Dual-band information helps constrain canopy water content and temperature retrievals from vegetation optical depth

Abstract. Vegetation canopy water content and temperature fundamentally control plant physiology. They usually have strong diurnal cycles with substantial variations on the hourly basis. However, due to the scarcity of high-frequency observations, a major obstacle has long been existing in deciphering the plant hydrodynamics. Vegetation Optical Depth (VOD) provides a promising satellite and ground-based proxy by quantifying microwave attenuation by canopy liquid water, but it co-varies with biomass and temperature. Temperature-dependent dielectric effects also complicate the isolation of water dynamics. To assess the information extent that can be inferred from VOD and to assess the contribution of a second wavelength (as available with Global Navigation Satellite System, GNSS) in disentangling influences from multiple processes, we developed a Bayesian inversion framework coupled to a semi-empirical electromagnetic model to retrieve canopy water content (M_g), vegetation volume, and canopy temperature from single- and dual-frequency data using both theoretical sensitivity analyses and GNSS observations. After testing with simulated VOD at two GNSS bands (1.575 and 1.176 GHz), we applied the method to in situ GNSS-based VOD and evaluated against field measurements. Simulations indicate that single-band VOD with 0.5–4% measurement error already constrains M_g and vegetation volume. Adding a second band substantially improves temperature retrievals, reducing posterior uncertainty to 2.3K, although the benefit declines rapidly as VOD errors grow. When applied to observations, incorporating scaled leaf area index as an additional constraint improves the M_g-leaf water potential relationship, underscoring the value of auxiliary information. Through a first direct comparison of the information provided by mono-band and dual-band, our results show that converting VOD into biophysical variables is strengthened by dual-frequency information and complementary constraints, emphasizing the need for additional validation to robustly retrieve canopy water content and temperature at high frequency.