Utilizing advanced and modified conventional trend methods to evaluate multi-temporal variations in rainfall characteristics over India

Adequate and consistent rainfall is essential for sustaining water resources, agricultural production, and overall economy of a nation. To explore the variability and changes in rainfall system, the most common and widely employed conventional trend methods are linear regression (LR) and Mann-Kendall (MK) trend tests. These methods are often subject to inconsistent results with respect to the extent of changes reported in rainfall patterns. This study utilizes the advanced LR i.e., quantile regression (QR) and the modified MK (m-MK) trend methods to investigate the retrospective rainfall characteristics and associated trends for multi-temporal periods i.e., long-term (1951–2020), bifurcated (pre-1985 and post-1985), and most-recent (2000–2020), over different climate zones of India. Furthermore, temporal evolution and trend consistency (stability) were examined by comparing multi-temporal slope coefficients at various quantiles ( τ ) of rainfall distribution. The long-term trends in general rainfall characteristics (GRCs) exhibited drying patterns, while opposite increasing trends were observed in extreme rainfall characteristics (ERCs) for most of the study region. The results of QR at median tail ( τ = 0.5) were more or less consistent with the results of m-MK test. Interestingly, an increase in the trend significance and magnitude was observed at higher quantiles ( τ > 0.8). The bifurcated and long-term periods showed contrasting results in rainfall characteristics, suggesting trend instability whereas during pre-1985, post-1985, and most-recent periods, the temporal evolution of GRCs revealed a systematic increment in positive trend significance. Altogether, the advanced and modified trend assessment in the present research compliments conventional trend methods with improved trend detection and trend consistency identification.