Reconstructing Atmospheric Chloride Deposition Using Chloride-Tritium Tracers Stored in Thick Loess

Abstract

The limited observation of atmospheric chloride deposition (ACD) hinders the analysis of the historical evolution of climate and hydrological processes. The objectives of this study are to explore a framework to reconstruct point-scale ACD using chloride-tritium tracers stored in soils and further extrapolate them to regional-scale ACD. Considering that the vertical variability of soil tracer profiles is related to the temporal variability of ACD, we inversely transformed two methods for groundwater recharge estimation (i.e., chloride storage method and chloride mass balance method) to estimate point-scale ACD for different periods. In particular, we combined soil tritium profiles to exclude fertilizer impacts on chloride deposition. Further, we employed the multimodel reference method and stepwise linear regression to identify controlling factors and predict regional-scale ACD. We collected 3348 soil samples from 43 profiles from 7 to 62 m deep in China’s loess region to measure soil water, chloride, and tritium contents. The proposed methods can effectively exclude fertilization impacts and perform satisfactorily in estimating ACD. Climatic and geographic factors had stronger influence on ACD than vegetation and soil. The predicted ACD concentration had high spatial variability in 1963-1993 (2.36±0.21 mg L-1) and the stable period before 1963 (1.53±0.68 mg L-1). The methods are superior for dry regions where short soil profiles include long-term ACD information. The methods and results provide a technical basis for ACD data acquisition and improve the accuracy of recharge estimation across the Loess Tableland and other unsaturated zones where tritium profiles still exist in the world.

Keywords: Atmospheric chloride deposition, Tritium peak method, Chloride storage, Chloride mass balance, Deep loess profiles