We use the global ionospheric maps(GIMs) provided by international GNSS service(IGS) to analyze the total electron content(TEC) changes during the storm on August 26, 2018. We use the sliding interquartile rang method to extract the disturbance characteristics of the global TEC, and discuss the response mechanism of the ionospheric disturbance caused by this storm. The results show: 1)The ionospheric TEC is a mainly positive disturbance except in the Arctic region during the main phase of the storm. TEC shows a long-term negative disturbance in the northern hemisphere during the recovery phase, but it is opposite in the southern hemisphere. 2) The ionospheric TEC disturbance caused by the storm is related to the change of thermosphere O/N₂ ratio. The eastward prompt penetration electric field(PPEF) has a great influence on the global daytime ionospheric TEC positive disturbance during the period when the Bz component of interplanetary magnetic field is southward; the sedimentation of high energy particles may be the cause of the positive disturbance of TEC in high latitude and polar regions. The seasonal circulation between summer and winter promotes the north-south asymmetric response of global ionospheric TEC during the recovery phase of the storm.

To carry out wave vector and Poynting vector analyses, we utilize ELF three-component waveform data of magnetic field and electric field in burst mode of over 6 000 number 1130 and 1135 orbital data recorded by the microsatellite DEMETER. These dataoccurred within 30 days before the earthquakes, ±10° above the epicenter for six earthquakes with M_S≥6 from January 2005 to December 2009 in northeast Asia (105°-145°E；38°-58°N).The study shows that the satellite recorded the arrival of the abnormal electromagnetic wave below the proton Lamor frequency originating before the M_S7.2 earthquake that occurred on August 16, 2005 near east coast of Honshu and the June 5, 2009 M_S6.6 earthquake that occurred in Hokkaido. The satellite did not record the arrival of the abnormal electromagnetic wave below the proton Lamor frequency originating from the earth before the November 10, 2005 M_S6.4 earthquake that occurred in eastern Russia and the August 2, 2007 M_S6.8 earthquake in Sakhalin.Though the satellite recorded the arrival of the abnormal electromagnetic wave below the proton Lamor frequency,the waves before the February 17, 2007 M_S6.2 earthquake in Hokkaido and the June 13, 2008 M_S7.3 earthquake near the east coast of Honshu, Japan were not from the earth.

Using the seismic waveform data since 2009 in southern Jiangxi and its adjacent areas, wecalculate the apparent stress of 249 M_L1.8 to 4.0 earthquakes and calculate b-value using the earthquake catalogue since 1978. The results show that the apparent stress value is between 0.002 to 2.070 MPa, and the average value is 0.152 MPa. From the time variation of apparent stress with M_L2.0 to 2.9 earthquakes, three of the five M_L≥3.5 earthquakes (aftershocks removed) show high b-value anomalies since 2009. From the perspective of spatial distribution, M_L≥3.0 earthquakes occur in the region of high apparent stress and near the boundary between high and low values. The apparent stress is positively correlated with the magnitude. The minimum magnitude of completeness in the study area is M_L2.0 since 1978, and the b-value is 0.78; since 2009, most of M_L≥3.0 earthquakes occur in the low b-value region or near the edge. The low b-value region is close to the high apparent stress region, which reflects the high seismicity and crustal stress state.

