To address the problems of complex and time-consuming computation of the diagonal matrix D, lower triangular matrix L, and transposed matrix L^T about LDL^T decomposed in the traditional least-square ambiguity decorrelation adjustment(LAMBDA), we propose the M-Cholesky decomposition method. The method uses the four-corner rule method to calculate each element of the synthesis matrix step by step, and uses at most three elements in each decomposition calculation, reducing storage space and improving computational efficiency. The simulation and experimental results show that the computational efficiency of the M-Cholesky decomposition method is about 15% better than that of the Cholesky decomposition method for solving the whole-period ambiguity of GNSS.

To verify the ambiguity fixing effect of the second generation SBAS experimental platform GATBP PPP service, we correct the satellite hardware delay using the signal deviation correction number associated with the observations, and fix the wide and narrow aisle ambiguities separately using the LAMBDA method to obtain the fixed single-difference ionosphere-free combined ambiguity. We solve the 7 d observations of 10 MGEX stations by PPP using GATBP precision products, and use the results of CODE precision products as reference. The results show that the mean fixed rate of ambiguity for wide and narrow lanes using GATBP products is 95% and 30.2%, respectively, the RMS of horizontal positioning error is better than 10 cm, the RMS of 3D positioning error is better than 15 cm, the RMS of phase observation residuals is better than 3 cm, and the RMS of tropospheric estimation error is better than 1.5 cm. The ambiguity fixed rates of both wide and narrow lanes are above 95% for CODE products, and the residuals of phase observation, positioning errors and tropospheric estimation errors are significantly smaller than those of the former.

This paper proposes an enhanced algorithm to estimate satellite DCB based on undifferenced and uncombined precise point positioning (UPPP). Our method uses UPPP and geometry-free linear combination of phase-smoothed range to extract slant ionosphere. Then we establish models to separate vertical ionospheric delay from ionospheric observations based on each single station, and assume that the sum of satellite DCBs is 0 to eliminate rank deficit. At last, DCBs and ionospheric delay are estimated simultaneously based on the weighted least squares adjustment method. However, because of the difference in the accuracy of the two types of ionospheric observations, the stochastic model suffers from the problem of weight inaccuracy. To overcome this problem, we introduce a variance component estimation method. The method iteratively adjusts the weight ratio of the two types of ionospheric observations based on the residual information, constructing a more reasonable stochastic model. The effectiveness of the algorithm is validated by observations from MGEX stations. Results show that, the satellite DCB estimation accuracy of the proposed method is 0.32 ns, which is 52.24% and 43.86% higher than that of the traditional geometry-free linear combination of phase-smoothed range method and the equal weight fusion model, respectively.

Using two-line element (TLE) ephemeris combined with SDP4 model, we predict the 30-day orbits of BDS IGSO, GEO and MEO satellites. We evaluate the orbit prediction errors using GFZ precise ephemeris. Then, we compare the orbit prediction errors of this method with the orbit prediction errors of broadcast ephemeris and almanac. The results show that the orbit prediction accuracy of our method is significantly better than that of broadcast ephemeris and almanac.

To solve the problem of divergence of Sage-Husa adaptive filtering algorithm in the integrated navigation system, we propose an improved Sage-Husa adaptive filtering algorithm. Firstly, we apply the compensation feedback correction algorithm to the linear Kalman filter algorithm to obtain the optimal estimation and simplify the innovation vector of filter. Then, we propose an improved, simple to implement, Sage-Husa adaptive filtering algorithm, avoiding divergence of state estimation. Finally, a GNSS/SINS integrated navigation experiment based on this algorithm shows that this algorithm can accurately track not only the variance of sudden change but also the variance of slow change of measurement noise, and the estimation accuracy of variance is equivalent to that of variational Bayesian method. Compared with the standard KF algorithm, the position accuracy and velocity accuracy can be improved by 20% and 21% respectively when the variance of measurement noise changes, thus effectively reducing the influence of unknown statistical characteristics of measurement noise on filtering accuracy.

