Improving GNSS-Based Tropospheric Delay Estimation for Airborne Quantum Gravimetry: First Results Using NWM Forecasting

Abstract

Airborne gravimetry is an alternative to traditional ground measurements for difficultly accessible areas like mountains or deserts and provides a valuable tool to link ground and satellite gravity measurements. In such applications, one of the main challenges is to separate the gravitational from the kinematic acceleration. As the vertical kinematic acceleration is the second derivative of the ellipsoidal height, it is typically estimated by GNSS-based positioning. Recent developments in the field of absolute quantum gravimetry will be part of the German Absolute Aero Quantengravimetrie (AeroQGrav) initiative. The AeroQGrav concept includes the fusion of multiple sensors to recover the gravity signal, aiming for resolutions of about 1 µm/s2 . The position of the gravimeter will mainly be estimated by a differential multi-frequency and multi-GNSS evaluation. Strongly correlated with the height of the computed position is the estimation of the tropospheric delay. Current differential GNSS positioning techniques rely on ground-based reference stations to provide accurate a priori tropospheric error estimates. These approaches are not sufficient for airborne gravity measurements as the altitude of reference stations and aircraft might differ significantly. Here, we suggest improving the a priori troposphere error information by utilizing numerical weather models (NWMs) with a high vertical resolution. A ray-tracing technique is implemented to compute the NWM-based satellite-dependent mapping functions to project the zenith delay onto the line of sight between the GNSS satellite and the receiver. The NWM-based tropospheric products are compared to GNSS-estimated delays. Finally, the benefits of adopting a NWM-based a priori model and NWM-enhanced mapping functions are evaluated in a precise point positioning (PPP) application. Results with and without using the a priori information and NWM-based mapping functions are compared and discussed in the context of the aircraft positioning for airborne quantum gravimetry to provide the rover algorithm with the best tropospheric delay possible.