Climate change is a worldwide problem caused by various anthropogenic activities, leading to changes in hydroclimatic variables like temperature, rainfall, riverine flow, and extreme hydrometeorological events. In India, significant changes are noted in its natural resources and agriculture sectors. In this study, we analysed the long-term spatio-temporal change in rainfall patterns of Madhya Pradesh, Central India, using Indian Meteorological Department high-resolution gridded data from 439 grid points. The coefficient of variance analysis showed low variability in annual and monsoon rainfall but significant variability in pre-monsoon, post-monsoon, and winter seasons, indicating considerable seasonal variation. Pre-monsoon rainfall exhibited an increasing trend (0.018 mm annually), while annual, monsoon, post-monsoon, and winter rainfall showed decreasing trends. Change point analysis identified shifts in rainfall patterns in 1998 (monsoon, annual), 1955 (pre-monsoon), 1987 (post-monsoon), and 1986 (winter). Spatio-temporal distribution maps depicted irregular rainfall, with some areas experiencing drastic declines in precipitation after 1998. The maximum average annual rainfall reduced from 1,769 to 1,401 mm after 1998 affecting water availability. The study's findings highlight a significant shift in Madhya Pradesh's seasonal rainfall distribution after 1998, urging researchers and policymakers to address water-intensive cropping practices and foster climate resilience for a sustainable future in the region.
The study analysed long-term rainfall patterns in Madhya Pradesh using high-resolution gridded data from 439 IMD grid points.
The results show a significant decrease in annual, monsoon, post-monsoon, and winter rainfall and a positive trend in the pre-monsoon season.
The study provides important information for policymakers to plan and implement climate-resilient regional strategies.
Climate change (CC) is not a small-scale phenomenon though it is a global problem that poses significant challenges to both human society and natural ecosystems (Kuniyal et al. 2021). It refers to long-term changes in the Earth's climate, causing variability in the patterns and frequency of hydrometeorological events that are primarily driven by anthropogenic activities such as fossil fuel and biomass burning and large-scale deforestation (El Kasri et al. 2021; Ampofo et al. 2023). CC has been linked to rising temperatures, one of this phenomenon's most visible and widely documented effects. Anthropogenic activities that release greenhouse gases into the atmosphere are primarily responsible for the 1.1 °C rise in the average surface temperature of the Earth since the pre-industrial era, as highlighted in a recent assessment report by the Intergovernmental Panel on Climate Change (IPCC 2021; Kumar et al. 2023; Singla et al. 2023). With the continuously rising temperature trends, CC has also led to significant changes in rainfall patterns. The recent IPCC assessment report notes that extreme hydrometeorological occurrences are predicted to become more frequent and intense across many areas, including regions of North America, Europe, and Asia. This trend may rise in the coming decades, increasing the risk of water-related disasters (IPCC 2021).
Rainfall is a crucial climatic factor, and disparities in the pattern can directly or indirectly affect agriculture production, human lifestyle, water resources management, and ecosystem function and structure (Kumar & Gautam 2014). Several scientific studies have also provided evidence of changes in rainfall patterns due to CC. A study by Trenberth et al. (2014) found that extreme rainfall events have intensified in many regions of the world, including North America, Europe, and parts of Asia. A similar study by Sheffield et al. (2012) showed that abrupt changes in precipitation patterns have increased water stress in many regions, including the Mediterranean and southwestern United States. In recent decades, significant changes in the global climate, characterised by unexpected temperature and rainfall pattern shifts, have emerged. These changes carry substantial implications for the social lifestyle and economy of rural populations, who primarily rely on agriculture as their main source of income. Their dependence on seasonal monsoons and environmental sustainability further underscores the potential impact of these climate variations (Nadeau et al. 2022). Various studies worldwide have revealed changes in rainfall patterns over different time frames (Das et al. 2022).
