Application of temporal stability analysis in depth-scaling estimated soil water content by cosmic-ray neutron probe on the northern Tibetan Plateau

Abstract

Soil moisture is a key limiting factor in grass growth and restoration in alpine meadow ecosystems on the northern Tibetan Plateau. In the deeper layers, soil moisture influences the processes of freeze-thaw, erosion and water cycle. Cosmic-ray neutron probe (CRNP) is a new method for continuously monitoring mean soil water content (SWC) at hectometer scale, which has been applied in an alpine meadow at a high accuracy. However, with CRNP measuring depth of only 30cm, depth-scaling is needed for sufficient insight into deep-layer soil moisture. This study evaluated the accuracy of CRNP measurement of SWC in the 2015 and 2016 growing seasons and the performance of temporal stability (TS) analysis in depth-scaling CRNP-estimated SWC. During the study period, 11 field samplings were done for calibration of CRNP-estimated SWC. Using 22 occasions of neutron probe measurements for each of 113 investigated locations, the TS of SWC was analyzed and its performance in depth-scaling CRNP-estimated SWC at five soil depths (10, 20, 30, 40 and 50cm) was evaluated. The results showed that the mean SWCs to the depth of 50cm were 12.9 and 17.0%, respectively in 2015 and 2016 growing seasons and were temporally influenced by precipitation and spatially by soil depth. The accuracy of the CRNP-measured SWC was high, with root mean square error and Nash-Sutcliffe efficiency coefficient (NSE) of 2.1% and 0.832, respectively. Representative locations for TS existed in all the soil layers, which increased with increasing soil depth. For the various soil layers, TS-estimated SWC was close to field-measured value. Only a relatively small error and high NSE were noted, suggesting that TS was reliable in application in CRNP depth-scaling. The study provided further scientific basis for the application of CRNP and an effective way of depth-scaling CRNP-estimated mean SWC in alpine meadow ecosystems.