Evaluating Long-term Climatic Trends in Agro-climatic Zones of Satara District, Maharashtra

1. Introduction

2. Study Area

Satara district is a historically located in the western part of Maharashtra and it lies on the western edge of the Deccan Plateau between 17° 5' N and 18° 11' N latitude and 73ᵒ33' E and 74ᵒ 54' E longitude (Figure 1). The east-west extension is 144 km and the north-south extension is about 120 km (10480 km²). Satara district is situated in the basin of rivers Krishna, Neera, Manganga, Yerala, and Krishna. Crops such as millet, sorghum, maize, onion, oilseeds, pomegranate, etc. are growing in the area. Indigenous cows, buffaloes, sheeps, and goats are the major livestock of the region. India has 15 agro-climatic zones, as noted by the Indian Planning Commission. The district encompasses four of these zones: Ghat Zone, Transition Zone - I, Transition Zone - II, and Scarcity Zone. The tehsils of western Patan, Wai, and Mahabaleshwar are all part of the Ghat zone. The tehsils of eastern Jaoli, eastern Mahabaleshwar, Wai, Western Khandala, and Middle Patan are all part of Transition Zone-I. In Transition Zone-II, East Khandala, Koregaon, East Satara, Karad, and East Wai, tehsils are included while Scarcity Zone covers areas from Phaltan, Man and Khatav tehsils.

3. Material And Methods

The non-parametric Mann-Kendall test, introduced by Mann in 1945 and further developed by Kendall in 1955, is widely used by researchers across various fields, including science, medicine and meteorological studies (Sneyers et al., 1990; Zhang et al., 2000; Yue and Hashino, 2003; Cannarozzo et al., 2006; Aziz and Burn, 2006; Wilks, 2006; Chandler and Scott, 2011; Kamal and Pachauri, 2018) have applied this test to detect monotonic trends in long-term time series datasets. The Mann-Kendall (MK) test developed by Mann in 1945 and Kendall in 1955 aims to statistically assess meteorological variables to identify trends, either upward or downward in a variable (Gocic and Trajkovic, 2013). The text mentions a non-parametric statistical test for detecting patterns in time series data (Kendall, 1955; Mann, 1945). Sen firstly created the Sen slope test to examine statistical linear correlations. This is used to calculate the magnitude of trends in long-term temporal data (Jiqin et al., 2023). In trend analysis, the null hypothesis for a trend test is H₀, which states no trend in the datasets. Alternative hypothesis H₁ shows a trend in the analyzed time series dataset. The Mann-Kendall test was applied to the annual rainfall trend (Cannarozzo et al., 2006). Kendall’s S (Kendall, 1962) is calculated using the following formula.

​\(S= ∑_{(i=1)}^{(n-1)}∑_{(j=i+1)}^nsign(x_j-x_i ) \)​       (1)

where, n is the data set length, and y denotes the data values at periods i and j.

​\( Sign \vartheta = \begin{cases} 1\ if \vartheta >0 \\ 1\ if \vartheta =0 \\ -1\ if \vartheta <0 \end{cases} \)​       (2)

Two parameters of the Mann-Kendall test are crucial for detecting trends. The slope magnitude estimate shows the direction and magnitude of the trend and the significance level, which indicates the test strength. Assuming that ​\(x_i\)​  is independent and randomly ordered, the statistic S (equation (3)) is normally distributed when n ​\(≥\)​  8 has zero mean and variance as follows:

​\(var= {1 \over18} [n(n-1)(2n+5)-∑_tf_t (f_t-1)(2f_t+5) ] \)​  (3)

where, t varies over the set of tied ranks and ft is the number of times i.e. frequency that the rank t appears. The Mann Kendall’s test uses the following test statistics:

​\( Z = \begin{cases} S-1/ \sigma \ if \ S >0 \\ 0 \ \ \ \ \ \ \ \ \ \ \ \ \ if \ S=0 \\ S+1/\sigma \ if \ S <0 \end{cases} \)​             (4)

A 95% confidence level was considered in this research. Trend analysis is critical for understanding past and present hydrological changes, forecasting future trends and taking necessary safeguards. It is commonly used in the study of hydrological time series (Zhang et al., 2000). Mann-Kendall test and linear regression test association amongst a reliant on variable Y and independent variable X were used to recognize the climatic trend using XLSTAT software.

