Applicability of cosmic-ray neutron sensor for measuring soil moisture at the agricultural-pastoral ecotone in northwest China

• Published in: Science China. Earth sciences Vol. 63; no. 11; pp. 1730 - 1744
• Main Authors: Tan, Xingyan, Zhang, Lanhui, He, Chansheng, Zhu, Yuzuo, Han, Zhibo, Li, Xuliang
• Format: Journal Article
• Language: English
• Published: Beijing Science China Press 01.11.2020; Springer Nature B.V
• Subjects: Accuracy; Arid regions; Biomass; Calibration; Concretions; Correlation coefficient; Correlation coefficients; Cosmic ray neutrons; Cosmic rays; Crusts; Earth and Environmental Science; Earth Sciences; Ecological studies; Estimates; Gravimetry; Habitat selection; Hydrologic studies; Hydrology; Interception; Parameters; Plant cover; Remote sensing; Research Paper; Root-mean-square errors; Semiarid zones; Sensors; Soil; Soil moisture; Soil moisture estimation; China; Cosmic-ray neutron sensor; The agricultural-pastoral ecotone in Northwest China (APENC); Soil moisture; observations
• ISSN: 1674-7313; 1869-1897
• DOI: 10.1007/s11430-020-9650-2

Accurate monitoring of soil moisture is crucial in hydrological and ecological studies. Cosmic-ray neutron sensors (CRNS) measure area-average soil moisture at field scale, filling a spatial scale gap between in-situ observations and remote sensing measurements. However, its applicability has not been assessed in the agricultural-pastoral ecotone, a data scarce semi-arid and arid region in Northwest China (APENC). In this study, we calibrated and assessed the CRNS (the standard N 0 method) estimates of soil moisture. Results show that Pearson correlation coefficient, R P , and the root mean square error (RMSE) between the CRNS soil moisture and the gravimetric soil moisture are 0.904 and less than 0.016 m 3 m −3 , respectively, indicating that the CRNS is able to estimate the area-average soil moisture well at our study site. Compared with the in-situ sensor network measurements (ECH2O sensors), the CRNS is more sensitive to the changes in moisture in its footprint, which overestimates and underestimates the soil moisture under precipitation and dry conditions, respectively. The three shape parameters a 0 , a 1 , a 2 in the standard calibration equation ( N 0 method) are not well suited to the study area. The calibrated parameters improved the accuracy of the CRNS soil moisture estimates. Due to the lack of low gravimetric soil moisture data, performance of the calibrated N 0 function is still poor in the extremely dry conditions. Moreover, aboveground biomass including vegetation biomass, canopy interception and widely developed biological soil crusts adds to the uncertainty of the CRNS soil moisture estimates. Such biomass impacts need to be taken into consideration to further improve the accuracy of soil moisture estimation by the CRNS in the data scarce areas such as agricultural-pastoral ecotone in Northwest China.