Based on dense array data in northwest Yunnan, using absolute positioning combined with the relative positioning method, we relocate the foreshocks, main shock, and aftershocks of the Yangbi M6.4 earthquake sequence. Simultaneously, we use the CAP full waveform inversion method to obtain the focal mechanism solution and centroid depth of M≥5.0 earthquakes. Combined with the accurate positioning results of the earthquake sequence, we analyze the Yangbi M6.4 earthquake. The results show that the Yangbi earthquake sequence presents a NW distribution with a length of 25~30 km and a width of about 5 km. The main shock is at the northern end of the whole sequence, and aftershocks are concentrated in the north. The moment magnitude of Yangbi M6.4 main shock is M_W6.03, the centroid depth is 5.8 km, and the strike, dip, slip angle of the nodal plane Ⅱ is 133°, 75°, -164°, respectively, which is consistent with the NW-SE direction of earthquake sequence. Both the focal mechanism solution and focal depth profile of M≥5.0 earthquakes show high dip angles, and there is a slowing trend from northwest to southeast. The M6.4 earthquake has the characteristics of unilateral rupture on the southeast side, and the nodal plane Ⅱ is the seismogenic fault plane. The Yangbi M6.4 earthquake is a right-lateral strike-slip earthquake with NW-trending high-dip. Its seismogenic structure is a northwest-trending hidden fault or a secondary fault on the west side of the Weixi-Qiaohou-Weishan fault zone. The fault may not be completely parallel to the Weixi-Qiaohou fault zone, and there is a certain angle.

Based on the GNSS velocity field from 2013 to 2018 in the Sichuan-Yunnan region, we calculate the regional strain rate fields, and analyze the current strain accumulation rates distribution of the eastern boundary of the Sichuan-Yunnan block. We then invert the degree of fault locking and slip-rate deficit on the eastern boundary of the Sichuan-Yunnan block, focusing on the analysis of the strain accumulation characteristics in the deep depth of the eastern boundary of the Sichuan-Yunnan block after the Lushan earthquake. Finally, combined with the results of historical earthquake rupture and fault Coulomb stress, we comprehensively analyze the risk of strong earthquakes in different segments of the eastern boundary of the Sichuan-Yunnan block. The results show that the eastern boundary of the Sichuan-Yunnan block presents obvious characteristics of left-lateral strain accumulation; the shear strain accumulation rate in the Sanchakou area is the largest, with a maximum value of 3.5×10^-8/a. Moreover, the eastern boundary of the Sichuan-Yunnan block also exhibits NS-NE directed extension and EW-NW directed compression deformation. The Kangding-Shimian segment has the highest EW-NW compression strain rate of -3×10^-8/a. The Anninghe-Zemuhe fault zone also exhibits E-W-directed compression; the E-W-directed compression strain rate is -2×10^-8/a to -3×10^-8/a, while the compression deformation of the Xiaojiang fault zone gradually slows down from north to south and turns into tensile deformation near the south of Dongchuan. The fault-coupling inversion results are consistent with the GPS strain rate results, and reflects the left-lateral shear strain and local compression or tensile strain accumulation characteristics of the eastern boundary of the Sichuan-Yunnan block. Comprehensive analysis shows that the Daofu-Bamei segment, the Moxi segment of the Xianshuihe fault zone, the Anninghe fault zone, and the Qiaojia-Dongchuan and Yiliang-Jianshui segments of the Xiaojiang fault zone currently have a high locking degree and a left-lateral slip rate deficit, resulting in high potential for the next strong earthquake.

Based on 1 255 seismic source mechanism solutions of M_W≥3.5 during 1976-01-01~2015-12-31 provided by Global CMT, we use the FMSI method to invert the stress field in the western Qinghai-Tibet plateau and its adjacent areas (western Xinjiang). We analyze the trend of fault sliding in the west Kunlun fault zone and Altyn Tagh fault zone. The results show that the maximum principal compressive stress axis σ1 in the western Tibetan plateau and neighboring areas is in the NNE-SSW direction and rotates clockwise from west to east; the minimum principal compressive stress axis σ3 is in the NWW-SEE direction. The dip of the maximum principal compressive stress axis σ1 is about 5°~9°, indicating that the tectonic stress field in this area is dominated by horizontal motion. The west Kunlun fault zone and Altyn Tagh fault zone have a strong sliding trend (T_s≥0.5), indicating that the possibility of future destabilization and sliding of the west Kunlun fault zone and Altyn Tagh fault zone is larger.

