15 January 2025, Volume 45 Issue 1
-
Select all|
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 1-4, 12. https://doi.org/10.14075/j.jgg.2024.03.090
Based on the images from Gaofen-1 satellite, taking the first section of the Three Gorges reservoir as an example, 160 landslide samples are visually interpreted and divided into training and validation samples according to the ratio of 9:1. Then, landslide automatic identification models are constructed using Mask R-CNN and Faster R-CNN algorithms. To further compare and analyze the performance of the two models under different sample ratios, calculations are performed using sample ratios of 8:2, 7:3, and 6:4. The results show that the recognition results based on Mask R-CNN model are better than Faster R-CNN in three indicators, including precision, recall and F1 score. Moreover, cross-validation proves that the performance of Mask R-CNN model is more stable.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 5-12. https://doi.org/10.14075/j.jgg.2024.02.071
In this paper, we collect 1 Hz high-frequency data from multiple GNSS ground reference stations near the epicenter of the Jishishan earthquake on December 18, 2023, and use PRIDE PPP-AR and TRACK software for dynamic PPP calculation and high-precision static data processing. The coseismic deformation displacement fluctuation and permanent deformation displacement of the GNSS reference stations are obtained. Then, we analyze the deformation impact of the earthquake on the GNSS ground reference station at the epicenter. The results show that PRIDE PPP-AR and TRACK software can both calculate the coseismic deformation displacement impact of the earthquake on the reference station. The static calculation results show that the earthquake caused a permanent deformation displacement of about 18 mm in the horizontal direction at LXJS station (5.3 km away from the epicenter), and the impact range of this earthquake on the stability of the GNSS reference station is about 50 km. The rapid deformation analysis method and results can provide reference and data support for disaster prevention and mitigation, as well as real-time navigation and positioning services.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 13-21. https://doi.org/10.14075/j.jgg.2024.03.094
To study the activity characteristics and patterns of the landslide, this paper uses stacked interferometric synthetic aperture radar (InSAR) based on ascending and descending track C-band Sentinel-1A data to obtain the deformation rates of the Cheyuping landslide from January 2020 to December 2021 along the slope direction and in the vertical direction. We obtain the landslide displacement time series evolution results. The results show that the maximum deformation rate of Cheyuping landslide is more than -90 mm/a along the slope direction and more than -40 mm/a in the vertical direction, and the deformation of the landslide is larger at the leading edge and middle part of the landslide, and smaller at the trailing edge, which is characteristic of the slip-type landslide. According to the analysis of rainfall data and time series results, rainfall has a certain influence on the landslide movement, showing seasonal accelerated deformation characteristics. At the same time, the geological and stratigraphic lithology of the landslide area and the change of the water level of the Lancang river also accelerated the landslide activities to a certain extent. This study can provide technical reference for the interpretation of landslide deformation mechanism and disaster monitoring and early warning in the canyon reservoir area.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 22-28, 35. https://doi.org/10.14075/j.jgg.2024.02.072
The water level change series of lakes in the Qinghai-Xizang plateau from 2003 to 2022 are extracted using the data of ICESat, Cryosat-2 and ICESat-2. Taking Qinghai lake as an example, the three retracking algorithms of Cryosat-2 LRM(Ocean-CFI, UCL Land-ice and OCOG) are compared and analyzed, and the results show that OCOG algorithm is the best. The accuracy of the time series of lake water level changes extracted from satellite altimetry data is evaluated by using the hydrological station data of Qinghai lake, Nam Co, Mêmar Co and Lumaqangdong Co, and the results show that the RMS of the difference between the water level change time series of Qinghai lake and the measured results of Xiashe station is only 0.092 m, and the correlation coefficient between the water level change time series and the measured results of Nam Co is 0.989, and the accuracy analysis of different measurement techniques in the same time period shows that the accuracy of the Cryosat-2 SARIn measurement mode is slightly higher than that of ICESat-2. The water level changes of 106 lakes in the Qinghai-Xizang plateau from 2003 to 2022 are extracted, and the water level change trend is analyzed, and the results show that the water level of 19 of the 106 lakes show a downward trend, and the water level of 83 lakes show an upward trend.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 29-35. https://doi.org/10.14075/j.jgg.2024.02.081
