“…Many other methods to identify discontinuities in a time series have been posited over the years (see Weatherhead et al ., 1998; Reeves et al ., 2007; Mudelsee, 2019; Muthuramu and Maheswari, 2019). Such methods usually seek to identify structural breaks or changepoints using a variety of different approaches including autoregressive and autoregressive‐moving averages (Karl et al ., 2000; Seidel and Lanzante, 2004; Davis et al ., 2006), regression‐based coefficients and methods (Jouini and Boutahar, 2005; Dixon and Moore, 2011; Aue and Horváth, 2013; Lyubchich et al ., 2013; Guo et al ., 2018; Elder and Fong, 2019; Tharu and Dhakal, 2020), t or F test with hypothesis testing (Lund et al ., 2007; Rienzner and Gandolfi, 2011; Gallagher et al ., 2013), higher‐order moments of the distribution (Hilas et al ., 2013; Xie et al ., 2019) and cumulative sums (Shao and Zhang, 2010), Bayesian methods (Ruggieri, 2013; Chen et al ., 2017; Yu and Ruggieri, 2019), bootstrap approaches (Bickel and Ren, 2001; Noguchi et al ., 2011; Lyubchich et al ., 2013), classification analysis (Anders et al ., 2013), manifold learning models (Xie et al ., 2013), functional data analysis (Alaya et al ., 2020), an informational approach (Beaulieu et al ., 2012), and nonparametric methods (Lanzante, 1996; Douglas et al ., 2000; Burn and Elnur, 2002; Xiong and Guo, 2004; McKitrick and Vogelsang, 2014; Xie et al ., 2014; Basarir et al ., 2017; Guo et al ., 2018; Hajani and Rahman, 2018; Zhou et al ., 2019; Bagniewski et al ., 2021). Most of these techniques work directly with the series in the time domain, usually searching for changes in parameters of the distribution or trend or shifts in the moments of the time series (e.g., mean, kurtosis) across a change point, although other researchers have explored the use of Hurst statistics to changepoint analyses (Outcalt et al ., 1997; Tan and Gan, 2017).…”