Abstract The Mann–Kendall (MK) and modified Mann–Kendall (for non-autocorrelated) trend tests have been applied to examine the significant positive and negative trends in rainfall at a 95% significance level. The trends of monthly, seasonal and annual precipitation in Khushkhera–Bhiwadi–Neemrana Investment Region (KBNIR), Rajasthan for the period 1962–2014 (52 years) were investigated based on the data from eight rain gauge stations. Sen’s slope (mm/year) test was used to detect the magnitude of change for the above-mentioned series. The Pettit–Mann–Whitney test and Standard Normal Homogeneity tests (SNHT) were also applied to detect homogeneity of the data and in order to determine the year of change points in the dataset. These analyses consistently indicated 1996 as the most probable change point across several datasets. To assess the spatial distribution of trends, change magnitude values of each station were interpolated using GIS software for pre-monsoon (MAM), monsoon (JJAS), post-monsoon (ON) and annual rainfall in KBNIR. The results indicate that only the Kotkasim station shows a significant increasing trend (+5.71 mm/year in annual rainfall), whereas other stations show declining trends, particularly in monsoon rainfall (up to −2.03 mm/year). The eastern part of the region demonstrated a relatively positive slope, in contrast to the western areas, where negative trends were predominant. These findings highlight critical changes in regional rainfall patterns and provide valuable insights for water resource management and climate-resilient agricultural planning in semi-arid zones such as KBNIR, helping to minimise the potential adverse impacts of climate variability in this key investment region. Research highlights <UnorderedList Mark="Bullet"> <ItemContent> Long-term rainfall trends (1962–2014) are analysed using multiple non-parametric statistical methods in the rapidly developing semi-arid region KBNIR of Rajasthan. </ItemContent> <ItemContent> Homogeneity and change-point analyses (SNHT, Pettitt, and Buishand tests) identified statistically consistent rainfall regimes across all stations, with 1996 emerging as the most probable regional rainfall shift year. </ItemContent> <ItemContent> Despite the absence of observation stations within the study boundary, the methodology relied on rainfall interpolation from well-distributed nearby stations to represent regional variability. </ItemContent> <ItemContent> Results show a contrasting rainfall regime, with declining monsoon rainfall and increasing pre-monsoon trends at most stations, indicating a shift in monsoon dynamics with direct implications for agriculture and water resources. </ItemContent> </UnorderedList>

Rainfall trend analysis and spatial distribution for implication of the agriculture water management in the semi-arid region

期刊论文
发表日期：

2026年7月21日

查看全文

Kuldeep Tiwari,
Rohit Goyal,
Archana Sarkar,
Raghuvesh Tiwari

摘要

Abstract

The Mann–Kendall (MK) and modified Mann–Kendall (for non-autocorrelated) trend tests have been applied to examine the significant positive and negative trends in rainfall at a 95% significance level. The trends of monthly, seasonal and annual precipitation in Khushkhera–Bhiwadi–Neemrana Investment Region (KBNIR), Rajasthan for the period 1962–2014 (52 years) were investigated based on the data from eight rain gauge stations. Sen’s slope (mm/year) test was used to detect the magnitude of change for the above-mentioned series. The Pettit–Mann–Whitney test and Standard Normal Homogeneity tests (SNHT) were also applied to detect homogeneity of the data and in order to determine the year of change points in the dataset. These analyses consistently indicated 1996 as the most probable change point across several datasets. To assess the spatial distribution of trends, change magnitude values of each station were interpolated using GIS software for pre-monsoon (MAM), monsoon (JJAS), post-monsoon (ON) and annual rainfall in KBNIR. The results indicate that only the Kotkasim station shows a significant increasing trend (+5.71 mm/year in annual rainfall), whereas other stations show declining trends, particularly in monsoon rainfall (up to −2.03 mm/year). The eastern part of the region demonstrated a relatively positive slope, in contrast to the western areas, where negative trends were predominant. These findings highlight critical changes in regional rainfall patterns and provide valuable insights for water resource management and climate-resilient agricultural planning in semi-arid zones such as KBNIR, helping to minimise the potential adverse impacts of climate variability in this key investment region.

Research highlights

Long-term rainfall trends (1962–2014) are analysed using multiple non-parametric statistical methods in the rapidly developing semi-arid region KBNIR of Rajasthan.

Homogeneity and change-point analyses (SNHT, Pettitt, and Buishand tests) identified statistically consistent rainfall regimes across all stations, with 1996 emerging as the most probable regional rainfall shift year.

Despite the absence of observation stations within the study boundary, the methodology relied on rainfall interpolation from well-distributed nearby stations to represent regional variability.

Results show a contrasting rainfall regime, with declining monsoon rainfall and increasing pre-monsoon trends at most stations, indicating a shift in monsoon dynamics with direct implications for agriculture and water resources.