Geodetic Network Design for Low-Cost GNSS

Abstract

Global Navigation Satellite System (GNSS) stations are a cornerstone of modern geodetic surveys, providing high temporal-frequency, sub-centimetric three-component measurements of surface displacement at fixed locations. However, the high cost of each instrument limits both spatial resolution and access for small-scale users.

Low-cost GNSS stations, in particular single-frequency instruments, provide a cheaper alternative to conventional systems, enabling the deployment of larger GNSS networks. Increased observation density around continental fault zones would improve our ability to recover distributed aseismic slip, in particular afterslip, on continental faults, which may be poorly constrained by other geodetic techniques such as InSAR.

To best recover aseismic slip using low-cost GNSS stations, a method for the estimation of optimal network layouts is required. For single-frequency GNSS stations, which present the greatest potential for increased GNSS network density, the reduced positional accuracy as a result of ionospheric delay must also be addressed.

In this work, a method for the automated design of single- and dual-frequency GNSS networks to recover distributed aseismic slip on continental faults is presented. Network layouts are generated using particle swarm optimisation and a criterion matrix technique to minimise the uncertainties on modelled slip values, relative to "best possible" values. These are estimated through non-uniform fault discretisation, in which a multi-objective genetic algorithm is utilised to explore the trade-off between the complexity of the discretisation and the associated model uncertainties. The reduced positional accuracy of single-frequency GNSS stations is mitigated through the network design, and an understanding of the spatial structure of the ionospheric delay.

Initial results demonstrate the potential of low-cost GNSS stations, in particular single-frequency GNSS stations, to recover distributed aseismic slip on continental faults. Future work should expand the methodology to included slip across multiple faults, and the generation of mixed GNSS networks.