Assessment of modernized GPS L5 SNR for ground-based multipath reflectometry applications

Abstract

Global Positioning System (GPS) signal-to-noise ratio (SNR) measurements can be employed to retrieve environmental variables in multipath reception conditions, whereby direct or line-of-sight transmission is received simultaneously with coherent reflections thereof. Previous GPS SNR multipath studies of soil moisture and snow depth have focused on the legacy GPS L1 and L2 bands. In the present paper, short-delay, near-grazing incidence multipath from the L5-band GPS SNR is assessed for its value in detecting soil moisture and snow depth. The L5 signal will become more important in the future because of compatibility and interoperability among the different global satellite navigation systems. The L5 results are compared with L2C estimates to determine whether the L2C-L5 differences (given their differing power budgets and their modulation properties) are significant. To address these questions, measurements and simulations were employed. A physically-based multipath simulator was enhanced to investigate the differences between parameter retrievals for the L2C and the L5 GPS signals. Parameter retrievals from synthetic observations for different scenarios were compared. Comparisons included varying reflector height, surface material, and surface roughness. Measurements from two UPS stations in Colorado, USA, were used to retrieve soil moisture and snow depth conditions. Over a 153-day period that encompassed the catastrophic 2013 Colorado flooding event, L2-derived volumetric soil moisture had an RMS difference of 0.042 cm(3)/cm(3) while the L5 RMS difference was 0.034 cm(3)/cm(3) with respect to in-situ data (values of volumetric soil moisture range between 0.04 and 0.34 cm(3)/cm(3)). In a separate 483-day campaign, L5-derived snow depth estimates were compared to L2C-derived values and found strongly correlated, deviating from a one-toone relationship by only 0.00001 +/- 0.0064 cm/cm. These results indicate the absence of any detectable biases in L5 as compared to L2C for retrieving soil moisture and snow depth from GPS SNR multipath observations.