In recent years, India has been affected and challenged by CC to its natural resources, notably its forest areas (El Kasri et al. 2021) and agriculture (Kumar & Gautam 2014). Between 1951 and 2015, rainfall decreased on average, but there was also a rise in the annual mean, maximum, and minimum temperatures, with warming rates of 0.15, 0.15, and 0.13 °C, respectively (Krishnan et al. 2022). India's economy, agriculture, and food security depend on timely rainfall, directly or indirectly (Abrol & Gupta 2019). Researchers have attempted to understand the scope and patterns of seasonal and annual rainfall throughout the country. In several parts of India, Kumar et al. (2010) and Machiwal et al. (2017) investigated long-term rainfall variations on a local geographical scale. They reported both regional and temporal heterogeneity in rainfall trends. However, as evidenced by several recent studies, India's rainfall patterns are already significantly impacted by CC. Katzenberger et al. (2021) found that the country has experienced a decline in total rainfall in recent decades, with some regions, particularly in the northeastern and southern parts of the nation, experiencing an increase in the number of severe rainfall incidents. Shah et al. (2021) observed that the state of Gujrat in India has experienced a significant decline in the proportion of rainy days throughout the monsoon season as well as a decrease in average rainfall over the past few years in their research on the impacts of CC on rainfall patterns in the state. Banerjee et al. (2020) assessed surface observatories in the catchment area witnessed a decreasing rainfall trend from 1983 to 2008. On average, the decline in rainfall was 15.75 mm per decade during this time period in the Uttarakhand state of India. These long-term changes in the environment will substantially impact the availability of food, water resources, and the general well-being of the people, highlighting the need to take immediate action to lessen the consequences of CC in the country. While regions of the central west, Madhya Pradesh (MP), a part of the northwest peninsula, exhibit a substantial decreasing pattern in monsoon rainfall, a sizeable decreasing trend was also seen in the northwest, central, and west coast peninsulas (Sam et al. 2020). Yadav & Singh (2023) investigated rainfall variation in East and West Madhya Pradesh from 1871 to 2016. They found monthly trends in West Madhya Pradesh indicated a significant decrease in June but a significant increase in August. However, seasonal and annual trends in West Madhya Pradesh are not significant. In East Madhya Pradesh, annual and seasonal monsoon rainfall decreases significantly, particularly in June and July. Som & Dey (2022) focused on rainfall trends in the Bundelkhand region of Madhya Pradesh. Their results were non-significant negative trends for annual rainfall patterns. But summer and winter seasons exhibited high precipitation variability. Sharma et al. (2021) indicated that the seasonal and annual rainfall increased in the Betul district of Madhya Pradesh, but the rate of increase was not statistically significant.
Limited studies have been conducted on seasonal variations and trends in rainfall patterns of Central India based on long-term gridded rainfall data. This study deals with a detailed analysis of the spatio-temporal variability and rainfall trends, which is essential to comprehending the regional effects of CC on Central India. The main objective of this study, which utilises long-term (1951-2021) gridded rainfall data, is to evaluate the trends and magnitude of seasonal (Annual, Monsoon, Pre-Monsoon, Post-Monsoon, and winter) rainfall in Madhya Pradesh. Descriptive and GIS analyses have been performed for better visualisation and understanding of changes in rainfall trends. The Mann-Kendall (MK) test and Sen's Slope estimator have been employed for trend analysis, and Pettitt's test to detect sharp transition points in annual and seasonal rainfall. For proper visualisation of change in rainfall pattern, the change point year of annual rainfall was used as the base year to compare seasonal and monthly rainfall, and the long-term time series data was separated into two periods – 1951–1998 and 1999–2021 – to understand the spatio-temporal patterns. The findings of this study offer insights into the historical changes in rainfall patterns in Central India from 1951 to 2021. However, further research with future scenarios would be needed to understand the region's response to CC. Researchers and policymakers can gain valuable insights into the region's response to CC by comprehensively analysing the spatio-temporal rainfall variance in Central India. This knowledge will allow them to assess the effects of changing rainfall patterns on critical sectors like agriculture and water resources. With this knowledge, they can develop effective strategies to mitigate the detrimental impacts of CC on the region.
