Forest plot soil moisture estimation using the cosmic-ray neutron (CRN) method and gamma-ray spectroscopy (GRS)

Water balance studies in forest is due to the vegetation particularly challenging. In a forest, precipitation will either fall directly to the ground surface (direct throughfall) or be captured by the trees (interception). From the canopy, the water will either return to the atmosphere through evaporation (interception loss) or fall to the ground as the storage capacity of the canopy is reached (indirect throughfall). Depending on the surface area and structure of the trees, the share of interception loss and throughfall vary considerably in space. The resulting water input to the ground is thus highly variable resulting in an uneven soil water content. In order to account for this variability, methods for estimating soil moisture at forest plot scale are therefore paramount to accurately assess and understand the forest water budget. For the last decade, the cosmic-ray neutron (CRN) detector has been used for soil moisture estimation at a spatial scale of hectometers. Recent studies from a farmland field site suggest that gamma-ray spectroscopy (GRS) can be used to estimate soil moisture at a spatial scale of decimeters to hectometers depending on the detection height. Unlike CRNs, gamma-rays are primarily emitted from the ground, and the signal detected at ground level is only affected by vegetation that is located in the space between the ground and the detector, not by the forest canopy. In this study, we test the recently presented procedure of GRS soil moisture estimation for a forest field site. For this, CRN soil moisture estimates are used. The signal of vegetation on the CRN measurements is considered by applying a site-specific conversion function for soil moisture estimation. A GRS sensor is installed on a mast at ground level (2 m height) and at canopy level (28 m height). At ground level the footprint radius is between 5 and 30 m, while it at canopy level is between 51 and 174 m. The canopy level sensor is both affected by soil moisture and the vegetation, and the signal of the vegetation is examined using both the ground level GRS and CRN measurements.