Near-surface vs. sub-surface soil moisture impacts on vegetation functioning

Soil water availability is a critical requirement for vegetation functioning in a water-limited regime. Vegetation takes up water from varying soil depths depending on their rooting location and soil moisture availability. The uptake depth varies spatially across climate regimes and vegetation types and temporally between seasons. Yet, a scientific consensus on the global relevance of near-surface and sub-surface soil moisture for vegetation functioning is still lacking and is the focus of this study. 

In particular, we calculate the correlation between the Near-Infrared Reflectance of Vegetation (NIRv) with both satellite-derived near-surface soil moisture from ESA-CCI and terrestrial water storage from GRACE. This is done globally and with monthly data during the growing season at each grid cell and accounting for the confounding effects of temperature and radiation. We analyze how these correlations vary spatially across varying vegetation types and climatic regimes, and temporally between all growing season months and particularly dry months. Finally, we repeat the analyses using Sun-induced fluorescence (SIF) data instead of NIRv. 

We find that NIRv and SIF correlate more strongly with near-surface soil moisture compared to terrestrial water storage in semi-arid regions with low tree cover. This suggests that the vegetation preferentially takes up water from near-surface soil moisture whenever available to meet its transpiration demand.   In contrast, in regions with more tree cover and in drier regions, the correlation with terrestrial water storage is comparable to or even higher than with near-surface soil moisture. This indicates that trees can make use of their deep rooting systems to access deeper soil moisture resources, similar to vegetation in arid regions. In particularly dry months, correlations with near-surface soil moisture increase while this is even more the case with terrestrial water storage, highlighting the relevance of deeper water resources during rain-scarce periods.

Overall, while direct observations of sub-surface soil moisture are scarce, this study employs different satellite-based data streams in order to estimate the relevance of near-surface versus sub-surface soil moisture for vegetation functioning. This can inform the representation of vegetation-water interactions in land surface models to support more accurate climate change projections.