MATERIALS AND METHODS
The three main seasons in Madhya Pradesh are summer, monsoon, and winter. The state's average maximum temperature is 34.6 °C during summer (March to June). The area often sees light dust storms and has very low humidity. The average annual rainfall is 1,194 mm. Due to high elevation, the western region gets 1,000 mm or less precipitation, whereas the southeastern areas receive up to 2,150 mm in specific locations. The majority of the state's rainfall happens from June to September. Mid-June marks the beginning of the southwest monsoon, which brings a typical monsoon environment to Madhya Pradesh. While some northwest regions receive less precipitation, the south and southeast regions witness significant rainfall. More than 1,500 mm of rainfall is recorded in areas like Mandla, Jabalpur, Balaghat, Sidhi, and other extreme eastern regions. However, less than 800 mm of rain falls annually in the western parts of Madhya Pradesh.
The required data for this study was taken from the Indian Meteorological Department (IMD), Government of India, from 1951 to 2021. IMD utilised 6,955 rain gauges across India to create high-resolution gauge-based gridded daily rainfall data with a spatial resolution of 0.25° × 0.25° (Bharti et al. 2016), which is responsible for providing weather and climate-related forecasts and warnings in India. It involved incorporating the data from the rain gauges to form a grid of precipitation values that spanned the entire country. The current study uses daily rainfall gridded data for 1951–2021, totalling 439 grid points. Since there is a lack of weather station data in the study domain, we used long-term gridded data from the IMD (https://www.imdpune.gov.in/). Observations from weather satellites (INSAT series) and gauge station records were used to generate this data series (Roshani et al. 2023). Several studies have utilised the IMD gridded dataset as an observed/reference dataset, and several hydroclimatological studies have evaluated this gridded rainfall dataset (Gupta et al. 2020). Collecting this data is crucial due to a shortage of weather stations, restrictions on observation, unequal distribution, and lack of data. Additionally, the year is segregated into four seasons: pre-monsoon (March to May), monsoon (June to September), post-monsoon (October to November), and winter (December to February) (Shree & Kumar 2018). The monsoon season has heavy rainfall, post-monsoon sees a decrease in rainfall, pre-monsoon is hot with occasional thunderstorms, and winter is cool and dry. The timing and length of each season can vary depending on the location and climate.
Data analysis techniques
Trend analysis is a key tool for effective planning along with the management of water resources. By analysing historical patterns in these variables, it is helpful to accurately forecast the future behaviour of hydrological parameters like rainfall, discharges, and direct runoff (Dinpashoh et al. 2014). This research evaluated annual and seasonal rainfall patterns using non-parametric methods. For this purpose, the SS estimator (magnitude of change) and the MK test (Mann 1945; Kendall 1975; Kumar et al. 2023) were used (Figure 2). These tests are often used in climate research to spot monotonic trends and sudden shifts (Chakraborty & Joshi 2016). The p-value determines how statistically significant the trend is, with a lower p-value suggesting a greater level of significance for the observed difference (Roshani et al. 2023). It should be noted that these tests need an independent data pattern. The 5% level of significance was used to analyse these patterns.
Mann-Kendall (MK) test
Sen's Slope (SS)
Rainfall variability test
Coefficient of variance (CV)
According to Asfaw et al. (2018), the level of variable climate fluctuation can be categorised as very high (CV > 40%), high (CV > 30%), moderate (20% < CV < 30%), and low (CV < 20%).
Change point analysis
Inverse distance weighted (IDW)
Statistical characteristics of rainfall
Table 1 displays descriptive statistics for Madhya Pradesh's annual and seasonal rainfall data from 1951 to 2021. This table contains five categories of rainfall measurements: annual, monsoon, pre-monsoon, post-monsoon, and winter. Table 1 includes the minimum and maximum values, mean, standard deviation, CV, and contribution percentage for each category. In this region, rainfall varies throughout the year, with different seasons bringing different amounts of rain. Annual rainfall for the region amounted to 1,041.36 mm, with 40 years experiencing below-average precipitation. It means that, on average, the region receives moderate rainfall annually. However, there is a significant variation in rainfall from year to year, as evidenced by the high standard deviation of 179.18 mm. It means that some years can be much wetter or drier than others. The recorded maximum and minimum annual rainfall values are 1,530.20 mm (1961) and 645.72 mm (1965), respectively, underscoring the variation in rainfall amounts within the specified time frame. This region received 488.84 mm of surplus rainfall in 1961 while a deficit of 395.64 mm in 1965. The distribution of rainfall in India is classified into four categories, namely, areas of heavy rainfall (over 2,000 mm), areas of moderately heavy rainfall (1,000–2,000 mm), areas of less rainfall (500–1,000 mm), and areas of scanty rainfall (less than 500 mm) (https://mospi.gov.in/sites/default/files/Statistical_year_book_india_chapters/ch34.pdf).
