Journal of the Geographical Studies of Mountainous Areas

Mendeley  

Zotero  

Mirhashemi H. Identification of structural breaks and change points of rainfall time series in mountainous areas (Case study: Khorramabad Synoptic Station). JGSMA 2021; 1 (4) :1-16
URL: http://gsma.lu.ac.ir/article-1-91-en.html

Identifying trends and structural breaks in the rainfall time series is one of the most important challenges in climate change studies. As the discovery of such structural breaks can help assess future scenarios and atmospheric phenomena such as floods and droughts. In this study, in order to identify structural breaks in the daily and monthly precipitation time series of Khorramabad synoptic station, two approaches of multifractal detrended fluctuation analysis (MF-DFA) and Maximal overlap discrete wavelet transform (MODWT) were used. The results of applying MF-DFA on the daily rainfall time series (1961-2018) showed that there are three structural breaks in this time series that the rainfall trend around these points has changed significantly. On the other hand, the results of this approach showed that the period 1985-2002 is the most fluctuating period of precipitation and the periods 1961-1969 and 2010-2018 are the periods with the lowest fluctuations in precipitation. Calculation of the fluctuation function at time scales of 2, 5, 10 and 20 years showed that in the period 2000-2018, fluctuations of 2 to 20 years have decreased. The results of the fluctuation function with q order also showed that the periods 1965-1970 and 1989-1994 were associated with large fluctuations in precipitation. If the two periods of 1999-2004 and 2008-2013 refer to periods with small fluctuations in precipitation. Also, the results obtained from the application of MODWT on the monthly rainfall time series showed that the wavelet coefficients changed in 2007. In this regard, by dividing the high frequency signals of the wave into two periods including 1961-2007 and 2008-2018, it was found that the differences between precipitation fluctuations during the scales of 8 to 16 months have the most variability in the precipitation time series.