In the present study, the Agro-climatic base map of India published by ICAR-ATARI, Pune has been used to prepare the Agro-climatic map of Satara district using GIS technology. The Planning Commission of India has considered factors such as physiography, soil, geology, climate, cropping practices, irrigation and minerals form different agro-climatic zones. Accordingly, ICAR-ATARI, Pune has developed an agro-climatic map. The climate data for 38 years (1984-2022) was taken from NASA’s Prediction of Worldwide Energy Resources POWER (DAV) agency’s Agro-Climatology Field. We used Mann-Kendall’s statistical test, Sen’s Slope and linear regression to examine the climatic trends from 1984 to 2022.

4. Result And Discussion

The Mann-Kendall assessment is a numerical method commonly used for analyzing trends in climatic and hydrological time series (Kamal and Pachauri, 2018). The significant level (0.05%) of this test demonstrates trends in climate variables such as rainfall, temperature, relative humidity and soil moisture. The Mann-Kendall test is used to show the normal distribution of trends and data outliers in a time series of hydrometeorological data. The Mann-Kendall test is a non-parametric test that evaluates the significance of trends using the statistical p-value (Duhan and Pandey, 2013). If the p-value is (˂ 0.05) specifies the trend is statistically important. Additionally, Sen’s slope is used to guess the magnitude of a trend with a confident value representing an increasing trend and an adverse value indicating a decreasing trend.  Typically, the impacts of climate change will exacerbate the existing vulnerabilities of poor people, placing additional strain on their livelihoods and coping mechanisms (Urquhart, 2009; Davis and Vincent, 2017; Masson-Delmotte et al., 2021; Thornton et al., 2014). The statistical method used in the present study consists of linear regression analysis and coefficient of determination R², and Mann-Kendall’s test is used to check the trend of climate factors. According to Mann Kendall’s test study area categorized into four groups such as Ghat Zone (GH Zone), Transition Zone- I (TR I Zone), Transition Zone- II (TR II Zone), and Scarcity Zone (SC Zone). Rainfall, maximum and minimum temperature, relative humidity and soil moisture are all averaged using the spatial interpolation method, which is also used to evaluate climate trends.

Annual climate data from 1984 to 2022 are used to analyze the climatic trends and pattern in the study area. The annual changes in rainfall, maximum and minimum temperature, relative humidity, soil moisture and wetness have been observed from 1984 to 2022. The spatial interpolation method is used to evaluate the climate parameters, patterns, and trends. The ghat zone has higher levels of average rainfall, relative humidity, and soil moisture but the distribution decreases eastward (Figure 2).

4.1 Ghat Zone (GH Zone)

If the climate series indicates an upward trend, the slope coefficient should be positive and statistically significant. In contrast, a downward tendency in a variable’s behavior should result in a negative slope value (Oza and Kishtawal, 2014; Gangarde et al., 2023).  Kendall Tau in the ghat zone is 0.216, the positive value of Sen’s slope is 3.495 and the p-value is 0.062 the R² value is 0.174 which indicates a statistically upward trend of rainfall (Figure 3). However, the R² value (0.174) is relatively high, indicating a positive relationship with rainfall.

This means that the data points show significant scatter around the fitted line and significant fluctuations in rainfall from year to year. Kendall tau of the ghat zone is -0.151, Sen’s negative slope is -0.016 and the p-value is 0.179 besides R² value is 0.174 which indicates the statistically downward trend of maximum temperature. However, the R² value (0.041) is relatively low, indicating a negative relationship with maximum temperature. According to Kendall tau, Sen’s slope, p-value and R² value, there is no trend in minimum temperature (Table 1). Relative humidity in this zone has a Kendall tau of 0.212, a positive value of Sen’s slope stands at 0.058, and a p-value of 0.059 besides an R² value of 0.150 indicating a statistically significant upward trend. The p-values ​​of soil moisture and soil wetness in this zone are 0.025 and 0.044, respectively representing that there is a statistically significant upward trend.

4.2 Transition Zone – I (TR I Zone)

In transition zone - I (Table 1 and Figure 2), Kendall tau is 0.205, the positive value of Sen's slope is 4.720 the p-value is 0.077 and besides R² value is 0.213 which indicates a statistically upward trend of rainfall.

