Investigating Diurnal and Seasonal Cycles of Vegetation Optical Depth Retrieved From GNSS Signals in a Broadleaf Forest

Abstract

Vegetation Optical Depth (VOD) has emerged as a valuable metric to quantify water stress on vegetation's carbon uptake from a remote sensing perspective. However, existing spaceborne microwave remote sensing platforms face limitations in capturing the diurnal VOD variations and global products lack site‐level validation against plant physiology. To address these challenges, we leveraged the Global Navigation Satellite System (GNSS) L‐band microwave signal, measuring its attenuation by the canopy of a temperate broadleaf forest using a pair of GNSS receivers. This approach allowed us to collect continuous VOD observations at a sub‐hourly scale. We found a significant seasonal‐scale correlation between VOD and leaf water potential. The VOD diurnal amplitude is affected by soil moisture, plant transpiration and leaf surface water. Additionally, VOD can help independently estimate plant transpiration. Our findings pave the way for a deeper understanding of response of the vegetation to water stress at finer temporal scales.

Emerging Methods to Validate Remotely Sensed Vegetation Water Content

“…These GNSS VOD dynamics have been shown to correlate well seasonally with leaf water potential in a forest in Ozark, MO, USA (Yao et al, 2024) and diurnally with relative leaf water content in a forest in Pasadena, CA, USA (Humphrey & Frankenberg, 2023). Similarly, a GNSS sensor was mounted on an unmanned ground vehicle which traversed underneath pine and oak tree canopies, and the spatial distribution of the computed VOD matched that of LiDAR vegetation height retrievals from an aircraft (Ghosh et al, 2024).…”

“…First, GNSS can be used to directly measure VOD by positioning a GNSS sensor under a vegetation canopy and another above the canopy (Ghosh et al, 2024;Humphrey & Frankenberg, 2023;Yao et al, 2024). GNSS sensors take advantage of receiving L-band (1-2 GHz, like SMAP and SMOS satellites) signals from four main GNSS satellite constellations (e.g., one is global positioning system).…”

“…A challenge is that the sensor field of view will capture different satellites at varying angles throughout the day, and thus different plants will contribute to the GNSS sensor at different times. Therefore, some integration over incidence angles and over time is needed to determine a full canopy VOD signal (Humphrey & Frankenberg, 2023;Yao et al, 2024). Nevertheless, such processing is tenable and, once completed, a VOD time series is obtained that represents hundred-meter spatial scales.…”

“…Components of the plant that have storages, such as those represented by a capacitance, can be translated into VOD (Hartzell & Bartlett, 2017;Liu et al, 2021b;Yao et al, 2024). These schemes are now being widely implemented within numerical land surface and terrestrial biosphere models (Christoffersen et al, 2016;Kennedy et al, 2019;Tai et al, 2017;Xu et al, 2016;Yao et al, 2024).…”

“…These include GNSS field measurements and VOD outputs from land surface models. Here, I summarize existing validation approaches, describe two newer validation approaches as inspired by Yao et al, 2024, and provide a vision for future work. I focus on validating satellite VOD itself and not derived products such as VOD-derived leaf water potential and above ground biomass.…”