|Categories .||Mean (μ) .||Standard deviation (σ) .||CV (%) .||Max (mm) .||Min (mm) .||Contribution (%) .|
|Categories .||Mean (μ) .||Standard deviation (σ) .||CV (%) .||Max (mm) .||Min (mm) .||Contribution (%) .|
Monsoon season refers to the amount of rain that falls during the monsoon season. It is the period when the region receives the most rain (949.02 mm), but in 39 years, rainfall was below the mean value. The standard deviation of 166.19 mm suggests more variation in rainfall from year to year but not as much as in the annual category. The maximum value of 1,367.22 mm (1961) indicated the occurrence of heavy rainfall in some years and observed a surplus of 418.2 mm of rainfall. The minimum value of 597.97 mm (1979) signified a significant rainfall deficit of 351.05 mm, suggesting that some rainy seasons may receive only half of the average rainfall. The surplus and deficit of monsoon rainfall influence the frequency and intensity of climate extremes such as floods and drought, which negatively impact agricultural activities.
The pre-monsoon season shows the rainfall received in the period leading up to the monsoon season. This period usually experiences less rainfall (20.57 mm) than the monsoon season. Among the observed years, 48 times had rainfall below the mean value during this season, indicating some variation (σ = 17.07) in rainfall from year to year. The maximum value of 84.05 mm (2015) shows infrequent considerable rainfall during this period, while the minimum value of 2.43 mm (1973) suggests that this period increased drought intensity. Notably, there was a surplus of 63.48 mm of rainfall in 2015 and a deficit of 18.14 mm in 1973 during the pre-monsoon season.
The post-monsoon season follows the monsoon and typically experiences reduced rainfall (42.71 mm) compared to the monsoon season, with 44 instances of rainfall occurring less than the mean value during the post-monsoon season and a deviation of 36.32 mm recommends a comparatively high variation in rainfall from year to year. The maximum value of 150.17 mm (2009) specifies occasional considerable rainfall during this period, while the minimum value of 0.63 mm (2011) proposes that this period is the driest. In 2009, it showed an excess of 107.46 mm of rainfall and a shortfall of 42.08 mm in 2011 during the post-monsoon season.
The winter season signifies the period of rainfall during winter and typically experiences less precipitation (29.05 mm) compared to the monsoon season with a moderate variation (σ = 23.19). Observations show that in 50 instances, the region received rainfall lower than the mean value during winter. The standard deviation of 23.19 mm suggests a high variation in rainfall from year to year. The region received 68.84 mm more rainfall (maximum value of 97.89 mm) than the mean value in 1997 and a deficit of 28.66 mm of rainfall (minimum value of 0.39 mm) in 2000, proposing Rabi crops were affected more due to rainfall uncertainty. The observed variations in rainfall across seasons affect the region's water resources, agriculture, and overall climate dynamics. Understanding these patterns is vital for formulating effective strategies for water management and CC adaptation.
Rainfall variability pattern
The coefficient of variation (CV) is utilised to assess the degree of variability in the rainfall data (Shree & Kumar 2018; Bharath et al. 2023; Kumar et al. 2023). Table 1 also shows the CV values (Equation (6)) for different rainfall variables across various seasons. The annual rainfall has a CV of 17.21%, indicating low variability. The monsoon season also shows low variability, with a CV of 17.51%. In contrast, the pre-monsoon season exhibits very high variability, reflected by a CV of 82.99%. Similarly, the post-monsoon and winter seasons display significant variability, with CVs of 85.04 and 79.82%, respectively. Overall, the results demonstrate the varying levels of seasonal rainfall variability. Low variability in annual and monsoon rainfall can provide some predictability and stability, but the high variability in pre-monsoon, post-monsoon, and winter seasons calls for adaptive water management strategies, infrastructure development, and CC mitigation measures. Policymakers and stakeholders need to consider long-term planning to cope with potential water scarcity and agricultural challenges arising from these fluctuations in rainfall patterns.
