ABSTRACT Graphical abstract summarizing a comparative trend analysis of nine hydroclimatic indices from 1971 to 2021 using MK, MMK, ITA, and MITA tests. The top row presents the workflow: input time-series data are processed through trend detection methods, MK/MMK significance testing with variance and autocorrelation adjustment, and ITA/MITA scatter-plot based trend assessment after splitting each series into two halves. The bottom row compares test outcomes using a radar chart of trend-detection percentages and stacked bar charts of grid counts across rainfall and temperature indices. Understanding hydroclimatic trends in semi-arid river basins is vital for sustainable water management under changing climate conditions. This study introduces a modified innovative trend analysis (MITA) to address the temporal ordering limitations of the conventional innovative trend analysis (ITA). MITA complements established nonparametric approaches – Mann–Kendall (MK), modified Mann–Kendall (MMK), and Sen's slope estimator – applied here in a four-method comparative framework. We applied these four methods to nine hydroclimatic indices – annual, monsoon, and non-monsoon rainfall; wet-day frequency and intensity; and temperature extremes – derived from high-resolution gridded data (1971–2021) across 62 grid points in the Manjira River Sub-basin (MRSB), India. Comparative analyses revealed distinct patterns: The Mann-Kendall test detected significant rainfall declines and notable warming trends (Z 6.0, p 0.01, slope: +0.014 to +0.017 °C year−1). Accounting for autocorrelation, MMK reduced the significance of rainfall trends. ITA, which disrupts chronology by sorting data halves, produced artificially rising rainfall trends. Conversely, MK/MMK indicated declining rainfall (Z = −2.32 to −2.56, p 0.01), while MITA – by correcting ITA's artificial sorting bias and preserving temporal order – revealed pronounced drying (−26.7 to −126.6 mm), consistent with the conservative but realistic estimates of MK/MMK. Cohen's kappa coefficients (κ = 0.89 for MK–MMK; κ = 0.76 for ITA–MITA) underscore that preserving temporal sequences yields more realistic trend detection, which is crucial for accurate water resources planning under climate change.
Kumar et al. (Wed,) studied this question.