Towards Estimation of Seasonal Water Dynamics of Winter Wheat from Ground-Based L-Band Radiometry

Abstract

Abstract. The vegetation optical depth (VOD) parameter contains information on plant water content and biomass, and can be estimated alongside soil moisture from currently operating satellite radiometer missions, such as SMOS (ESA) and SMAP (NASA). The estimation of water fluxes, such as plant water uptake (PWU) and transpiration rate (TR), from these Earth system parameters (VOD, soil moisture) requires assessing potential (suction tension) gradients of water and flow resistances in the soil, the vegetation and the atmosphere, yet it remains an elusive challenge especially on global scale. Here, we used a field-scale experiment to test mechanistic models for the estimation of seasonal water fluxes (PWU and TR) of a winter wheat stand including measurements of soil moisture, VOD, and relative air humidity (RH) under a controlled environment. We utilized microwave L-band observations from a tower-based radiometer to estimate VOD of a wheat stand during the 2017 growing season at the Selhausen laboratory in Germany. From VOD, we first extracted the gravimetric moisture of vegetation and then determined subsequently the relative water content (RWC) and the vegetation water potential (VWP) of the wheat field. Although the relative water content could directly be estimated from VOD, our results indicate this may be problematic for the phenological phases, when rapid biomass and plant structure development take place in the wheat canopy. The water uptake from the soil to the wheat plants was estimated from the difference between the soil and vegetation potentials divided by flow resistance from soil into wheat plants. The transpiration rate from the wheat plants into the atmosphere was obtained from the difference between the vegetation and atmosphere potentials divided by flow resistances from plants to the atmosphere. For this, the required soil matric potential (SMP), the vapor pressure deficit and the flow resistances were obtained from on-site observations of soil, plant and atmosphere and simple mechanistic models. This pathfinder study shows that the L-band microwave radiation contains valuable information on vegetation water status that enables the estimation of water dynamics (up to fluxes) from the soil via wheat plants into the atmosphere, when combined with additional information of soil and atmosphere water content. Still, assumptions when estimating the vegetation water potential from relative water content as well as when estimating the water flow resistances between soil, wheat plants and atmosphere had to be made. Moreover, validation of water flux estimates for assessing their absolute accuracy could not be performed due to a lack of in situ PWU and TR measurements. Nonetheless, our estimates of water status, potentials and fluxes show the expected temporal dynamics and intercompare reasonably well in absolute terms, providing confidence in further developing the proposed approach. Our findings support that passive microwave remote sensing techniques allow for the estimation of vegetation water dynamics next to traditionally measured stand-scale or plot-scale techniques. This might shed light on the potential capabilities of monitoring water dynamics in the soil-plant-atmosphere system using wide-area, remote sensing-based Earth observation data.