The road in the open-pit mining area is narrow, with many curves, rubbles, and bumpy roads. The special surface closed loop leads to system model anomalies and measurement noise uncertainty problems during positioning of transport vehicles. To address the above issues, we propose an SVD-AUKF algorithm with a noise estimator. The algorithm suppresses the pathological matrix that occurs during the filtering process by singular value decomposition of the covariance matrix, estimates the measurement noise using the Sage-Husa estimator, reduces the system state error, and introduces an adaptive factor to correct the system model error in real time to suppress the filtering divergence. We apply and test the proposed SVD-AUKF algorithm to the GPS/INS integrated navigation system model on a vehicle in a mining environment. The results show that the new algorithm has good applicability, with position accuracy improvements of approximately 50.69%, 51.20%, and 48.54% compared to the AUKF algorithm; speed accuracy improvements of approximately 31.24%, 33.83%, and 31.30% in E, N, U directions, which can effectively improve the filtering performance and increase system positioning accuracy.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 36-40. https://doi.org/10.14075/j.jgg.2024.02.083
By utilizing satellite time-monitoring data to mitigate the impact of satellite-side and transmission path errors in standard single-point positioning, the timing stability and positioning accuracy at the user end are improved. At the same time, the timing deviation of the user relative to the satellite navigation system is corrected to be relative to the national standard time. Taking GPS as an example for a short baseline experiment, the results show that in terms of timing, the mean timing deviation of the L1 frequency band was improved from -42.8 ns to 5.4 ns, and from -53.6 ns to 2.8 ns for the L2 frequency band; in terms of positioning, the mean positioning errors in the X, Y, and Z directions for the L1 frequency band were improved by more than 55%, and by more than 77% for the L2 frequency band. This proves that by using satellite time-monitoring data, users of the satellite navigation system can not only enhance the pseudo-code positioning accuracy to the sub-meter level but also obtain the deviation relative to the national standard time, achieving ns-level timekeeping of the national standard time.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 41-47. https://doi.org/10.14075/j.jgg.2024.02.053
In response to the imprecise localization of moving radiation sources in complex scenarios and issues related to the accuracy of the observation station's position and the geometric relationship with the target using the two stage weighted least squares(TSWLS) algorithm, we propose a passive radar centroid passive localization algorithm for moving targets based on Taylor expansion and the joint application of TSWLS. This method utilizes a TSWLS algorithm to determine the target's position and velocity. The obtained target parameters serve as initial values for Taylor expansion, constructing the positioning error equation. Through iterative optimization, the algorithm refines and solves for optimal target parameters. The proposed algorithm and TSWLS independently provide estimations, with the final result obtained through centroid localization. Simulation experiments show that, regardless of target speed, in comparison with traditional methods, the presented algorithm significantly improves robustness and positioning accuracy while reducing sensitivity to geometric relationships between the observation station and target.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 48-54, 59. https://doi.org/10.14075/j.jgg.2024.02.076
On August 23, 2023, a M4.6 earthquake hit the Pulandian, Liaoning province. To describe the source characteristics of this earthquake, we explore its incubation and seismogenic mechanisms. In this paper we determined the focal depth of the Pulandian M4.6 earthquake, and calculate the focal mechanism solution, moment tensor solution and centroid depth of the earthquake, obtain the center of focal mechanisms of the earthquake. In the same, we analyze the relationship between focal mechanism and tectonic stress field, and fit the fault plane according to the results of small earthquake relocation. We preliminarily determine the seismogenic fault plane of this earthquake. The results show that the initial rupture depth of the Pulandian M4.6 earthquake is 12.0 km. The focal mechanism solution is strike: 50°, dip: 75° and rake: －169° for nodal planeⅠ, strike: 317°, dip: 80°and rake: －15° for nodal plane Ⅱ, the moment magnitude is MW4.8, while the centroid is 12 km. The seismic moment M0 is 1.796×1016 Nm, the moment tensor solution is Mrr: －0.004, Mtt: 0.946, Mpp: －0.942, Mrt: 0.017, Mrp: －0.305, Mtp: －0.125. The center of focal mechanisms solution is strike: 47.03°, dip: 79.04° and rake: －168.15° for nodal planeⅠ, strike: 314.75°, dip: 78.37°and rake: －11.19°for nodal plane Ⅱ. The relative shear stress and relative normal stress generated by the tectonic stress field acting on the nodal plane Ⅰ are 0.877 and －0.544; on nodal plane Ⅱ they are 0.911 and 0.161. The fitted seismogenic fault plane is strike: 148.91°, dip: 89.85° and rake: 26.47°. The analysis shows that the Pulandian M4.6 earthquake occurred on the Pulandian-Changhai NW tectonic belt. It is a natural earthquake that after stress accumulation occurs along the optimal nodal plane of the stress field and takes left-lateral strike-slip as the dislocation mode.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 55-59. https://doi.org/10.14075/j.jgg.2024.03.089
