Multi-scale volumetric soil moisture detection from GNSS SNR data: Ground-based and airborne applications

As multipath effects still represent a major problem for reaching precise GNSS positioning, the mitigation of their influence has been widely investigated. Previous studies have lately proposed to use these interferences of GNSS electromagnetic waves to estimate parameters related to the reflecting surface (e.g., antenna heights, roughness...). Variations in the nature of the surface are likely to modify the properties of the reflected waves, and consequently lead to fluctuations of amplitude, phase and frequency of the signal-to-noise ratio (SNR), e.g., recorded at 1 Hz by a GNSS receiver. By analyzing the time variations of these three SNR metrics, linked to the dielectric constant of the surrounding soil, it is possible to recover the local fluctuations of the soil moisture content. The aim of our study is to validate this SNR technique in various conditions of reception and reflection. Three different configurations are under study. First, we use a static geodetic Leica AR10 antenna set up within an agricultural plot with rich clay content. Correlation between our estimates and independent soil moisture records reaches 0.95 during the 1.5 month period of measurements. The experiment is then repeated under the same conditions of reception but on another test site, presenting different conditions of reflection, namely, the surrounding soil is only sand. This leads to drastic changes of the penetration depth of the GNSS waves, and therefore different behavior of the three SNR metrics. Last, a low-cost device and airborne solution is tested for the first time, with a U-Blox antenna embedded on a UAV (Unmanned Aerial Vehicle), flying over a clay field.