1 Introduction

There are about 16.7 million reservoirs larger than 0.01 ha in the world, and this number will continue to increase (Lehner et al., 2011). Currently, about half of the stream and river flow is regulated by reservoirs and dams, and by 2030 this number will climb up to 90% (Grill et al., 2015). Reservoir’s interception and regulation of river flow will obviously weaken peak flow and increase low flow (Vorosmarty et al., 1997). This will have a significant effect on the hydrological process many hundred kilometers downstream of the dam (Grill et al., 2015), as well as on the matter and energy cycle with streamflow as the medium (Pringle, 2003). Qualitative or quantitative research on the impact of reservoir construction on global or regional hydrological process and ecological environment is of profound significance.
There are many researches about the impact of reservoir on the hydrologic alteration and matter cycle: Grill et al. (2015) assessed the patterns and trends in river fragmentation and flow regulation by global dams at multiple scales; Hecht, Lacombe, Arias, Dang, and Piman (2019) reviewed the hydrological impacts of the hydropower dams of Mekong River basin; Taylor Maavara, Dürr, and Van Cappellen (2014) studied the worldwide retention of nutrient silicon by river damming; T. Maavara et al. (2015) studied the global phosphorus retention by river damming; Van Cappellen and Maavara (2016) studied the global scale modifications of riverine nutrient fluxes by damming; Meanwhile, many researchers have also studied the impact of reservoirs on downstream hydrological drought (Rangecroft, Van Loon, Maureira, Verbist, & Hannah, 2019; Tijdeman, Hannaford, & Stahl, 2018; Anne F. Van Loon et al., 2019; Wanders & Wada, 2015). In addition, reservoir’s regulation of streamflow can also influence the downstream baseflow recession characteristics. However, we have not found some quantitative global or case studies, this study may be the first attempt.
The shape of baseflow recession curve is an important feature of a basin. Baseflow recession analysis can be understood as a statistical analysis and description of the recession curve for a specific basin (Stoelzle, Stahl, & Weiler, 2013; Thomas, Vogel, & Famiglietti, 2015). The detailed baseflow recession analysis method will be described in Section 3. Baseflow recession analysis is widely used in hydrological research, water resources planning and management, and watershed hydrogeological research (Sujono, Shikasho, & Hiramatsu, 2004; Zhang, Chen, Hickel, & Shao, 2008). The main application objectives include: estimating long-term groundwater storage trends (Brutsaert, 2008), estimating watershed groundwater balance (H. Wittenberg & Sivapalan, 1999), baseflow separation (Huyck, Pauwels, & Verhoest, 2005; Hartmut Wittenberg, 1999), baseflow regionalization (Beck et al., 2013), and determining basin-wide hydrogeological parameters (Mendoza, Steenhuis, Walter, & Parlange, 2003; Oyarzún et al., 2014). The influence of reservoir on the baseflow recession characteristics will obviously affect the above application results, so we should first make clear the influence mechanism of reservoir on the baseflow recession characteristics.
At present, the main methods used to study the effects of human activities on hydrological processes include scenario modelling method (Veldkamp et al., 2015; Wada et al., 2017), paired catchments method (Best, Zhang, McMahon, Western, & Vertessy, 2003; Brown, Zhang, McMahon, Western, & Vertessy, 2005), pre‑post comparison method (Liu et al., 2016), upstream‑downstream comparison method (Rangecroft et al., 2019), and observation‑modelling comparison method (A. F. Van Loon & Van Lanen, 2013). Anne F. Van Loon et al. (2019) has made a more detailed summary. Among them, the simpler and more widely used method is the pre‑post comparison method, which selects two series before and after human activities to compare. Peñas, Barquín, and Álvarez (2016) pointed out that this method is hard to separate human activities from climate change, and it is usually necessary to select a control basin to eliminate the impact of climate change. However, we must also be wary of the spatial differences in climatic characteristics between the control basin and the impact basin. These spatial differences may lead to incorrect estimates of the impact of climate change.
The main purpose of this paper is to study the impact of Chaersen reservoir (Northeast China) on the downstream runoff and baseflow recession characteristics using the pre‑post comparison method with two upstream sub‑watersheds as control basins. The main contents include: 1) Before and after the construction of the reservoir, two streamflow series with similar rainfall‑runoff characteristics are selected to compare the differences of runoff and baseflow recession characteristics. 2) The impacts of climate change are estimated based on two sub‑basins in the upper reaches of the reservoir. 3) The influence mechanism of the reservoir is further explored.