Uncertainties in chloride-based tracing methods for deep drainage estimation under shallow- and deep-rooted plants

Abstract

• Estimated potential recharge vary with methods and vegetation types. • Recharge estimated by CMB and CP is problematic under deep-rooted plants. • CMB performs the best in estimating recharge under shallow-rooted plants. • CWMB present the best results for recharge estimation under deep-rooted plants. • Uncertainties in methods originate from the requirements of methodological theory. Chloride (Cl)-based tracing methods are popular for deep drainage (or potential groundwater recharge) estimation in the unsaturated zone, but uncertainties in various such methods have rarely been investigated. We collected 1827 samples from 24 soil profiles under plants with shallow or deep root systems in the thick unsaturated loess deposits of China, and measured soil water, Cl, and tritium contents. With the drainage estimated with the tritium peak method as reference, we evaluated the applicability of several Cl-based approaches for drainage estimation under shallow- and deep-rooted plants, including the chloride mass balance method (CMB), chloride storage method (CS), chloride peak method (CP) and chloride-water mass balance method (CWMB). The estimated deep drainages varied with methods, and the uncertainties were large with a coefficient of variation of 0.71. Compared to the drainage obtained from tritium-based methods (33.6 ± 11.2 and 9.6 ± 14.5 mm year −1 under shallow- and deep-rooted plants, respectively), CMB and CS (38.0 ± 16.6 and 36.7 ± 17.6 mm year −1 ) performed better than CP (23.2 ± 17.8 mm year −1 ) under shallow-rooted plants, while CWMB (7.6 ± 13.9 mm year −1 ) performed better under deep-rooted plants. The large discrepancy in estimated drainage originates from methodological theory such as the time scales of representation and sensitivity to exogenous Cl inputs. The root water uptake-induced soil water depletion can enrich soil Cl concentration and greatly influence CMB and CP. Our findings provide insights for how to estimate deep drainage through appropriate methods.