Multi-level trend analysis of extreme climate indices by a novel hybrid method of fuzzy logic and innovative trend analysis

Abstract

Multi-level trend analysis of extreme climate variables is an efficient method for in-depth investigation of the climate change impacts on ecohydrology. However, most of existing statistical methods do not reveal potential trends in different levels of data. In this study, a new approach namely Fuzzy Innovative Trend Analysis (FITA) was introduced that takes the advantages of fuzzy logic to improve and facilitate Innovative Trend Analysis (ITA) abilities to multilevel trend detection at Extreme Climate Indices (ECIs). Regarding the graphical nature of the proposed method, two new indices, namely Grow Percent (GP) and Total Grow Percent (TGP) were suggested for quantifying the power of trend at distinct levels. The FITA was utilized for trend detection at three levels of four important ECIs related to precipitation and temperature. To this end, long-term (1960–2021) daily temperature and precipitation observations at six meteorology stations across diverse climatic zones of Iran were used. The multilevel trends attained by the FITA were further compared to those of ITA, Mann-Kendall (M-K), and Sen’s slope (SS) tests. The results indicated that the FITA provides promising results with higher interpretability and reliability than its counterparts at all stations. The underlying high-resolution trends detected at certain stations also pointed out that the M-K and SS tests may yield in misleading interpretations when they are used for identifying trends in ECIs.