Using 54 GNSS reference station observations in and around Yunnan from May 20, 2019 to May 20, 2021, we obtain the velocity field and strain field in Yunnan based on the least squares configuration method.We combine the time series of GNSS baseline length changes in the near-field and far-field regions of the 2021-05-21 Yangbi M_S6.4 earthquake to analyze the crustal deformation characteristics before the earthquake. The results show enhanced motion in the SSW direction at the western boundary of Yunnan before the earthquake, and a dispersive motion along the Nujiang fault. The near-field velocity values near the epicenter did not change much, and the anomalous changes in the regional velocity field are not obvious. The value of the surface expansion rate of the study area as a whole before the earthquake dropped to about 0.37 times its original value, while the maximum shear strain rate and principal strain rate also dropped sharply, indicating that the crustal strain accumulation in the area before the earthquake weakened significantly, and the epicenteris located on the zero value line of alternating tension and compression.The GNSS baseline length time series results show that the GNSS baseline anomalies in the near-field region before the earthquake changed insignificantly, and the GNSS baseline in the far-field region has trend anomalous changes that deviate from the long-term background before the earthquake. The baseline KMIN-YNYM accelerates backward in tension against the background of trend compression, deviating from the trend motion by 8.4 mm. The baseline YNLP-KMIN accelerates in compression against the background of trend compression, deviating from the trend motion by 6.4 mm; the baseline YNLP-YNDZ accelerates in compression against the background of relatively smooth trend deformation, deviating from the trend motion by 8.1 mm. The baseline YNSM-YNLP accelerates in tension against the background of trend tension, deviating from the trend motion by 7.8 mm. Weinferthat the stress equilibrium state in the region before the earthquake was broken and induced the earthquake.

Using Sentinel-1 ascending and descending data and D-InSAR technology, we obtain the LOS deformation field of May 22, 2021Maduo,Qinghai,M_W7.3 earthquake. As the earthquake rupture reaches the surface, the interference fringes along the fault section are discontinuous, broken and overlapping. In order to accurately extract the geometric distribution of the seismogenic fault of the earthquake, combined with the Landsat-8 optical image data before and after the earthquake, we extract the horizontal deformation field and geometric distribution characteristics of the Maduo earthquake based on the frequency domain cross-correlation algorithm. The results show that the surface fracture track of the earthquake is 155.6 km, and there are branch fractures at the head and end. The main surface fracture zone can be divided into three segments, with great differences in strike between each segment. Its overall distribution is NWW. The seismogenic fault is mainly sinistral strike slip. It is speculated that this earthquake is controlled by the Kunlunshankou-Jiangcuo fault.

BDS Ground-based Augmentation System(GBAS) is an important infrastructure to promote high-precision BDS+ applications. For the first time, this paper studies the positioning performance of three nationwide GBASs: Qianxun, Sixents Technology and China Mobile. Through the analysis of two 8-9-hour observation sessions on July 21 and 22, 2021, we infer: 1) Both Qianxun SI and China Mobile support BDS-2 triple-frequency signals and BDS-3 dual-frequency signals(B1I and B3I), whereas Sixents Technology supports BDS-2 and BDS-3 dual-frequency signals(B1I and B3I); 2) All three BDS GBASs can provide 100% data availability; 3) Static baseline results show a <2×10^-6 relative error of the closed loop with the three virtual base stations; 4) Single-epoch RTK results show that all three BDS GBAS can provide less than 3/9 mm horizontal/vertical precision. However, cm-dm level positioning bias exists among different BDS GBASs. As the result, it is not recommended to use inconsistent BDS GBASs in a single RTK task.

