Field scale soil water prediction based on areal soil moisture measurements using cosmic-ray neutron sensing in a karst landscape

Abstract

Cosmic-ray neutron sensing (CRNS) is a newly-developed method for continuously measuring soil water content (SWC) at the hectometer horizontal scale. However, it is unknown whether CRNS-based SWC could be used for soil moisture prediction at field scale. In this study, we try to predict field scale SWC using the CRNS combined with the temporal stability analysis in a karst watershed. The CRNS was installed in a karst watershed as well as capacitance-based SWC sensors located in five land uses. The CRNS was calibrated by eighteen manual sampling procedures. A weather station was established about 200 m away from the CRNS for monitoring basic meteorological variables. The mean footprint radius and effective depth of the CRNS were 311.51 m and 11.98 cm, respectively. The CRNS can accurately estimate intermediate scale soil moisture, with a coefficient of determination (R2) of 0.89. Significant correlations (P < 0.01) were observed between SWC measured by the CRNS and Em-50 instruments. Though the SWC derived from the CRNS was significantly lower than that from Em-50 instruments in different land uses, the SWCs at field scale can be predicted accurately. The influences of landforms and land uses on prediction accuracy were not obvious from our data. The CRNS has great potential to improve water resources management in areas with greater heterogeneity.