Rebecca Grace Chapman, Andy Baker , Liza Kathleen McDonough, Monika Markowska, Shawn Laffan

Understanding the mechanisms controlling spatial heterogeneity of drip water percolation into caves is essential for interpreting karst aquifer recharge and speleothem isotopic and geochemical records for paleoclimate analyses. Here we present the first analysis of drip rate variability using a novel time-varying Functional Principal Component Analysis (FPCA), validated against drip water stable isotope composition. Twenty-six drip sites were monitored across Harrie Wood Cave, south-east Australia, over a 2.5 year period. A positive relationship with cave drip water hydrology and rainfall and soil moisture was identified, with soil moisture recording the strongest relationship. FPCA was used to classify drip-water flow (percolation) pathways based on temporal shifts in the drip rate time series. Our results reveal that three percolation classes can be used to explain water movement within the cave: storage baseflow, fracture baseflow and overflow. The successful application of FPCA in this study suggests that this statistical technique will be useful for the analysis and interpretation of other large, discontinuous hydrological datasets.