The positioning accuracy of medium-long baseline RTK has not been determined in the newly added B1C and B2a signals of BDS-3 MEO satellite and IGSO satellite, so to resolve this problem we analyze the data quality of new signals of BDS-3, B1I and B3I signals of BDS-3 and L1 and L2 signals of GPS. We further study the positioning of dual-frequency RTK using four groups of measured data of medium-long baseline. The results show that, in terms of data quality, the number of visible satellites and PDOP of BDS-3 are better than GPS, and the SNR ratio and multipath error of new signals of BDS-3 are equivalent to B11 and B3I signals of BDS-3 and L1 and L2 signals of GPS. In RTK positioning of medium-long baseline, the ambiguity of the new signal B1C+B2a combination of BDS-3 is better than B1I+B3I combination of BDS-3 for the first time, and the positioning accuracy of new signal combination is slightly better than B1I+B3I combination of BDS-3 and L1+L2 combination of GPS. This can provide users with centimeter-level positioning accuracy.

In this paper, we evaluate the performance of BDS-3/GPS/Galileo triple-frequency carrier ambiguity resolution(TCAR) based on geometry-free(GF) model with raw data. We first present triple-frequency linear combinations that are suitable for the BDS-3/GPS/Galileo TCAR with short baselines. Then we evaluate the performance of BDS-3/GPS/Galileo single-epoch extra-wide lane(EWL), wide lane(WL), and narrow lane(NL) ambiguity resolution with raw data from short baselines of different lengths. Experimental results demonstrate that for the baselines of 5 m, 3.6 km and 13.1 km, the fixing rates of EWL ambiguities for BDS-3/GPS/Galileo are similar(more than 99.7%), and the fixing rates of WL ambiguities for BDS-3/GPS(more than 97.2%) are slightly higher than Galileo(more than 93.0%). The fixing rates for BDS-3 NL ambiguities(more than 92.2%) are slightly higher than GPS/Galileo(more than 89.4%).

To make full use of the advantages of BDS-3 quad-frequency observation information, we develop a cycle-slip detection and repair method for quad-frequency observations. First, we properly select the combination coefficients of the geometry-free phase combination and geometry-free ionosphere-free combination. We combine three geometry-free phase combinations as well as one geometry-free ionosphere-free combination to detect different cycle slips. Then, we use the least-square method to estimate the cycle-slip float solutions. Finally, we adopt the LAMBDA method to further obtain the cycle-slip integer solutions. The BDS-3 quad-frequency observation data verifies that the proposed method can effectively and quickly identify various cycle slips and repair them.

Using GF combination of adaptive threshold model in static and dynamic environments, we develop the method of BDS peri-hop detection for different orbiting satellites. The results show that the method can be applied to circumferential hop detection under unfavorable observation environments, such as low altitude angle and active ionospheric period of the satellites in each orbit of BeiDou, avoiding misjudgment and omission in each sampling rate as much as possible, significantly improving detection accuracy when the sampling rate is low and the multi-path error is large.

For the BDS/GPS integrated positioning system, each sub-system needs proper weight allocation, which, however, can be inaccurate due to the difference between systems. To improve the accuracy of the integrated positioning, this paper proposes an integrated positioning algorithm based on the robust Helmert-variance-component estimation, namely robust Helmert(RH). Firstly, we establish the Helmert-variance-component-posteriori estimation model. We distinguish the observed values with multi-type and different accuracy to achieve dynamic allocation of sub-system weight. Then, to adjust the weight of the observed values with gross errors, we construct an improved equivalent power function based on IGGⅢ scheme. This solves the convergence distortion problem in Helmert model owing to gross observation errors. Finally, by collecting and testing the actual data observed by the dual system, we verify the effectiveness and accuracy of the algorithm.

We propose a PWV estimation model considering the water vapor attenuation factor, and obtain the PWV by inputting the surface atmospheric water vapor pressure and the water vapor attenuation factor. We use the 1a observation data of 85 radiosonde stations and 7 IGS stations in China in 2018 to verify the accuracy of the new model. The results show that when there is a water vapor attenuation factor of the day, the model’s estimated PWV accuracy is about 2 mm; the water vapor attenuation factor at any position can also be obtained through GPT2w grid interpolation. The accuracy is equivalent to that of the traditional first-order polynomial model, but the newmodel has a wider range of action and stronger applicability.

