The research, conducted by Hiren J. Chavda and Kalpesh L. Kapadiya, civil engineers based in Junagadh and Surendranagar, rests on one of the most valuable climate datasets available for the Indian subcontinent: the India Meteorological Department’s gridded daily rainfall product, delivered at a spatial resolution of 0.25 degrees by 0.25 degrees. Rather than relying on a handful of weather stations, whose records are often patchy and prone to relocation artifacts, the gridded dataset interpolates observations onto a uniform mesh, allowing district-level averages to be computed consistently across the entire 1901 to 2025 period. The authors aggregated the daily values into annual and seasonal totals, then subjected those series to a battery of statistical tests designed to interrogate the data from multiple angles.

That multi-method approach is the study’s methodological backbone. The Mann–Kendall test, a non-parametric workhorse of hydrology, detects monotonic trends without assuming that the underlying data follow any particular distribution, a crucial property for rainfall series that are notoriously skewed and noisy. Sen’s slope estimator complements it by providing a robust estimate of the magnitude of change, computed as the median of slopes between all pairs of points, which makes it resistant to the distorting influence of a single catastrophic flood year or an exceptional drought. Linear regression offers a conventional parametric cross-check, while Innovative Trend Analysis, introduced by Zekai Şen in 2012, compares the means of low, medium, and high rainfall sub-groups visually, revealing whether the trend is driven by wet extremes, dry extremes, or a uniform shift across the distribution. When four methods that make different assumptions converge on the same answer, the conclusion is far harder to dismiss as a statistical fluke.

The numbers themselves tell a striking story. Long-term mean annual rainfall across the six districts ranges from 570 to 773 millimeters, but with coefficients of variation between 35 and 43 percent, meaning that year-to-year swings of several hundred millimeters are entirely normal. Against that noisy backdrop, Sen’s slope estimates for annual rainfall range from roughly +0.9 to +4.0 millimeters per year. Compounded over 125 years, the upper end of that range implies an additional several hundred millimeters of annual rainfall, a substantial fraction of the regional mean. The southwest monsoon season, which delivers the overwhelming majority of Saurashtra’s water, shows corresponding gains of about +0.8 to +3.8 millimeters per year, with statistically significant increases in Amreli, Bhavnagar, Botad, Junagadh, and Rajkot. Surendranagar, the driest and most inland of the six, shows a weaker, non-significant upward trend, hinting at a coastal-to-inland gradient in how the strengthening monsoon manifests.

The seasonal dissection is where the study earns its relevance for water managers. While the monsoon strengthens, winter rainfall generally declines across the region, and pre-monsoon rainfall decreases in four of the six districts. Post-monsoon rains show weaker changes that are mostly not statistically significant. Taken together, the pattern suggests not simply more water but a reshaping of the hydrological calendar: a shorter, more intense wet season bracketed by longer dry periods. For rain-fed agriculture this is a double-edged development. More monsoon rainfall could support higher yields and fuller reservoirs, but if the same water arrives in fewer, heavier episodes, a larger share will run off rapidly rather than infiltrate the soil, increasing both flood risk during the wet months and irrigation demand during the dry ones. This tension between total volume and temporal distribution echoes broader findings in the climate literature, including influential work published in Science in 2006 documenting an increasing trend of extreme rain events over India in a warming environment.

To translate these statistical trends into something a water planner can act on, the authors devised a simple rainfall-based irrigation-stress proxy built on annual rainfall variability. The proxy classifies the districts into three tiers. Amreli and Junagadh emerge as high-stress locations, where the combination of rainfall amounts and variability leaves agriculture most exposed to deficit years. Bhavnagar, Rajkot, and Surendranagar fall into the moderate-stress category, while Botad registers comparatively low stress. The proxy is deliberately preliminary, a first-order screening tool rather than a hydrological model, but it demonstrates how trend statistics can be converted into a spatially differentiated picture of vulnerability. In a region where groundwater depletion has driven farmers toward increasingly deep wells and where check dams and farm ponds have become fixtures of state water policy, knowing which districts face the greatest stress is a meaningful input for prioritizing storage investments and crop advisories.

The choice of Saurashtra as a study area is far from incidental. The peninsula juts into the Arabian Sea and depends almost entirely on the southwest monsoon for its water budget, with irrigation drawing heavily on groundwater aquifers that recharge slowly in the semi-arid climate. Agriculture in the region spans groundnut, cotton, wheat, and horticultural crops, all of which are sensitive to the timing of rainfall as much as to its quantity. A monsoon that arrives late, pauses mid-season, or withdraws early can damage crops even in a year with normal total rainfall. By analyzing seasonal rather than only annual series, the study captures exactly this dimension of risk, and its finding of a concentrating wet season aligns with the broader expectation, articulated by Kevin Trenberth and colleagues in 2003, that a warming climate is changing the character of precipitation toward more intense but less reliably distributed events.

The authors are careful about the limits of what their analysis can claim. Trend detection in a 125-year rainfall record establishes that change has occurred, but it does not by itself attribute the change to any specific cause, nor does it project future rainfall. The irrigation-stress proxy, being based solely on annual rainfall variability, omits the soil moisture, evapotranspiration, crop water demand, and aquifer dynamics that a full hydrological assessment would require. The study explicitly frames its results as a preliminary basis for discussing adaptive irrigation and water-storage strategies, and calls for future work that couples rainfall trends with hydrological models and crop models. That caveat matters for policy: a district flagged as high stress by a rainfall proxy should be treated as a candidate for deeper investigation, not as a final verdict on its water security.

Nevertheless, the study’s scale and methodological rigor give it weight in a region where planning decisions have often been made with shorter or less systematic records. The use of open-source tools, the R statistical environment and QGIS for mapping, and the reliance on a publicly documented, high-resolution meteorological dataset make the analysis transparent and reproducible, and the authors report no funding and no conflicts of interest. For the farmers of Saurashtra, the practical message is nuanced. The monsoon that sustains them appears to be delivering more water over the long run, but it is also becoming a more concentrated, less forgiving resource, with the shoulder-season rains that once buffered the dry months in decline. Adapting to that reality, through storage that captures intense monsoon pulses, irrigation scheduling that assumes longer dry spells, and crop choices tuned to a compressed growing season, may define the next chapter of water management in this semi-arid corner of India.

Subject of Research: Long-term rainfall trends and irrigation water stress in semi-arid Saurashtra, India

Article Title: Multi-method rainfall trend analysis for irrigation-relevant water resources planning in semi-arid Saurashtra, India

Article References: Chavda, H. J., & Kapadiya, K. L. (2026). Multi-method rainfall trend analysis for irrigation-relevant water resources planning in semi-arid Saurashtra, India. Theoretical and Applied Climatology, 157(11), Article 702. https://doi.org/10.1007/s00704-026-06634-9

Image Credits: AI Generated

DOI: 10.1007/s00704-026-06634-9

Keywords: rainfall trends, monsoon, Saurashtra, Gujarat, semi-arid climate, Mann-Kendall test, Sen's slope estimator, Innovative Trend Analysis, irrigation stress, water resources planning, India Meteorological Department, climate variability