Temporal analysis of rainfall trends
|Categories .||MK test (Z) .||Sen's Slope (Q) .||P-value .||Trend .|
|Categories .||MK test (Z) .||Sen's Slope (Q) .||P-value .||Trend .|
The negative trend in annual rainfall suggests that the region may be experiencing a decline in overall rainfall patterns, which could adversely affect agricultural production and other water-dependent activities. Similarly, decreasing monsoon rainfall could result in water shortages during the critical monsoon season, reducing water supply. On the other hand, the pre-monsoon (Z = 0.288) season shows (Figure 3(c)) a non-significant increasing trend with an annual magnitude of 0.018 mm. The rising (positive) trend in pre-monsoon rainfall suggests that this season may receive more rainfall and that changes in rainfall patterns will positively impact water availability to agricultural production and other sectors. The post-monsoon (Z = −0.764) and winter (Z = −0.735) seasons have shown (Figure 3(d) and 3(e)) a non-significant decreasing trend with an annual magnitude of −0.116 and −0.067 mm, respectively. This trend is particularly concerning since these periods are critical for recharging groundwater and maintaining water levels in rivers and reservoirs.
Monsoons are complex phenomena influenced by various factors, including sea surface temperatures, land-sea temperature gradients, and atmospheric circulation (Singh et al. 2019). Alterations in any of these factors can lead to changes in monsoon patterns and reduced rainfall during the monsoon season. Land use changes, deforestation, and urbanisation are examples of local and regional climate factors that can affect rainfall patterns and cause a reduction in precipitation during specific seasons (Lal et al. 2021). The decreasing trends in annual, monsoon, post-monsoon, and winter rainfall are putting pressure on water resources, affecting agriculture, drinking water supply, and industrial needs. Effective water management with conservation, rainwater harvesting, and efficient irrigation is crucial to cope with potential scarcity during dry periods. Declining annual and monsoon rainfall may negatively impact agricultural productivity, reducing crop yields, and economic losses. Conversely, increasing pre-monsoon rainfall offers opportunities for crop diversification and improved output. Pre-monsoon rainfall may increase due to variations in ocean currents and sea surface temperatures, such as those caused by the El Niño-Southern Oscillation (ENSO) (Loo et al. 2015). These findings underscore the need for CC adaptation and mitigation. Policymakers should plan for infrastructure development, sustainable water use, and long-term strategies to adapt to changing rainfall patterns.
Change point analysis
The outcomes indicated that the trend pattern abruptly changed after the 1980s, except for pre-monsoon rainfall. As mentioned earlier, annual, monsoon, post-monsoon, and winter rainfall are declining. So, in graphical representation (Figure 4(a), 4(b), 4(d), and 4(e)), applying Pettitt's test, we can identify that after detecting the change point, it has a declining trend, represented in the red colour line (Figure 4(a), 4(b), 4(d), and 4(e)). Notably, after the change point detection, the pre-monsoon season exhibits a positive trend, represented by a green line in Figure 4(c). In Madhya Pradesh, identifying abrupt changes or change points in rainfall trends using Pettitt's test carries significant interpretations and implications for various sectors. The state has experienced notable shifts in its precipitation patterns over the past few decades (Duhan et al. 2013). The detected change points can have hydrological impacts, affecting riverine flow and water availability, as the Madhya Pradesh Water Resources Department (MPWRD) reported in their State Water Resources Plan (http://mpwrd.gov.in/wp-content/uploads/pdf_upload/44b776f6-34ef-4467-b13b-e808202edd83.pdf). These shifts in rainfall may also influence agricultural resilience, impacting crop yields and planting seasons (Raza et al. 2019). Additionally, the occurrence of change points in rainfall trends could have implications for ecosystem dynamics and wildlife habitat, highlighted in a report by the Madhya Pradesh Forest Department (https://www.devalt.org/images/L2_ProjectPdfs/(15)ForestsDiversity.pdf?Oid=150).
