Abstract
Abstract. Altimetry by using GNSS (Global Navigation Satellite Systems) reflectometry is regarded as a new promising technique. One important step to utilize this technique is modeling the interferometric path (the difference between the direct and the reflected signal paths). This paper evaluates three models: the planar model, the planar model with tropospheric correction and a model based on ray-tracing. If decimeter level accuracy for water surface fluctuation is required, the planar model cannot be used when a receiver is at an altitude of a few hundred meters and observations are taken at low elevation angles. On the other hand, depending on the mapping function and zenith total delay in the tropospheric correction, the planar model with tropospheric correction can provide decimeter level accuracy for low altitude stations. If simulated observations with a 1 cm accuracy have been employed to estimate Sea Surface Height (SSH) by the model based on ray-tracing the numerical results present 1 cm as RMSE for phase retrieval and 5 cm for Doppler retrieval. The planar model with tropospheric correction does not yield RMSE better than some decimeters for the same condition.
Highlights
Sea level fluctuations have been monitored for a very long time
Three interferometric path models, the planar model, the planar model with tropospheric correction and the ray-tracing model are studied at a station in North Greenland
Interferometric path lengths are computed and noise is added to mimic observation errors
Summary
Sea level fluctuations have been monitored for a very long time. Many techniques and instruments have been employed and each one has its own benefits. The potential of water level altimetry by using GNSS signals that are reflected off the water surface was first proposed by (Martin-Neira 1993) and introduced as a new bi-static technique that is called GNSS Reflectometry (GNSS-R). In GNSS phase altimetry for the estimation of instantaneous sea level height, an accurate interferometric path model is a prerequisite (Martin-Neira, Caparrini et al 2001). The Doppler retrieval does have no ambiguity but is sensitive to noises Both simulated phase and Doppler observations are used in this study where, we make use of simulated observations that are generated by the raytracing with adding random noise.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
