Principal component analysis of watershed hydrochemical response to forest clearance and its usefulness for chloride mass balance applications

Abstract

Application of the simple watershed chloride mass balance (CMB) method over areas of historical forest clearance is highly uncertain because the CMB method requires a steady state assumption, while the beginning, duration, and end of the transient period between preclearance and postclearance steady states is unknown. To address this difficulty, principal component analysis (PCA) with groundwater chemistry data is explored for the Piccadilly Valley in the Mount Lofty Ranges in South Australia, incorporated with hydrological modeling. The results indicate that PCA provides a potentially useful tool to identify postclearance dynamic hydrochemical response. Processes during the postclearance transient period, such as the release of historical soil chloride together with the charge‐balance cations, alter ionic ratios from a steady state condition, which make the hydrochemical signals from the transient processes distinguishable from other processes by PCA. For the studied watershed, the effect of postclearance transient processes accounts for 24% of major‐ion data variance, in comparison to 43% resulting from the effects of precipitation diluting and evapotranspiration concentrating processes. The PC scores associated with the transient processes are shown to be useful for estimating the timing of the postclearance transient, and estimating chloride concentration of groundwater recharge at the preclearance and postclearance steady states, even under conditions where the new steady state has not yet reached or the preclearance hydrochemical signal is lost.