Profile distribution of soil-water content and its temporal stability along a 1340-m long transect on the Loess Plateau, China

- Strong persistence of spatial patterns of SWC was observed vertically and temporally.
- Both direct and indirect methods can accurately estimate the mean SWC for each depth.
- The accuracies of prediction of the two methods were not depth-dependent.
- The driest moisture locations were more likely to be the most time-stable locations.

Information on the profile characteristics of soil-water content (SWC) and its temporal stability is essential for an accurate understanding of hydrological processes. This study investigated changes of spatial variation and temporal stability of SWC in soil profiles and estimated mean SWC based on direct and indirect methods. SWCs were measured at 20-cm intervals in the soil profiles to a depth of 3 m using neutron probes at 135 locations along a 1340-m long transect on 18 sampling dates from 2012 to 2013 on the Loess Plateau in China. The coefficient of variation over space (CVS) of SWC first decreased and then increased vertically. The coefficient of variation over time (CVT) of SWC decreased along the soil profiles. The spatial patterns of SWC strongly persisted vertically and temporally. Mean Spearman's rank correlation coefficients decreased from a depth of 10 to a depth of 20 cm, fluctuated from 20 to 180 cm, and then increased below 200 cm. Temporal stability increased along the soil profiles based on the standard deviation of mean relative difference (SDRD) and the mean absolute bias error (MABE). The number of locations with an SDRD and/or MABE < 5% increased along the soil profile, and the number of locations with a mean relative difference within ± 5% and/or representative locations were variably dependent on depth. Both direct and indirect methods could accurately estimate the mean SWC for each depth based on the mean absolute relative errors and root mean square errors. The driest and wettest locations tended to remain representative for more depths than did locations with mean-moistures. The driest locations were more likely to be the most temporally stable. These findings should improve our understanding of soil-water dynamics in soil profiles.