Study On Modeling Method And Model Quality Control Of Ionosphere Based On GNSS

Because GNSS has the advantages of global coverage, uninterrupted service, high precision, except the main function of navigation and positioning, to study the ionosphere through the GNSS observations, also has received more attention in many subjects. The significance of using GNSS to study the ionosphere, on the one hand, it can provide one effective way for monitoring the ionosphere activities and research on the solar activity rules,on the other hand, it can in turn be used to improve precision of GNSS users especially single-frequency real-time navigation and positioning users. The research method is usually with the help of mathematical modeling technology to remodel total ionospheric electronic content. But in practice, many model error sources are involved and the highest theoretical precision can not be reached, that means there are still plenty of rooms for improvement on the existing modeling methods. Paying attention to how to improve existing ionosphere modeling methods and how to improve the modeling quality as far as possible, the concrete works, innovative ideas and main conclusions of this article are listed as follows:1. From success rate and misjudgment rate of cycle slip detection, the usually detecting methods for cycle slips are studied in-depth again, which shows that the Melbourne-Wübbena method alone or Geometry-free method alone has blind spots and insufficiency , but if combined, complementary advantages are obvious, and with some certain hypotheses, success rate can reach 100%. The least-square essence of Blewitt method for cycle slip detection is pointed out along with the redundant freedom problem when using posteriori precision to make hypothesis testing interval.2. The probability when repairing cycle-clip is analyzed. The formulae of single-frequency and double-frequency pseudo-ranges smoothed by carrier phases are derived from the point of bias or un-bias and the whole covariance matrix acquisition problem is focused on. The characteristics of four basic GNSS ionosphere combined observations are compared and three combined plans of using ionospheric observations are determined from the point of the precision of data-processing and the reliability of modeling results. Some related software such as for cycle clip detection of zero-difference phase observations are programmed to extract the ionosphere double-frequency code and phase combined observations.3. Rigorous expressions of non-difference phase observations and ionospheric combination observations are derived when considering the ionosphere higher-order items, geometric bending correction and TEC bending correction.The order of magnitude of that three correcting items are preliminarily studied by using existing models, which shows the ionosphere second-order correcting item generally should be taken into consideration, but other items may be considered according to the ionosphere activities when reconstructing high precision absolute ionosphere model.4. The rigorously formulae to calculate IPP coordinates are given firstly and then the error rules of IPP coordinates caused by three types of approximate formulae are studied. It is shown that three approximate formulae may cause certain IPP coordinate system error but not constant systematic deviation. The influence of satellite ephemerides error and station coordinate error on IPP coordinates is also studied and it is shown that in bad condition (a high-latitude station, small altitude angles), 100m station coordinate component errors will cause about 20" errors for latitude or position and about 33' errors for longitude, and in same case the influence caused by satellite ephemerides is one tenth but the influence of approximate formulae is much bigger. The necessity of using rigorous formulae is analyzed from theory and practice.5. The ionospheric monitoring area and the distribution law of IPPs of single station are systemically studied. For earth-fixed coordinate system, the ionosphere monitoring area of a single station diminishes quickly with increasement of altitude restriction angle and there are "obvious holes" in IPP distributing area when considering GPS constellation only. For sun-fixed coordinate system, the upheld window of a single station is actually the window when upholding different time observation windows of the single layer ionosphere on the sphere and "obvious holes" phenomenon disappears after being upheld. The distribution of IPPs of the station near the equator is more even and will help to improve the precision and reliability of ionosphere modeling results.6. Two mathematical unification forms of expressions of ionospheric models are summarized and one common problem how to select the model coefficients in practices is pointed out. The basic principle of GNSS ionospheric modeling is introduced. Three kinds of standards to evaluate the ionospheric model equalities are summarized and the rigorous formula for evaluating outside precision is derived.7. Two kinds of modeling software suitable for two kinds of basic GNSS ionosphere combined observations, namely IONSP4 and IONSL4, are programmed and used in ionospheric modeling for global different stations.8. Selecting main satellites method for modeling is put forward to solve instabilities of the solutions of integrated instrument bias parameters or integrated phase ambiguity instrument bias parameters.9. Moving window method for modeling is put forward to solve discontinuous values and extreme non-uniformity in accuracy between the neighbor intervals.10. Optimum searching method for model coefficients is put forward to solve how to select one and only optional group of model coefficients.11. Phase restriction method for ionospheric VTEC modeling by use of phase restricting pseudo-range is put forward. Two restriction plans are given, one is varying distance restriction and the other is equal distance restriction. Compared with modeling by smoothing pseudo-range, this method not only improves accuracy of the modeling result with pseudo-range alone but also keeps its reliability because it can include strict stochastic model.12. Three types of ionospheric modeling software with comprehensive advantages of selecting main satellites method, moving window method and optimum searching method for model factors are programmed, namely IAMWOP4, IAMWOL4, IAMWOLP4. Data processing for different IGS stations has shown their availabilities and good prospects of future in-depth studies.