Use of reflected GNSS SNR data to retrieve either soil moisture or vegetation height from a wheat crop

Abstract

Abstract. This work aims to estimate soil moisture and vegetation height from Global Navigation Satellite System (GNSS) Signal to Noise Ratio (SNR) data using direct and reflected signals by the land surface surrounding a ground-based antenna. Observations are collected from a rainfed wheat field in southwestern France. Surface soil moisture is retrieved based on SNR phases estimated by the Least Square Estimation method, assuming the relative antenna height is constant. It is found that vegetation growth breaks up the constant relative antenna height assumption. A vegetation-height retrieval algorithm is proposed using the SNR-dominant period (the peak period in the average power spectrum derived from a wavelet analysis of SNR). Soil moisture and vegetation height are retrieved at different time periods (before and after vegetation's significant growth in March). The retrievals are compared with two independent reference data sets: in situ observations of soil moisture and vegetation height, and numerical simulations of soil moisture, vegetation height and above-ground dry biomass from the ISBA (interactions between soil, biosphere and atmosphere) land surface model. Results show that changes in soil moisture mainly affect the multipath phase of the SNR data (assuming the relative antenna height is constant) with little change in the dominant period of the SNR data, whereas changes in vegetation height are more likely to modulate the SNR-dominant period. Surface volumetric soil moisture can be estimated (R2 = 0.74, RMSE = 0.009 m3 m−3) when the wheat is smaller than one wavelength (∼ 19 cm). The quality of the estimates markedly decreases when the vegetation height increases. This is because the reflected GNSS signal is less affected by the soil. When vegetation replaces soil as the dominant reflecting surface, a wavelet analysis provides an accurate estimation of the wheat crop height (R2 = 0.98, RMSE = 6.2 cm). The latter correlates with modeled above-ground dry biomass of the wheat from stem elongation to ripening. It is found that the vegetation height retrievals are sensitive to changes in plant height of at least one wavelength. A simple smoothing of the retrieved plant height allows an excellent matching to in situ observations, and to modeled above-ground dry biomass.

Deriving surface soil moisture from reflected GNSS signal observations from a grassland site in southwestern France

“…4. VSM retrievals from the Zhang et al (2017) method are closer to the in situ VSM observations at 5 cm depth than retrievals from the benchmark method. Day to day changes in VSM retrievals are also smaller using the Zhang et al (2017) method.…”

“…Normalizing ϕ time series ensures compatibility among different satellite tracks (Zhang et al, 2017). Here, ϕ is normalized with zero minimum in order to obtain the scaled wetness index (ϕindex) as the following:…”

“…The GNSS-IR technique allows relating the reflected signal to the characteristics of the reflecting surface and to retrieve geophysical variables. Over land, variables such as soil moisture, snow depth and vegetation parameters can be observed using this technique (Larson et al, 2008;Small et al, 2010;Larson and Nievinski, 2013;Wan et al, 2015;Larson, 2016;Roussel et al, 2016;Zhang et al, 2017). GNSS satellites emit active Lband microwave signals (between 1.2 and 1.6 GHz).…”

“…Small et al (2016) compared different algorithms of GNSS-IR soil moisture retrieval in 10 the presence of vegetation. Zhang et al (2017) used the GNSS-IR technique for a wheat field throughout the growth and senescence period, from January to July 2015. They showed that VSM could not be retrieved when the vegetation canopy is too dense, i.e.…”

“…SNR analysis using the GNSS-IR technique is used to retrieve VSM over a field covered with grass using two approaches: the method proposed by 30 Chew et al (2016) and the normalization method based on the newly-established scaled wetness index proposed by Zhang et al (2017). The retrievals from both methods are compared.…”