Spaceborne GNSS reflectometry: remote sensing of ocean and atmosphere
[article]
Milad Asgarimehr, Technische Universität Berlin, Jens Wickert
2020
Global Navigation Satellite System Reflectometry (GNSS-R) is a novel remote sensing technique that exploits the GNSS signals after being reflected off the Earth's surface. Monitoring of ocean wind is one of the applications and the main objective of recently launched satellite missions. This thesis aims at the development and characterization of these geophysical data products. Using the UK TechDemoSat-1 (TDS-1) measurements, a GNSS-R wind speed dataset is developed. The resulting data are
more »
... ated in comparison to those obtained from the Advanced Scatterometer (ASCAT). Wind field estimates of European Centre for Medium-range Weather Forecasts (ECMWF) reanalysis-Interim (ERA-Interim) and in situ observations from the Tropical Atmosphere Ocean (TAO) buoy array in the Pacific are taken as reference. The evaluation using ECMWF winds results in a root-mean-squared error (RMSE) and bias of 2.77 and -0.33 m/s, being comparable to those derived from ASCAT estimates, as large as 2.31 and 0.25 m/s, respectively. The derived winds show a higher level of robustness against rain with an RMSE and bias of 2.94 and -0.21 m/s over oceans under precipitations, in comparison to those obtained from ASCAT measurements, which are 3.16 and 1.03 m/s, respectively. Nonetheless, the BRCS reduces to lower values during rain events at low wind speeds. The signal attenuation by rain is investigated simulating GNSS-R delay-Doppler maps at different rain rates and reflection geometries. It is shown that the resulting bias is smaller than approximately 0.35 m/s (1%) at a wind speed of 30 m/s and an incidence angle of 30 degrees. At the same wind speed and incidence angle, the examination reports that a continuous rain at every cell of the signal propagation path, at rates of 10, 15 and 20 mm/h, could lead to overestimation of wind speed not larger than 0.65 m/s (2%), 1.00 m/s (3%), and 1.3 m/s (4%), respectively. It is concluded that rain attenuation is ignorable within the current GNSS-R applications. Despite a commonly made conclusion that [...]
doi:10.14279/depositonce-10974
fatcat:zkwvkzac6rcx3mdpbaeirfppy4