Disaster preparedness and infrastructure planning could also benefit from the insights gained from these change points, as demonstrated in the Madhya Pradesh Disaster Management Plan (https://erc.mp.gov.in/Documents/doc/Guidelines/ERC/State_Action_Plan_MP.pdf). Overall, understanding these abrupt changes is crucial for sustainable water management, agricultural planning, and climate adaptation in Madhya Pradesh, providing valuable guidance for policymakers and stakeholders to enhance climate resilience across the state. However, these changes influence the frequency and intensity of floods and droughts, directly impacting agriculture and human livelihood. So, some of the planning done by the government in the agriculture sector in Madhya Pradesh State Action Plan on Climate Change, 2012 includes the following:
Promoting soil and water conservation technologies,
Planning cropping systems suitable for each agro-climatic zone,
Management of risks for sustainable productivity,
Enhancing dissemination of new and appropriate technologies developed by researchers and strengthening research,
Agriculture information management,
Additional impetus to mechanisation and accessibility to markets,
Creation of rural business hubs, and
Capacity building for sustainable agriculture.
Spatial change analysis of rainfall patterns
The maximum average monsoonal rainfall decreased from 1,623 to 1,311 mm after 1998. Less than 25 mm of rainfall area is increasing after 1998 during the post-monsoon season. Furthermore, the winter season has seen significant changes, with less than 25 mm of rainfall area increasing and more than 50 mm of rainfall area decreasing after 1998, as shown in Figure 5. After assessing all seasons of rainfall distribution, the average annual rainfall distribution was also changed after 1998. Only the western and north parts of this region were getting less than 1,000 mm of annual rainfall during 1951–1998, but the scenario changed after 1998, and this area (<1,000 mm) increased. The 1,000–1,250 mm of rainfall area was shifted, and more than 1,250 mm of rainfall area decreased. The maximum average annual rainfall was 1,769 mm from 1951 to 1998, which changed to 1,401 mm after 1998. These findings suggest that Madhya Pradesh experienced a significant shift in its seasonal rainfall distribution after 1998, which requires further attention and investigation.
Figure 6 provides valuable insights into the spatio-temporal distribution of monthly rainfall in Madhya Pradesh from January to April. Interestingly, while January experiences a declining trend in rainfall, February, March, and April all demonstrate a positive trend. The map for January indicates that the area receiving more than 20 mm of rainfall has sharply declined, while the region receiving less than 5 mm of rainfall has increased significantly after 1998. However, in February, March, and April, the area receiving less than 5 mm of rainfall decreased, while the area experiencing more than 5 mm expanded remarkably after 1998.
Figure 7 presents a clear visual representation of the changes in the rainfall trends of Madhya Pradesh during the crucial monsoon period of May to August. Significant transformations were observed in the northern and southwest regions of the state in May, with a considerable shift in the rainfall pattern. During June, there was a variation in the rainfall patterns of the northern part, with a decrease in the 50–100 mm rainfall area and an increase in the 100–200 mm area after 1998. Moreover, the maximum average rainfall in June reduced from 204 to 186 mm after 1998. In July and August, the area with over 400 mm of rainfall experienced a sharp decline, while less than 300 mm of rainfall increased remarkably after 1998. These changes indicate the complexity of the monsoon season in Madhya Pradesh and highlight the need for continued monitoring of rainfall patterns to gain a deeper understanding and reduce the impact of these changes on the region's ecosystems and communities.
The changing pattern of rainfall is highlighted in Figure 8, which reveals the spatio-temporal distribution of monthly rainfall in Madhya Pradesh from September to December. Unfortunately, the trend is not positive, with all months showing a decrease in rainfall. The September map shows a reduction of more than 200 mm of rainfall area and an increase in less than 100 mm of rainfall area after 1998. The October map also highlights significant changes, with 20–100 mm of rainfall covering most of the state area from 1951 to 1998; this scenario changed after 1998, especially in the north and northwest regions. The November map shows a decrease in 10–20 mm of rainfall area, and less than 5 mm of rainfall area increased after 1998. Finally, the December map shows a significant variation, with most of the area getting only 5–20 mm of rainfall from 1951–1998. Still, the scenario completely changed in recent years, and the entire state is getting less than 5 mm of rainfall in December.
