;

(5) the three-dimensional Ionospheric Tomography model of multi-source data is resolved, inverting region ionospheric electron density.

2., as claimed in claim 1 based on the three-dimensional Ionospheric Tomography method of multisource data fusion, it is characterized in that, described step (1) specifically comprises: determine the longitude of target area, latitude and altitude range; Determine the time period of inverting; Calculate ionized layer TEC according to the observation data in region, and extract the related data of satellite and survey station coordinate.

3. as claimed in claim 1 based on the three-dimensional Ionospheric Tomography method of multisource data fusion, it is characterized in that, described step (2) is calculated as follows: be nonlinear between ionospheric electron density and ionized layer TEC, adopts Discrete Inversion to treat the ionosphere spatial discretization of inverting; Inverse model is analysed by pixel basic unit for discretize, selected pixels target function b _jas basis function, if ray is through certain pixel, then b _jbe 1, otherwise be 0; And ionosphere is turned to three-dimensional graticule mesh by discrete in longitude, latitude and short transverse, its equation expression is:

$b_j=\{\begin{array}{cc}1& \left(\stackrel{\→}{r}\∈V_{voxel}\right)\\ 0& \left(others\right)\end{array}---\left(1\right)$

$Ne(\stackrel{\→}{r},t)\≅\underset{j=1}{\overset{n}{\Σ}}{x}_j\left(t\right)\·b_j\left(\stackrel{\→}{r}\right)---\left(2\right)$

In formula, n is the graticule mesh number of discretize, namely total pixel count; x _j(j=1 ..., n) be model parameter, the ionosphere graticule mesh electron density namely after discretize.

4., as claimed in claim 1 based on the three-dimensional Ionospheric Tomography method of multisource data fusion, it is characterized in that, described step (3) is calculated as follows: be expressed as the TEC measured value in every bar raypath:

$\begin{array}{cc}{\mathrm{TEC}}_i\≅{\∫}_l\underset{j=1}{\overset{n}{\Σ}}{x}_j\left(t\right)\·b_j\left(\stackrel{\→}{r}\right)ds=\underset{j=1}{\overset{n}{\Σ}}{x}_j\left(t\right){\∫}_l{b}_j\left(\stackrel{\→}{r}\right)ds=\underset{j=1}{\overset{n}{\Σ}}{a}_{ij}\·x_j\left(t\right)& (i=1,...,m)---\left(3\right)\end{array}$

In formula, m is ionized layer TEC observed reading sum, a _ijfor projection matrix element, i.e. the intercept of i-th ray in a jth graticule mesh; Consider the impact of observation noise and discretization error in measurement, and suppose that in graticule mesh, electron density is constant in certain hour section, then the ionized layer TEC measurement data on every bar ray travel path is expressed as:

$\begin{array}{cc}{\mathrm{TEC}}_i=\underset{j=1}{\overset{n}{\Σ}}{a}_{ij}\·x_j+{\mathrm{\ϵ}}_i& (i=1,...,m)---\left(4\right)\end{array}$

By above formula matrix representation, as follows:

y _m×1＝A _m×n·x _n×1+e _m×1(5)

In formula, y is the m dimensional vector of ionized layer TEC observed reading composition, and A is projection matrix, the row vector that m the n that namely intercept of ray in respective pixel is formed ties up, x is the n dimensional vector of unknown parameter composition, and e is the m dimensional vector of observation noise and discretization error composition.

5. as claimed in claim 1 based on the three-dimensional Ionospheric Tomography method of multisource data fusion, it is characterized in that, described step (4) is calculated as follows: utilize the Observation of Occultation data on GNSS observation data, low orbit satellite, ionosonde data and Jason-1 and Jason-2 seasat to survey high data aggregate inverting ionospheric electron density, its expression formula is expressed as:

$\left\{\begin{array}{c}{A}_{GNSS}x=y_{GNSS}\\ A_{COS}x=y_{COS}\\ A_{ALT}x=y_{ALT}\\ A_{ION}x=y_{ION}\end{array}\right.---\left(6\right)$

In formula, y _gNSS, y _cOS, y _aLTand y _iONrepresent the TEC value that the TEC observed reading of ground GNSS, the TEC value of LEO spaceborne GNSS observation, the TEC value surveying the acquisition of high satellite and altimeter obtain, A respectively _gNSS, A _cOS, A _aLTand A _iONrepresent corresponding matrix of coefficients, x represents electron density value to be asked.

6., as claimed in claim 1 based on the three-dimensional Ionospheric Tomography method of multisource data fusion, it is characterized in that, in described step (5), carry out tomographic inversion by multiplication algebra restructing algorithm (MART), solve ionospheric electron density.