High Accuracy Atmospheric Spatio-Temporal Modeling In Small Scale Area Based On BDS/GPS

Ground-based GNSS atmospheric remote sensing is an important application of GNSS technology. Today’s GNSS atmospheric remote sensing object is using CORS network to establish a wide range of atmospheric temporal and spatial model within centimeter-level accuracy, but using GNSS to model high spatial and temporal resolution atmosphere in small scale area in real time with millimeter-level is a research gap.In the deep space exploration field, the antenna array technique is an effective way to improve the measurement and control capability to deep space probe. When the deep space probe signal transmit through the earth atmosphere, it suffers in ionospheric and tropospheric delay. Different electron density in ionosphere and different water vapor density in troposphere result in different signal delay in each transmit path, meanwhile the different signal delay led to low signal noise ratio in antenna array synthetic signal. Therefore we need to monitor the atmospheric phase variations. At present it mainly use interferometer, water-vapor radiometer (WVR) and Global Naviation Satellite System (GNSS) to analyse the temporal and spatial characteristics of atmosphere.Until now, the research of atmosphere with GNSS is focused on medium and large scale (range from several kilometers to hundreds of kilometers). Based on university cooperation project’the effect of atmospheric phase fluctuations on uplink arrraying and modeling technology’, this thesis research on the BDS/GPS CORS data processing theory in small scale area, and building the high accuracy, high spatial and temporal resolution atmospheric model, then solve the problem of atmospheric phase delay calibration in antenna arraying. The specific content and results of this thesis are summarized as following:(1) In the small scale area CORS network, we adopt single difference strategy, single difference can take advantage of the earth-centered earth-fixed (ECEF) repeat period of each satellite to mitigate the multipath effects. Meanwhile the single difference atmospheric delay can be easily used to modeling the atmospheric delay in antenna arraying. In the experiment, the continuous several days’residuals weighting is used to obtain multipath model.(2) With the comparison of the atmospheric delay of BDS/GPS dual-frequency, the ionospheric dealy of single-differences compared with tropospheric delay is much smaller, and considering the frequencies of antenna arraying system are much higher than GNSS signals, the ionospheric delay will be smaller. Therefore, the tropospheric delay is mainly being discussed.(3) The correlation of atmospheric delay between receivers in small scale area is reduced with the length of baseline increasing. The correlation of tropospheric delay between receivers is enhanced with the satellite elevation angle increase. For instance, when the elevation angle is greater than 10°, the Root Mean Square (RMS) of tropospheric delay between receivers is under 2.6mm with the normal weather and 8.6 mm with unnormal weather; and when the elevation angle is up to 30°, the RMS are 1.7 mm and 4.5 mm corresponding to the normal and unnormal weather. Using the known delay of the receivers, the tropospheric delay distribution of the whole area can be interpolated, and in this experiment the performance of the Linear Interpolation Method (LIM) is better than Low-order surface model (LSM). The average RMS of the values of observation and interpolation is 1.5mm, and LIM can correct 29% tropospheric delay.(4) Because of the multipath effect, the troposphere delay is not accurate enough, it caused the weak correlation of tropospheric delay between satellites. It needs other methods to improve the precision of tropospheric delay, then analyse the correlation of tropospheric delay between satellites.(5) Temporal Structure Function (TSF) is used to characterize the temporal correlation. With the interval increase, the values of TSF increase firstly then tends to be stable. When the interval is 60 s, the values of TSF are 0.87 mm in normal weather and 1.9 mm in unnormal weather. And when the interval is 360 s, the corresponding values are 1.33 mm,3.1 mm. The elevation angle of satellites also affect the temporal correlation of tropospheric delay. When the interval is 120 s, elevation angle is between 40°～50°, the values of TSF are less than 1.5 mm in normal weather and 3.8 mm in unnormal weather, when the elevation angle is between 20°～30°, the values are 2.3 mm and 5.2 mm coorespondly, and when the elevation angle is between 10°～20°, the values are 3.2 mm and 6.4 mm cooresponding to the normal and unnormal weather.(6) The modeling strategy of atmospheric phase delay on antenna arraying:firstly, using GNSS and WVR to get the tropospheric delay on the target direction, then interpolate the tropospheric delay of the antenna at any position. The frequencies of antenna arraying system are much higher than GNSS signals, so the ionospheric delay is smaller. Therefore, the ionospheric phase fluctuations can be neglected.