Search

Orthometric height datum systems are critical in the geodetic community. Nonetheless, individual nations or local regions maintain their distinct height systems. These systems, primarily determined by conventional spirit leveling, exhibit inhomogeneity, as the mean sea level does not represent an equi-geopotential surface. As such, accurately unifying global height systems is a persistent challenge. The gravity frequency shift equation provides a promising approach to this issue, offering a means to unify global height systems by comparing the frequencies of two distant clocks via optical fibers. In this study, we develop a model based on the Optical Fiber Frequency Transfer (OFFT) method, recognized for its unparalleled precision. We simulated a network of optical clocks interconnected by optical fibers to unify the height system across Southeast Asia. Our study suggests that the optical fiber clock network can realize height propagation between any two height datums at the centimeter level, providing a favorable opportunity for achieving regional height system unification.

Based on the general relativity theory, the geopotential difference can be determined by comparing the change in time difference between precise clocks using the precise point positioning (PPP) time transfer technique, referred to as the relativistic PPP time comparison approach. We focused on high-precision time comparison between two high-performance clocks for determining the geopotential difference using this approach, and conducted the simulation experiments to validate this approach. In the experiment, we consider three cases for evaluating the performance of this approach with different types of atomic clocks, namely, the fractional frequency stabilities of the clocks equipped at three selected ground stations (BRUX, OPMT, and PTBB) are $4.0 \times 10^{-13} / \sqrt {\tau }$ (Case 1), $2.3 \times 10^{-14} / \sqrt {\tau }$ (Case 2), and $2.8 \times 10^{-15} / \sqrt {\tau }$ (Case 3) at averaging time ${\tau }$ , respectively, and the accuracy of these clock have been evaluated to be $5.3 \times 10^{-16}$ , $7.8 \times 10^{-17}$ , and $8.6 \times 10^{-18}$ . Two main conclusions can be drawn from the experimental results. First, high-performance clocks can significantly improve the precision for GNSS PPP time transfer. Compared to Case 1, the long-term stabilities of the time link OPMT-BRUX as well as PTBB-BRUX are improved in Cases 2 and 3. Second, the geopotential difference between any two stations can be determined at the decimeter level, and the accuracy of geopotential difference is consistent with the stabilities of the time links in Cases 1-3. In Case 3, the determined geopotential differences between OPMT and BRUX deviate from the EIGEN-6C4 model values by -0.64, m2/2 with an uncertainty of 1.11 m2/s2, whereas the deviation error between PTBB and BRUX is 0.76 m2/s2 with an uncertainty of 1.79 m2/s2. The results of this study suggest that a one decimeter-level geopotential difference between two arbitrary stations can be determined based on this approach.

Recent studies have shown that a super-resolution generative adversarial network (SRGAN) can significantly improve the quality of single-image super-resolution. However, existing SRGAN methods also have certain drawbacks, such as an insufficient feature utilization, a large number of parameters. To further enhance the visual quality, we thoroughly studied three key components of SRGAN, i.e., the network architecture, adversarial loss, and perceptual loss, and propose a DenseNet with Residual-in-Residual Bottleneck Block (RRBB), called a residual bottleneck dense network (RBDN), for single-image super-resolution. First, to improve the utilization of features between the various layers of the network, we adopted a dense cascading connection between layers. At the same time, to reduce the computational cost, we added a bottleneck structure to each layer, greatly reducing the number of network parameters and accelerating the convergence speed of the training process. Second, the proposed RRBB, as the basic network building unit, removes the batch normalization (BN) layer and employs the ELU function to reduce the opposite effects in the absence of BN. In addition, we applied an improved overall loss function during the model training process to stably train the model and further improve the realism of the reconstructed high-resolution image. To prove the superiority of our proposed model, we conducted a comprehensive and objective evaluation of the Peak Signal-to-Noise Ratio, structural similarity, learned perceptual image patch similarity, and other evaluation indicators obtained from the three test sets, i.e., Set5, Set14, and BSD100, from the recent state-of-the-art model. Finally, we conducted qualitative and quantitative analyses of the results obtained in terms of the evaluation indicators, the authenticity of the restored HR images, and textural details, which show the superiority of the RBDN model.