Based on the seismic ambient noise cross-correlation method, we calculate the relative velocity variation within the underground medium. We utilize continuous waveform data from the CCH and BDS stations at Dongtan mine, spanning from 2 a.m. on February 8, 2022, to 7 a.m. on March 8, 2023. The analysis focuses on the correlation between these velocity changes and both the mining operation and seismic activity at the 63up06 working face. The findings reveal a notable impact of the mining operation at the 63up06 working face on the wave velocity within the underground medium. Additionally, a clear correlation exists between the changes in wave velocity and the incidence of mine earthquakes, highlighting the interplay between variations in underground stress levels and seismic events. Furthermore, the release of underground stress during mine earthquakes can lead to slight fluctuations in wave velocity.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 60-65, 71. https://doi.org/10.14075/j.jgg.2024.03.100
Based on the measured ground motion data of the MS4.5 earthquake in Luding, Sichuan in 2023, we conduct an asymmetric slope time-frequency domain data analysis. The results show that: 1) The PGA amplification factor at monitoring point 1# on the right bank is highest, with horizontal and vertical values of 3.16 to 15.08 and 4.37, respectively. 2) The maximum Arias intensity on the right bank is 2.24 cm/s, and the maximum Arias intensity on the left bank is 0.79 cm/s, indicating that the seismic energy on the right bank is stronger. 3) The dominant frequency distribution of the monitoring points on the right bank of the slope is 2 to 9 Hz, while the dominant frequency on the left bank is 2 to 17 Hz. The energy presented by the seismic amplitude is more concentrated in the horizontal than in the vertical direction. On the right bank of the asymmetric high steep valley slope, the protruding peninsula shaped slope terrain has a stronger response than the straight shaped slope.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 66-71. https://doi.org/10.14075/j.jgg.2024.03.123
Utilizing the envelope cross-correlation method, we localize the very low frequency earthquakes(VLFEs) before the Jishishan MS6.2 earthquake. We conduct preliminary localization over a larger area, followed by precise localization within a smaller area. Analysis by increasing or decreasing the number of stations in different directions from the epicenter reveals that an even distribution of stations in all directions is crucial, effectively reducing localization errors. The localization results indicate that the VLFEs before the Jishishan earthquake were primarily distributed near the fault zone, with their distribution orientation is consistent with the fault direction.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 72-79. https://doi.org/10.14075/j.jgg.2024.03.097
Aiming at the problem of missing data and gross errors in GRACE Level1B observations, we propose an optimization method to complete the missing data of SCA1B. At the same time, we use an optimization strategy to eliminate gross errors for KBR1B and kinematic orbit data, then use these data sets to invert the time-varying gravity field model. In addition, based on synthesized data and real data respectively, we analyze the impact of the error from ACC1B in the Y direction on the inversion results and propose an optimization correction strategy. The optimized completion method can recover the missing data of SCA1B effectively, and fully considers the variational characteristics of the observations throughout the arc. The accuracy of the result derived from ACC1B data after correction using the optimization strategy is 3.7 mm, which is higher than the result derived from the uncorrected data. It can be seen from the results that the overall accuracy of the gravity field model solved by the data processed using the optimization method is like that of the three official centers. However, the detailed signals performance of the results from different institutions in regionals are different. It shows that the optimized data preprocessing strategy is effective and feasible.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 80-85. https://doi.org/10.14075/j.jgg.2024.03.092
We calculate local terrain and isostatic correction using spectral combination and nested different topographic resolutions and determine the free air gravity anomaly grid by the gravity reduction method. The results show that the accuracy of terrain correction calculated by the spectral combination method is the best in the test area, and the accuracy of isostatic correction calculated by nested different resolutions in the central and edge regions is best. When the integral radius is 1° and 2°, there is little difference between the local terrain correction, and the isostatic correction results differ significantly. The accuracy of free air gravity anomaly grid results obtained from the complete Bouguer and isostatic gravity anomalies by gravity reduction is the same, and the accuracy of the results is much higher than that obtained from the direct grid of free air gravity anomalies at ground points.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 86-90. https://doi.org/10.14075/j.jgg.2024.02.085