The feasibility of GNSS in cis-lunar space is simulated and analyzed from three aspects based on the classical cis-lunar orbit and circumlunar orbit: number of visible satellites, PDOP value, and dynamic characteristics. The results show that it is appropriate to set the carrier-to-noise ratio threshold of the receiver on the cis-lunar orbit and circumlunar orbit to 12 dB-Hz. Based on this analysis, we propose the technical issues of using GNSS signals in cis-lunar space spacecraft.

We propose a new adjustment method for rationally using prior information. Firstly, we use the idea of balanced estimation to control the influence of inaccurate prior information on the parameter solution by giving the prior information a weight. Using the error magnitude of observation information and prior information in the residuals, we give the calculation method of balance factor. Finally, an example is given to illustrate the effectiveness of the algorithm.

Based on the laser energy attenuation model, we construct the influence coefficient of laser beam direction accuracy using the variation trend of laser beam direction accuracy under different scanning characteristics. Combined with the influence of ranging, angle measurement and light spot on the laser beam direction accuracy, we verify the accuracy of the constructed influence coefficient. Simultaneously, we give the laser scanning ranging accuracy after inversion.

We use the data from 51 GNSS stations, 116 meteorological stations, 2 radiosonde (RS) stations and ERA5 in Henan province spanning from UTC July 15 to 25, 2021. By combining GNSS derived atmospheric PWV with multiple meteorological elements, we analyze  the spatial and temporal variation characteristics and internal relationship of precipitation, PWV, relative humidity and total cloud cover during the “21·7” heavy rainfalls in Henan province. The results show that the correlation coefficient between GNSS-PWV and RS-PWV is more than 0.95, the absolute bias is less than 0.7 mm, and RMS is less than 2.9 mm. The absolute bias of GNSS-PWV and ERA5-PWV are less than 3 mm, and the RMS range is 2.5 to 4 mm. During the period of extremely heavy rainfalls, PWV firstly increased with the increase of precipitation and then decreased with the decrease of precipitation. At 4 to 6 h before the peak of the precipitation, PWV began to increase sharply to the extreme value, ΔPWV with a range of 7.35 to 9.02 mm. When the precipitation reached the extreme value, the PWV content over the GNSS stations also reached the peak value (above 70.58 mm). The spatial distribution of PWV, relative humidity, total cloud cover and precipitation are consistent with each other. The PWV content over central, northern and western regions with rainfalls in the province was constantly at a high level (above 65 mm), while the PWV content over the eastern region without rainfalls was lower (below 65 mm).

We propose a spherical harmonic analysis method to calculate the impact of non-tidal ocean loading in Guangdong and its surrounding areas using high temporal and spatial resolution sea level change data. This method can fully consider the global impact of the ocean and overcome the incompleteness caused by the direct use of Green’s function method for local integral calculation. We find that the impact of non-tidal ocean loading on ground elevation in Guangdong shows obvious regional characteristics, more than 10 mm along the coastline. The effect decreases as the distance from the ocean increases.

In order to solve the problem of abnormal disturbance of dynamic model and gross error of gravity anomaly observation data in SITAN algorithm for underwater gravity matching navigation, we propose a robust adaptive matching algorithm. We construct the adaptive factor and the robust factor to adjust the weight of the contribution of state prediction information and measurement information to the filtering, which can effectively suppress the influence of gross error and prediction information anomaly, and improve the stability and reliability of the matching algorithm. In the South China Sea, a sea area with rich variation of gravity anomaly features is selected for simulation experiments. The results show that compared with the ordinary SITAN algorithm, the proposed algorithm can significantly reduce the influence of abnormal observations, and the positioning accuracy and robustness of the system are improved.

