Research on Improved Strategies of Regional Ionospheric Tomography Algorithm
Yun Sui,Haiyang Fu,Denghui Wang,Shaojun Feng,Feng Xu,Yaqiu Jin
DOI: https://doi.org/10.33012/2021.18139
2021-01-01
Abstract:Three-dimensional computerized ionospheric tomography (CIT) is an important tool for the reconstruction of electron density distribution. Due to the limited number of ground receivers and observations, ionospheric electron density inversion is typically an ill-posed problem. Limited by this problem, it is difficult to further improve the reconstruction resolution and accuracy. We proposed a model/data-driven hybrid sparse ionospheric tomography method to achieve high spatial and temporal resolution in Reference [1]. In this paper, several strategies have been proposed to modify the cost objective function in order to further improve ionospheric tomography performance. We mainly compare the spline interpolation and the inverse distance weighted (IDW) interpolation algorithm. These methods have been also tested on observation of Beidou Ground Based Augmentation System (BDGBAS). The test results show that the use of the spline interpolation algorithm significantly improves the resolution, the use of the IDW interpolation algorithm significantly improves the inversion accuracy, and the computational cost is significantly lower than that of direct inversion of high-resolution grids. The first strategy we proposed is using a spline interpolation algorithm to fit the electron density of the redefined high-resolution grid. The second strategy is using the IDW interpolation algorithm to obtain the Slant Total Electron Content (STEC) value of an independent reference station, which can be used as a constraint in the objective function of ionospheric tomography modeling. The above strategies are both used in the third strategy. Through differential STEC (dSTEC) analysis, it is verified that these strategies can improve the performance of tomography model while increasing the accuracy of user STEC estimation. By using more than 2200 proprietary CORS stations and more than 300 global reference frame stations, Qianxun SI have already announced the satellite-based augmentation system, "Universe Voice Plan", which is a global SBAS system, broadcasting the correction data in L6 satellite signal and Internet connection. Both simulation and observation data from BDGBAS are used to verify the effectiveness of proposed strategies. Based on data of BDGBAS in Yunnan area, this paper adopts the data-driven method we have proposed for ionospheric tomography modeling, and compares the results of this method with strategy 1, strategy 2 and strategy 3. Firstly, based on the NeQuick2 model, STEC simulation data can be constructed according to the real geometric relationship between satellites and stations. Tomographic models with different strategies are constructed, and the dSTEC error of independent reference station are calculated. Then, the three-dimensional electron density of tomography models and the truth value of NeQuick2 model are compared, it can be seen that the use of these optimization strategies improve the resolution of the ionospheric model while maintaining the characteristics of the electron density distribution. Finally, models can be constructed based on the measured data of BDGBAS, the dSTEC analysis and electron density distribution are also used to compare the impact of each strategy on the tomographic modeling. It can be seen that after adding Strategy 1, Strategy 2, and Strategy 3 to the tomographic modeling, the dSTEC error estimate of the reference station is reduced, and the improvement effect brought by strategy 3 is the most obvious. Through BDGBAS-based simulation and measured data, the effectiveness of the above model improvement strategies are proved. That is, the data-driven tomography algorithm optimized by both optimization strategies can not only rely on sparse observation to achieve high-resolution and high-precision three-dimensional electron density distribution, but also can be used as an atmospheric constraint in the positioning algorithm to further improve the positioning accuracy of the user station. This data-driven 4-D tomography is an important tool for space weather, which can be adopted to achieve large regional quasi-real-time tomography and can be further evaluated to reconstruct dynamic ionospheric regimes, such as equatorial ionization anomaly.