Introduction

Hydrologic systems are dynamically influenced by both anthropogenic activities and natural variability (Pathiraja et al., 2016). Among them, bushfire is an important cause of hydrological change in fire-prone landscapes (Shakesby & Doerr, 2006). Bushfire can result in vegetation loss and post-fire microenvironment temperature rise, which can change hydrological processes such as interception, evapotranspiration and soil water storage. Effects of bushfire on streamflow can be estimated using the paired catchments method (Liu et al., 2004). Many studies have been conducted to understand the hydrological change due to vegetation cover changes using the paired catchment method (Brown et al., 2005; Loon et al., 2019; Zégre et al., 2010). However, the paired catchment analysis has drawbacks since it only involves small catchments and is expensive to conduct (Zhang et al., 2011). An alternative method is to use the time-trend analysis which does not require control catchments and can be applied to small to large catchments.
Apart from anthropogenic activities, climate variability can significantly affect streamflow and should be considered when evaluating the streamflow response to vegetation change due to bushfire (Lane et al., 2005; Zhang et al., 2011). Several methods have been developed and applied to separate the effects of vegetation change and climate variablity (Tuteja et al., 2007; Zhang et al., 2008; Zhao et al., 2012). These methods firstly identify key factors affecting streamflow and estimate the total streamflow changes using observed data. The second step is to determine the changes in streamflow due to the key factors such as vegetation change caused by bushfire and then to attribute the remaining changes to other factors such as climate variability. These methods can provide consistent estimates of vegetation change and climate variability impacts on streamflow in the catchments being considered (Zhao et al. (2010)).
Many attempts have been made to quantify the impacts of bushfire on hydrology in the past five decades (Anderson et al., 1976; Brown, 1972; Daniell & Kulik, 1987; Neary et al., 1999; Robichaud et al., 2000; Tiedemann et al., 1979). Some recent studies mainly focused on hydrological modeling in bushfire impacted catchments, aiming to better estimate the streamflow variation under bushfire impacts (Folton et al., 2015; Heath et al., 2014; Mannik et al., 2013; Papathanasiou et al., 2015; Seibert et al., 2010; Zhou et al., 2015). To date, most of these studies mainly investigated or evaluated the bushfire impacts on streamflow but few of them quantified the impact of climate variability.
This study aims to quantify the climate variability and bushfire impacts on streamflow for 14 catchments located in Victoria, Australia that have been severely burnt by the 2009 Victoria bushfire, the second severest national bushfire disaster. The specific objectives are: i) investigating annual streamflow trends from the 14 catchments; and ii) separately estimating the vegetation change caused by bushfire and climate variability impacts on streamflow using time-trend analysis and statistic sensitivity-based model