Using the RMS of the difference between the observed data and the DDW tidal model and the gravity residuals as constraints, we improve the calibration method of gPhone gravimeter scale factor based on M₂ wave. Using the improved method, we calibrate 9 gPhone gravimeters in the continuous gravity network. We evaluate the accuracy of the calibration results based on the relative error of M₂ tidal wave and the RMS of the residual amplitude spectrum, and compare them with the FG5 comparison method and the M₂ tidal wave calibration method. The results show that the average relative error between the M₂ tidal wave tidal factor and the theoretical solid tide model is improved to 0.215 2%, and the RMS of the residual amplitude spectrum reaches 1.690 2 μGal under the improved method of this paper, which is comparable to the accuracy of the FG5 comparison method and better than the M₂ tidal wave calibration method. At last, we discuss the distribution regularity of the scale factor and the longitude dependence of the gravity residual vector on the location of the station.

The stray current generated by three subways in Guangzhou is monitored by an orthogonal electrode device, and its source is analyzed.Results show that the spontaneous potential generated by subway startup or braking exceeds 100 mV, while its background value is less than 3 mV at a distance of 1 km. The main period of spontaneous potential sequence is 119.2 s, and the other significant peaks are 51.16 s, 178.9 s and 358.5 s. The main energy of its power spectrum is below 200 s, which is significantly different from that without subway operation. Through template matching of the pulse details of the subway signal, we pick up the number of pulses close to the operating frequency of subways. Locating these pulses can identify the direction of source. We can develop filtering technology based on the data of this test and other in operation geomagnetic field and geomagnetic field data interfered by subway to effectively remove the stray current signals of geoelectric field and geomagnetic field.

To discuss the rapid identification of active faults using airborne LiDAR technology, the Tieluzi fault, which borders Weihe basin and Qinling mountains in Guanzhong, is taken as the practice object. After accurate extraction of topographic point cloud data, we give a 2 m resolution DEM, and construct the geometric distribution characteristics of the scanning area. The test results show that the method based on the high-resolution DEM data and the interpretation of fault characteristic troughs and triangles are helpful to improve the accuracy of the airborne LiDAR technology in the rapid identification of active faults in complex terrain. The airborne LiDAR provides a strong technical support for the rapid detection of active faults.

We investigate the calculation of seismic hazard with 120″, 60″, 30″, 15″, 9″ and 6″ grids respectively, and then interpolate the risk results of smaller grids. We compare the direct calculation of small interval results with those obtained by interpolation of larger spatial grids, and then evaluate the reliability, applicability and requirements of the results obtained by the larger spatial intervals. The results show that the calculated values of 30″ grid gives the results of various small grid results with an error of less than 1 Gal, which is a grid interval considering both accuracy and efficiency.

In view of the insufficient threshold estimation in the traditional wavelet threshold denoising algorithm, this paper proposes a new wavelet threshold estimation algorithm and applies it to the denoising of deformation monitoring data after deriving that the noise influence amplitude in the high-frequency signal is gradually reduced by half. Theoretical analysis and engineering examples show that this method can effectively overcome the constant deviation of the soft threshold method, and further improve the accuracy and reliability of denoising.

We introduce variational mode decomposition(VMD) and bidirectional long short-term memory (BiLSTM) neural network for dam deformation prediction research. Firstly, we use VMD to reduce the influence of nonlinearity and non-smoothness of the original dam data on the prediction results; secondly, we combine the hunter-prey optimizer(HPO) with the parameter optimization of BiLSTM to construct a dam deformation prediction model based on VMD-HPO-BiLSTM; finally, we compare the results of this model with LSTM, BiLSTM, and VMD-BiLSTM models using a hydroelectric dam as an example. The experimental results show that the VMD-HPO-BiLSTM model are 0.446 mm, 0.264 mm, and 18.593% in the three accuracy evaluation indexes of RMSE, MAE, and MAPE, respectively, which are better than other three models and has highest prediction accuracy.

