Abstract
To assess the potential of the COsmic-ray Soil Moisture Observation System (COSMOS) for measuring soil moisture on the slope of volcanic ash soil, we analyzed the response of its epithermal neutron counts (N) to the volumetric water content (θ ) recorded by soil moisture sensors installed at five locations and three depths. By comparing θ across these locations and depths, we estimated the maximum effective observation depth of COSMOS is 17 cm. Based on this estimate, we constructed four datasets comprising the mean θ values from TDT sensors at a depth of 10 cm (θAvg) and the corresponding N values, along with their 3-hour, 12-hour, and 24-hour averages. In all combinations, the correlation between N and θAvg was low, suggesting that the presence or absence of vegetation cover significantly influences N. Consequently, we classified the data into periods with and without vegetation cover — determined using camera imagery — and applied distinct calibration equations for each period. This approach significantly improved the agreement between COSMOSestimated θ and θAvg. Our findings demonstrate the effectiveness of COSMOS for monitoring surface soil moisture on volcanic ash slopes characterized by exceptionally high porosity of volcanic ash soil.
https://ror.org/04f4wg107 