For acquiring a better knowledge of the consequences of changing climate and for improved management and planning of resources and infrastructure, climate variability, trend detection, and spatial distribution analysis are essential. Meteorological variables are a primary source of information for evaluating these trends and understanding the changes happening in the climate system. Non-parametric tests are often used in climate research (Ampofo et al. 2023) because they can detect trends in time series data without making assumptions about the underlying distribution. These tests analyse multiple meteorological variables at various scales, including temperature, precipitation, humidity, wind speed, etc. Spatial distribution analysis is also essential in climate research as it helps to understand the regional differences in climate variables and how they change over time. This study attempted such efforts with rainfall data and statistical techniques. The high CV values for the pre-monsoon, post-monsoon, and winter variables suggest that the amount of rainfall received during these seasons can vary significantly from year to year. The low CV values for the annual and monsoon variables indicated that the amount of rainfall experienced within these periods might also change but not to a significant degree. The findings of a recent study on rainfall variability concur with those of other studies, such as Chandniha et al. (2017), Shree & Kumar (2018), and Warwade et al. (2018). Specifically, the study found high pre-monsoon, post-monsoon, and winter rainfall variability.
In contrast, annual and monsoonal rainfall exhibits a low degree of variability. According to earlier research, the region is vulnerable to floods and droughts (Pandey & Ramasastri 2001). The study also discovered considerable interannual variance. Applying the MK test and Sen's Slope reveals a significant shift in rainfall patterns in Madhya Pradesh, with a general negative trend in annual, monsoon, post-monsoon, and winter precipitation, while pre-monsoon rainfall shows a positive trend. However, these trends are statistically non-significant. The inaccessibility of century-scale data for the same study area might be one of the critical causes of non-significant trends. No previous study is available that considers the same period and study area. However, our results are consistent with several Central and Central West India research. For instance, Kundu et al. (2017) reported a negative trend in monsoon, post-monsoon, winter, and rainfall annual over the past 111 years (1901–2011), while pre-monsoon reported an increasing trend. Jain et al. (2023) found a non-significant declining trend in annual and monsoonal rainfall in Madhya Pradesh during the last 146 years (1871–2016). Similarly, Kumar et al. (2010) identified a moderate coefficient of variation (CV) of 18% and a declining trend of annual and monsoonal rainfall, and a positive trend of pre-monsoonal rainfall over the past 135 years (1871–2005). Pal & Al-Tabbaa (2011) also reported a negative annual and monsoonal rainfall trend in central India during 1951–2003. Rai et al. (2014) and Devi et al. (2020) also have observed similar trends of decreasing rainfall across Madhya Pradesh.
A crucial feature of climatic time series data is changing point analysis, which identifies when the data trend abruptly changes. There are many methods to detect change points, but in this study, Pettitt's test was applied to detect sharp transition points in Madhya Pradesh's rainfall data. Pettitt's test result (Figure 4) showed that the annual rainfall trend abruptly changed from 1,054.69 mm in 1998 to 1,013.53 mm in 1999. Similar studies were also done in different regions to detect sharp transition points in climatic variables. Zarenistanak et al. (2014) used time series data (1950–2007) of Iran and observed that the mutation point occurred in 1973. Kumar et al. (2023) used three stations' data, i.e., Mukteshwar, Hawalbagh, and Almora, from 1980 through 2019. They identified that annual rainfall trends were abruptly changed in 2004, 1998, and 1991, respectively. The spatio-temporal distribution of seasonal and monthly rainfall maps created based on the average for each grid point's rainfall data and time series data were split into two parts based on Pettitt's test result for the comparative study. These maps (Figures 5–8) represent how seasonal and monthly rainfall changed and shifted from one place to another. The maximum average annual rainfall decreased from 1,769 to 1,401 mm after 1998. Yadav & Singh (2023) also reported that extreme rainfalls have decreased in the last few decades. Sengupta & Thangavel (2023) used meteorological data to examine the changes in the distribution patterns of rainfall, temperature, and severity of drought (Standardised Precipitation Index – SPI) over the study period (1990–2015). Their MLR results conclude that the drought severity is high due to changing precipitation patterns in Maharashtra, significantly threatening the cotton yield level.