The vertical pendulum tiltmeter and seismometer are used to observe the same point, the output signal of the sensor is sampled uniformly, and the data is quantified uniformly with the time domain background noise envelope as the standard. The amplitude spectrum is calculated, and the detection ability of the two instruments for different signals is judged according to the amplitude changes with frequency. On this basis, background noise, seismic wave and solid tide are used as natural test signals to test the detection capabilities of the two instruments. The results show that the amplitude recorded by the vertical pendulum tiltmeter is larger than that of the seismometer for the signal in the long period frequency band of solid tides, and the data recorded by the seismometer in this frequency band is mainly from the instrument noise. For far seismic surface waves, both of the two instruments have good recording effect, and the amplitude recorded by the vertical pendulum tiltmeter is greater than that of the seismometer. The high frequency ground motions caused by near and local earthquakes are outside the response passband of vertical pendulum tiltmeter design and the data are not very comparable.

The borehole strainmeter can be calibrated by seismometer through joint observation of teleseismic surface waves. The basic idea is to calibrate the shear strain sensitivity coefficient B with Love waveform first and then calibrate the surface strain sensitivity coefficient A with Rayleigh waveform. However, we find that the B and A calibrations of different teleseismic surface waves are quite different. Therefore, validity is introduced to indicate the degree to which the surface waves can be accurately calibrated to B and A values, and the calibration with the highest validity is used as the final calibration result. This method is applied to calibrate the borehole strainmeter of Renhe seismic station, and the results are as follows: A=0.36, with a validity of －0.71; B=0.73, with a validity of －0.92.

By comparing the observation curves of the seismometer and the borehole strainmeter at Renhe station, the Pg wave velocity of several local earthquakes is calculated by sliding analysis method. Considering the influence of reflection wave and the measurement validity, the Pg wave velocity of Renhe station is finally measured to be 5.98 km/s, which is very close to the P wave velocity of the upper crust of the Yunnan 2015 crustal velocity model. The work in this paper shows that the borehole strainmeter can be used as a seismic instrument after accurate calibration to make up for the deficiency of the seismometer in the observation of low frequency events.

Based on geological structure data, we use the finite element analysis to construct the geometric model of the Wuzhishan station. Combined with the GPS and gravity observation results, we analyze the physical significance of the trend change of the fixed-point deformation observation data of Wuzhishan station. The study found that under the action of NNW-directed pressure, the deformation amplitude of the upper part of the model is significantly larger than that of the lower part, and when the lower part continues to shorten, it is easy to form an upper arch. Because the tendency of the rock formation tends to the southeast, the deformation trend also tends to dip to the southeast. This is consistent with the changing background trend of the existing observational data.

We scan the coseismic response of Three Gorges well network to the global earthquakes of magnitude 7 and above during 2013 to 2022, evaluate the reflecting ability of each well in Three Gorges well network, analyze the influencing factors of the reflecting ability, and discuss the conditions for generating step type coseismic response. We take the 2008 Wenchuan 8.0 and the 2021 Maduo 7.4 earthquakes as examples, and use the coseismic response of well water level to inverse the stress adjustment in Three Gorges reservoir area. The results show that there is a positive correlation between seismic energy density and coseismic response amplitude; the threshold values of seismic energy density that can cause the water level of each well to produce coseismic response have an order of magnitude difference, which reflects the different earthquake reflecting capacities of each well; the seismic reflection capacity of the wells in Three Gorges well network are positively correlated with the hydraulic conductivity, and negatively correlated with the bulk compression modulus and the distance from the well to the fault; when the seismic energy density is greater than 10^-3 J/m³ and the regional tectonic stress accumulation is strong, the coseismic response of the well water level in Three Gorges well pattern is mainly stepped. Through inversion of the stress changes in Three Gorges well network after the Wenchuan 8.0 and Maduo 7.4 earthquakes, we find that the stress adjustment is obviously controlled by regional faults.