In view of the complexity and poor performance during the real-time conversion process GNSS precipitable water vapor (PWV) in the southeast coastal area of China, based on the data of 18 GNSS stations in the area from 2017 to 2018, we analyze the linear relationship between GNSS-PWV and zenith tropospheric delay (ZTD), ground temperature (T_s) and ground atmospheric pressure (P_s). We apply the multiple linear fitting method to establish the direct conversion model of PWV,which provides a simple and effective method for predicting GNSS-PWV in the area. Experimental results show that GNSS-PWV has good correlation with ZTD, P_s and T_s, their related coefficients are 0.98, -0.65 and 0.78. The multi-factor GNSS-PWV model based on ZTD, P_s and T_s has the highest precision, and its RMS is 0.33 mm. It is much better than the single-factor PWV model based on the ZTD (RMS=4.66 mm). For the double-factor GNSS-PWV model based on ZTD, P_s has the second highest precision, and its RMS is 0.50 mm.

We use the relative calibration method, the four-component self-test analysis, the relative error calculation of M2 tidal factor, and the correlation comparison method with the observation data of TJ-2 volume borehole strain at the same station to calculate the internal accuracy of the instrument, the self-test situation, and the accuracy of M2 tidal factor before and after the probe replacement of RZB-2 borehole strain instrument at Zoucheng station from January 2020 to July 2021. We further comprehensively analyze the stability of the borehole component strain instrument observation system and the reliability of data quality. The analysis shows that after the sensor probe change, the data of RZB-2 borehole strain instrument at Zoucheng station have high precision. The two groups of surface strains conform to the four-component self-test equation, and the correlation coefficient is close to 1. The correlation between surface strain and volume borehole strain surface strain data is also close to 1, that the instrument observation system is normal and stable, and the data are reliable. However, due to the increase of a new factory workshop around the station in 2021, pumping interference increases, and the accuracy of the tidal factor of the data is poor, the average value is about 0.07. We suggest that environmental protection work should be actively done to reduce the pumping interference and improve the observation quality of the data.

To quantitatively study the similarities and differences between ground and borehole background noise, we set up a seismograph on the ground of the Taole Seismic Station and use borehole seismograph for simultaneous observation. Firstly, sample data are selected to calculate the power spectral density of background noise. We analyze the power spectral density and average noise RMS values of different seismograph at different depths. The near and far earthquakes recorded by three seismographs are compared and analyzed. The results show that the near seismic waveform recorded by surface broadband, short period and underground short period seismograph are relatively clear and easy to identify seismic phases. Moreover, the teleseism waveform recorded by surface broadband and underground short period seismograph is relatively clear, while the teleseism waveform recorded by surface short period seismograph is poor, the waveform distortion is serious, and identifying the seismic phase is difficult. At the same time, in analysis and comparison of the recorded seismic source spectrum, we find that the near earthquakes recorded by the borehole seismograph are seriously disturbed in the frequency band of 7 to 12 Hz compared with the surface seismograph. The recorded teleseisms are seriously disturbed in the frequency band of 4 to 7 Hz, so they need to be handled carefully in seismic phase analysis.

The electronic tilt sensor generally has a phenomenon of tilt drift and output data jump. Therefore, detecting the inclination drift law of the tilt sensor and the natural frequency of the data jump can effectively improve the data quality of the tilt sensor. Based on the working principle of high-resolution electronic tilt sensor HRTM, we analyze the factors affecting the tilt drift of the tilt sensor. We design the experimental platform for testing the natural frequency of the tilt sensor output data. We then propose a natural frequency detection method based on Fourier transform. Based on 10 h experimental data collected at constant temperature in the basement, the results show that the tilt output data of the sensor has obvious tilt drift and output data jump phenomenon, the tilt drift is highly linearly related to the internal temperature of the instrument, and the inherent frequency of the x-axis and y-axis jump is 2.773 Hz and 2.813 Hz, respectively.