Large earthquakes can simultaneously excite numerous normal modes of Earth's free oscillations. Due to the Earth's ellipticity, rotation, and internal anisotropy, these normal modes undergo splitting, with frequencies of splitting modes becoming very close(only a few μHz apart). This imposes higher demands on the detection of Earth's free oscillation modes. This article derives and verifies a new method for detecting free oscillation modes based on normal time-frequency transform(NTFT). Taking the detection of 3S1 splitting modes as an example, compared with the classical FT spectrum method and the latest OSE method, our method has higher frequency resolution for detecting free oscillation modes.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 91-96. https://doi.org/10.14075/j.jgg.2024.02.073
Based on the observation data from regional fixed geomagnetic stations, this study quantitatively investigates the large-scale geomagnetic daily variation anomalies that occurred in the Sichuan-Yunnan region before the Yangbi MS6.4 earthquake on May 21, 2021. This article is based on Biot-Savart's law and the Monte-Carlo model to invert the distribution state and depth of underground current. The results show that: 1) The spatial distribution position and depth of underground current are consistent with the distribution of underground electrical high conductors, which is consistent with the crustal current position given by previous studies. 2) There is non-uniformity in both the horizontal and vertical interfaces below the research area, and the distribution of surface currents may be related to tectonic activity. 3) The upwelling of deep thermal fluids may be one of the reasons for the concentrated distribution of underground currents.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 97-103. https://doi.org/10.14075/j.jgg.2024.03.124
Tonga volcano violently erupted at 12:26:30 on January 15, 2022. The atmospheric pressure observation station of Yunnan Geophysical Network clearly recorded the intense pressure change process caused by the eruption. The atmospheric pressure change forms and characteristics of all observation stations were highly consistent, with the average peak amplitude of 2.16 hPa and the average propagation velocity of 314.30 m/s. The wave train has the properties of Lamb wave propagation, and the atmospheric pressure observation data are basically reliable. The water level observation in Yunnan deep well clearly recorded the water level change caused by the short-time pressure change. The static pressure efficiency of the observation well was estimated by the ratio of peak water level change to the amplitude of pressure change. We analyze by regression the short-time pressure change and well water level data, and take the determination coefficient, regression line slope and static pressure efficiency as indicators. Then we analyze the bearing property and stress-strain sensitivity of the borehole water-bearing system. The results show that the water level and pressure observation data of 25 wells, such as Kunming station, have high goodness of fit, small lag time, and good permeability of aquifer. The atmospheric pressure efficiency of 7 wells in Kunming station, Daguanxijiagou, Tengchong station, Ludianciyuan, Nanhua, Shiping and Zhaotong SK3 is slightly higher, and the sensitivity of well water level to stress and strain may be low. The larger the coefficient of determination for an observation well, the higher the probability that its monitoring and forecasting performance will be assessed as qualified.
-
Journal of Geodesy and Geodynamics. 2025, 45(1): 104-110. https://doi.org/10.14075/j.jgg.2024.03.125
In this paper, after preprocessing and various corrections of the Zigui station gPhone gravimeter data for the whole year of 2017, we use four methods to carry out a comparative study of drift correction including polynomial correction order selection, drift removal effect evaluation of each method, whether the BEADS algorithm is effective and other issues. The research results show that when the polynomial fitting method is used to remove the drift problem of gPhone gravimeter data, it will become stable when the order reaches about fourth and fifth orders. For general research, it is recommended to use fifth-order polynomials to remove drift, which can meet data processing requirements. When filtering to remove drift, we need to pay attention to the marginal effect. The EEMD method is highly operable and has more advantages than wavelet filtering. The BEADS algorithm can realize the function of fitting trend terms to remove drift, but can it be used in various studies? This requires further research. The research results of this article can provide an effective reference for drift correction of related spring-type gravimeters.