CC and variability provide a severe challenge to human livelihood. Due to the changing pattern of rainfall, at the same time, some regions are affected by floods, and other regions are facing droughts. As per our findings, it is visible that the rainfall pattern was changed after detecting the change point. The major outcomes of this research emphasise the need for continuing climate and rainfall pattern monitoring and analysis in the area to create efficient policies and strategies for controlling and adapting to the changing climate and its influence on the environment and human activities in the region. These findings highlight the region's challenges in managing water resources and dealing with extreme weather events. Researchers and policymakers must consider these trends to address water-intensive cropping practices and ensure climate resilience. The study provides valuable insights for informed decision-making and resource allocation, aiming for a sustainable and climate-resilient future in the region.
This study analysed variability patterns, temporal trends, change point detection, and spatio-temporal rainfall distribution in Madhya Pradesh. The non-parametric tests were utilised to find trends in time series data of different meteorological variables.
The CV analysis depicts low variability in the annual and monsoon rainfall data, while there is significant variability in the pre-monsoon, post-monsoon, and winter seasons. It suggests significant seasonal variation in the amount of rain received during the pre-monsoon, post-monsoon, and winter seasons.
Pre-monsoon rainfall showed an increasing trend with an annual magnitude of 0.018 mm, while annual, monsoon, post-monsoon, and winter rainfall all showed an overall decreasing trend. Although these trends are statistically non-significant, a reduction in overall average rainfall in the monsoon season may considerably affect water availability and agricultural dependents.
Change point analysis was applied to detect when the trend of the rainfall pattern changed abruptly. The change points for the monsoon, pre-monsoon, post-monsoon, winter, and annual were 1998, 1955, 1987, 1986, and 1998, respectively. We had taken change point results of annual rainfall as a temporal reference and depicted the seasonal and monthly spatial distribution of rainfall in various maps.
The spatio-temporal distribution of seasonal and monthly rainfall maps showed irregular rainfall in MP. The main concern is that some areas in the Hoshangabad and Balaghat districts of MP, which received higher precipitation, more than 1,300 mm earlier, completely decreased after 1998. Additionally, the maximum average monsoonal rainfall decreased from 1,623 to 1,311 mm after 1998, implying that water availability drastically declined. It is an important issue to note the policymakers because, on the one hand, the water-intensive cropping area, namely wheat, rice, and soyabean is increasing, and on the other hand, the reduction in average rainfall. These findings suggest that Madhya Pradesh experienced a significant shift in its seasonal rainfall distribution after 1998, which requires further attention and investigation.
The study emphasises the need to continue monitoring and analysing climate and rainfall patterns in the region to develop efficient strategies and programmes for controlling and combating changing climate and its consequences on the environment and anthropogenic activities in the region.
Thus, these findings highlight the importance of understanding the variability in rainfall data across different seasons for effective water resource management and agricultural planning.
The first and second authors are thankful to the University Grant Commission (UGC), New Delhi, for supporting this research with their Senior Research Fellowship (No: 190520123449/(NET-DEC 2019) and 3585/(NET-DEC 2018)). The authors also want to express their gratitude to the Indian Meteorological Department (IMD) for providing the high-resolution gridded rainfall data, which allowed them to analyse the changes in rainfall patterns in Madhya Pradesh, Central India, over a long period. The authors also acknowledge the School of Humanities and Social Science at the Indian Institute of Technology Indore, the Department of Computer Science & Engineering at the Indian Institute of Information Technology Ranchi, the Department of Civil Engineering at the Indian Institute of Technology Indore, and the Department of Geography at Central University of Tamil Nadu for their valuable contributions to this study. Together, they hope that their findings will contribute to a better understanding of the region's rainfall variations and help address water-related challenges for a sustainable future.
It is noted that at stage 3 ( data analysis), different stationarity tests have been conducted (ADF, KPSS, etc.) and the annual series of different rainfall seasons found to be stationary at level, i.e., I (0).
DATA AVAILABILITY STATEMENT
The data utilised in this study was sourced from the Indian Meteorological Department (IMD) Gridded Rainfall Data (https://www.imdpune.gov.in/). The availability and access to the data used in this research are subject to the policies and regulations of the Indian Meteorological Department. Researchers interested in accessing the data are advised to contact the Indian Meteorological Department directly for further information on data availability, access, and any required permissions.
CONFLICT OF INTEREST
The authors declare there is no conflict.