The rainfall in the transition zone- I increased during the period 1984 to 2022. However, the R² value (0.213) is relatively high, indicating a positive relationship with rainfall. The negative value of Kendall tau in the transition zone- I position is -0.186, Sen’s negative slope is -0.020 and the p-value is 0.097 besides the R² value is 0.057 which indicates the statistical downward trend of maximum temperature (Figure 4). According to Kendall tau, Sen’s slope, p-value and R² value, minimum temperature shows no trend. The relative humidity in this zone has a Kendall tau of 0.217, a positive value of Sen’s slope of 0.062, and a p-value of 0.053 indicating a statistically significant upward trend. The p-values ​​for soil moisture and soil wetness in this zone are 0.073 and 0.042, respectively, indicating a statistically slightly upward trend.

4.3 Transition Zone – II (TR II Zone)

In transition zone- II, Kendall tau is 0.198, the positive value of Sen’s slope is 4.761, and the p-value stands at 0.092, which indicates a statistically upward trend of rainfall. From this, rainfall in transition zone-II, is increasing during the period 1984 to 2022. However, the R² value (0.213) is relatively high, indicating a positive relationship with rainfall. The negative value of Kendall tau in transition zone- II is -0.251, and Sen’s negative slope is -0.021, which specifies the statistically downward trend of maximum temperature. In this zone, no statistically significant trend is observed in minimum temperature (Figure 2). The relative humidity in this zone, Kendall tau is 0.188 and the positive value of Sen’s slope is 0.070 which indicates a statistically significant upward trend. The p-values for soil moisture and soil wetness in this zone are 0.044 and 0.122, respectively, indicating a statistically slightly upward trend (Figure 5).

4.4 Scarcity Zone (SC Zone)

In the scarcity zone, Kendall tau is 0.199, and the positive value of Sen’s slope is 3.495, which indicates a statistically slight upward trend of rainfall (Table 1 and Figure 2). The negative value of Kendall tau in the scarcity zone is -0.251, and Sen’s negative slope is -0.018, which indicates a statistically decreasing trend of maximum temperature. In this zone, no statistically significant trend shows in minimum temperature. The relative humidity in this zone, p-value is 0.116, and R² value (0.119) which indicates a statistically slightly upward trend. The p-values for soil moisture and soil wetness in this zone are 0.081 and 0.348, respectively, indicating a statistically slightly upward trend (Figure 6).

5. Conclusions

This study examined long-term climatic trends and spatial variations across the agro-climatic zones of Satara district using data from 1984 to 2022. Key climatic parameters, including rainfall, maximum temperature, minimum temperature, relative humidity, soil moisture, and soil wetness, were analyzed using statistical methods such as Mann-Kendall’s tau and Sen’s slope to determine trends and their significance. The results indicate a steady increase in rainfall across all four agro-climatic zones (GH, TR-I, TR-II, and SC), with statistically significant p-values (<0.5), confirming the presence of a notable trend. The ghat region showed the most significant increase in rainfall, with a Sen’s slope of 3.495. However, transition zone II and the scarcity zone exhibited more pronounced variations due to differences in topography and rainfall distribution. In the last decades of rainfall analysis, the graph shows the pressure of extreme events as the peak values are more than the mean normal. These areas, experiencing inconsistent rainfall, require reliance on artificial irrigation for stable agricultural production.

Regarding temperature trends, all zones showed a decline in maximum temperature with statistically significant p-values (<0.05). The SC and TR-II zones, in particular showed a declining trend as indicated by a negative Sen’s slope (-0.021). Despite the overall decrease, temperature variations were observed in annual climatic patterns. Relative humidity exhibited an increasing trend across all zones with statistically significant p-values (<0.5). This rise in humidity was accompanied by an increase in soil moisture and wetness both showing significant upward trends (p < 0.05). These findings suggest that higher rainfall and humidity contribute to lower temperatures and improved soil moisture conditions, which are essential for agricultural productivity.

Understanding these climatic changes is crucial for long-term planning in agriculture and water resource management. To address these shifts, it is necessary to implement climate-resilient farming strategies such as improved water management techniques, integrated pest control and the cultivation of drought-resistant crops. Moreover, enhancing community awareness through training and accessible information can strengthen adaptive capacity. Sustainable land management practices, including soil conservation, afforestation, and watershed management, should be promoted to mitigate the adverse effects of climate change on agriculture.